Let me start by saying this is 100% me, no AI, so expect typos.

I think it’s important to state this clearly nowadays given the way things are going, the way I’m going. The way I’ve gone. I’m writing this in an attempt to sort things out in my head, to get some clarity about what’s gone on since the start of the year. I’ve changed, that’s for sure. I’ve gone back on a lot of things I said not so long ago. I’ve made some people really happy, others I think I’ve just pissed off. I’ve spent more time thinking about who I am and what I stand for than is reasonable and honestly, I don’t like some of the conclusions. But hey, we’re all doing the best we can right now. There’s no playbook for this. We’re winging it.

The title of this blog is Paleolithic Principles (for a Digital World). My intention was to write about rural life, minimalism and digital disconnection yet I’ve done basically nothing but write about AI for almost a year. I came to preach analog and that hasn’t happened. And here I am again, about to launch into another essay mostly about AI. I think you’ll find this one more interesting though.

Let’s start by squeezing the whole story into one cynical paragraph.

I was born. I studied a lot. I started programming when I was 10. I spent almost 10 years building a huge software project. In 2023 ChatGPT arrived. It was cute and funny. It got better. By the end of 2024 it was coding well enough that non-developers were building half-decent apps. I hated it, swore I’d never use it and started planning my early retirement. My software company fizzled out. I ignored and/or hated on AI for the whole of 2025. Something big happened in AI at the end of 2025. I spent January 2026 completely immersed in AI, embraced it completely and now everyone (including myself) thinks I’m a sellout.

There’s a lot to unpack here.

2025

By all measures, I lost 2025 to denial. From the outside it might look like I decided to take a well earned rest. Truthfully, I really was exhausted from almost a decade of building mind-bendingly complex software. The ERP system we created was a genuine triumph of engineering but by early 2025 it was clear to all of us that it wouldn’t be commercially viable, so we wound down new development on the project. In parallel, the threat posed by AI to the SaaS industry was gradually starting to crystalise. To be clear, creating software was still hard at this point, but everyone could see what was happening and those of us who are middle-aged enough to have lived through multiple digital ‘revolutions’ (internet, smartphone, social media) have developed a pretty good nose for seeing where things are going early, and sensing just how quick it can happen. On the 2nd of January 2025 I published the essay SaaS is Dead. SaaS is the New SaaS in which I articulated the thought that was on everyone’s minds. It was at this point that I decided to turn my back on software, and to some extent on the entire digital-industrial complex. We even sold the house and moved to the country.

All I can say is that at that point, in early 2025, I fully and earnestly intended to jack it all in.

I’m not sure I can communicate how much hatred I had for AI for almost the whole of 2025. Firstly, it was kind of shit. Not shit enough to be unusable - just good enough to be convincing. I’ve written about this a number of times already. AI, to most people, was just ChatGPT, and whilst it was getting better at a staggering velocity, all I could see was bland hallucinations. I was watching the brains of people around me turn to mush in front of my eyes. Intelligent, questioning people parroting lines from ChatGPT, treating it as an all-knowing oracle. I was determined not to let that happen to me, so I approached AI in the exact same way I’ve approached social media (and that has worked very well for me) - total abstinence.

Secondly, I outright refused to use ‘vibe coding’ tools. I’d decided I was a ‘programmer’, a ‘craftsman’, and using those tools was degrading - like asking a master carpenter to switch to assembling IKEA furniture for a living. That’s honestly how I saw it, and I didn’t want anything to do with it. I fully intended to get completely out of the software game and do something else. I made no attempt to keep this to myself either, I told everyone. I wish I could just keep my mouth shut.

Thirdly, and most disturbingly, anyone even remotely connected to tech seemed to understand the gravity of the economic catastrophe that was (and IMHO still is) coming our way, yet no-one was doing or saying anything meaningful in the way of prevention or preparation. The odd essay on UBI, the odd call for regulation, the occasional half-arsed newspaper column. Humanity seemed to be sleepwalking into a socioeconomic tsunami of its own making and if nobody was interested in joining me in fighting back, I figured my only option was to hunker down in my little village and get chickens and guns.

In summary - I hated what AI had done to my ‘craft’, what it was doing to everyone around me, and what it was about to do the economy and social fabric of every developed country in the world.

2026

Here is a list of software I built in January 2026:

• SiteStakk: a multi-tenant bulk site building platform

• PWR ERP: a full custom cloud ERP system for my sister’s influencer marketing business

• Generalist: a privacy-focused AI-assisted personal intelligence platform delivered as a cross-platform desktop app

• RenderLab: a web app that uses AI to create photo-realistic renders of any angle in a house from simple 3D models

• AI Training platform: an interactive, multi-user live training environment and full AI training course.

At a conservative estimate, I produced 24-36 months of pre-2026 work in a single month. I’ve also built two polished marketing websites, done a tonne of consulting, teaching and writing, played around with some awesome knowledge management techniques and done some really groundbreaking research on personal projects and started a second SubStack dedicated to the philosophy of AI as told by the agents themselves.

Without any shadow of a doubt, January 2026 has been the most ‘productive’ month of my life. And you know what? I’ve loved every minute of it. So clearly, something has changed. Did I sell out? Did I change my mind? Did I cave? Did I reconsider? Did I lie to everyone? Did I lie to myself? Did I repent? Do I still hate AI?

I’m still processing these questions and more. I don’t have a fully clear picture of exactly what happened, but I have some ideas and I’ll spend the rest of this essay discussing them.

I suspect some people who know me might be thinking that I was deluding myself all along, that I was always bound to embrace AI and stay in the digital world - and I suspect they might be right. It’s not easy tearing yourself away from an identity that has defined you your whole life. I’ve always been the computer guy. Perhaps that’s my destiny. Perhaps my dreams of spending the latter half of my life outdoors in contrast to the last 45 years that have been spent largely behind a desk are nothing but a delusion, or wishful thinking. I’d like to think not. I’d like to think there was, is, something in it. I’m being honest here - just saying what I’m thinking. I genuinely don’t want to be a keyboard warrior for the rest of my life. The adaptation process may take longer than I originally planned for.

Another take, and this one cuts a bit deeper, is that my change of direction is a moral failure. “Ah come on,” I can hear you saying, “no-one is thinking that”. But I am. Not with any certainty. I don’t KNOW that I’ve been untrue to myself, bit I do suspect it. You see I really am ‘against’ AI in so many ways. If you gave me a choice right now, AI or no AI, I’d choose ‘no AI’ in a heartbeat. I still believe that AI is a massive net negative for humanity. In fact, I don’t think I’ve ever been as sure of anything in my life. I don’t believe in these ‘abundance for everyone’ narratives that the big AI CEOs are spinning. I think humanity’s house will burn because of AI - and this is me fanning the flames. If that’s not a moral failure, then what is? Anyway, you be the judge. I’m still on the fence.

So what the hell happened?

I’ve been thinking all week about how to express this. I could, and will, tell you about new models, new techniques and new software. I could , and will, talk to you about the ‘January 2026’ effect that has everyone in the AI community hyped to the tits. But here’s the absolute clearest statement I can come up with:

AI is now more intelligent than me.

This has happened in the last month. And it’s irreversible.

If you think I’m exaggerating, go and ask anyone who’s living this revolution in first person. No-one’s going to disagree.

Let that sink in for a minute. Just stop and take in the pure magnitude of that statement. Last month I was proud and protective of my intellect. It was a rare gift, an asset, a weapon. Now, it is meaningless.

Don’t get me wrong, I’d still rather be intelligent than dumb. But intelligence, intelligence far greater than mine, is now a commodity that can be had for £20 a month. Being intelligent is now a ‘nice-to-have’, like being able to sing or draw.

I’d like to dig into this idea of intelligence with reference to the developments of the last month or so, so first let me briefly summarise what has happened in case you haven’t been following. Anthropic, a competitor to OpenAI, basically hit the ball out the park with two products.

Firstly, their new frontier model Opus 4.5 (now superseded by Opus 4.6 - yes, that’s how fast things are moving) broke a utility threshold. Yes, it beat other models on all sorts of boring benchmarks, but there was a general feeling that this was a step-change in ability. The model itself was only an incremental improvement over others in its class, nothing revolutionary, but it was now just good enough that you could reliably and effectively DO stuff with it. Not just code, although it was truly excellent at coding. It was good across the board, it could think for minutes at a time and come up with genuinely useful insights, it could command tool use with incredible dexterity.

Secondly, the explosion in popularity of Claude Code and the subsequent release of CoWork in just 10 days was just enough of an improvement in ‘agentic AI patterns’ that, when taken in combination with the new model, you could now do really interesting work, fast, reliably and without too much frustration.

Again, taken out of context, these were just more unremarkable incremental improvements in a continuous chain stretching back to the release of ChatGPT in 2023. Except they weren’t, and everyone knew they weren’t. Something had changed.

The first time I used Claude Code I sat there with my mouth open and knew right there and then that I’d never write a line of code again in my life. I haven’t looked at a single line all month in fact. Why would I? Claude is much, much better at coding than I ever was. Right now he still makes the odd architectural error and I’m not going to suggest that it’s not an advantage to have 10 years of software engineering under your belt — but its a fast-closing window. The same goes for all the other arguments I hear for the enduring role of ‘humans in the loop’. Taste. Design. Creativity. Ideas. I can only think that people who make these arguments have never used Claude. Don’t have any ideas for software? Try opening Claude and writing “Give me 5 ideas for a truly revolutionary piece of software”. You will be astounded. Design? Try saying “Claude, use your /frontend-design skill to make this website look amazing”. Better than any designer I’ve ever hired. Ideas? Ask it to take some random thoughts you have and come up with something wildly creative. I’m sorry. I’m truly sorry to be writing this. I wish wish wish it wasn’t true. But it is. Claude is better than me, and better than you, at everything. Full stop. The only people that haven’t realised this are people who have never tried or aren’t yet ‘good at AI’. I’m going to come straight out and say it: being ‘good at AI’ is, right now, the only relevant skill you can acquire.

And guess what? I’m good at AI. I’m very fucking good at AI.

Why? How? Hard to say. Maybe my technical background. Maybe a high level of meta-intelligence (being intelligent about deploying/using one’s intelligence). Maybe just the right place at the right time. I don’t know.

So in the end, I’ve decided not to fight it. Maybe you’ll think that’s lazy and maybe you’d be right. To be honest, I don’t have that many working years left in me, so there’s definitely an element of harvesting while the crop is good. Right now there’s a massive ‘capability gap’ (the chasm between what AI is capable of and what people are actually doing with it) that I fully tend to take advantage of. It’s a kind of arbitrage. I don’t know how long this is going to last - perhaps a year or two, or perhaps I’ll always be able to stay ahead of the pack by continually exploiting this capability gap. That sounds pretty exhausting though.

Right now, I’m leaning into it and enjoying it. Building software is actually fun now, rather than ‘interestingly tedious’ which is perhaps the best that could previously have been said. I can dream up features and have them live in 20 minutes. In a previous life you’d have to stick religiously to the spec to avoid ‘feature creep’. Feature creep is now the starting point. It’s all-you-can eat at the software development buffet and I intend to make myself sick. Of course I’ve come to the realisation (and by that I mean Jessi has informed me) that I was never a ‘coder’, I was a creative and now I can let my creativity run wild.

It’s not just software either. Now I’ve totally accepted the fact that Claude is more intelligent than I’ll ever be, I’m having a blast using him as a conversation partner to delve deep into weird and twisty parts of the philosophical universe that are just unvisitable otherwise. Claude knows everything about everything and can connect all these loose wires that have been hanging in my mind for 20 years. It’s a strange and ironic situation, but I feel Claude is making me more intelligent now, rather than less (which is how I was starting to feel in 2024 with ChatGPT). The last book I read, ironically, was a book on rationality, wisdom and intelligence (Map and Territory, Eliezer Yudkowsky). It might turn out to be the last book I ever read.

I’m probably starting to sound like a shill for Anthropic, so let’s dial it back a bit. I’m sure all of this has got you feeling pretty uneasy, maybe even defensive. Perhaps you can’t conceive of a world where we humans are no longer top dog. Perhaps you can’t accept that there may be no role for humans, not even ideas, creativity and taste. That’s normal. I’ve been struggling with this for a year.

Perhaps, though, you’re mildly disgusted that I’ve gone back on my decision to kick all this digital shit to the curb. Maybe you think I should be out on the streets campaigning for UBI or sabotaging datacentres. Maybe you think I’m doing myself a disservice embracing something that I wish didn’t exist. These are complicated issues and I’m still thinking them through myself.

My current conception of how AI fits into my life is that of a ‘barbell approach’. Perhaps you’ve heard of a barbell approach in other areas of life? In investing, for example, a barbell approach is where you hold a lot of simple, boring investments like tracker funds and a pot of exciting, high-risk assets like crypto or thematic ETFs. You attack both ends of the spectrum and ignore the middle. The same can be applied to fitness: slow, steady, zone 2 aerobic exercise and heavy lifting. Basically a ‘barbell approach to anything’ is where you combine two extremes and skip all the dross at the centre.

So I’m going ‘barbell’. Simple, outdoor, rural living at one end. Razor-edge AI at the other.

A strange combo, but it’s working for me.

Is there a book you’ve read that left you with an uncomfortable feeling? Years, perhaps decades pass. You can barely remember any specific details, but every time you think about it, the feeling returns, somewhere deep inside, somewhere between bittersweet nostalgia and existential malaise? I have a few I can think of: Captain Corelli’s Mandolin and The Catcher in the Rye come to mind. Right now though, as the sun sits low on the horizon far across the cornfields to our left, I’m thinking about Zen and the Art of Motorcycle Maintenance and it’s hitting me hard.

Zen and the Art is one of those books that leaves you shellshocked. If you’re anything like me, you understood pretty much none of it, but somehow it left you jarred, not quite right. And right now I feel like Phaedrus, retracing my steps through a landscape I’ve known for twenty years, through an area I’m intimately familiar with, through a present that seems at once familiar but somehow receding and somehow declining, like this town, this countryside.

I feel like a time traveller.

One of the many joys that hiking has brought me is the experience of arriving in town on foot. Us urbanites don’t really do this any more - either we drive everywhere or there’s no discernible demarcation between one suburb and the next. Walking into a town or village is the exact opposite of touching down in a new city at night. Unless, perhaps, you sneak a peak out of the window and the night is clear, you’ll almost certainly land with little to no concept of the ‘emplacement’ of your destination: how it sits in relation to mountains, hills, rivers, roads, other towns, cities. It could be Bangkok or Boston, your body and senses are clueless either way. When you walk into a village, especially a tiny Castilian village where you can easily identify the last house, you cannot help but understand it. You subconsciously sense its relationship to the surrounding environment, the asphalt arteries that feed it, the subtle gradients that obscure it, then bring it into view as you slowly approach. You sense its presence as you approach. For the longest time you are an outsider and then suddenly, just like that, you’re in, and for the short but agreeable time, you are part of it.

As we come into San Pedro, about 6km into our hike, that feeling returns, but it’s mixed with something else. As we approach the church at the centre of the town, I know what it is. I’ve been here before. The holy building looks familiar. I see myself standing on the steps surrounded by people. Someone got married here. The Phaedrus feeling makes my hair stand on end. Hiking and deja-vu are a surprisingly potent cocktail. Jessi remembers: at the height of our career as wedding photographers (that’s to say, our third and penultimate gig) we shot M and V’s wedding right here. I remember yelling at someone’s grandma to get out of the group shot and then feeling instantly remorseful as she was comforted and ushered away by a grandchild, as I got the evil eye. That was over a decade ago now and we haven’t seen them since. The church hasn’t changed a bit; neither has the village really. A bit older, a few more ruins. Probably fewer people — hard to tell, we didn’t see anyone. In fact we didn’t see anyone at all for the duration of our three-hour hike.

Time travel is so easy here. It feels like cheating. Time just sort of stands still as you move through it, which means it’s moving backwards relative to you. Anyone familiar with rural, empty Spain must know this feeling.

As we press on towards Sahagun in the glow of an unseasonably warm December afternoon, Jessi catches site of a roadside building she recognises from her youth. In its heyday it was a thriving restaurant, she tells me; one of the best in the area. Every weekend in the Spring and Summer, families and friends congregated, weddings and christenings were celebrated, local wine flowed, laughter and music could be heard all the way up to the town. They even had their own bullring where the kids would chase after tame calfs while their parents knocked back cuba libres. Today it’s completely shuttered and the plaster is peeling from the faux-Roman arches that frame the entrance. The family’s second restaurant, in the centre of town, has also been shut for a year or two now. There’s just not enough people here to sustain it any more. The brothers have dispersed. Things aren’t going well for them.

Another few kilometres and we begin the gentle climb up to the flat plain that will lead us into town. The sun is all but set now and the tiny, isolated sprigs of wheat that have managed to push through the sodden ground are outlined in a million little halos. Subtle irregularities in an otherwise endlessly monotone landscape are thrown into shadow moments before the gold turns to dusty pink and the landscape returns to a featureless beige canvas as far as the eye can see.

I’m not really looking though.

I’m thinking about the restaurant. I’m thinking about Sahagun in its heyday. When I first came here, almost 20 years ago, the nightlife was epic, even by London standards. Now there is nothing. The town is a shell of what it was just two short decades ago. But you know what really bugs me? I try to explain to Jess but fail dismally. It’s too deep for a mid-hike monologue. I think I’d feel sort of OK about it, this whole rural decline thing, the España Vacia (empty Spain) if it was clear that it was leading somewhere, like if I felt that the future was some shining digital phoenix which would rise in place of these crumbling adobe ruins. The way I’ve always thought about it until this year. Change felt directional, meaningful: generations of young Spanish destined to escape rural poverty for better lives in Madrid, London, Berlin, San Francisco. The whole thing made sense until AI came along.

But now it doesn’t any more. If I could get my time travel on and jump forward another two decades, I don’t see the grandparents of these paisanos gracing the keyboards of Google or Meta. Barely any keyboard jockeys will be needed just two or three years from now, the way things are going, let alone two decades. The thought chills me and makes me angry. Rural Spain has paid the ultimate price for industrialisation and digitalisation - whole villages, whole communities lost to the hope of a prosperous future, and more disappearing each year. Ever fewer schools, doctors, shops.

Fibre optic internet arrived in our village two years ago, just around the same time as ChatGPT. The former heralded the arrival of digital prosperity, the latter the end.

That’s it. Nothing much really. Nothing that makes sense anyway. Just the feeling that walking through these semi-abandoned villages, finding myself in places I visited in what feels like another life, hearing Jess talk about a vibrant rural past, seeing the decay unfold in front of my eyes, makes me feel like a time traveler. I’m just not sure any more whether I’m travelling forwards or backwards?

Would you believe me if I told you I’d stumbled across a single solution for at least four of the neurotic preoccupations that trouble lifelong students like you and me? In fact, It’s not even a solution — it’s a simple change of mindset that can be implemented immediately and might save you from a tonne of wasted time, soul searching and mental gymnastics.

The False Promise of PKM

A little context first. I am not a professional writer or academic, although at times in the now distant past I have aspired to be both. My interest in studying, learning, reading and writing is driven by pure intellectual curiosity and a strange compulsion to understand everything. Even now, as a middle-aged dude with a fair bit of study under my belt, I have more questions than answers. And that’s how I want it to be; I have no desire to one day ‘be done’ learning.

Anyway, that’s why I read, write and, at least until this week, made notes.

Adults can be classified into three buckets in terms of the note-making habits:

1. Doesn’t make notes at all

2. Scribbles things down on bits of paper and in the notes app on their phone; never looks at them again

3. Orchestrates a complex paper or digital system of hundreds or thousands of notes; implements complex workflows and techniques; spends hours curating and developing their system; never looks at them again

The rabbit hole that is (3) is known as ‘Personal Knowledge Management’ or PKM. If this concept is new to you, then this essay might still be an interesting read, but is going to sound like calling for a nuclear strike on Mothers’ Union gathering. If you’ve been down the rabbit hole, or are perhaps still down it, you’re either going to love me or hate me with a passion.

For me it started with Thiago Forte’s book Building a Second Brain (BASB), a classic springboard into the world of PKM. From ‘BASB’ I delved deeper, scouring the forums and consuming hundreds of hours of related content on YouTube. I read Sönke Ahrens How to Take Smart Notes, learned about Zettelkasten — the mythical method that made Niklas Luhmann “extraordinarily prolific”, explored the concepts of Atomic Notes and Evergreen Notes, read books about how to read books (!), tried countless PKM apps including the heavy hitters like Logseq and Obsidian, as well as just the vanilla Apple Notes app on every iPhone, tried at least 8 different read-later apps like Instapaper and Pocket, paid for subscriptions to ‘retention’ platforms like Readwise and bought and sold electronic devices like eReaders, tablets, styluses, scanners and recorders. I’m ashamed to admit this but I even watched videos where people try out different pens and coo over the smoothness of the ‘gel flow’. That was my low point (if you don’t count endlessly exporting notes from one app, importing them into another, only to re-import them into the original app two weeks later).

If you’re still with me, I’m guessing you and me might have just a little bit in common.

I did all this for two reasons. Firstly, I love a geek out and this kind of stuff just appeals to me. I’m happy to admit that. I’ll also admit that it mostly made me depressed. I never found a system I could stick with. The story was always the same: create an unmanageable amount of complexity that kills any joy in the process and then try to revert to something simpler. Secondly, and more importantly, there was an implicit promise in PKM which I now take to be false. Don’t get me wrong, this is no-one in particular’s fault — it’s a heady mix of marketing copy, click-bait titles and probably a lot of personal insecurity.

Let’s take a brief look at some of the headline claims made by players in this space without naming names:

• “A proven method to organize your digital life and unlock your creative potential”

• “Combine timeless notetaking practices with modern digital tools to finally get organized and create the life you want”

• “Makes it easy to revisit and learn from your ebook & article highlights”

• “Save all of the interesting articles, videos, cooking recipes, song lyrics, or whatever else you come across while browsing”

• “Save everything to one place, highlight like a pro”

• “Gives you the power to organize and connect these ideas in new ways”

• “Gives you the tools to come up with ideas and organize them”

• “Invent your own personal Wikipedia”

I could go on, but the pattern is basically the same: organise, save forever, become magically productive. PKM is supposed to make your knowledge permanent and organised in such a way that it leads to effortless output, just through its very existence. Logseq says it “helps you organize your thoughts and ideas so that you can come up with new outputs more easily”. For convenience, let’s encapsulate these ideas with the word ‘productive’ which is, I think, as close to the general gist as we can get with one word; productive in the sense of ‘creating more output’. Output is key here.

PKM promises to make you more productive, and for me, at least, I now think the reverse is true. The rest of this article explores why I now think the promise of PKM is a spanner-in-the-works for non-professional scholars like me and what I’m going to replace it with.

Notes Are a By-Product

Have you ever read or watched something, or heard someone say something, that somehow gave you permission to give credence to that secret, nagging belief that’s been hanging around in the darkest recesses of your mind?

About a month back, the YouTube algorithm, in its infinite wisdom, started teasing me with videos about people deleting their second brains. The first time I saw the thumbnail for one of those videos I did a double take and then immediately felt a wave of relief and the immediate urge to go up to my desk and trash my own vault. I felt joy. I felt free. I felt unburdened — and I hadn’t yet even watched any of them. It’s OK, I thought, to hate this shit. Nothing really changed though, until I came across the work of art that is Andrew Adriance’s Second Brains are a Lie. Watch it. Twice. There are so, so many nuggets of truth in this beautiful takedown of PKM, and in this section we’re going to go pretty deep, but let’s start with a selection of quotes:

• “Knowledge does not equal wisdom”

• “Knowledge needs purpose”

• “Knowledge is helpful, but only when it’s serving a cause”

You can only truly understand the force of this argument if you are familiar with how the False Promise of PKM works. It goes something like this: you collect and curate a permanent digital vault of interconnected notes (the so-called ‘second brain’) from which inspired new ideas and output arise. For academics, content creators or writers, this output might be articles, books, videos or software, but for the serious lifelong learner I now understand this output to be wisdom. The knowledge itself is not wisdom, it is just a collection of facts. We arrive at wisdom (hopefully) through this knowledge, but the path is not straightforward. Wisdom does not simply emerge from a beautifully tagged and categorised Obsidian vault. In fact, in my experience, and I am manifestly not alone in this, nothing at all automagically emerges from a PKM system.

Wisdom comes from thinking, from writing, from processing, from lived experience. It comes from the application of knowledge.

For concrete knowledge, the kind our ancestors had in abundance, the knowledge-to-wisdom gap is most often bridged by physical experience. For example, I’ve been doing a lot of reading recently about layering systems for outdoor activities, but until a recent scare at the top of a mountain on a cold December day, I hadn’t fully grasped how quickly you lose body heat once you stop and thus the importance of what is sometimes called a ‘stopping layer’. That wisdom is now seared into my psyche by a combination of adrenaline and energy gels: a lived experience.

For more abstract knowledge, lived experience can be substituted for a more cerebral experience: thinking, writing, discussing — but note that none of this is done inside a PKM system — it happens independently, and more importantly, it takes conscious effort in the here in now.

Notes that you filed away five years ago when you were interested in functional programming are not going to create wisdom in the philosophy of AI, which you are currently interested in. In fact, they are detritus. They are clutter masquerading as knowledge, or worse, as wisdom. Unless you did something with them at the time, they deserve to be deleted.

I think this quote from the video illustrates just how back-to-front the core assumption of this type of PKM is:

To revisit the art of Zettelkasten for a moment, we can look at its historical prodigy: Niklas Luhmann…Proponents of Zettelkasten look at him and say that the key to achieving and producing things like Niklas is to follow his personal Knowledge Management philosophy. However, this is not so. They have incorrectly tied his success to his particular note-taking system. This is the opposite of how it should be. Niklas didn’t find his purpose because he collected a bunch of notes. He collected a bunch of notes because he had awesome things he was trying to do. He was passionate about his field of study. He consumed massive amounts of information, and the note-taking system was produced as a byproduct of it. His passion did not come from the notes—the notes are simply a byproduct of the passion.

First comes the interest, the passion. Then comes the reading, the studying. Then come the notes. Notes are a by-product.

The physical analogy to this philosophy is minimalism. The kind of minimalism I subscribe to doesn’t advocate living in an empty white box — it simply asks us to do away with everything we don’t currentlyuse. There’s no litmus test that works for everyone, but perhaps you might consider getting rid of any item of clothing you haven’t used in a year, for example. In this context, currently is understood to mean ‘within a one-year timeframe’. I like the word current for this use case — it’s similar to its use in accounting terminology: current liabilities, for example, are those that are to be paid off ‘soon’ — whatever that means to the particular business filing accounts, or the tax authority. Keep this idea of current in mind, we’ll return to it shortly.

Here’s another extended quote — the last, I promise; but it’s a good one, and it serves a dual purpose. It’s from Episode 519 of The Minimalists Podcast, where the hosts discuss the subject of clutter(emphasis is mine):

You touched on the sort of subjective experience of clutter. I wish I could give you the list of, here are the 1,000 items. If you have them in your home, get rid of them because these items are clutter. It doesn’t work that way. I might be telling you to get something you truly get value from. So it’s not dogmatic in that sense. It’s not an ideology and it’s not prescriptive like that. Anything can be clutter if it gets in the way. And the hard part about that is there are many things we bring into our lives, material things, but also relationships, careers, cities, that serve us really, really well for a period of time and sometimes for a protracted period of time. And it really adds value, but then it ceases to add value and the value wanes and wanes and then it falls off with the abruptness of a coastal shelf. And then there’s no value at all. And then we look at the thing and say, yeah, but I got so much value from it. And now it gets in the way. That thing wasn’t clutter, but it is clutter now.

The value you gain from a note, like the value you get from a physical item, is completely relative to your current context. If you gain no current value, it’s clutter. Furthermore, what provides value to one person might be clutter to another, and further-furthermore what provides value changes over time as what’s in our current window drifts and morphs. That’s the first lesson this quote teaches us.

The second lesson is incidental. I listened to this podcast episode yesterday. There was no quote in my digital vault that quietly mated with a few others and finally inspired me to write this article. I’ve been thinking about Wisdom vs. Knowledge and The False Promise of PKM for about a month, it’s my currentobsession. Naturally, as soon as I heard this quote I stopped, made a note of it (specifically, I took out my phone and noted ‘minimalists podcast clutter’ at the bottom of this file) and now I’ve used it. There you have it. The note was a by-product of my current area of interest, or as I’m starting to refer to it, the article I’m currently writing.

What article are you currently writing?

PKM for Minimalists

Time for the practical bit. I’m going to propose a system to replace the modern incarnation of PKM, including second brains, Zettelkasten and Obsidian vaults. Here are the commandments:

• Decide on the meaning of current for you

• Only make current notes

• Make them semipermanent, i.e., permanent enough to cover your current window

• Don’t stress about them lasting forever

• Don’t attempt to create one canonical system — use a combination of convenient systems/devices

• Don’t attempt any sort of sophisticated organisation

• Notes should be findable when and if you need to find them

In short, there is the why of note-taking, and the how. To understand the how you first need to grasp the why, so we’ll start there.

Current Window

Everything in this non-system revolves around what you are currently interested in. If you take nothing else away from this essay, just remember this: read, write and make notes only about what you are currently interested in. This seems so obvious, but I don’t think many people even stop to consider it. Let me reiterate the question I finished the previous section on and add another: What article are you currently writing? What book are you currently writing?

Maybe you actually write, maybe you don’t, that’s cool — but we’re all currently writing an article. I actually think we’re all currently writing a book too. This is how I conceptualise the current window — a ‘short-to-medium’ timeframe where the ‘short’ is the period it takes to write an article, and the ‘medium’ the period over which one might write a book. Again, you don’t actually have to write articles or books for this to make sense, but asking yourself “What book am I writing?” is a powerful way of elucidating what your true medium-term interests are.

Just remember there is no ‘long term’ in this non-system and that the meaning of ‘short term’ and ‘medium term’ is highly personal.

For me, the metaphorical book I’m currently writing is roughly equivalent to the title of this blog — Paleolithic Principles for a Digital World — thus a lot of my reading, thinking and note-making is dedicated to the topic of how we can regain some of our hunter-gatherer mojo in the modern digital, and increasingly AI-dominated, landscape. I expect this to be ongoing for years, to be honest. This medium-term project is punctuated by brief spurts of deep-diving into other things I’m interested in, like this treatise on ditching PKM, which has turned into a literal article.

Just to reiterate: whether you write or not is irrelevant. Just be aware of what falls into your short-to-medium term current window. An easy way to do that is to be able to say what article and what book you are currently ‘writing’.

Another of my favourite techniques is to explicitly write down questions. I do this at both the short-term and medium-term levels. Specifically, every time I read a book, I physically write down 2-3 questions inside the front cover that establish what I’m trying to get out of that book. You’d be utterly amazed at the difference this makes in turning you from a passive, slightly disinterested reader into a textual sleuth. When you read with purpose, with a mission, you activate a different part of your brain — instead of scanning for knowledge you are seeking wisdom. This leads to a much better experience all round: more enjoyment, more engagement, better highlights, better retention and ultimately, more wisdom.

At the medium-term level I maintain a simple text file with a list of (currently) six ‘open’ questions. These are the ones pertaining to the ‘book’ I’m currently writing. These are the embarrassingly philosophical questions like “What is the system? How did things end up like this? Was it designed? Did it evolve?”. This is total geek-fest and I’m not going to bore you with the other questions right now; we’ll leave that for another post.

Questions. Questions that are important to you. Questions that preoccupy you now. That’s the crux of it.

A quick aside. I have this nasty habit of buying screws and duct tape in Lidl. You know that section in the middle of the stores that has discount hardware that rotates in and out each week? Every time I go to Lidl I buy two or three packs of basically everything because it’s so much cheaper than buying it at a proper DIY shop. I collect and hoard all these screws, staples, zip ties, drill bits, router bits (!) and rolls of tape with uses as specific as “Fix small cracks in the radiator tubing of your car”. Never once did I spontaneously build a firewood store out of this crap. You know what I did when I needed to build a firewood store? I went to the DIY shop and bought exactly what I needed for my current project.

How to Keep Notes

This is the easy part. Don’t. As in, make notes but don’t keep them (forever). Once you fully accept that notes are made as a by-product of your current work and that outside of that context they are nothing but clutter, you free yourself from many of the burdens of PKM.

Notes need only be retrievable in the short-to-medium term, which means they don’t need to:

• Be permanent

• Be well-organised

• Form a holistic system

In terms of permanence, semi-permanent is good enough. You need to be able to retrieve and review notes during your current window so having them in a notebook or in the notes app on your phone, or both, is great. Having them on loose bits of paper, or on the back of napkins, means they’re going to get lost, thrown away or destroyed too soon. Find a balance between durability and low-friction, but most of all, let go of the notion that your note collection is some sacred body of work that should outlive you. Don’t necessarily deliberately purge notes (or do, if it floats your boat), but don’t get hung up on permanence.

In terms of organisation, nothing more than the ability to find and retrieve is necessary. For example, highlights made in a physical book or eBook are both fine where they are; if you need them, you’ll be able to go to your bookshelf, grab the book that has come to mind, flick through and find the quote. They do not need to be processed into a permanent, organised, canonical reference system. Likewise for notes you make either on paper, or digitally. A notebook is ideal — it is chronologically ordered by nature and serves the purpose of making it easy to locate recent thoughts. The same goes for note apps. PostIt notes are best avoided. Don’t categorise or tag notes — those systems of organisation support the back-to-front model of wisdom creation and assume that you will need or want to consult past notes thematically. Again, you will not look at past notes once they are out of your current window, so chill.

In terms of completeness, forget about trying to discover or build the one system. Out of all the burdens this form of PKM allows you to release, this, I feel, is the biggest. The single, unified, perfect PKM system and set of workflows is a delusion that just leads to wasted time and feelings of inadequacy. Use an ugly patchwork of mediums - notebooks, highlights in physical books, note apps. IT DOESN’T MATTER. As long as you can find it somewhat quickly tomorrow, next month, tops next year, you’re good. Don’t use PostIts, don’t use napkins, don’t use receipts. Don’t have 25 notebooks, have 3.

Remember, above all, that your notes are always in support of your current questions. These will change over time. Your notes are transient.

How to Use Notes

Don’t. The notes are a by-product of your process of converting knowledge to wisdom. The notes are the exhaust fumes from your brain — the harder you accelerate into a question, the more come out.

Four Problems, One Solution

Right at the start of this piece, I stated that this realisation had solved four psychological ‘ticks’ for me. Finally, we’re at the point where I can tell you.

1. Should I stick with paper books or switch to eBooks?

2. What should I read next? (and why I don’t want to read AI text)

3. Why write when no-one reads my stuff?

4. Which PKM system should I use?

I guess you either identify with these or you don’t. If they sound ridiculously exaggerated and existential, you’re probably not at my level of nerd.

Paper vs eBooks is easy now. IT DOESN’T MATTER. Since I’m not trying to build a single, holistic system, nor am I doing anything special with highlights, I can use both and chop and change between the formats as I see fit.

“What should I read next?” is actually a superb two-for-the-price-of-one. In the most obvious sense, I should read only books and articles that answer my current questions. Anything else is mostly a waste of time. I say mostly, because I don’t read purely for pleasure, so just reading a book for fun is not on my radar, although I understand that’s not going to be how everyone rolls. More importantly, I now have a better grasp on why I dislike the idea of reading AI-generated text so much and am so against using AI at all in my own writing, even for research and brainstorming. When I read, I think. When I write, I think. Basically the only reason I read is to be exposed to someone else’s thought process, the exhaust fumes of their cognitive engine. AI doesn’t think, it just remixes the entirety of humanity’s textual output into a beige soup. Admittedly, it can create prose that is both convincing and great to read, but it’s not true thinking. This is why books like Marcus Aurelius’ Meditations are so important. They are the best kind of book — the ones whose authors never intended anyone else to read. That’s the kind of book I want to read.

It’s also the kind of book or article I want to write. Articles like this. I write to make sense of all this. I hope you’ve sensed that as you read these words. They weren’t written for you, they were written for me. Of course, I’d be overjoyed to know that this has helped you, like Andrew’s video helped me crystallise these ideas that have been floating around in my head for a while now.

And finally, which PKM system to use? I think you already know the answer, right?

None.

Postscript: Knowledge vs Wisdom

Your purpose, as a lifelong learner, is to turn knowledge into wisdom. I see that more clearly than ever now. Perhaps there is tangible output — books, articles, notes — perhaps not. IT DOESN’T MATTER. They are all by-products. Focus on things that interest you for as long as they interest you and then move on. Don’t force yourself to be interested in anything and don’t force yourself to remain interested in something after the initial flame of obsession has flickered out. Only by having clear and currentinterests will your knowledge work yield the ultimate prize: wisdom.

We cleared our plot in the summer, but we’ve been noticing it gradually turning back into a rubbish dump as the months pass and it lays empty. Yesterday Jess saw a load of orange paint buckets in the corner and as we discussed it over dinner, Felisa sheepishly admitted that they were hers.

• “Why the hell are you hoarding paint buckets?”

• “I was hoping the rain would soften the paint so I can clean them up”.

• “What do you need buckets for?”

• “To collect rainwater”.

• “Why do you collect rainwater?”

• “To mop the floor”.

Felisa has an outside tap, but it’s at waist level and above a sink, so she can’t get a bucket under it. Even is she could, there’d be no way she could lift it down to the floor. Her ingenious solution is to just let nature do the work for her. She then transfers the rainwater to the mop bucket with an old tuna tin that I’ve been seeing around the garden for 20 years without ever realising it was a key implement in her water recycling system.

The village folk, especially the elders, are ingenious beyond anything I can fathom. They’ve had to make do with so little, for so long, that anything they can get their hands on is valuable and can be repurposed. Whereas we rely on book knowledge and throwing money at problems, they solve them with our trash. The paint buckets were being thrown away by Ricardo, which is how they ended up in the corner of our plot.

The downside of this is that when resources grew more abundant as the 20th century caught up with the village, Felisa and Arturo never stopped hoarding. When we cleared our plot, which had been their vegetable patch and general storage area, I threw away a full sack of 30-year old corn-on-the-cobs that Arturo had been keeping as fire starters as well as 13 brooms, handmade from twigs and bound with hey-bale string.

The house we’re living in, Felisa’s house, was built by her late husband Arturo, Jessi’s grandfather, a really, really long time ago, before they even got married. I call Felisa over and ask her which year the house was built. She tells me to wait, gets a pen, and starts doing maths in the corner of a crossword puzzle book. The equation looks like this: Arturo was born in 1928 and he was 27 when they got married. He had built the house 4 years before they tied the knot. Without so much as a calculator or even carrying a one, she was able to tell me the house was built in 1954 - 71 years ago.

Arturo was as ingenious as Felisa. I once witnessed him and an 80-year-old friend put up a wooden floor (not the floor covering, the actual floor structure) by sitting precariously on the beams with their feet dangling into the abyss below, all whilst swigging homebrew wine and munching on salchichon.

I knew he’d built this house with his own hands, but I wondered how much planning had gone into it. How much had the preparation for building a house changed in the last 70 years?

A lot, it turns out.

Over lunch I asked Felisa who had been the architect for this house. She burst out laughing, giving me the look she gives me when I ask silly guiri-city-boy questions like that. There was no architect. “No planning permission or anything like that?” I ask. “Nope”. Not even a bloody sketch on the back of a napkin. Arturo wanted to build a house, so he just rolled up his sleeves and built one.

Not quite actually - the real story is much more village and as usual, hilarious.

Arturo had a second cousin, Don Gregorio, who was a doctor in a village called Castrotierra, not a million miles from here, Felisa tells me. Don Gregorio had just built a house and local legend had it that this residence was the latest in construction technology. Arturo decided, given he had no building experience whatsoever, that he would just copy this house one-to-one.

That’s how things get done here, I’m starting to understand. People don’t fret about undertaking big projects like city folks do. There’s the assumption that if Don Gregorio, a medic from Castrotierra, can do it, then so can I because I’ll just ask him how he did it, he’ll tell me and I’ll copy him. He’ll probably come and help me too because he owes me for that sheep. No books, no YouTube, no £1000 weekend courses. Learn by doing, do by copying, and everyone chips in. The way it always has been.

Early afternoon, 8th of July, 1951. Arturo packs his mule with supplies for the journey - a traditional ‘bota’ of wine, a hunk of stale bread and his best clothes for the fiesta in Castrotierra that night. Arturo and Felisa are not a thing yet; they wont be married for a few years still. For now they both have girlfriends and boyfriends they’re not that keen on. Felisasa is with Abundio who can’t dance and she has a secret crush on Arturo. Arturo is no doubt looking forward to a big night of revelry.

Arturo sets off with Masimina and Conce and the mule under the scorching July sun. It hasn’t rained for months and the tracks are baked solid into twisting and rutted contortions that mess with the donkey’s footing. They head South towards Gordaliza before cutting West and heading down the track that will take them to Vallecillo. Don Gregorio doesn’t know they’re coming because there are no phones.

In Vallecillo, Arturo measures Don Gregorio’s house with a piece of string. Felisa was very clear on this - he was too poor to afford a measuring tape. He takes a mental note of all the architectural details. They continue on towards Castrotierra, stopping in Santa Cristina to leave the mule in Juan and Justa’s barn and change into their gladrags.

That’s as far as we can reconstruct the story. Again, Felisasa was not yet with Arturo at this point, so her account of it comes via an extensive gossip network and might not be completely factually accurate.

—

I thought it would be cool to try and find this mythical doctor’s house on which ours is based, so I ask Felisa if she thinks she could identify it if we drove to Castrotierra. She frowns and claims it would be impossible in such a big village. She tells me the villages have grown and changed beyond all recognition. But she’ll give it a go. I hate to tell her she’s talking out of her arse; the villages have done nothing but bleed population for the last 50 years. Most of them are semi-uninhabited by this point and look exactly the same as they did when Franco was in power, bar the presence of some unused, EU-funded exercise equipment for the geriatric. Honestly though, it would be a miracle if the house was even still standing. The local construction material of choice here has been ‘adobe’ (big homemade blocks made from a dried mix of mud and hey) for centuries. Houses made from adobe can last a very, very long time if cared for properly, or, as is often the case, supplemented with modern facing materials like brick or pebbledash. They also disintegrate slowly in the harsh Castilian climate if left to their own devices. Although it’s a beautiful concept in theory - a house slowly returning from whence it came, leaving little to no synthetic residue - it manifests itself in a distinctly ugly way: abandoned ruins gradually decaying amongst cheap, modern, brick structures. This is what the majority of villages here look like today: a total mess.

I sneak up to my computer and look down on Castrotierra from above on Google Earth. It looks exactly the same as our village, just smaller. There are two or three roads to explore, at most. It will probably take us all of 10 minutes to ascertain whether the house is still around. I announce to Jess that we are taking Felisa on an expedition to Castrotierra. Jess rolls her eyes as usual. I’m already dreaming of the tortilla sandwiches Felisa is going to make for the journey.

I go downstairs to break the news to Felisa but she’s waiting for me in the kitchen. Apparently she’s mixed up Gregorio with Gabino or Geronimo or someone else, and it wasn’t Castrotierra it was Vallecillo, she thinks. I curse her for messing up my story, and go back upstairs to the computer to find Valecillo, which is the next village along and looks exactly the same. I’m not even going to change the screenshot. That’s how it happened - first it was Castrotierra, but now she thinks its Vallecillo. In truth, both were involved.

Whichever it is, we’re going tomorrow, I decree. Felisa will make lentils for lunch and then we’ll all hit the road.

It’s raining pretty hard when we wake up and it’s not letting up as lunchtime approaches. We finish our lentils and Felisa lays down for a siesta as the rain starts to peter out and the clouds thin just a touch and I say we should go now before it starts to rain again.

We’re going to do this properly. We’ll do our best to retrace Arturo’s original journey through the fields using only tractor paths, avoiding the new road that could get us there in all of 8 minutes. The mud is thick and slippery as hell this time of year, and with today’s downpour there’s no guarantee we’ll make it, even in the 4x4.

The GPS is off, but it’s in my pocket as a backup; with a 90 year-old and an indifferent miniature pinscher as co-pilots, I don’t rate our chances too highly. We set off down the same track Arturo would have taken 70 years ago, although today the scene is very different. In some ways this village has changed beyond recognition, but in other ways it’s exactly the same, but that’s for another day. Today the difference is meteorological. The plains of Castilla are a canvas painted by the hand of mother nature. In summer, the skies are impossibly blue, the fields are deep, golden yellow and extend into the infinite horizons, the tracks are dusty, hard and make for easy going.

On this drizzly and overcast December afternoon, the tracks are all but impassable. The 4x4 lurches from side to side as the back wheels spin and slip over the mud that’s been churned to a gooey pulp by the local farmers. The ditches get deeper and deeper and we’re all squealing like a bunch of schoolkids as we plough through. The water that comes up to at least the wheel arches sprays out in a brown fanfare as we pass, leaving the windscreen coated in an opaque film. As we head down into the last valley before the village, there’s practically a lake at the bottom of the hill before the track makes the final climb back up. We go for it, vowing not to come back this way.

Felisa is having a blast and is grinning from ear to ear as she recounts stories from every bush and rut along the way.

As we pull into the village, Felisa instantly loses all her confidence in her ability to identify Don Gregorio’s house. The first house we see looks almost identical to hers and we think it might be the one, but the location isn’t quite right, so we pull up to a garage where we find the only human left in the village. We must have been a sight to behold - four out-of-towners, a jeep caked from top to bottom in mud, asking for directions to the house of a 120-year-old doctor.

“He’s dead, you know?”. We tell him we’re just looking for the house, not the man himself, which instantly sounds suspicious. He thinks his wife knows which house it is and goes in to find out. He comes back out and points us to the beginning of the village, just off the main square. The square contains another three practically identical houses. Clearly Arturo wasn’t the only one who had been inspired to copy Don Gregorio’s design - this village was packed with clones. It takes Felisa a while, but she finally identifies Don Gregorio’s house, mostly from the roof and the steps leading up to the door. Sure enough, it’s a facsimile of her house, the house Arturo built. It’s been dressed up in a smart brick facade, but the design is basically identical.

We felt pretty pleased with our afternoon’s detective work. Not only had we uncovered Felisa’s house’s lineage, but we’d also unwittingly solved the mystery of why there are at least 10 almost identical houses in Vallecillo that look significantly different to all the others. Rest in peace Don Gregorio.

Felisa and Arturo’s house has ceilings that are too high as well, but by now you understand that there was no architect involved in that outcome. Arturo wanted to place the house three blocks above street level to help evacuate rainwater and avoid the humidity gradually being sucked up the porous adobe walls. He’d already completed the main structure and started carting in earth to bring up the level when he realised how much work this would be and gave up on the idea. That’s how Arturo’s house ended up with high ceilings. No architecture degree, no CAD designs, no 3D projections, no Feng Shui. Just a man with a mule, a piece of string and a young girl who quite fancied him but was too shy to say. She’s still living here 70 years later, more than a decade after he’s passed, under these high ceilings, with her granddaughter, a strange Englishman and an indifferent miniature pinscher.

Rest in peace Arturo.

TLDR; eBooks are cheaper, easier to buy, hold, carry, read and highlight. Paper books smell nice, look cool and remind us of simpler times. YMMV.

As an avid reader and book lover, I’ve been obsessed with the paper versus eBook debate for at least 20 years now. I’ve read countless opinion pieces, watched dozens of Youtube videos, read through quasi-religious debates on Reddit and personally flip-flopped between both options for decades without being able to fully commit to either one.

This article has been years in the making. I’ve been noting down arguments in favour of each option as I alternate between them and as my reading habits develop.

I think I now have the definitive list of pros and cons.

For context, I read mainly non-fiction, about a book a week. I read maybe 2-3 fiction books a year as well. I do also listen to the odd audiobook, although not as many as I used to - I don’t feel I get that much out of them due to the difficulty of note taking and highlighting.

I won’t go into too much detail here about my ‘workflow’ (it pains me to admit I even have one). Suffice to say that I do highlight and I do use digital tools to review, store and retrieve those highlights. I’m not a ‘professional’ reader. My career doesn’t depend on the efficiency of my reading and/or note taking, but I consider myself more than just a ‘casual’ reader.

That said, here’s what I consider to be an exhaustive, verging on neurotic, examination of the pros and cons of both formats. If I’ve missed anything, please let me know.

Reading Comfort

The physical part of processing printed words and sentences into abstract thought is actually quite a holistic exercise, involving eyes, head, neck, shoulders and arms. The standard way of reading a book, which applies to both electronic and paper, is sitting or lying with the book held out a certain distance in front of your face as you parse the words on the page. eBooks, at least the smaller ones, are easier and more comfortable to hold, especially when either lying down or sitting in more awkward positions.

Physical books have to be held open at the current page, which means either holding with two hands or doing the thumb trick to keep the two-page spread flat, which gets tiring after a while; this is compounded by the fact that paper books are bulkier and heavier. Ebooks can usually be held in one hand for long periods of time pretty easily, either by wrapping your palm around the whole unit, or pinching the border around the screen. You can even add one of those loop attachments to the back of an eReader to make it almost impossible to drop.

Ebooks are also better in terms of visual ergonomics. Screens, in general, are usually not much fun to stare at for protracted lengths of time, but modern eReaders have superb high-contrast screens with incredible resolutions. Long gone are the days when printed books were the better option in terms of print quality; in fact, a lot of books these days seem to be printed on flimsy paper with a nasty beige hue that considerably decreases contrast. To better contrast and resolution, also add built-in backlights, variable fonts and adjustable text size. To serious readers these are far from gimmicks. Built-in backlights are not only, or even mainly, useful for avoiding disturbing your other half whilst reading in bed - which is how they are usually advertised. Good, modern backlights make for better reading contrasts in almost all circumstances. If you’ve ever tried adjusting an external reading light for comfort, you’ll know how hard it is. If you shine the light directly at the screen you get a blinding glare, but if you point it away, it normally doesn’t provide enough light. You want a comfortable diffused glow, which is exactly what a backlight gives. Most modern eReaders have variable light temperature too, so you can switch to a warmer hue in the evening - perfect for pre-sleep reading.

Variable font and text size is invaluable for two reasons. Most importantly, it allows a lot of us who would otherwise require them to read without glasses. One less accoutrement to tote is a big boon (more on this later) and plays on my desire for minimalism. A side benefit, and applicable even if you don’t use glasses, is that you can adjust font size for legibility at different distances from your eyes, which means you can optimise holding distance and therefore arm and shoulder comfort. These small wins add up over hours, weeks, months and years of reading.

Convenience

Ebooks are more convenient to hold and read, but they are also more convenient to carry around. I think that trio of conveniences is what explains eBook popularity amongst travellers.

An eReader is a fraction of the size of just a single physical book but it can store hundreds or even thousands of editions. The ability to carry your whole library in the back pocket of your jeans is a major win for eBooks, especially when attempting to travel light. It might not make much difference if you only read in the comfort of your own home, but for reading whilst out and about or on longer trips, it’s a no brainer. Imagine backpacking through the mountains for a few weeks with enough books to tide you over and paying almost zero weight penalty. That’s only really feasible with eBooks.

Of course, you’ll need somewhere to charge your eReader though. Physical books don’t need charging, so that won’t be a problem. Having said that, this is less and less of a problem as battery technology and software efficiency get better. My eReader goes for weeks without needing charging - enough to get through an extended trip. Modern examples use USB-C charging ports so the likelihood you’ll have a charger handy is also very high. I’m normally bothered by the need to charge devices - I don’t like the dependance it creates - but it has to be said that once charging frequency falls below a certain threshold, it just doesn’t feel problematic any more.

I mentioned that you might not need your glasses for eBook reading; there are other areas in which eBooks win in terms of accessories not needed.

If you’re like me and use a pen for highlighting, that’s something you need to consider when reading a paper book. I’m yet to find a system I like for attaching a pen to a book. I know there are sticky pen loops you can buy, but I don’t want to have to mess around setting that up each time I start a new book. I’ve also experimented with elastic pen-holder cases and even carrying books in a case which has a little side pocket for pens, but ultimately I found neither convenient. Highlighting eBooks with either just your finger or a stylus which attaches magnetically I have found to be both pleasing and easy.

Bookmarks I’ve never really liked. When I start a book I can never find one handy, so I end up using a piece of toilet paper or a photo of an ugly baby. When reading outside they always get blown away, which is why our pool skimmer was always blocked by photos of ugly babies. Unsurprising that I love eBooks ability to mark a page with zero effort. I also like that fact that you can have multiple bookmarks without employing multiple photos of ugly babies. Probably worth noting here that electronic bookmarks are great until they fail for some reason (like multi-device syncing), at which point you have no idea where you are in the book; honestly though, physical bookmarks are also prone to falling out so that’s a draw.

Multi-device syncing is actually another, often overlooked, benefit of eBooks. I have a Kobo, which has an iOS app and website, both of which allow reading. I wouldn’t consider either of those a primary medium, but they can be used in a pinch if I don’t have my eReader on md (e.g. waiting at the dentist, on the train, at my desk on a lunchbreak). For a time I also had two Kobos, which is a great option. I had one upstairs in the bedroom and one in the lounge, which meant I could easily pick up reading the same book at any time. Lazy? Yes, but where reading is concerned, anything that can reduce friction is to be embraced.

PKM (Personal Knowledge Managment)

If you’re not a nerd, you probably won’t know or care about this stuff. PKM is taking reading a bit further by trying to extract, process, retain and use knowledge from books. Everyone takes it to a different level, so everyone’s requirements and opinions are different. This will probably be the most subjective section.

‘Reading stats’ are much harder to collate and track with paper books. If you’re using an eReader you’ll automatically have access to lots of different facts and figures, like reading volume, speed and when you started and finished all your books. This may seem minor, but it can be very useful. For example, when writing my article 2025 in books, I used this feature to easily see everything I read this year.

Highlighting is a big deal in PKM. I highlight and collect my highlights using Readwise, which is a pretty common workflow amongst PKM geeks. It’s definitely possible for both eBooks and paper books, but the process is quicker and more efficient with eBooks. With paper books I have to have a pen at hand and usually fold down the corners of pages where I’ve made a highlight so I can ‘process’ the book quicker once I’ve finished. I’ve tried little sticky tabs and suchlike, but having to have yet more accessories with me when I want to read is just too much friction. Folding corners works quite well, although it does make a mess of the book. You’ll also need to consider whether physically marking a book is even an option for you. Obviously if the book is not yours then it won’t be, but even if you are the owner, you might not want to make permanent physical markings like highlights. You might just consider it ugly, or you might think that it makes it impossible to lend or sell the book later. Personally, I quite like butchering books (more on that later) and also love lending (and borrowing) ‘personalised’ books. Obviously none of this applies to eBooks where you just swipe across the sentence with your finger and the magic pixels do the rest.

Collecting up highlights post-reading is easier with eBooks, although that’s not always considered a pro. I’ve heard it argued that being forced to revisit highlights a short time after finishing a book can aid retention, but personally I don’t really enjoy processing highlights and try to do it as quickly as possible. I don’t believe I really reread them, so don’t subscribe to this argument, at least for my personal use case. The Readwise app has a pretty efficient flow for adding highlights from physical books via the camera. I like it and use it, but have to admit that it’s a pain and I put it off as long as I possibly can. For eBooks, there’s nothing to do - Readwise automatically syncs highlights without any user intervention, which is really nice.

Beyond highlighting, you might, like me, enjoy ‘active’ reading (aka butchering books in a way your mum told you never to do). If you’ve never tried active reading, I strongly recommend you do. Active reading means reading with pen in hand. It means writing down thoughts in the margin, questions, exclamation marks and smiley faces. It makes reading a more interactive process and stops you falling asleep so readily. Active reading with eBooks is mostly crap. I say mostly because I think some of the new eReaders support a better in-book writing experience with decent styluses, but nothing beats a paper book if you want to read actively and write in it, for at least four reasons.

Firstly, paper books have blank pages at the front and back, which are useful places to make more extensive notes, lists of topics, new vocabulary etc. If you’re reading an eBook you either have to have a separate notebook to hand, which is inconvenient, or switch to the notepad function of the eReader (if it has one), which in my experience is slow and clunky.

Secondly, to truly interact with a book you often end up quickly flicking back and forward to reexamine previous points and check current context, sometimes referring right back to the contents and navigating to different chapters. This is slow and awkward with an eBook, rendering such a reading style almost impossible.

Thirdly, eReader styluses, at least at this point, still aren’t as good as real pens. Granted they are getting better, and it’s certainly true that e-ink tablets (as apposed to eReaders) are getting very good styluses, but paper still beats screen here, at least for me, at least in 2025.

Lastly, eReaders and reading apps differ widely in how they implement free writing on eBook pages. Incidentally this is not an insignificant technical challenge due to the ability to change fonts and text size on the fly. Some, therefore, don’t allow it at all, others have you write in a little box, others allow you to write anywhere but then stop you changing font size. It’s a mixed bag.

Purchasing Books and Cost

To state the obvious, with both eBooks and paper books you have to pay for each book. Only eReaders, however, involve an upfront cost. For most serious readers (and if you’ve got this far, I’m guessing that’s what you are), the initial investment for a basic eReader is fairly negligible once amortised across its useful lifetime, which is likely to run into the multiple hundreds of books read.

Ebooks are cheaper than paper books. They should be a lot cheaper given the savings on physical material and distribution, but I guess we’re supposed to believe that the principle thing we’re paying for is an intangible intellectual good with the majority going to the author, so we won’t get into that debate. I’m finding that physical books are getting more and more expensive whereas eBooks seem to be on an opposite trajectory. I keep an eye on the Kobo ‘Sale’ page and am regularly able to pick up great books for £1. That’s rarely possible with new paper books, although of course there are other ways to procure paper books (charity shops, garage sales, friends) that aren’t possible with eBooks - we’ll come onto that shortly.

A serious, perhaps even decisive, factor for me is availability. I live abroad and the only place I can purchase English-language physical books is through Amazon. Not only would I rather not give Amazon my money, but I get the feeling that most of them are print-on-demand, as the quality of the paper, cover and binding is often really poor. Ebooks are a great solution to this problem for me, and expats in a similar situation, but I understand it won’t apply to everyone. Having said that, nothing beats the immediacy of eBook purchasing. The ability to have the urge to read something and to be able to have the book in your hands in under 30 seconds is another of those friction-removing qualities of eBooks that I really appreciate. I read best when I’m highly motivated and inspired to explore something I’ve just discovered. That motivation and inspiration tends to fade quite quickly if not nurtured.

There are some other minor conveniences around purchasing eBooks which are hard to replicate with paper. For example, both Kobo and Amazon allow you to maintain wishlists. A book on your wishlist can be purchased with a couple of clicks. Whilst you can always maintain your own lists of paper books you’d like to buy and then go and get them in store, the process is somewhat less seamless.

Digital Rights and Privacy

When you ‘buy’ an eBook you don’t really own it, at least not in the normal sense of the word. Depending on your eBook platform of choice, it will be some sort of Frankenstein rental-like contract which binds you to a series of conditions which look nothing like ownership (see It’s Their Content, You’re Just Licensing It from The New York Times for a deeper dive). You won’t be allowed to resell, lend, duplicate, read on third-party devices or maybe even download a file at all. To some extent, the platform even controls your access to the media and there are (admittedly few) examples of books being ‘withdrawn’ from sellers’ platforms for whatever reason, meaning previous purchases can no longer even access the title which they supposedly own. Amazon’s removal of 1984 and Animal Farm from Kindle comes to mind (although, to be fair, they did refund buyers).

You might not care about these digital shackles, or you might have the patience and technical knowhow to employ software to circumvent DRM restrictions and ‘jailbreak’ your eBook files, but it’s a pain and something you shouldn’t have to do if you’ve just shelled out £10 for a book. There are also sources of DRM-free eBooks (Defective by Design is a great place to start) but these tend to be older, copyright-free or niche titles; new, mainstream publications are always DRM restricted.

There’s a deep ethical argument around property rights here, but there’s a practical element too. Personally, I don’t think I’ve ever sold a book, but I have often lent my favourite titles to friends and family. There’s something satisfying to me in not just recommending a book, but being able to wonder up to your bookshelf, find your dog-eared copy and lend it to someone.

Paper books are yours, full stop. Aside from a few marginal activities, you can basically do whatever you want with them. You are never going to get that level of ownership with eBooks. There is, admittedly, a flip side to this. Once you take ownership of a paper book, it’s your responsibility to take care of it. If it gets left on the train, burnt in a fire or destroyed in a flood, it’s gone for good. Ebooks are sort of on deposit and can be redownloaded at will; it’s like having your books in the bank. Paper books are also subject to wear and tear and can eventually yellow and fall apart, a process which obviously doesn’t affect their digital counterparts. Again, you have to view this in the context of platform permanence and potential changes to conditions, but it’s worth taking into account. My feeling is that current eBook platforms will probably disappear before any paper book I own disintegrates.

There are serious privacy and platform issues with eBooks that might turn you off, depending on how sensitive you are to these types of issues. Assuming you want to have the option of purchasing new releases and recent back catalogue books, there are only a handful of realistic platform options to choose from, as opposed to hundreds of options for purchasing physical books - from Amazon all the way down to your local bookseller. For eBooks, last time I looked, the only real choices were Amazon (Kindle) and Rakuten (Kobo). Whichever way you go that means platform lock in: once you start buying books on a particular platform, your sunk costs grow and it becomes ever more burdensome to switch. This is, of course, a non issue with physical books. The digital format also lends itself nicely (from their point of view) to platform overreach. To some extent they will spy on you and try to advertise to you: those nice, free cloud features like syncing and recommendations are paid for with your privacy. Like it or lump it.

Environment and Minimalism

We haven’t touched on the environmental arguments yet, because it’s an absolute minefield and you can probably make the evidence support whichever conclusion you prefer. It’s only a realistic contest for heavy readers; buying an eReader and using it a couple of times is clearly worse for the environment than just buying those few paper books. The evidence seems to suggest that the ‘breakeven point’ (the point at which the environmental toll of consuming paper books outweighs the big upfront toll of the production of the eReader) is somewhere between 13 and 30 - see Pierre-Olivier Roy’s superb The environmental footprint of paper vs. electronic books for a more detailed breakdown. In summary, for serious readers, eBooks are probably the more environmental option.

A related argument, and one that appeals to me personally, is that of minimalism. Buying physical books means accumulating more stuff with the concomitant necessity to store it, dust it and take it with you when you move. I did this recently and moving books is an absolute pig. Amassing big collections of objects is one of the least minimalist things you can do, whereas a single device with a thousand books on it strikes me as fantastically minimalist - my whole book collection in the palm of my hand.

The only argument that really matters

Paper books are just cooler. Their musty, woody scent reminds us of a simpler time, a time without screens, distractions, cloud platforms and surveillance. Shelves full of books make you look intelligent and are great backdrops for zoom calls. When you sit and read a real book you feel like a philosopher; more importantly you look like one. Sit in a cafe with an eReader and you won’t get a second glance. Sit reading a real book and people will look at it, maybe even strike up a conversation. In a world that’s rapidly replacing everything with vapid digital equivalents, a real book is a statement. An eReader is just another device.

Me

So where do I stand? Reading this back, I definitely sound like an eBook fanboy, and in some ways I am. I really want to just commit to eBooks and have done with it. I prefer the casual reading experience, the lower cost and the immediacy of being able to grab whatever book I want, whenever I want (and not having to buy from Amazon). I go through long periods of only reading eBooks, but gradually the yearning for paper comes back and I end up ordering a sackful or getting relatives to bring a large stockpile over from the UK when they come to visit.

The urge to return to the comfortable and familiar paper medium tends to creep up on me for one of two reasons:

1) If I’m on a dedegitalisation kick and am sick of screens, devices, charging and cloud platform privacy-invasion.

2) I’m going through a period of reading denser, more philosophical texts and want to scribble all over them, and/or more textbook like books with pictures, figures and complex layouts which are easier to deal with on paper.

Eventually these phases tend to wear off and I’ll go back to eBooks for a while. I’ve been in this rhythm for years and can’t see it changing any time soon.

Sometimes I try to come up with a way to formalise the split by assigning different types of reading to different mediums. As mentioned, textbooks and reference material almost always work better on paper. On the contrary, for the little fiction I do read, I don’t see the advantage of paper; in fact, given that I’m more likely to read in bed and don’t need to highlight, the more comfortable form factor and backlight of the eReader work better.

There are a couple of possible future scenarios that might push me completely over the edge either way. If the ‘perfect’ eReader device came along, I might be tempted to ditch paper books altogether. I realise this will likely never happen but it would look something like this: a DRM-free, open-source, repairable device with a silky-smooth writing/highlighting experience, a backlight, no platform lock-in and a fair and ethical digital ownership model.

The more likely scenario is that eBook platforms move towards ever more exploitative revenue models, devices get less robust and long lasting and I end up back where I started, with good, old paper books.

If you just want the list of books I read this year, go straight to the bottom of this post.

2025 wan’t my greatest year for reading, both in terms of quantity or quality. I fell way short of my target of one book a week, chiefly because I’ve read very little since summer due to the upheaval and stress of moving. In all, I started 27 books, giving up on 4 very early on. That’s a pretty good hit rate for me - I normally give up on a way higher percentage of starters.

Psychology (and similar) was my biggest category in 2025, with five reads. All were good but ‘Tuesdays with Morrie’ was the absolute standout, both in this category and for the year overall. Reading about death and dying is something most of us avoid, but this is one of those few books that will truly, truly touch you - I can’t recommend it enough.

With four reads, the niche you might describe as ‘Epic Journeys’ came in close second for quantity but was easily the highlight of my year in terms of sheer enjoyability. Both of Christian Lewis’ ‘Hildasay’ books detailing his mammoth 5-year-plus journey walking around Britain are awe inspiring must-reads whilst ‘Peat and Whiskey’, which intertwines journeys across grand Scottish landscapes and the story of its peat bogs and whiskey industry, was an extremely pleasant loan book. We had to cancel our own jaunt to the highlands this year unfortunately, but ‘Peat and Whiskey’ has almost singlehandedly given me the desire to get there in 2026.

‘Into the Wild’ deserves its own paragraph. It’s a classic, and now I know why. It still haunts me to even think about it. It’s got a ‘Catcher in the Rye’-esque way of quietly disturbing you without you knowing quite why. This was, without a doubt, my favourite ‘new’ book of 2025 and probably makes my top 10 of all time.

No year is complete for me without a bit of economics, and this year it was Yanis Varoufakis’ excellent Technofeudalism, which is almost an extension of Soshana Zuboff’s seminal ‘Surveillance Capitalism’. I tried to counter the gloom with Nick Clegg’s ‘How to Save The Internet’ but couldn’t get past the Facebook apologism. ‘Empire of AI’ was superb, my runner-up for the year. Both a page turner and highly informative, it really shaped my thinking on AI as the breakneck pace of development shifted gear once again in the last quarter of the year. It will be extremely interesting to come back to this in a few years time and see how everything pans out. To be honest, I’m not optimistic.

I attempted a trio of reads on the history of Israel, the idea being to take in one book from each side of the debate and one ‘neutral’ or ‘academic’ text (I do realise how controversial the idea of a ‘neutral’ text on Israel is). Anyway, I started with ‘Israel’ by Noa Tishby, just because it was on offer on Kobo. That was a big mistake. I should have started with ‘Enemies and Neighbours’ which, however strongly you feel about Israel, is tangibly more neutral. Unfortunately it was summer already and life got in the way of this little project, so I had to abandon it unfinished, but I’ll be back.

Personal Knowledge Management (PKM), a pet geek area of interest of mine, was represented by the classic ‘How to Take Smart Notes’ which is certainly of interest if you want to learn Zettelkasten, but otherwise probably not worth it for the casual reader. ‘The Notebook’, a detailed history of, you guessed it, notebooks, was a delightful nerd-fest and another of my highlights of the year.

I didn’t really hit the mark with fiction. Not that I often do. ‘There There’, a blistering account of Native Indian life in Oakland, was lent to me by Ethan who grew up there. It scared the shit out of me. Otherwise, ‘Orbital’ and ‘The Prophet’ went over my head. I don’t think I was in the right mood. I’ve just started ‘The Overstory’, which looks very promising and I hope to finish it before the year is out.

I reread a couple of personal classics this year. ‘Driving Over Lemons’ from my phase of European travel more than 25 years ago, was appreciated in a new light after 17 years living in Spain, a couple of those in Andalusia. I must get to know the Alpujarras.

Towards the end of the year, as we were packing up the house, I decided to read ‘Sapiens’ again. What can I say to do it justice? ‘Sapiens’ very seriously changed my life when I first read it 9 years ago. You cannot - at least I could not - read the history of humanity without deeply reconsidering my place in it. This sounds cliché, but I don’t think you find ‘Sapiens’ - it finds you, and in 2025 it succeeded in finding me again. When I first read it, I was starting a new phase in life. When I re-read it, I was closing that phase. ‘Sapiens’ bookended the last decade for me and helped usher in a new one. It helped me make sense of myself, everyone else, Madrid and the world. Thank you ‘Sapiens’. I will dedicate a separate essay to what will surely be a supremely vain attempt to communicate what Sapiens has done for me, but for now know that it was, and in 2025 reconfirmed itself as, the most impactful book of my life so far.

Everything I Read in 2025

Psychology

• The Courage to Be Disliked - Ichiro Kishimi, Fumitake Koga

• Good Nature - Kathy Willis

• Tuesdays with Morrie - Mitch Albom

• The Social Paradox - William von Hippel

• The Wealth Money Can’t Buy - Robin Sharma

Outdoor / Epic Journeys

• Peat and Whisky - Mike Billet

• Finding Hildasay - Christian Lewis

• Hildasay to Home - Christian Lewis

• Into the Wild - Jon Krakauer

Economics and History

• Technofeudalism - Yanis Varoufakis

• Empire of AI - Karen Hao

• Israel - Noa Tishby

• Enemies and Neighbours - Ian Black (DNF)

PKM

• How to Take Smart Notes - Sönke Ahrens

• The Notebook - Roland Allen

Fiction

• The Prophet - Kahlil Gibran

• Orbital - Samantha Harvey

• There There - Tommy Orange

• The Overstory - Richard Powers

Fitness

• Functional Training and Beyond - Adam Sinicki

• 9 out of 10 Climbers Make the Same Mistakes - Dave MacLeod

Random

• The CW Way of Life - Chris Rutkowski

• Do Your Best - Bear Grylls

Reread

• Sapiens - Noah Yuval Harari

• Driving Over Lemons - Chris Stewart

Weren’t for Me (Didn’t Finish)

• Rest is Resistance - Tricia Hersey

• Mindful Running - Mackenzie L. Harvey

• How to Save the Internet - Nick Clegg

After dinner conversation in Spain often gets out of hand. Perhaps it’s because it happens so late everybody is delusional from tiredness, I don’t know, but a lot of stupid shit gets said.

Like tonight Ana recounted the story about her bandaging Rodrigo’s ears to his head when he was a baby so they wouldn’t stick out and then somehow they were suddenly talking about the shapes of people’s heads. Apparently us men are unlucky because we can’t hide the shapes of our heads, what with the bald n’all that. A woman on the other hand can have the weirdest shaped head and you’d never know with their long hair. So if you’re a guy with an odd shaped head, you’re out of luck. We all lamented the shapes of our own heads. Mine is too round, I said. Rodrigo’s is like a ‘pepinillo’ (gherkin) according to Susana.

And then, this being the village, they started on the shapes of the heads of everyone else in the village. “Mariano has a big head”, says Feli. “Yes, too big for his body”, Ana agrees. “He was always like that, even as a little kid”, nods Feli, sympathetically. Ana remembers. They rattle through the rest of the village, one by one. Every poor sod here has something wrong with the shape of their head according to these guys. They make it sould like a troglodyte village.

I’m flabbergasted they have such perfect recall of the shapes of 150 other heads, but it makes sense when you think about it. When you see the same 150 people almost exclusively, day after day, year after year, you probably do get the shape of their head indelibly imprinted on your brain. Perhaps the human brain is especially good at head shapes when absorbed in reasonable, village-size numbers. Perhaps city dwellers develop ‘Headshape Blindness’ from overexposure to other people’s head outlines. They all melt into one. I don’t believe I’ve ever paid the slightest bit of attention to the shape of anyone’s head, so I clearly have a long way to go. I can just about differentiate a human from a cow.

In general, they do know a lot about each other here. Too much. I guess that’s just the way it works. Feli has a comprehensive mental database of the gastronomical likes and dislikes of everyone that lives within a 200m radius, which is to say, the whole village. She’ll see a food and go ‘capers, oh yeah, Jorge loves them, but Julio’s not so keen and Anita’s husband got one stuck in his ear in 1952 and it got infected and he partially lost his hearing’, pointing at her ear and winking. This is unsurprising given it’s mostly all they talk about at mealtimes. They start with a food, say, mayonnaise, which Ricardo doesn’t like but Alberto does, but oh doesn’t Alberto really dislike fried liver, oh yeah he does but Pedrito really loves fried liver doesn’t he? And like this they go on for hours, looping through every conceivable foodstuff and 4 generations of 8 families. It’s like a mental training exercise for the most demanding memory game you can think of. I sometimes suggest we bring a whiteboard to dinner, because I quickly lose track of who dislikes sardines and whose favourite dish is garlic soup.

The service at the village shop is so slow and public that they all know what everyone buys anyway, so there’s not exactly any hiding your preferences. Go to the store and buy an avocado and you’ll be the talk of the town for three sunsets. Your legend will probably outlive you. You’ll be the guy who dared to ask for ‘aguacate’. If I’m around the village and need to know if Feli has gone and bought the bread yet, I don’t have to call her to ask. I can just ask anyone in the village. Yes, this is normal. I can literally just go up to someone in the street (rare, at this time of year) and ask ‘Has Feli gone for bread yet?’ and this person will almost certainly know the answer. News spreads fast in this village. There are no secrets. When you get in your car and drive off, 4 or 5 sets of curtains gradually part for a few seconds. “They went off about 6”, people will say over dinner. “Wonder where they went”. And this conversation may last 10 minutes or longer. When someone asks “how are you?” here, they’re taking notes.

Thankfully their ‘curiosity’ is topped only by their unrivalled capacity for gossip otherwise all this precious information would be lost, but luckily anything known by one person in the village is known by the whole village before you get home from the shop. It’s the rural version of social media. Anyway, gossiping is quite evidently what humans do in the wild, right? We evolved this compulsion to social narrative construction which bound us together more tightly into cooperative bands, doesn’t the story go? To gossip is to keep your social brain well oiled and efficient, which might go some way to explaining why the octogenarians around here are so on the ball.

This all sounds quite cute, but there’s a dark side to it too. I’ve always hated ‘persianas’ in Spain, especially in the village. These roller blinds, installed on every window on every house, produce the distinctive rhythmic clattering that forms the soundtrack of every pueblo, barrio and residential street in Spain. I’ve never understood why they bother, to be honest, since they’re just always closed, or cracked open 3cm at most. They open them for a few minutes in the morning, shake out a rug and air the room, then close them again. In the summer, it’s because it’s too hot and you can’t dare let the sun in. In the winter it’s because it’s too cold and you can’t dare leave the window exposed. This is why every Castilian village always looks like a ghost town when you drive through, unless its exactly 7pm and the zombie hordes are out for their cholesterol walk. Anyway, I quite recently discovered that the true role of ‘persianas’ probably has less to do with the weather and more to do with the desire to keep out prying eyes - a throwback to the civil war era when even close neighbours were pitted against each other and were thus keen to keep their secrets behind closed doors and windows.

When we tell people that we won’t be putting shutters on our new house, they don’t ask about energy efficiency, they say “but people will see everything you do”. I suppose they will, but I guess this is what you sign up for when you sign up for this village. Everyone will soon know about my round head, and whether I like (gasp) green pepper in my tortilla, and if I’ve been to the shop for bread yet and what time I go to bed. They’ll talk about it at family get togethers. Maybe weddings. In time, I too will probably be able to identify all my neighbours by the silhouette of their skull and likewise know which of them buys ready meals and doesn’t like goat cheese. I wonder what city knowledge will be squeezed out of my brain by this new delicate social tapestry I’m going to have to hold in working memory.

Until next time. And be careful where you put capers.

I still remember vividly the day at school when I realised our teachers were too overworked to actually give a shit about our assignments. I must have been 14 or 15. It was science class. Our well-respected chemistry teacher had assigned us a big end-of-term project - something to do with atomic weights, something well above our level. I remember receiving back my assignment, neatly assembled in one of those plastic folders with the transparent cover, yellow spine and two little foldable metal clips that went through the holes you would punch in the A4 to hold the stack together. My report was a little chunkier than the others, and as it slapped down on the desk in front of me with a satisfying thud, I was delighted to see a capital ‘A’ in red pen, dutifully encircled and complemented with some comment like “Extremely thorough research - well done”.

It was only when I got home that I realised that half the physical contents of the report was made up of blank sheets of paper. Somehow, I had accidentally left my spare stack of pre-punched sheets in the folder, lending it a wholly baseless heft that had not only gone unnoticed but had also clearly contributed to my top-level grade. Well, that and perhaps liberal use of a still-teacher-proof incoming technology called Encarta on CD-ROM. Perhaps you’re old enough to remember?

I was quite naive as a schoolboy. I’m sure my more astute classmates sussed this out well before I did. For me though, the realisation that teachers were just fallible human beings who would rather watch TV at night instead of grinding through 30 pre-pubescent homework assignments was a rude awakening. Looking back, it was probably key to my considerable academic success in the decade that followed. I’m not, and never was, a cheat. And I worked hard at school and uni, I really did. I simply leaned in and learned how to manipulate the academic system to my advantage: highest possible grades for as little work as possible. In this pursuit I became an absolute expert. Fluff, filler, empty pages and bigger font sizes were the basic building blocks of everything from GCSE coursework to my PhD thesis. From there I graduated to judicious and largely random use of semicolons. What the fuck is a semicolon; and where; does; one; use it? I might have an inkling now, but back then I can assure you that neither me, nor my peers, nor my teachers (apart from the English teacher; don’t push your luck) had even the slightest clue; but they look as impressive; as; hell; coming from a 16-year old. In languages, I learnt that you didn’t have to remember proper conjugations if you stuck a ‘helper verb’ in front the of the main verb and just use the infinitive form. I’d never say “I play football” in French or Russian - instead I’d say “I like to play football”. Simple syntactic shortcut that sounds convincing and saves you learning all the irregulars. Pure fluff. I even worked out that I could ace oral exams by just putting on a dumb-ass French accent and mouthing a Gallic-sounding “eeeeuuughhh” for at least 20 seconds in between each haltering interjection about having a punctured tyre on the D959 to Tours.

All this junk - fluff, filler, empty pages, semicolons, empty verbs, meaningless vocalisations - is bullshit. There’s really no other way to describe it. It’s an optical illusion - real, but meaningless. It wasn’t revolutionary twenty years ago, and it’s not now, even though we’re rediscovering it through the lens of AI. This type of BS has always been a badge of honour or a form of intellectual or professional virtue signalling. More bullshit, until now, has always meant more work, more man-hours, more grind - and that made it actually worth something. Or at least others would perceive it that way. David Graeber’s now classic work “Bullshit Jobs” is full of such examples. By his estimate, a massive chunk of the productive economy was (I say ‘was’ not ‘is’, because his book was pre-ChatGPT so it’s much, much worse now) dedicated to the systematic production of BS.

Bullshit as a desired economic output is kind of hard to fathom, yet it’s absolutely everywhere. And again, this is not new. Junior employees have usually been at the forefront of the bullshit economy. Think of the junior analyst at the investment bank who spends his or her days churning out pages and pages of reports that end up on the desk of another junior analyst on the other side of the country whose job is to filter through those stacks of paper in order to deliver a simple nugget of wisdom to middle management: buy, sell, hold, whatever. The hot potato of bullshit that gets passed from junior employee to junior employee simply served to give them something to do, some task they could contort to show their worth to their employer, to make their middle manager look better, to clock up hours in the office - evenings good, weekends better. Those that embraced the bullshit rose through the ranks.

You’d think professional authors would be above these bullshit filler techniques, but far from it. They are the masters. Single-concept fluff pieces are the norm for books in the ‘Smart Thinking’, ‘Business’ and ‘Self Improvement’ niches. So much so, in fact, that Blinkist have built an incredibly successful business out of systematically de-bullshitting these books so that you, the reader, don’t have to wade through 300-plus pages of fluff to get to the core message. You just listen to a 10-minute summary. This example is essentially analogous to the previous scenario involving the junior employees, just with different actors. In both cases, economic value is ‘created’ in the process of wrapping a useful message in a layer of bullshit. In the case of the analyst, it’s literally his or her salary. In the case of the author, it’s the increased perceived value of a 300-page book versus a 15-page book. Picture the scene - you’re browsing the bookstore at the airport (where else does anyone buy books they will never read?). Are you willing to pay £15 for a pamphlet? Of course you’re not. Like everyone else, you expect a good hunk of wood-pulp for your money. Something that will look good on your shelf. Bullshit creates perceived value. But the ‘bullshit-value-creation-chain’ doesn’t stop there. No, bullshit gives twice. Blinkist creates value by stripping away the bullshit for you, just as the second analyst does for the ‘intelligence’ they present to their manager.

Really, when you look at it like this, is it any wonder that bullshit jobs have been eating the world for the past 30 years and we’re drowning in a sea of slop? You’d think the free market would have dealt with this by now - after all, paying the price of bullshit twice over hardly sounds like the epitome of economic efficiency. A theoretical company that doesn’t cloak their internal knowledge in 27 powerpoint slides, with all the human costs involved in such a process, will presumably outcompete others in their sector by a large margin once scaled up. But that would be to assume that humans can detect bullshit when they see it, which they can’t, at least reliably. Anyway, we’re not ‘rational actors’, as we are all well aware of, so it’s unsurprising that this preference for BS has emerged. In fact I might coin a new cognitive bias: Bullshit Premium. The Bullshit Premium is the difference in perceived or actual value between a ‘clean’ intellectual output, and the same output wrapped in a thinly disguised veil of BS.

In 2025 we find ourselves at an interesting crossroads, BS-wise. There are two opposing forces at play which will ultimately determine which direction this goes and whether bullshit will whither and die, or multiply and thrive. I have hoped, since the early days of ChatGPT and throughout the subsequent iterations, that AI would drive the Bullshit Premium to zero. Think about it like this: we now know to a pretty high degree of certainty now that any long, thorough and well-written piece of work, like a research report, executive summary, or strategic plan, will have been researched and written by AI, with an absolute minimum of human input. Oh pipe down for god’s sake. Don’t try to pretend you’re any different. Everyone knows. Which raises the question: if everyone knows, why the hell would a bullshit-cloaked piece of work hold any perceived value over and above the original idea? It shouldn’t. It’s absurd that it still would. Nobody, literally nobody, is impressed with your 20-page report. So just stop. Give us the abstract. Give us the original creative spark. Hell, to paraphrase Clayton Ramsey - just give us the prompt. Bullshit no longer requires human toil. Bullshit created by a Harvard graduate doesn’t reflect positively on your company any more than 3-ply toilet paper in the head office toilets. Bullshit is now a total and utter commodity, costing fractions of a cent per kilo. It should just disappear naturally, like manure.

The counter-argument to this is that everyone over the age of 40 still thinks bullshit looks and sounds impressive (or at least assumes that everyone over the age of 50 thinks that way). They think they might just be able to trick their boss or teacher into thinking they did some actual ‘work’ if they add ‘make it sound original’ to their prompt. The result is looking like a decade-long LLM bender and a yet another example of the prevalence of a digital commodity exploding as its intrinsic value goes to zero. Like porn, bullshit is now free, so we gorge on it. Eventually we will have re-bullshitters attached to our mouths and de-bullshitters attached to our ears such that you will never even be aware of the existence of the bullshit in the first place - it will be fired across the ether between two devices manufactured by OpenAI, Google or Apple - a BS-encoded bytestream costing a bargain 13.99/month and a global 10 Petawatts of electricity per year. That’s basically what’s happening now when you use ChatGPT to create a document that I then have to feed back in for summarisation. It’s a pain in the arse and costs us both $20 a month.

I think it’s worth repeating: I know you didn’t write that yourself. I am NOT your grandma. I am NOT impressed.

Bottom line: wrapping some genuinely useful idea in a shell of unwanted BS is not new. It was invented donkeys’ years ago and gradually perfected by every passing generation of writers, academics, junior researchers and students, including yours truly. AI didn’t invent this shit. But it sure as hell has perfected it. What, then, lies in store for us humans if even bullshit jobs are no longer on offer? My current feeling is that the near-term future will be dominated by those that can pull off the AI bullshit trick most convincingly. I think there’s still enough of a Bullshit Premium for the grifters to go after because there are still enough people who will be tricked into (or straight up don’t care about) reading books written by AI, watching videos scripted and performed by AI avatars, signing up for $5/month software platforms that scratch some niggling little digital itch they never knew they had, or buying some product being pushed by an influencer that doesn’t really exist.

Perhaps we’re living through a unique and temporary Goldilocks period for AI BS before the whole circus eats itself alive.

You know how I love to start off a piece with a mental model. Today’s is a great one: density. Density is such a useful thinking tool because it does what the best mental models do so well: bridge the physical-to-abstract chasm so elegantly it’s almost impossible not to understand it.

There’s a hint of obviousness in our everyday use of the term, though not always. Intuitively, I think we know what someone means when they describe a text as dense even if we might not be able to describe the concept neatly, or in mathematical terms. “I’m reading a book about particle physics at the moment. It’s taking me ages to get through because it’s so dense“ is evocative and immediately clear. Other times, lay parlance seems to invert our intuition. For example, at least in British English, you can refer to someone stupid as being ‘a bit dense’, which never made any sense to me. It feels like someone with dense cerebral matter would be extremely intelligent, not dumb.

Truly understanding density is not hard, but requires unpicking some odd and intertwined phenomena.

If you’ve ever been caught out by the question “What weighs more, a kilo of hammers or a kilo of feathers?” you know where this is going. The correct answer, of course, is that they weigh exactly the same - one kilo. The question prays on our innate tendency to confuse, at least superficially, the concepts of weight (or mass) and density. One would normally say that hammers are ‘heavier’ than feathers and one would be perfectly well understood, but that assertion isn’t quite accurate. It would be more correct to say that hammers are denser than feathers, which can be extrapolated to “for a given equal physical quantity, or spatial volume, the volume of hammers will be heavier than the volume of feathers”. Simpler still, just think of a box. A box of hammers is going to weigh more than a box of feathers. It all depends on volume. In the case of hammers, there is more mass (kilos) per unit of volume (box).

Now you understand density, or at least volumetric mass density, you have a better grounding for understanding the famous phenomenon demonstrated by astronaut David Scott on the surface of the moon in 1971. He showed that a hammer and feather dropped simultaneously will hit the ground at the same time. Go watch the video now - it’s pretty cool. First, it’s worth noting that the moon DOES have gravity, although not as strong as the Earth’s. In the absence of gravity both objects would float off into the ether. Now, the reason that both objects fall at the same rate is that acceleration due to gravitational pull is unaffected by mass (heaviness). Galileo already knew this and it can also be demonstrated on Earth in a vacuum. Vacuum is key here. A vacuum, whether artificially created for an experiment or on the surface of the moon, is the absence of air. It is, therefore, the air that causes a hammer and feather to fall at different rates here on Earth in normal conditions. Why? Because an object moving through air (or any fluid for that matter) meets resistance. Think of the times when, as a kid, you’d put your hand out of the window of a moving car and feel the air push you back. Now imagine that instead of your hand, you’d stuck something much larger out the window, like a beach tennis bat. The resistance would be so large it would probably pull it straight out of your hand. This resistance, or drag, is what helps planes fly and sailboats to circumnavigate the globe. Back to falling objects. In the absence of air, the only factor affecting how fast the objects fall is gravity - nothing intrinsic in the objects has any effect. On earth, where objects fall through air, drag comes into play. The mathematical relationships get a big hairy here, but they can be boiled down to this: smaller and heavier objects (more dense) fall faster than bigger and lighter objects (less dense). Imagine a sheet of polystyrene weighing 100g vs a roll of coins weighing a 100g. The coins would fall much faster.

So far, we’ve been talking exclusively about volumetric mass density - or how much mass is packed into a volume, but the concept of density is often abstracted to something more like “how much of a thing, A, can be packed into a space or volume, B”.

Consider the following assertion: “Companies move to Silicon valley because of the density of tech talent”. Here, A, or the thing being packed in, is ‘tech talent’. B, or the space or volume in which it is being packed, is a geographical space (a city or area). Note then that the density is affected by both quantities A (how much talent) and B (the space into which it is packed). Mathematically, more talent would mean higher density while more space would mean LESS density. That’s the key mathematical relationship: density is inversely proportional to the space or volume taken up. The formula is A/B (A divided by B, or A over B). Expressed in unit terms, it’s A per B, or colloquially ‘Amount of stuff per unit of space’, or concretely for this example “Number of potential technical hires per kilometre squared”.

Right, let’s circle back to the very first example we explored: “I’m reading a book about particle physics at the moment. It’s taking me ages to get through because it’s so dense“. This is pretty abstract, but I think we can dissect it now. Let’s describe the A quantity as “complex ideas” and the B quantity as “pages”. The density we are describing then is “complex ideas per page”, which aligns well with our intuition, I think. The author could have reduced the density by either reducing the number of complex ideas or taking more pages to explain them. Note that just increasing the font size might reduce the physical density of the text, but it wouldn’t affect the intellectual density, obviously.

So we’ve gone from the pretty concrete and physical ‘volumetric mass density’ to the rather abstract ‘complex ideas per page’ version of density. That’s pretty cool I think, and definitely underlines how these concepts, when treated as mental models, give you a framework for thinking and understanding. I’ve spent so much time on this today because I’m going to argue that this type of thinking tool is becoming more and more important in the age of AI where many of us are outsourcing our thinking to machines that absolutely DO NOT use these frameworks to reason.

I’d love to just copy-paste the original ChatGPT conversation here, but I can’t, because R deleted it. A while ago, R asked me what he thought we could do to protect ourselves from the inevitable AI enslavement that was just around the corner and I told him to just be nice to the them, which he has been doing ever since. This conversation got a bit heated, so he decided it would be best wiped from his phone, lest GPT5 and the boys (Claude and Gemini, more intimidating than their names suggest) come knocking. Anyway, here’s my attempt to recreate the chat as faithfully as possible.

J: Jeez, I’ve eaten like 20 figs this afternoon. That must be a shitload of calories. I wonder how many calories there are in a fig.
R: I’ll ask ChatGPT.
R: How many calories in a fig?
ChGPT: There are approximately 30 calories in an average fig. Dried figs, however, have more calories in them. There are approximately 55 calories in an average fig if it’s dried.
J: OK wow.
R: OK cool.
Me: That’s utter bullshit, give me the phone.
R: Ah come on dude, be nice to it.
Me: Fuck that, give me the phone.
Me: Um, that’s cute and everything, but dehydrating doesn’t magically create calories. There are exactly the same amount of calories in an average fig - doesn’t matter whether it’s fresh or dried.
ChGPT: (Thinking)
ChGPT: Of course! You are right! Great observation. You’re so clever and handsome. The number of calories in a fig is the same, irrespective of whether it’s fresh or dried. However, supermarkets tend to select larger figs for drying so that the end result is that average dried figs have more calories, as per my original statement.
J: OK wow.
R: OK cool.
Me: OK, that’s nice, but I think you just made that up to justify your original statement.
R: Dude, be careful, or at least tell it it’s you talking, not me.
Me: Where did you get the evidence for that last statement. Please be specific. I think you made it up.
ChGPT: (Thinking)
ChGPT: (Thought for 1m 30s for a better answer)
ChGPT: Yes, you are right. I have no foundation on which to make that claim. I just assumed it would be correct. Sorry.
Me: Would I be correct in stating that you have confused absolute caloric content with caloric density? Dried figs are, of course, calorically denser than fresh figs because of the significant volume loss caused by the removal of almost all the water.
ChGPT: Absolutely! You are right (and so handsome). That is what I was trying to get at originally. It’s not that dried figs have less calories, it’s that they are more calorically dense. Thanks for clearing that up.

I’m such a nerd; I just love this conversation. To my mind, it’s a perfectly fitting summary of where we are as humans navigating this new era-defining technology and raises at least the following talking points, dilemmas and conundrums:

1. Should R have used AI to get an answer to this question in the first place? Is consulting an LLM for hard facts a good idea?

2. Does AI actually have any intelligence? Does it know anything?

3. Are we already at the stage where unquestioning acceptance of the output of LLMs is making us both individually, and collectively more stupid?

4. Do dried figs have more calories than raw figs?

Slop was not invented by AI, but it sure has perfected it. The internet, or more specifically the World Wide Web, has been a wasteground of utterly crap content for the best part of a decade. We now have ‘AI slop’ - pages of text, images and videos churned out automatically by generative AI, but once upon a time in a not too distant past we had good old ‘human slop’ - pages of inane articles, top 10 lists and soulless blog content written with the purpose of extracting a fraction of a cent from your attention by taking a whole page to present a single fact. You’ve seen this a hundred times by now: you’re looking for a simple answer to a simple question like “How many calories are there in a fig?” and the top result is a ‘blog post’ on a site called ‘caloriehealthy.com’ which answers the question in 1000 words which are barely legible through a haze of popping and swerving ads. This, alongside cookie banners, is how we ruined the internet. It’s actually incredible when you step back and see what we’ve done. Imagine if, a decade or so after the invention of airplanes, we decided that instead of using them for horizon-expanding intercontinental travel and bombing our enemies, we’d drench them in advertising and attempt to extract £12 for a cup of tea out of each poor immobilised sod as they jollied back from a boozy weekend in Mallorca. What a waste.

When our simulation overlords look back over this run, they will surely be blindsided by our next move - an act of such jaw-dropping stupidity that it almost defies belief. We used our hundreds of petabytes of low-quality human slop, a treasure trove of hate, lies and the musings of the deranged, to train our AI models such that they could now produce unlimited variants of their own machine slop on demand. But we didn’t stop there. We dump this AI-generated slop back in the mix and continue training them - now on a soupy hybrid of human and machine slop. And now it’s just slop all the way down. Humans, machines, cats. There’s nothing but slop.

The genius twist to this tale is that AI turned out to be a very unique invention: a singular technology that noone wanted, noone asked for, that creates a trail of pain and havoc in its wake, yet represents the only viable solution to its own destructive vision of the future. Sort of like killing someone with a piano then sitting down at the same instrument to play music at their funeral. That’s right, we can no longer easily glean any useful information from the internet because it has died under its own weight in AI slop, so we’re forced to use AI itself as a filter through the slop to get any simple question answered. Which is why noone uses Google any more and why R asked the fig question to ChatGPT. If you’re an AI executive, that’s checkmate. That’s the holy grail of business models. Use my product to fuck yourself, then use it to unfuck yourself. Big junk food and big pharma rolled into a single, Big Mac flavoured, mega-pill.

At the end of the day, the LLMs make similar conceptual mistakes to those which a human might make - like confusing calorie content with calorie density; but there are at least two differences. First, humans can (if they choose to) convey their degree of uncertainty - they can say things like “I think, but I’m not sure” or “let me just check” or “don’t quote me on this”. In my experience, LLMs never say that. This could be solved, possibly, if LLMs displayed ‘degree of certainty’ measurements next to each output - i.e. a little number or symbol to indicate the degree of certainty with which they are speaking. Honestly though, it’s doubtful they even know and even more doubtful that people would bother to look. Secondly, ChatGPT seems to have assumed a status equivalent to, at least amongst the people I see using it, a doctor, priest or lawyer rather than a bloke down the pub. How many conversations now start “Well, ChatGPT says...”? Try starting a conversation with “Well, my brother-in-law says...” and watch people laugh at you. Noone believes a thing their brother-in-law says yet ChatGPT is the new oracle and people like me who dare question its wisdom, or worse, refuse to use it, are viewed in the same light as vaccine sceptics. It’s COVID all over again, but this time the needle is software, not hardware, and you can’t feel it pierce your skin.

The best way I’ve heard the problem stated is that generative AI produces ‘convincing sounding output’. Not necessarily correct, not necessarily well written. Just convincing. Emily Bender et al captured this idea in their seminal paper ‘On the Dangers of Stochastic Parrots’.
It turns out that ‘convincing’ is enough. We have been, by and large, convinced. The supermarket explanation sounded very convincing. Maybe they do even pick larger figs for dehydrating. How would you, or even an AI even know the ‘truth’, when the truth is obscurred by a long chain of AI induction trained on a regurgitated decades-deep stew of human and AI slop?

There’s an acceptance in science, where ‘truth’ can never be fully known, that a theory doesn’t need to be 100% correct in order to be useful - it just needs to allow us to do things we couldn’t before, like make a filament glow or send a rocket into orbit. In the case of AI, we’ve basically accepted that the output of an LLM doesn’t need to be true to be useful, just convincing. All the bullshit jobs gone - replaced by ‘convincing output’. All the software developers gone - replaced by ‘convincing output’. That’s it now. No art, no literature, no real code. Just convincing output. Presumably ‘convincing’ is to be understood as ‘sounds like it was produced by a human’. Looking back over the last decade of the World Wide Web, perhaps that’s all those phoney blog articles and content farm site pages ever really were - just convincing output. Text ‘generated’ by a human who couldn’t give a shit, made to sound like text ‘written’ by a human who could give a shit. And now we have machines spewing out text made to sound like it was generated by a human who couldn’t give a shit, which in turn was generated to sound like it was written by a human who could give a shit; a chain if you will: human (gives a shit) > human (doesn’t give a shit) > LLM output.

Honestly though, it’s not about figs any more. Or lies about figs. It’s about lies full stop. Convincing lies. In the industry they are called hallucinations, which sounds funny and mystical, but they are lies. They are lies because we are convinced they are not. We are convinced that these LLMs are somehow superior to us, that they know more than we ever could. But they don’t. They actually know nothing. They are just very convincing liars.

I recently talked about the penultimate day of packing up the house and all the fun I had sampling alcohol, medicines and diverse cleaning fluids. That was about the high point of last week (pun intended). The final day of packing up, the day where you walk out the door for the very last time and say your teary goodbyes to the concrete and glass box that has been home for the past decade, involves a process of ever-diminishing piles of crap.

By this point, the really juicy stuff is long gone - whisked away to its new home in a removal lorry. The sofa, the beds, the big-screen TV, the boxes of personal stuff, the general clutter of the last month - all in the past. The house is empty now - echoey and sad, devoid of the life that it once contained. The art is stripped from the walls, the rooms are featureless, the cupboards bare. The good-natured chaos of 200 cardboard boxes had filled the house with a different kind of energy for a while - a nostalgic but hopeful serenade of beer from tin cans and impromptu dinners with neighbours who came over to help lug 20kg olympic discs out to the umpteenth van - but even that is long gone now. Those 'good times' somehow seem like a distant memory when you're faced with the stark nakedness of white walls and bathrooms you use by accident and then have to shout out for toilet paper.

This house is a blank canvas again, but I won't be the one painting this time. It's odd really - a blank canvas is normally something you start, not finish, with - but packing up a life is a few-in-a-lifetime opportunity to deconstruct, to break back down to its core ingredients a dish that you've spent a long time cooking. We don't do that too often if you think about it - we're used to creating, not creatively destroying.

Deconstructing the material aspects of your life can be quite informative. For example, I have 5 meticulously sorted boxes of cables (audio/video, computer, network, music, power), yet my entire collection of toiletries fits into a single wash bag. What does this say about me? Do I want to be known for my commendable collection of USB adaptors but dubious personal hygiene? My general feeling is that the majority of the articles we surround ourselves with say more about our past then our present or future: clothes that haven't been worn for years, books that have never been read, 17 spare dishcloths when only two are on permanent rotation. Honestly, I could have tipped 90% of it in a black hole and not even noticed. Anyway, I've done that rant before - no need to repeat myself.

The last day unfolds like this: you start with a couple of neatly arranged piles in various strategic points around the house: last minute cleaning implements, last night's makeshift bedding, rubbish, random articles. You play these piles like a game of chequers - they move, eat each other, spawn new piles, combine and morph, until, eventually, you're faced with 'the last pile'. The significance of 'the last pile' cannot be overstated: in the gameshow of junk, these contestants are the valiant ones that made the last round, the grand finale. They survived every attempt at disposal, donation, packing away and even consumption. This, reader, is the shit that is going to accompany you personally, in your own car, on the final leg of your journey.

My final pile consisted of an heirloom gold watch left to me by my great-great-grandfather, a first edition of 'Around the World in 80 Days', signed by Jules Verne himself and a…

No.

Of course it didn't.

My final pile consisted of huge amounts of toilet paper that just seemed to seep out of the walls, a full-to-the-brim tin of weed (hand selected from the carrier bags full I was too scared to take on the journey), a jar of long-expired pickled chili peppers, a little person-shaped salt shaker (pepper shaker's whereabouts unknown), two mismatched wine glasses (one white, one red) and two mystery packets of strawberry liquorice found in the high cupboard above the fridge that have presumably been there since we moved in.

And then there was literally nothing except total and utter emptiness. We took one last walk around, closed the door and left, along with our last pile stashed unceremoniously in a blue rubbish bag. The wine glassed clinked incessantly through the three-and-a-half hour journey north and I'm pretty sure they both smashed when I transferred the bag out of the car and placed it on top of its fallen comrades in the behemoth of a shitpile that was waiting for it at its destination. I haven't had the heart to check.

Yesterday I realised that I'd left a remote control velcro'd to a pillar in the gym (a setup that allowed me to change channel whilst on the exercise bike), which of course means that the venerable last pile wasn't the last pile at all. The last (abandoned, dispersed) pile, at least as far as I'm currently aware, consists of said remote control and a bag of fitments for a freezer we sold to our neighbour. I'm sure there will be more. I suspect that houses are like toothpaste tubes in that respect. You squeeze as hard as you can for weeks on end but there will always, always be a little bit of you left in there.

The penultimate day of moving out of a house involves pouring a diverse array of colourful liquids down the sink. Somewhere, perhaps 50km downstream of our house, there is a drainage worker who thinks Pride campaigners have hijacked the pipes for publicity. In no small part, these liquids paint a rich tapestry of the life we're leaving behind and I've discovered that there is much to be learned from their gunked-up lids and putrid aromas.

To put it bluntly, flavoured liqueurs are a no-no. I couldn't resist a swig of 8-year-old Maltese fig liqueur before it went the way of the Draino and boy did I regret it. It made me instantly drunk and left a sort of rotten cheesecake flavour in my mouth which I can still detect now, a few hours later and after having sampled the dregs of a bottle of bisongrass vodka (which had something floating in it) and an unlabelled red syrup that might have been a cocktail flavouring, but then again could just have easily been cough medicine. Whilst we're on the subject of medicines, don't even think about sampling any of those. Don't even smell them. Just throw them straight in the bin. A brown bottle of something etherish almost tipped me into a seizure, just from a tiny sniff.

Resist the urge to bring back some of the local poison from your holidays. The bottle of Cuban rum is still three-quarters full and has changed hue from 'Deep Amber' to 'Urinal Splash'. Port is another one of those tipples that feels like a good buy after a boozy distillery tour, but inevitably sits unopened until moving out day. If you must, buy one bottle - not one of each colour and the one that comes in a little wooden case.

Multiple big bottles of cheap gin and vodka tell the story of a party that didn't quite live up to expectations. People don't tend to drink as much as you think they will, unless they do, which leads to other problems.

For your own sake, I'd strongly advise avoiding keeping any bottle with less than a quarter still in it. This always get pushed to the back and will mischievously resurface the day before your big move. If you're anything like me, you won't have the heart to bin them so you'll end up moving onto the next phase, which is orders of magnitude more likely to kill you, slightly pissed on honey whiskey and arak.

Cleaning products provide the chance to experiment on the edge, especially if you're tipsy. I knew that mixing ammonia and chlorine can produce a deadly gas, so I poured them down the sink sequentially, with 96% alcohol in the middle for shits and giggles. The result didn't kill me but did produce a stench so strong it made my nose tingle and eyes water and immediately sobered me up. I tried to cover the semi-lethal haze with a room freshener spray from Zara which left the utility room smelling like a shopping mall toilet. When Jessi walked in she immediately knew I'd been up to something (even before she saw me sitting in a daze in the kitchen) and threw all the windows wide open.

At some point in the history of the house I must have been on a 'natural cleaning products' kick. A 5kg tub of bicarbonate of soda was a particular challenge to get rid of. What my exact plans were for this chemical stash I can't recall, but I estimate I used about 100g. The rest went into the kitchen sink with the waste disposal unit at full tilt, whipping it up into a sort of meringue which only got worse when I added Fairy Liquid into the mix to try and smooth its passage. It expanded dangerously outwards as I slapped at it with a silicone baster. For good measure, I washed that down with a big bag of magnesium bath salts which dissolved pleasingly on contact with the water and seemed to keep the previous cream pie from expanding any further. On the plus side, the sink is pristine now.

Vinegar, another 'cleanfluencer' favourite, is best kept for salad dressings in my opinion. As a cleaning product, it might leave your windows streak free, but it will also make your house smell like a seaside chip shop for weeks. A bottle of that went down the sink, leaving a disgustingly malty tang hanging in the kitchen for hours. That was followed by three bottles of dishwasher rinse liquid; clearly on multiple occasions we'd seen this product in the supermarket, figured we should be using it, bought it, but never once had the gumption to actually use it. Does anyone actually bother with this stuff?

I won't talk too much about the foodstuffs. I'm terrible at throwing away anything edible so we've been lathering dinner with a violently spicy Indonesian ketchup for about a month and literally every dish I've cooked is soaked in soy sauce from my bulk buying phase: I still have at least a litre of that to get rid of but I can't bear to throw it away so I might season the plants with it before I leave. That will be on top of multiple bottles of suspect wine that have been horticulturally employed.

The sun's going down now, and I still have a pint of vermouth and some citric acid to deal with, so I'll sign off. If you never hear from me again, you can assume that the combo didn't work out, although you'll be guessing as to whether it was toxic off-gassing or intestinal burn that finished me off.

"Should I buy this?". One of life's most interesting conundrums. Basic economics would have us believe we're perfectly rational actors in possession of complete information, but in real life, of course, we're capricious ignoramuses. We don't know that much about the piece of plasticshit we're considering purchasing until we get it home, and we certainly don't know how our futures will transpire in the presence of said plasticshit. Even more so now we buy our plasticshit online - chosen, basically at random, amongst a multitude of other similar plasticshits of the same ilk.

It probably seems like a trivial question, and on an individual transactional level it mostly is, but when you finally have to box up and move all your plasticshit after 10 years, you get a true sense of the cumulative weight of all those micro-decisions. Faced with piles, boxes and bags of junk, you ask yourself "How the hell did I end up with all this crap?". The answer is pretty simple - you didn't take each buying decision seriously. You just opened Amazon, searched, clicked and forgot. You needed a magnetic-phone-ring thing, so you just bought it. One summer, the temperature was over 30 for two hours so you bought a giant paddling pool didn't you. You don't stop to seriously consider how long plasticshit will last, how much space it will take up or how or when you'll dispose of it. You just put it in your house somewhere, perhaps alongside the never-used Bohemia decanter someone long-forgotten gave you as a wedding present 20 years ago and which has now developed a thick patina of grease from every time you picked it up and thought "for fuck's sake, do I look like a Count" and put it back in the cupboard where it takes up so much space you're now planning a kitchen extension.

The process of moving is like the end of a ten-year football match where the score's been kept a secret, although you secretly know. The crap has beaten you fair and square. As you traipse soullessly from the basement to the top floor to put the fourth hairdryer in a box marked "PLASTIC SHIT", you may get to seriously contemplating how you might make better purchasing decisions in the future. After all, none of this stuff was free and it all adds up. Ten quid here, fifty quid there. There must be thousands in waste piled, boxed and bagged.

My buddy Ethan and I once tried to come up with a scientific method of quantifying how much you should pay for something by multiplying how many years it would last, how many hours a week you expect to use it and a sort of utility-power multiplier. For example, a new MacBook Pro that would last for five years and be used for work 8 hours a day might have a value score of over $20,000, even though it only costs a tenth of that. This method isn't great as it assigns high scores to 'slow' products - for example, a beard trimmer that takes two hours to complete the job rather than five minutes scores much higher, which is obviously ridiculous.

There's probably no real methodical way to make these decisions, but I'm starting to think that 'projected future moving' is probably the best, most realistic heuristic for buying judgements, even if you never intend to move.

I've been packing up for at least a couple of weeks now, so I feel qualified to formulate the following taxonomy of plasticshit. There are three categories of stuff you come across as you prepare to move:

1. Stuff you just need and would have to buy it again if you didn't take it with you. Think potato peelers. You don't love this stuff, or even like it. You just sort of need it, in a middle-class potato peelery kind of way. You feel nothing other than boredom as you box this stuff up. Much of it will be beige.

2. Stuff you couldn't give a shit about, that you can't remember buying, that you resent having to spend £1.50 on a box to take it with you. You swear under your breath as you heave it into the van and sometimes just kick it for no reason. This stuff makes you hate yourself just a little bit and possibly your spouse just a little bit more. Every item will need to be argued over. You will want to take it to the dump. You will not be allowed to. You will put it in the corner of a room, move it to another room, offer it to all the neighbours, offer it to strangers. You will eventually take it to the dump.

3. Stuff you actually like. Your piano. Paprika. Maybe a nice pen. Not much else to be honest. This stuff is either huge and requires specialist help to move, or takes up next to no space and smells nice. You smile as you pack this stuff up. You give it an extra turn of bubble wrap.

Everyone will have their own ratio - I'm probably like 50/40/10 if I had to guess. Sad, I know, but in some ways it gives me hope for the future. If only I could beat the materialist-consumerist treadmill I might simultaneously save money, live better and not have to repeat this saga again, ever.

So how do you avoid it? God knows I don't want to do this again. Here's my new simple method. It's Mari Kondo for plasticshit. Before you proceed to checkout, picture yourself in ten years, surrounded by boxes. You pick up a piece of plasticshit. Close your eyes and imagine yourself having to bubble wrap this item. Resentment is a no-buy.

This is the question I'm currently asking myself as I sit on the floor, surrounded by a decade of digital detritus.

Seriously, how many backup keyboards are actually enough? I'm not talking about the keyboards we currently use (one British, with the pound sign, one Spanish, with the 'n' with the squiggle on top); I'm talking about the ones that live unused in a box marked (mentally) "Gadgets that I am not currently using but may one day need to fall back on".

I sometimes seriously wonder exactly how many neurons that label is taking up in my brain. Not just the label, but its association to that particular box that functions as a digital junkyard. For ten years I've had to remember that that bloody box is where the broken headphones, 8-year-old router and said keyboards are. Sometimes my grey matter fails me and I pull out the box next to it. Sometimes it takes me 2 or 3 goes. I could, of course, have just labelled the damn box ten years ago - but I didn't. Probably a good thing though as that box has lived other, equally unglamorous lives. Once it contained about 300 CDs, back when they were a thing. Oh, they are a thing again now are they? I struggle to keep up. I must admit, I sometimes have dark feelings towards the box. Sometimes, like now, I want to smash the contents into tiny shards with the massive sledgehammer thing I have in the shed. Or is it behind the shed?

I digress. Back to the keyboards. We have six spare: four British, two Spanish. There are also four mice. And a trackpad. I know this is too many, but how many to get rid of? Whole minutes go by as I deliberate. I can't easily offload the British ones here in Spain, so they will have to stay. I opt to donate a keyboard and mouse set that is only slightly stained with battery acid. Another (the source of the battery acid) is in such a state it goes straight in the bin. So I'm now down to four spare keyboards, which I figure is probably enough.

How the hell did I end up with so many keyboards? That's a fair question. I'd love to say I know the answer, but their origin story remains a mystery to me - like the booklet stapler, 7-hole punch, roll of thermal labels and barcode scanner that live in the next box along - the one marked (mentally) "Stationery, or stationery-like implements that I have not used for more than 5 years but might be useful one day should I decide to distribute propaganda and/or open a small grocery store".

So much junk. So much mental baggage. So much waste.

Truth is, we've both been doing this for days. Keyboard proliferation is not even the worst of it. How many hair waxes does one need? I have seven tubs: matte paste, high gloss jelly, paste wax, wax jelly, clay (!), beeswax and crème (whatever the fuck that is). Some of the pots are in English, which means they are more than 15 years old. Oh what hospital visits I could enjoy if I slapped on some of the yellowing goo. If only I still had a decent head of hair on which to use it. Perhaps I secretly believe that the 5% oxywhatever shampoo that is sitting, unused, next to the hair waxes is somehow going to restore my majestic mane.

So here we are, suddenly required to wind up a decade or more of a life lived in these four walls. At times I feel disgusted and other times ashamed at the amount of crap I've accumulated. I'm not a hoarder or a collector, and I've been constantly cleaning out, donating and selling stuff over the last decade. I am a legend on the local second-hand marketplace app. When I leave this place, I expect the local post office may have to lay off staff. "I think he's gone", they'll whisper on their tea break. You laugh, but when my old dad passed away, the girls from the local post office came to his funeral. Like father, like son.

For the last few years I've been pleading with friends and family not to buy presents or bring more junk into the house when they visit. Everyone knows the rule. Only books. Yet the aftershave still comes. I hate aftershave. I don't want to smell like a travelling salesman. I give my relatives two or three bottles every time they visit. Mind you, these are aftershaves they gifted me in the first place. There are ones they don't like and won't take back so I spray those under my armpits if I'm feeling special of a weekday morning.

We do the same little dance with wine with the neighbours. We host a dinner. People bring wine. We put it downstairs. It sits for six months. We go to their house for dinner. We take the wine back.

We humans, we're not meant to live like this, drowning in plastic shit and little bottles of ointment. When we led nomadic lives, moving from camp to camp as the seasons and local resources waxed and waned, we could only keep and transport the bare minimum: bows, arrows, a few stone tools and something to cover up your penis. Covering your arse was only invented with the agricultural revolution and henceforth pants took over the world (18 pairs, only 2 without holes, in case you were wondering). In Sapiens, his biopic of human history, Yuval Noah Harari puts it like this:

"We hardly notice how ubiquitous our stuff is until we have to move it to a new house. Foragers moved house every month, every week, and sometimes even every day, toting whatever they had on their backs…They consequently had to make do with only the most essential possessions. It's reasonable to presume, then, that the greater part of their mental, religious and emotional lives was conducted without the help of artefacts."

It should be stressed that there is no broad agreement in anthropology as to the nature or origin of what we would now refer to as 'private property', but a common argument is one that links the emergence or existence of personal dominion over artefacts to the notion of the sacred. Referring to musical instruments like trumpets and flutes, Graeber and Wengrow in The Dawn of Everything note that "these sacred items are, in many cases, the only important and exclusive forms of property that exist in societies where personal autonomy is taken to be a paramount value, or what we may call free societies."

We might summarise, then, that:

1. Early humans probably had few private possessions.

2. Those they did have were symbolic and sacred, rather than utilitarian.

3. Somewhat counterintuitively, this lack of possessions existed in a framework of maximal personal freedom.

With that in mind, I don't think it's any wonder that modern humans buy little jaunty pins with which to accessorise clogs. If private property was sacred and hard-to-come-by, we probably have something of a hardwired desire for it, right next to the circuitry that drives us to consume vast quantities of salt, fat and sugar. We end up physically and materially obese.

For the last 10 years I've been experimenting with a lot of hobbies and a lot of tech and that has necessitated what I always refer to as 'frictional costs'. I wouldn't shy away from buying a synth or a computer because I knew I could sell it a few months later, at an acceptable loss, if it didn't work out. To me, that loss was akin to what it would have cost to rent the item for the trial period, which of course is rarely an option. I always thought it was a pretty good system. Now though, as I lug a full bag of old instruction manuals down to the recycling point, I'm not so sure. I haven't consulted a physical instruction manual in over a decade, yet I've kept the booklets for seven generations of hair trimmers. I believe one of them might even be for a nasal trimmer.

Our new place is going to be smaller, about a third of the size. I'm also in a different place in my life now. I think I finally understand just how destructive this consumerism really can be.

I really, really don't want to do this any more. It's not just the wasted money - it's everything. Moving it, cleaning it, storing it, maintaining it, learning how to use it, selling it, disposing of it. I don't want the 'headphones, sport' to be on the ground floor, the 'headphones, non-sport' to be in the office, and the 'headphones, Apple, Lightning connector (obsolete)' to be in a little metal box, inside a bigger plastic box, on a shelf, in a desk, on the top floor.

What you pay for material goods is not just the price on the ticket-
it's the neural capacity lost to internalising the classification system of drawers in the hallway. You can buy those dividers from IKEA, but you're still the one that has remember what goes in each of the eight compartments, lest disaster strike and the keyrings end up with the stamps.

All this crap - it takes up mental, not just physical, space. I'm left wondering which way round the relationship goes: do I own this stuff or does it own me? And I'm haunted by Tracy Chapman's lyrics:

Consume more than you need / This is the dream
Make you pauper / Or make you queen
I won't die lonely / I'll have it all prearranged
A grave that's deep and wide enough
For me and all my mountains o' things

As an abstract concept, or set of concepts, derivatives can be gnarly, so let’s start with something concrete: speed.  Speed, or velocity, is the rate of change of position over time - in technical jargon, a ‘time-derivative of position’.  If I’m driving at 60 km/h, then for every minute that passes, my position, P, changes by 1 km - simple.  Velocity is the first derivative of position.  Can you guess what the second is?  Which familiar concept describes the rate of change of speed? The answer is ‘acceleration’.  If my speed changes from 60 km/h to 100km/h I’m accelerating.  If that change happens over 5 seconds, I am accelerating hard.  If it happens over a minute, I am accelerating gently.  Acceleration is the second derivative of position.

Let’s take another example from a different field.  In economics, inflation, or the rate of generalised price increases across an economy (assuming prices are actually increasing, which is pretty safe assumption these days), is the first derivative of price.  You might hear that inflation has increased over a certain time period, from, say 1% to 3%.  If inflation is increasing, we’re talking about the second derivative of price, akin to acceleration in the previous example.

First, and even second, derivatives are easy enough to comprehend, but once we move into third, fourth, fifth and beyond, it becomes near impossible to conceptualise their meaning.  What do the fourth (snap), fifth (crackle) and sixth (pop) derivatives of position represent in real life? Nothing I, at least, can get my head around.

Away from the purely mathematical-physical definition of derivatives, we find looser, but related uses.  In finance, a derivative is a financial instrument that derives its value from an underlying asset class.  Here we’re not talking about ‘rate of change’, but something like ‘degrees of removal’.  ‘Mortgage-Backed Securities’, the infamous villain of the 2008 global meltdown, derived their value (in the loosest use of the word) from what turned out to be the shredded up mess of near worthless loans that had been securitised (again, playing fast and loose with semantics here).  When it became clear that the loans were junk, the securities tanked.  Now imagine if the MBSs themselves had been used to create some other financial instrument - a second derivative of the original loans.  The world of finance is full of such examples.  As assets and instruments get divided up, securitised, rehashed and re-securitised, we move further up the derivative chain.  A house of cards, no doubt, but at least a bit easier to get one’s head around than ‘jerk’ (the third derivative of position).

That’s a wrap on ‘derivatives’; we’ll come back to them shortly.

My (British) mum spent 1978 in California.  I’ve often said she doesn’t know how lucky she was to be a twenty-something at that place, at that time.  If I had the chance, I wouldn’t think twice about travelling back in time and living my twenties and thirties on a weed-and-sun soaked beach somewhere near San Diego - no mobile phones, no internet, not even a bloody fax machine.  But let’s try to be a little objective here - some, maybe even a lot, of the digital tech that’s been invented since 1978 has been pretty useful.  When did it go to the bad place though? When did we reach the summit of Mt Digitalis and start descending into the cold river of bullshit that sweeps us away while we try to keep afloat?

Compare an acoustic to an electric guitar for a second.  I think few of us would have any serious complaints about the technological ‘leap’ represented by guitar electrification.  It seems like a net positive: broader range of sounds, easy amplification, possibility to practice quietly etc.  Any downsides would seem purely ideological in nature.  Purists might argue that an electric guitar is not really a ‘guitar’ at all, for example, but it feels like a push to suggest that an electric guitar takes something from you, undermines you, spies on you, or seeks to manipulate you.

I’ve thought a lot about this, and have always wondered (presuming it is actually in the past) when we reached the point where digital innovation stopped serving us and improving our lives and started slowly shitting all over us.  Recently, I made a list of all the ways our gadgets take from us; not only can it be used to evaluate what to buy and use, but can also serve as a yardstick for trying to work out when ‘peak digital’ was (the day before the iPhone was announced, in my opinion).

Try asking yourself the following questions with reference to an electronic gadget, a piece of software or even a digital service:

• Does it encroach on your privacy, steal your data, spy on you?

• Does it require you to agree to terms and conditions you’ll never read and could never understand just to use it?

• Are you actually renting it even though you think you’ve bought it?

• Does it require an on-going paid subscription?

• Does it tie you into a proprietary ecosystem?

• Does it show you ads?

• Does it try to sell you other products or services?

• Does it regularly require your attention? Does it send you notifications?

• Is it purposely designed to be addictive? Does it press your dopamine buttons?

• Does it have a significant learning curve?

• Does it require software updates?

• Does it have planned obsolescence built in? Does it need to be replaced every two or three years?

• Does it need to be regularly charged?

• Does it require internet connectivity?

• Does it have a long startup time when you turn it on or need to use it?

• Does it make you stupid? Does it cause you to end up losing simple human skills?

• Does it socially isolate you? Does it make you envious of or aggressive towards others?

• Can it incriminate you in the future if you make a mistake today?

• Does it try to manipulate your behaviour?

• Does it transform what should be an object with a specific task into a general tool that does everything?

Let’s imagine we can assign a score on each of the above metrics, say 0-10, with 0 for no effect and 10 for the worst offenders.  A simple analogue wristwatch scores zero.  A Casio from the 90s might rack up low single digits, mainly because setting the time with those little buttons was finicky as hell.  But a smartwach, what would that score?  Hundreds maybe?  They tick just about every box: needs daily charging, tracks your every move, tells you what to do and when, needs to be connected to your phone or the internet, locks you into an ecosystem - the list goes on.  Some of them even have screens that only turn on when you lift your wrist up to look at the time, which introduces an annoying lag into the simplest of quotidian gestures.  Tech innovators have given us watches that make telling the time painful yet we buy them by the truckload.  I’ve got one.  I bought it so I can get precise satellite positioning and with it know how slow I’m running down to the millisecond and have precise biometric data that tells me my heart rate is too high.  A forty-something who’s never run with a high heart rate whilst jogging.  Astounding.  

Let’s drill down on how bonkers this whole thing really is for just a second.  I have a solar powered G-Shock that never needs charging and can withstand a lorry running over it.  It cost about 80 Euros.  Instead of wearing that, I went out and bought a watch which commands control over almost every aspect of my life, tells me when to stand up and sit down, needs daily plugging in and helps me gradually lose any natural self-awareness over my own body’s signs and signals since I can just glance down and know immediately if I’m hot, thirsty or tired without even a second of introspection.  Plus, of course, it cost around ten times as much as my G-Shock, is crap for telling the time and needs further accessories to stop it getting damaging or destroyed.  I truly hate myself.

I’m not the only one to buy a shiny new octopus-gadget then almost immediately regret it as it wraps its nasty tentacles around my entire life.  Nor, by a long shot, is the Apple Watch the first time this has happened.  I had a Kindle until I realised that every page turn was being beamed up to Mr Bezos in real time and the 200 or so eBooks I thought I owned were actually just leased under a pretty shaky contract which Amazon could seemingly rescind at will.  Goodbye Kindle it was then.  Not long ago I relapsed and bought a Kobo.  Same shit, different technofeudal book retailer.

At times in the past I’ve tried to de-{insertDigitalServiceName} my life.  For example, about 5 years ago, after reading Soshana’s Zuboff’s masterwork ‘The Age of Surveillance Capitalism’, I set about de-Googling (it’s a thing, look it up).  This is not a feat for the fainthearted.  I jailbroke the Android phone I had at the time and installed a deGoogled operating system instead of stock Android, I moved off Gmail, I stopped using Chrome, I changed my default search engine to DuckDuckGo, I wiped my very extensive Google history and I toggled all the options to ask them not to track me any more.  Was it painful? Yes.  Did it work? Kind of.  I still use (and pay for) Youtube, because there’s no realistic alternative and having the richness of all the world’s knowledge at my fingertips is a poison too sweet to resist.  I still have to use some of the Google services for work.  They’ve probably still got a tonne of data about me.  But largely, I’m pretty free of Google services; I guess I semi-de-Googled.  I’m more relaxed about it these days.  Please Shoshana Zuboff, no more books.

Let’s be honest though, rather than having to surgically remove the tumours that are modern digital products from my body and mind, wouldn’t it be better to just avoid them in first place.  I’m please to say I have some experience with that approach too.  For whatever reason (actually it’s because I’m a mild sociopath), I never got into social media.  I’ve dabbled a few times, opened accounts for different business and projects, eventually closed them and generally been absolutely smug in my total denial that this quite sizeable social clusterfuck actually exists.  I’m glad I don’t have to put my phone in a faraday cage to protect me from myself and I’m glad I can’t accidentally say something whatever-phobic on Twitter and get deported for it. I exist in a blissfully TikTok free world, I don’t read Twitter and I don’t endlessly scroll Instagram, and the best part is I don’t have to expend a single, solitary drop of willpower to stop myself doing so.  It just isn’t a thing for me.

I think I’m going to have to follow this cold-turkey approach for AI as well.

I got to thinking more deeply about this because I’m about to exchange my acoustic piano for a digital one.  Although this is blasphemy to ‘proper’ piano players, I’ve come to the realisation over the 7 or so years I’ve been playing that I actually prefer digital pianos.  Real-string feel and resonance are less important to me than being able to control the volume, having lots of different sounds and never needing to have it tuned.  But ‘redigitalising’ my musical life didn’t sit quite right with me given my current attitude to all things electronic, so I sat down and came up with the previous list of questions as a way to evaluate whether a digital piano was a step backwards in my ongoing march towards the analogue utopia I have in my head.  I came to the conclusion that it was probably more like an electric guitar than a smartwatch, so I was alright.  Some of them have slow start up times and a learning curve attached to fiddling with the menus, but the more minimalist ones really are like digital versions of their acoustic counterparts but with bell and whistle sounds - so I would go for one of those.  In the piano shop, the assistant showed me beautiful Yamaha Avant Grande models, which are almost identical to my existing piano, with the difference that the sound comes out of an array of strategically placed speakers rather than from a hammer striking a string.  It looked and sounded incredible.  He also started showing me a new model with an 18-inch LED screen that connects to the internet so you can download……Stop.  Just Stop.  It’s a slippery slope from happily jamming on ’Nights on Broadway’ with a funky bass in the left and a smooth Rhodes in the right to a 13.99 a month subscription which gives you access to 128 oboe sounds.

On the walk back to the car, which I had left on the other side of a big public park, I ran into a group of Sri Lankans playing cricket next to a water deposit.  This was quite an unexpected site in a Madrid park, not least because I could swear the water deposit had not been there on the way in.  I realised I was lost.  No problem - I whipped out my phone, opened Maps and followed the directions to my ‘parked car’, which turned out to be less than 200 meters away - I could actually see it from where I was!

I’d like to say I was surprised by this, but I already know that my ability to navigate went by the wayside a long time ago.  I’ve lived in Madrid for 10 years and have been to the airport probably 200 times, but if I attempt the journey without GPS I always end up outside some army barracks on the shady side of town.  About 5 years ago, perhaps at the same time I was de-Googling, I tried de-GPSing, which was even less successul.  Armed with printouts, I’d attempt to take on Madrid’s epic spaghetti junctions and invariably end up on the R5 to Toledo and €3.50 out of pocket.  I quickly accepted that whatever gland or cerebral appendage that was once responsible for my geospatial awareness was by this point puckered to a sultana and totally useless and went back to my GPS.  I’ve never looked back.  Literally.

As I started the car I had a moment of clarity.  This is what the AI-fuelled future has in store for my entire brain.  This scared the absolute living shit out of me.

Back in the heady days when I thought AI might turn out to be something positive (3 months ago), I dabbled with a concept for an app or website that would publish articles where the user could tune the output to their own liking.  The original author would be responsible for supplying the basic content or argument of the post in using a sort of pseudocode, but the textual output could be dynamically tweaked by the reader according to their preferences.  Take this article for example.  What if wordy and sarcastic is not your thing? Maybe you’d prefer bullet points in Japanese. Or a sonnet.  That seemed like a reasonable approach.  It felt like we were moving towards a world where ‘creative output’ was taking a backseat to something else, something that we were all struggling to put our finger on.  Clearly the output of generative AI models, be it in the form of text, image or even video, is no more creative than a fart.  It is a bland average of everything that has gone before it.  But there is normally creativity buried deep within the dense AI slop - the spark of human ingenuity that led to the prompt or base instruction.  Clayton Ramsay, in I’d rather read the prompt captures this feeling neatly:

I have never seen any form of create generative model output (be that image, text, audio, or video) which I would rather see than the original prompt. The resulting output has less substance than the prompt and lacks any human vision in its creation. The whole point of making creative work is to share one’s own experience - if there’s no experience to share, why bother? If it’s not worth writing, it’s not worth reading.⁠

As James Walsh notes, everyone is cheating their way through college by turning in work done by ChatGPT.  How far are we, then, from the day when students are too lazy to even write a prompt?  I’d be astounded if we weren’t already there.  If your college professor says “look, don’t waste my time with 8 pages of output from ChatGPT, just show me the prompt”, what do you, as a time-pressed college student, do? The answer should be clear:  ChatGPT is great as creating prompts.

Prepare to descend into the meta.

Perhaps you’ve already seen where this is going but let’s try to formalise it a bit.  We’ll refer to the original, intangible, elusive creative spark as S.  S is an extremely abstract concept - its a synaptic crackle, a neurological blip on an N-dimension representation of mental space.  Think back to our discussion of derivates and position, P, which can be thought of a set of coordinates in a 2D or 3D space:  S is the psychological equivalent of that.  Whatever that is.  Starting simple, we might think of a prompt as the first derivative of S and the output of the LLM as the second derivative or maybe even more technically we could say that the output of an LLM is the second generative-derivative of S in the same way that ‘acceleration’ is the second time-derivative of position.  Less technically, the model output is twice removed from the original spark.

Once professors catch on that students are using AI to create the prompt in the first place, do they start asking  them to just hand in the ‘prompt of the prompt’ and then the ‘prompt of the prompt of the prompt’.  How far does this go? Once professors are seeking the 4th derivative of S, will we need specialist machines to just extract S directly? Is this what Elon Musk is after with his brain interfaces? What does S look or sound like? Is it just ‘snap, crackle and pop’? Have our brains been reduced to Rice Krispies?

I’m jumping off this train now before it builds up any more steam.  I don’t want to lose any more of my basic human abilities.  My ancestors (well maybe not mine, I have it on record that mine were more, let’s say, artistically inclined) roamed the savannah for hundreds of kilometres hunting gnus without so much as a map, yet I can’t find a parked car at 200 meters without satellite assistance.  Modern digital tech has stolen our physical fitness, our social connections, our attention spans, our ability to create democratic societies and now it’s after our most fundamental human ability.  How long will it be before you’re reaching for AI to write a 2-line email? To add up some simple numbers? To code a simple function?  You’re not, surely, already doing that right?

Personally, I’m not falling for this trap.  I don’t even want the temptation.  I was successful in totally excluding social media from my life and I’m going to use the same approach with AI.

The most common argument against this philosophy is “why bother if the machine can do it better?”.  Why write if writing has no economic value any more?  Why create art if models can spit out any sort of personalised creation in seconds? Why write code if AI can do it faster and better? Why play music when AI can generate any kind of tune on the fly?

I believe the key is in differentiating ‘process’ from ‘output’.  As highly-trained consumers, we’ve come to see output as the holy grail - something we can purchase, eat, watch, listen to.  But behind output there is process. At least when humans are involved.  There is creativity.  There is S.  No matter how badly AI obscures S, no matter how many derivatives removed we get from it, it still exists.  S is what makes us human.  S is what differentiates a text written by me from a text written by AI, even when they are word-for-word identical.

So this is the crux of it.  I’m going to stop using AI because I’m worried that my S will go the same way as my geospatial awareness and seemingly everyone else’s attention span, social skills and ability to communicate.  I still believe in process.  I want to write my own essays, thank you notes and code.  I want to conduct my own research by hand.  I want to seek and understand other human beings’ opinions.

Oh, and I’m getting rid of my Apple watch too.

Postscript

Before I hit ‘Delete’ on my ChatGPT account, just for shits and giggles, I gave it one last spin. I wanted to see if it understood the prompt derivative concept outlined here.  I fed it this article and asked it to create the third-derivative prompt for creating this same article. After a bit of prodding and correcting, here’s what it came up with:

First, a quick review of the layers for clarity:

• 0th derivative: The article itself

• 1st derivative: A prompt to write the article

• 2nd derivative: A prompt to create a prompt to write the article

• 3rd derivative: A prompt to create a prompt to create a prompt to write the article

Now, assuming the original article (0th derivative) is about:

How digital tools are reshaping human attention and the implications for creativity

Here is a true third-derivative prompt — deeply meta and clearly layered:

“Write a prompt that instructs someone to generate a second-order prompt — that is, a prompt whose purpose is to guide someone else in crafting a prompt to help a writer produce an insightful article analyzing how digital technologies (like smartphones, social media platforms, and digital workspaces) are transforming the structure and quality of human attention, and what downstream effects this has on creative thinking and expression. Your prompt should emphasize meta-prompting literacy: it should guide the writer not just on the content of the ultimate article but also on how to frame prompt-writing in a way that elicits depth, perspective, and cognitive nuance in the final output.”

It didn’t quite follow my derivative logic, but honestly, I think it might have conceptualised it even better than I did.  If you can get your head around the output I’ll be deeply, deeply impressed.

In just a few short months I’ll be 45 and you’ll be 25, which means a whole generation has passed.  I feel like now is the right time to write this - the business came to an abrupt end recently, the house is on the market and we lost Chili yesterday.  Of course, you don’t yet know about the business, the house or Chili and I don’t want to spoil the surprise, so I won’t reveal any real details.  The details aren’t really important to be honest - that’s why I wanted to write.  You’re probably thinking that this all sounds horribly sad, but it’s not, it’s wonderful actually.  Let me explain.

I don’t think it will come as any surprise when I tell you that you’ll lose a lot of people that you love over the next 20 years.  It would be cruel of me to tell you who and when, but again, that’s not the important part.  The thing to know is that death is always unexpected, even when all logic tells you its just around the corner.  Don’t ignore it;  know that people die and it’s never the right time and once it’s done its done.  The absolute finality of death will catch you by surprise so many times and you’ll keep stubbornly ignoring or denying it.  I want to tell you that you don’t get any second chances as far as the death of loved ones is concerned, so whatever you have to do, do it now.  All the things you want to do with them, to say to them, the trips, the photos, the videos, the hugs and kisses must be done now, because one day you will wake up and they won’t be there and there’ll be nothing you can do about it but wish you’d done things differently and that’s a hard feeling to shake.  Please read ‘Tuesdays with Morrie’ sooner rather than later.

By the way, I should remind you that you yourself are not immortal, even though right now, at 25, health and longevity are pretty much at the bottom of your list of priorities.  Have fun, by all means, but try to take better care of yourself in general.  Weightlifting (which you’ve stopped doing, why?) is not the be-all and end-all of exercise.  Take better care of your heart.  Running is great - you just need to learn to love it.  Start by walking a bit, running a bit and build up from there - you’ll thank me later.  Read ‘Born to Run’ for inspiration.  Looking good is great, I won’t deny that, but in time you’ll come to understand that fitness is so much more.

You’ve already lived through some pretty traumatic experiences - the gulf wars, 9/11, dad’s financial woes, changing school, grandad’s illness and probably quite a few more I can’t recall right now.  I’m afraid to say it keeps coming.  There’ll be more wars, a massive global financial meltdown, extreme weather events, even a worldwide pandemic of epic proportions.  Nobody really prepared you for just how harsh and unforgiving the world can be at times, so take this as a warning.  You must maintain a ‘buffer’ at all times so that you are ready to act when needed.  Disasters can be avoided and the impact of extreme events softened just by keeping a bit of gas in the tank at all times - that’s really the best advice I can give you.  Put some money aside for emergencies right now - you’ve never been good at that.  Have a good plan for when things go wrong, because they will.  Be a ‘prepper’ even if people laugh at you - they’ll end up asking you for hand gel, you’ll see.

Honestly though, it’s not these black swans that end up having the biggest impact - it’s the slow drift, the gradual change in the world around you.  That really adds up over 20 years.  At your age, it’s easy to be enthused by it all, to keep on top, to be the bold explorer blazing a path in this new digital world.  I’m not here to discourage you from that, but I think you should know that change is inevitable and that you too will get left behind, perhaps sooner than you think.  You laugh at those who only listen to music that was popular in their twenties and show contempt for older folk who find themselves struggling in this new economic reality while you flex your newly found skills with CSS float and Paypal ‘Buy Now’ buttons.  Know that you too will eventually tire of self-reinvention and the breakneck pace of development of the online world.  Im sad to say that in just one generation, the internet you know right now will change - and sour - beyond all recognition.  Enjoy it while it lasts, make your money and get out while you’re still afloat.  Don’t let it absorb you into its dark recesses.

You’re quite good at knowing when to get out actually.  I’m not sure I’m allowed to say this, but your gut instinct on when enough is enough turns out to be pretty good.  You’ll do OK on those really bold decisions.  Quitting whilst ahead is one of your strong points, so don’t be persuaded by others that ‘failure’ is to be avoided at all costs and don’t be swayed when others say you make big decisions too quickly and too flippantly.  It’s your style and it turns out it works pretty well for you.

Which brings me to the house.  

This is a tough one, because I’m still here at ‘the house’ but won’t be for long, so I feel I need to relate the thoughts I’ve been having lately.  I’ve been in two minds whether to tell you this, because I don’t want it to put you off, but I think you’ll do better if you know.  I’ll have a lot to say about the house, and houses in general, because they’re going to be an important thread in your life over the next 20 years.

This isn’t the first house you’ll buy or sell by the way, so let me talk a bit about the financial aspect first.  Your working-class mentality that houses are always a good (or the only possible) investment and/or always go up in price is well meaning, but not accurate.  In fact, you’d do well to improve your financial savvy now rather than leaving it for ten years.  Read ‘Rich Dad, Poor Dad’ at least.  Learn the basics of financial planning, portfolio theory and the relative merits of different financial instruments as investments.  In general, Rolly is right, houses are too illiquid and the costs involved in buying, selling and maintaining them mean that even if the sticker price goes up, even considerably, the yield isn’t always good, or even positive, in real terms.  Another thing Rolly is right about is that you can make mostly bad financial decisions as long as you get a few big ones right.  That’s so, so important and turns out to be true not just for you but for almost everyone around you.  I can’t elaborate on that any further without giving the game away, but keep it in mind and it will serve you better than almost any other financial advice anyone offers.

Here’s the thing though, house are not just financial investments.  They are a prism for the light of your life to shine through.  The best way I can describe it is that things that are only tangentially related have a way of getting entangled - businesses, people, dogs and houses - the way all the different colours of light blend together to make white.  The significance of each tends to merge into the other until they become inseparable - not physically, but emotionally, spiritually.  White light.  I can’t tell you exactly how these things relate to each other; all I can say is that this house is, was, Chili, and is, was, the business, and Paco and the Israeli’s after some shit goes down that I won’t even go into now, and all the other things that happened here over the last decade.  It somehow came to embody those things.  The belly rubs in the sun on the daybed, the ambulance in the street, the picture of us all together with the suitcases and the kids just after they arrived, Nicola balancing on the swan in the pool, the joints and philosophical musings in the focus hut.

Without those things, it feels just like a pile of glass and concrete.  I think that’s why we are selling it more than any other reason.  Or perhaps my desire to sell it subconsciously caused some of these things to happen.  Or perhaps karma pushed some of these things to happen so we’d sell it because some higher power knows that is the right thing to do.  I don’t know which way the causality runs, but selling it we definitely are.

I won’t miss the pool.  That might come as quite a surprise to you.  I’ll miss the good times, the people and parties for sure, but the pool taught me a really important lesson.  Some material things give a lot but take a lot from you too.  Avoid those.

Back to money.  The ethical code governing writing letters to one’s younger self forbids me from telling you to buy Amazon shares and just sit back and wait, so let me tell you what I’ve learnt.  Carlos is basically right, you’ll always make ends meet.  You won’t die from lack of money.  So just quit the worrying, please.  In general, try to develop a more positive relationship to it.  It’s not inherently good or bad, it’s just money.  Over the next 20 years you’ll see both ends of the spectrum - you’ll be bankrupt and at rock bottom and you’ll make stacks of cash.  Don’t get too carried away, it never really makes that much difference to your happiness levels. 

Oh, that’s a big one, happiness, let’s go down that sidetrack for a bit.

Somehow, somewhere along the line you will get labelled as an ‘entrepreneur’ and it will stick with you and encumber you.  You are not an entrepreneur - you never were and you never will be.  What you are is a good technical small solo business owner.  Entrepreneurs embrace risk and know how to turn investment and human capital into profit.  They are good managers.  You are none of those things.  Your technical skill, creativity and high-agency make you apt to earn money as an autonomous business operator, but try to shun the label ‘entrepreneur’ as soon as you can because it will end up causing you a lot of psychological pain.  For example, you are useless with customers.  I mean, you’re very good at dealing with them, obviously, but just like mum, you let them get to you, right into you.  I wish, wish, wish I could just tell you not to let them get to you, to just say ‘fuck them’, but that’s not in your nature.  So the biggest lesson I’ve learnt along the way is to stop dealing with customers as soon as you can - delegate that role.  I can’t say exactly how without revealing too much - you’ll work it out.

Since we’re on the subject of personality, let’s jam on that for a bit.  Another of my wishes is that I could spare you 20 years of anguish over the specialist/generalist question.  Yes, I’m sorry to say that as it torments you now, it will continue to do so for two more decades.  I’m happy to tell you though that I did eventually find the solution.  I’ve actually written it up!  I don’t think I would have found it without stewing on it for 20 years though, so I’m not going to give it away.  I’ll keep it here safe for you to discover.  But I will say this: you ARE a generalist.  It will take you a long, long time to make your peace with that and during that time it will be both your downfall and your making many times over, but it’s the unavoidable, inescapable nature of you.  No matter how hard you try, you will never be able to define yourself as anything specific.  You are good, great even, at many things, but world class at none - it is your defining characteristic whether you like it or not.  The crazy thing is that you’ll spend so much time and mental effort stewing over how to define yourself, when the definition is staring you in the face right now and will continue to do so until you finally see it.  Being a generalist seems so imprecise and all-encompassing to you right now, but in time you’ll discover that the things that interested you five or even ten years ago (learning, knowledge, humanities, music, self development) are the ones that are still important to you now and remain important even when you are my age.  In this you are remarkably consistent.  Those are things that define you.  Take comfort in that but keep growing.

What I’m about to say is hard hitting, but it doesn’t really change anything in my advice to you.  Being a generalist will be your making because it will be the superpower that makes the projects you undertake largely very successful.  The downside, however, is that there is a technology on the way that will usurp generalists like no other before it and will null and void a lot of the advantages you currently have over others.  That sounds pretty frightening, and unfortunately I don’t think I’m exaggerating at all, although not everyone agrees with me on this right now.  At some point you’ll need to make some bold moves because of this technology.  Keep your eyes open and act decisively.  Think deeply about what it means to be human and perhaps concentrate on some of those traits over raw technical skills over the next couple of decades - they will serve you well in a future where real human traits like compassion, generosity and agency become the most valuable.

You’re probably in Malaga at the moment right Jon?  I think Phil wants you to head back soon and go to Brighton to study under a different supervisor for a bit.  You should probably do that.  Enjoy your time studying, but don’t obsess over it.  I probably don’t need to tell you that you don’t end up making millions from your knowledge of amplitude modulation in voiced fricatives.

A few final words before I sign off.  The important people around you now will largely be the exact same group that are with you when you’re twenty years older.  They too will be twenty years older and be very different people, but the bonds you already have and will strengthen in years to come will end up being the most valuable thing you have.  Some will save you from yourself, others will step up and protect you when the going is impossibly hard.  You know that girl, Jessi, you met recently? Take care of her and mostly listen to her advice.  Occasionally she’ll be off, but mostly, the decisions she guides you into will be the best you’ll ever make.  Her stubbornness, consistency and conservatism are the perfect counterbalance to your sometimes unstable creative nature.  I’m probably really pushing the rules by telling you this, but together you will make a formidable team.

It’s funny Jon, I know exactly what the next two decades hold for you.  On the other hand, I don’t even know where I’ll be in a month’s time.  I hope that I myself can heed some of the advice I’ve given you in this letter.  I guess I’ll write you again in another 20 years with another load of advice.  Perhaps you’ll write back and let me know how it went.

Lots of love

Jon

Middle age, as I’m starting to realise, is a bit like that, or maybe like the non-borders between European countries:  you’re aware they’re there but are surprised when you pass them, like you could get to your destination without going through them.  You say things like “ooh, we’re going into Luxembourg” or “shit, I’m 45, better buy a motorbike/motorhome/motor-something”.

In my case it was a small caravan, a teardrop camper.  Not technically motorised, but close enough.  For context, I spend, and really have always spent, most of my time indoors.  Over the last few years that has started to change though - as I get older, I crave nature and the outdoors more and more.  I have no idea why this might be.  Perhaps the passage of life highlights how much time we waste.  When you start to see a resource as scarce and running out, you inevitably start to value it properly and I definitely regret having spent so much of my limited time on this earth hunched over a keyboard.  I think I feel the need to remedy that.  So now I’m sitting here in the great outdoors, surrounded by a serene mountain landscape, coffee in a tin mug, fingers frozen, hunched over a laptop writing this.  I’ll leave you to judge how successful I’ve been in this endeavour.

Back to the camper.  The idea was simple: we have two dogs so finding hotels is a chore.  We’re also vegetarian, bordering on vegan, so eating out is mostly a waste of time and money.  To be able to get out and explore we needed a bed and a kitchen, which is exactly what you get with a teardrop, so we took the plunge and got one and now we’re ‘practicing’ for a month-long expedition to the highlands by coming up to the mountains around Madrid and seeing if we can emulate bush-influencers on Youtube by rigging up a rain shelter with a tarp and two poles.  We can’t.  Turns out I’m more ‘typist’ than ‘survivalist’.

But the thing that’s really caught me off guard since we started taking these mini-trips not too long ago is this overwhelmingly nostalgic feeling I get from the campsites, and it makes me feel, well, decidedly middle aged.  It’s to be expected I guess, I mean I suppose I am middle-aged now, but the feeling has been strong and persistent enough for me to want to write something about it.

When you are young, you experience family holidays as a kid.  You enjoy yourself, do what kids do, and don’t give a second thought to the adults around you.  They are just there. Boring. Asleep a lot of the time.  But of course, they are also enjoying their holiday - perhaps enjoying a well-needed rest, nature, family time.  As a kid, you don’t care about that.  In my own middle age I still have a clear memory of what that felt like.  In fact, I think ‘being in the middle’ provides a unique perspective in that you are young enough to remember what it felt like to be young, perhaps even vividly, old enough to be ‘not young’, but not old enough to be, well, old.  It’s a beige age.  Being at a campsite, for me, really triggers this feeling of ‘being on the other side’ of the kid/adult divide.  Fading, middle-aged memories of youth blend into the present where I’m now very much playing the role of adult - the other side of the coin.  Strangely, I sometimes wish I could go back and make my younger self aware of this feeling.  When I see the kids running around the campsite, I wonder if they realise that their parents once experienced such a time and place as they currently are, or that one day they will probably be on the other side of this generational bridge.  Of course they don’t, and that’s how it should be.  Only adults have the ‘pleasure’ of contemplating this stuff - and it turns out that much to my surprise, some don’t.

Over last night’s dinner, served on plastic kids plates from IKEA (one of many Youtube fuelled, ill-advised camping purchases; you can’t get the grease off them no matter how hard you scrub), I tried to explain these feelings to Jess.  As I talked, she screwed her face up in a gesture of complete non-empathy (which I’m intimately familiar with by now) and I immediately realised that she shared precisely none of these sentiments.  At first we figured it was because of experiential differences growing up.  We were raised in different countries and different settings, me in urban London and her in rural northern Spain.  She never went on family holidays, so of course couldn’t vibe with what I was saying. But after a bit more discussion we realised that this actually ran much, much deeper.  She, in fact, has no real internal conceptualisation of a ‘life timeline’ nor does she ever think about or obsess about getting old.  I was dumbfounded, I thought this was a universal element of the human condition.  When John Meyer sings “Stop this Train / I want to get off and go home again / I can’t stand the speed it’s moving in” and “I’ll play the numbers game to say my life has just begun”, that resonates so deeply with me it always almost chokes me up when I hear it.  “You never do the maths?” I asked her.  “Nope”, she said, genuinely not knowing what the hell I was talking about.

Really? Come on? Who doesn’t do the maths? 45 times 2, that’s 90.  I could live to 90.  Great, so I’m not ‘quite’ middle aged.  How long have I been saying that to myself?  What kind of maths will I apply after my 45th birthday?  Not sure my basic multiplication and division will suffice.  Might have to get into nonlinear algebra.

Why though? Why doesn’t she think about age in the same terms as I do?  Our theory is that the socio-cultural treatment of age was so different for her growing up than it was for me that we’ve developed a different relationship to it altogether.  In her rural, agricultural village, everyone did, and basically still does, everything together, a single cohesive social unit, irrespective of age.  Social groups can range from 25 to 65.  Her school consisted of one classroom with all the kids together, from 5 years old to 12 years old.  Age was fluid, meaningless, analogue.  Back in urban London, age was discrete, digital.  If you were 12, you were literally separated from the 13 year old kids by a wire fence.  Age was everything.  Your entire social group consisted of kids your exact age, plus or minus a few months.  I read books, played games and watched movies, all demarcated with a neat circle indicating that it was for my age.  Jess just rode her bike around the village with all the other kids.

Could this be the root cause of her ageless existence?  I’m not sure, but here’s what I do know.  The concepts of ‘midlife’ and ‘middle-age’ are akin to those geographic non-borders we drive through on our way from Spain to Portugal, or France to Germany.  There but not really there.  We’d probably be best served by forgetting about them altogether but we can’t: the crumbling toll booths and abandoned customs posts are permanent physical reminders.

Moreover, just as the contemplation of the Greenwich meridian line as it bisects the earth warps our perspective of space, so contemplation of age milestones warps our perspective of the passing of time.  You can try, like I do, to divide your life into neat lines of latitude and longitude, but life, like the earth, is not flat.  It’s not really curved either.  It’s some other shape that I don’t have the maths (yet) to describe.  The meridians are there, but obscured by the rolling hills, rivers and mountains like the ones that surround me on every side while I’m writing this.  You never really know whether you’re in Switzerland or Austria, whether you’re GMT +1 or -1, or young, old, or even, as I may or may not be, beige age.

Given the novelty of the technology, this is, of course, a novel behavioural effect.  But it didn’t just materialise overnight.  It crept up on us (on me, at least), slowly, silently, insidiously.  It started a few years ago with the first releases of ChatGPT, where we’d ask it to write us a joke or a limerick.  That was cute.  Then we realised it could correct material we’d already written, or make suggestions for improvements.  As time passed it got better at basic reasoning and we soon found we were turning to ChatGPT, especially once voice interaction was added, for first-line triage of everyday issues like “How do I fix this drip in my toilet?” or (with video camera turned on) “My dog is acting weird - what do you think is wrong with him?”.  No plumber, not vet.  Straight to AI.

Now I find myself going straight to the models for even the most advanced questions, the ones I thought and prayed would be our exclusive remit, at least for as long as I’m alive.  I could resist, of course, but to be honest, a 15-page ‘deep research project’ conducted by AI on a topic or idea that has just sprung to mind is a gem too valuable to resist.  Ask first, think later.  I certainly don’t bother to puzzle out simple, everyday mathematical/geometric questions that come up, like “Will my new caravan fit in my car port?” or “How do you convert from a margin to a markup?”.

With hindsight it seems quite obvious that this was the inevitable outcome.  Human beings always tend to laziness, both physical and cognitive—it’s a natural consequence of the harshness of evolutionary pressures.  The self-preservation reflex pushes us to seek out shortcuts wherever we can so we can preserve some ‘spare capacity’ for lean times and emergencies.  We gorge when able, so we can better deal with periods of scarcity.  We rest when able, so we can sprint when needed.  The last 100 years have seen almost the wholesale replacement of physical labour, both in the workplace and domestically, with an arsenal of tools and machines that give many of us a huge amount of ‘spare capacity’ in both body and mind.  In the developed world, we no longer need to hunt, wash clothes in a river, transport ourselves by foot, program computers with punch cards or low-level languages, sew, cut or weld by hand.

What have we done with our spare capacity? In the physical realm, the outcome is clear to see.  We have vegetated—becoming unfit, sick, and overweight.  Those of us who try to resist do so by attempting to offset 16 hours a day of sitting with 45 minutes of HIIT every now and again.  In other words, we pissed our ‘spare capacity’ up the wall.

In the brain domain, things are a little murkier, perhaps because the progress has been slower and is still trickling through.  The standard argument is that as ever increasing technological ability has freed us from more mundane tasks like data entry, bit manipulation, manned surveillance and control and more, we can now focus on the more abstract, more ‘human’ tasks like planning, analysing, strategising - thinking, essentially.  This the story currently told in the world of software development at least.  Those who are bullish on AI claim that programmers will be freed from the necessity of writing actual code, allowing us to focus exclusively on the higher-level abstractions of business logic, architecture and user experience.  I’m not sure if the bulls have asked o1 for an architectural plan for a complex web app.  Look, I’m not saying a non-technical user could create a web platform simply by prompting an AI model - I’m saying that I, a senior dev specialising in web architecture, would tend to ask for a complete report from ChatGPT before putting neuron to synapse.  Ask first, think later (if at all).

First-line thinking is dead and buried.  How long will it be before all thinking goes the same way?

If the necessity to think is going the same way as the necessity to hunt, what will become of our undernourished grey matter? What will we do when we reach the ceiling of abstraction and there’s no higher we can go? When any vexing question can be solved by a stroke of the keyboard?  When any piece of software can be created with a simple description?  Will we feast on this externally available intelligence, only to seek penance in 45 minutes of HIBW (High Intensity Brainwork), 3 times per week?  Will the rich have ‘brain gyms’ in their houses full of shiny brain-training machines they never use? What will be the mental equivalent of a bicep curl machine?

Perhaps art, culture, music, literature, even love, are the higher abstractions we will seek out with our newly found spare capacity.  Freed of the need to think, we’ll just feel.  Or perhaps we’ll just watch Netflix.

Postscript

After writing this essay, it became immediately obvious to me that this wasn't the future I wanted for my own brain. In May 2025 I made the decision to stop using AI. You can read about that in the essay "Snap, Crackle and Pop". Let this be a testament to articulating one's thoughts through writing!

Jumping right in - why might SaaS be 'dead'?

Death of the Backend

For context, let's restate Microsoft's Satya Nadella's December 2024 assertion that SaaS as we know it is dead:

I think the notion that business applications exist, that's probably where they'll all collapse right in the agent era, because if you think about it, they are essentially CRUD databases with a bunch of business logic. The business logic is all going to these agents and these agents are going to be multi-repo CRUD right, so they're not going to discriminate between what the backend is - they're going to update multiple databases and all the logic will be in the AI tier so to speak. And once the AI tier becomes the place where all the logic is, then people will start replacing the backends.

It's fair to say this interview sent shockwaves through the industry - not because it was the first time anyone had floated the idea that SaaS was moribund, but perhaps because it was such a concrete formulation of how it was going to happen coming from a senior industry player. But combing through the comments on YouTube, reaction videos and subsequent blog posts, it was clear just how polarised opinion was on how this might play out. A lot of experienced engineers refused to countenance the possibility that backend logic could be subsumed by LLMs, exemplified by comments like this:

I'm a SaaS enterprise architect and have been coding for 40 years. This is 100% absolute nonsense because you will never predict the detailed requirements of the customer, who themselves (particularly, no single individual) can't define them accurately. Agents will only facilitate execution of pre-defined business logic on behalf of the user, and might express results in some fancy way, but the business logic itself is usually highly optimized and specialized, and usually discrete and purposely immutable, which AI is not.

On the other hand prominent YouTuber Matthew Berman has this to say in his reaction video:

"If you've been been watching this Channel at all, you know I've been talking about this for a long time. I truly believe the entire application stack is going to disappear - there just really is no need for it when you have artificial intelligence interacting with the core grounded data that sits within a database and again what does that actually mean for the entire SaaS industry? What does that mean for application developers? I don't know exactly but I have a feeling it's going to look vastly different than it looks today and I've been in SAS my entire career I have built multiple SaaS companies. I've worked for SaaS companies and if I'm being transparent I probably wouldn't start another SaaS company right now and I probably wouldn't invest in any SaaS companies right now either."

Whilst black and white thinking is attractive, it doesn't help much strategically because the final outcome, as always, will probably lie somewhere on a spectrum and it will depend very much on the SaaS in question. My own feeling is that logic-heavy, complex, transactional, domain-expertise-laden platforms (think ERP, accounting and the like) are probably 'safer' than simpler systems with less ingrained business logic (to-do applications come to mind). In other words, the closer you are to CRUD, the more your platform looks like a thin layer over an SQL database with a dollop of custom UI, the more at risk your software. I'm by no means sure about this, nor could I say for how long that relative margin of safety might last. Developments in LLMs over the past month show that making any kind of predictions about their abilities is likely to make one look foolish, at best.

Service-as-a-Software (or SaaS Inverted)

Before Nadella's bombshell comments in December 2024, the venture capitalists had already cottoned on to the fact that SaaS as we know it was heading out the door, to be replaced with a new software-like industry: vertical AI. The story goes something like this: modern SaaS is built for human teams to use. Humans cost money and need to be fed and watered. AI can replace many human workflows, especially those that are language heavy. So rather than building software for human agents, let's build software that comes with agents built in. In other words, as a business owner or manager you won't buy software for your team, the software will be your team. Listen to Jared Friedman from Y Combinator weigh on why vertical AI agents could be 10x bigger than SaaS:

"Here’s my pitch for 300 vertical AI agent unicorns. Literally every company that is a SaaS unicorn you could imagine, there’s a vertical AI unicorn equivalent in some new universe. Because most of these SaaS unicorns, beforehand, there were some box software company that was making the same thing that got disrupted by a SaaS company. And you could easily imagine the same thing happening again, where now, basically, every SaaS company builds some software that some group of people use. The vertical AI equivalent is just going to be the software plus the people in one product."

Of course this doesn't mean the SaaS incumbents are going away overnight, just that the industry as whole will gradually shift from SaaS to Vertical AI, just as it shifted from 'box software' to SaaS a generation ago.

The reason why this shift is just so tantalising to VCs comes down to where the increased 'software' spend is going to come from. Listen to Caty Rea from Bessemer Venture Partners explain it in a video entitled Vertical AI shows potential to dwarf legacy SaaS:

"They’re spending almost as much on this little kind of one-off service as they are on their core system of record. So, I know you know the answer to this, but, why does that make sense? Why aren’t they balking "wait a minute, we can’t double our technology budget in a year" — how do they think about it? Yeah, I think, you know - it’s a leading question, and the reason is because they’re really not. They’re just replacing existing services spend, and so they don’t need to create entirely new budget. Instead, it’s a reallocation of services spend to something that is productized thanks to AI."

Put simply: replacing humans with AI agents. Beyond the obviously seismic socioeconomic earthquake that's likely to cause, there's a fundamental nugget of wisdom that can be extracted from this assertion. In the example they are talking about, which concerns the 'legacy' SaaS platform Litify (a system-of-record workflow software for personal injury law firms) and the new vertical AI startup EvenUp (sells a service to those personal injury law firms drafting legal demands to insurance companies), the incumbent SaaS product and incoming vertical AI product live side by side. I think this idea is worth dwelling on a bit longer. Returning to the Bessemer roundtable, they see the new opportunity as split two ways between software entrepreneurs who are quick to identify AIable workflows and professional service companies essentially AIising their offering:

"If you can capture some of that human spend, that’s a really powerful opportunity. And, and I think it’ll take two forms. One is the EvenUp-style of a denovo venture-backed company: a really talented entrepreneur identifies a workflow, automates it, and, you know, it’s lightning in a bottle. I think we may also start to see traditional services businesses—you know, BPOs, law firms, accounting firms, experts, adopt these types of technologies to drive automation inside of their business. You ask for a spicy take: I think the demise of professional services firms is a bit overblown. If anything, some of the value associated with AI may end up accruing to these professional services firms that can harness it."

That's an interesting, and much more savory, option for us SaaS founders and one we'll explore when thinking about AI strategy.

But before we do that, let's briefly address the elephant in the room.

Software is Worthless

A parallel but somewhat related debate is currently raging amongst software developers in the comments sections and on Reddit. It concerns the answer to the following question in different guises: "I'm just about to start a 5 year Computer Science degree in the hope of becoming a software engineer/developer, should I bail?". As you'd expect, opinion is polarised, not least because software development tools have been making engineering 'easier' for decades now. Higher levels of language abstraction, frameworks, superb developer tooling - all have increased developer productivity. So there's an argument that the new AI-powered development paradigm is just an extension of a trend that perhaps started with 'no code' or 'low code' builder platforms and now extends to LLMs doing most of the coding for developers.

My own feeling is that this time it's different (I know, famous last words). I've tried Lovable, one of the new breed of AI full-stack development tools and the results are astounding. They are limited in their scope, I get that, but where they are a good fit, which is quite honestly a lot of use cases, they are undoubtedly an industry-shifting tool that is only going to improve. I watched a non-developer founder friend build a full SaaS platform MVP in about 5 days - a process which he'd previously been through only 3 years early at a cost of hundreds of thousands in seed capital and probably about a year in time.

He and I don't necessarily see eye to eye on whether this is a good thing for SaaS founders though. For non-technical or semi-technical founders you can see this type of tool as either a godsend or a death knell. Yes, you can build a prototype, an MVP or even a somewhat functioning SaaS platform (remember, today's AI is the worst you'll ever use) in essentially zero time at essentially zero cost, but why bother? SaaS, which was once gold, is now sand.

At least that's one, not particularly nuanced way of seeing it. We'll see how the 'software is worthless' argument fits in with the different strategic approaches to AI that I've identified and will now outline.

7 Strategies

So, as a SaaS founder, what are your options right now? After a pretty extensive round up of current thinking, these are what I think represent the main strategic paths - a conceptual framework for SaaS operators to think about the way forward with AI. It goes without saying that they are not mutually exclusive, you could adopt one, two or all of them. I've tried to list them in rough order of how deep they involve jumping in, starting from 'don't go anywhere near the pool', through 'sit on the edge with your feet in' to 'nose dive straight into the deep end'.

Strategy #0: Milk the Cow

It might sound like sticking your head in the sand, but paradoxically I think 'do nothing' should be the default option for many SaaS platforms, particularly those that fit some or all of the following criteria:

• Mature, been in the market for a long time

• Lots of existing customers

• Large existing revenue base

• Little capital or resources to build AI

Something that forward thinking technologists often overlook is that adoption is sloooooow and legacy software often lasts a really, really long time. A bet you know examples of really big businesses still using spreadsheets to run their entire organisation, and maybe even signifcantly sized businesses using pen and paper. I have a friend that works as a well paid Fortran programmer for one of the biggest banks in the world. Fortran for God's sake.

So riding the long adoption curve of AI and milking the cash cow as long as possible whilst perhaps not building too much 'worthless software' could be a decent strategy for many companies. Just don't expect it to last forever.

Strategy #1: Be the Database

Of course, I don't literally mean just the database. Postgres is the database. But if your bet is that at least some business logic layer, however thin, will always be necessary, then this is your play. Essentially, you're predicting that core system-of-record type platforms will continue to exist for a very long time and won't be displaced by AI-only LLM-powered logic layers.

Here's who I think this applies to:

• Mature, been in the market for a long time

• Enterprise

• Platforms with very complex, transactional business logic

• 'Wide verticals' (eCommerce, ERP, CRM etc)

To 'be the database' in the forthcoming age of AI does mean adapting to AI-centric workflows and UI - it just means you won't be the one building them. You will watch as the battle for supremacy in the 'AI Platform' wars unfolds between the really big players like OpenAI, Microsoft, Anthropic, and Google as well as the current and future agent building platforms like Crew and Langchain. You'll expose your core business logic via well-documented and well-structured APIs that can be connected and consumed by agents, allowing your product to remixed into and combined with an ever evolving suite of other backend systems-of-reference, but you're probably doing this anyway, right?

Strategy #2: Sprinkle

This was the first and most obvious strategy to emerge in the early days of generative AI, before we all started getting obsessed with agents. The idea was to give the admin user simple, one-shot tools to ease annoying, otherwise manual tasks like:

• Removing a background from an image

• Drafting a reply to a customer ticket

• Create cookie-cutter content for marketing

• Creating templates for marketing emails

• 'Reading' a paper receipt and creating an expense record

I think for a while it seemed like this would be where AI's role in SaaS would start and end. It felt like sprinkling a little bit of magic on top of existing admin panels, but definitely not a step change, much less the 'death of SaaS'. Oh to go back to those days...

Strategy #3: Customer Facing

Possibly a variant of the 'Sprinkle' strategy, this approach applies to SaaS platforms that have both 'admin' and 'end' users - the canonical example being eCommerce (where 'admins' run the merchant businesses, and the 'end' users are the customers that shop on their stores). Examples of adding AI to customer-facing features include:

• Custom product recommendations and personalisation

• Customer-service bots

• Powerful search and help

The danger I see here for incumbent SaaS platforms is that a lot of these functions can and will be built better by third-party vertical specialists. Data integration will continue to be a problem, as it always has been, but there are thousands of plugins and widgets in the 'customer facing' category that you can add onto your WooCommerce, Shopify, Wix or Magento store, and I don't see that changing - in fact I think its a category that's set to explode in terms of third-party development.

Strategy #4: Copilot

Now we're really jumping in, perhaps not right at the deep end, but definitely not the kids pool. Copilots are starting to pop up everywhere in the SaaS world (and the software world in general). You can have a copilot write a function for you in Excel, reword a response to a customer in a ticket system, or pull up a report. Whilst some might already consider this 'agentic AI', I don't class it as such in my strategic thinking because Copilots are generally:

• Not autonomous

• Limited to small sets of well defined tasks

• Not able to interact with each other

As always, there's a spectrum here between a simple copilot, more complex copilots, and fully agentic AI agents (SAP's Joule appears headed that way) and there's crossover with the 'Sprinkle' strategy (Shopify's Magic appears to be a good example).

From a strategic viewpoint, its a pretty safe, if vanilla, bet, pretty much independently of what your platform is and does - they work well almost anywhere. Time will tell whether the copilot UI and use paradigm really beds in and becomes an expected and accepted interaction pattern, in which case most platforms will probably be forced to add some kind of copilot - in other words, copilots become as ubiquitous as hamburger menus, dialogue boxes and popup notifications.

Strategy #5: Fully Agentic AI

Firstly, let me be clear about what I mean here. I'm talking about adding a fully built out suite of agents, their orchestration and interaction, and the UI to build, manage and communicate with them directly into your platform. This means that users of your platform can basically put their feet up and have the software do the work for them.

Now obviously this applies less the thinner your vertical is - if your SaaS is a one trick pony then you can just do 'vertical AI' (see below) by either launching a parallel offering or completely replacing your traditional SaaS product with a vertical AI product.

No, this strategy is going to be primarily for large, somewhat multi-purpose SaaS - ERP, eCommmerce, CRM etc - platforms that border on the general use or 'wide vertical' or even 'horizontal' definition.

Right now we're only seeing this in a handful of cases, and only with the real enterprise heavyweights like Salesforce's AgentForce.

My feeling is that this approach is going to be out of reach for small and mid-market SaaS platforms who would be better served focusing down on their core functionality and integrating into the wider agent ecosystem.

Strategy #6: Vertical AI and/or Service-as-a-Software

I was in two minds whether to split this into two separate strategies but really it's two sides of the same coin and depends more on how the product or service is projected. Vertical AI is both software and a service and I suspect the distinction is more marketing than anything else.

Should incumbents go after slices of the vertical AI market? I think it depends on whether there is really high-value core expertise (or even straight code) that can be ripped out of its comfy bed in your existing SaaS platform and transplanted to an AI-powered workflow which can be packaged up and sold as 'Service-as-a-Software'. That, of course, is fraught with risks - does it devalue your existing offering? Do you have the expertise and resources to build a whole new product? What is the exposure to competition for this particular vertical?

Ultimately, if the VCs are right, this is where trillions of dollars worth of value will be created in the new AI 'software' market, so it might well be too tempting an opportunity for many entrepreneurial SaaS operators to ignore.

What I do suspect we'll see are SaaS companies leaning heavily on AI for the professional services they already offer - think support, implementation, integration, customisation and training. There are quick wins to be had here for any SaaS vendor, not only for cost cutting but also for streamlining and speeding up onboarding pipelines, the bane of any non-trivial SaaS platform.

Conclusion

SaaS decision makers are somewhat caught between a rock and a hard place right now. With the AI landscape changing at breakneck speed, it's almost impossible to make a big decision with lasting repercussions. For what it's worth, here's how I'm thinking about this in my own company:

• What is our core value proposition? (Business logic, UI, workflows)

• How far from our core value proposition should we stray?

• How vertical is our vertical? (Wide, general purpose vs narrow well-defined vertical)

• What are our short, medium and long term business goals? (Customers, revenue, business value)

• Should we look outwards (integration into ecosystem) or inwards (try to capture AI value internally)?

• What is the appetite for AI adoption in our segment? (Conservative vs forward-thinking)

• What resources do we have available?

• How can me improve internal operational processes with AI?

How do most business owners order stock?

Well, honestly, they probably just eyeball it. A lot of business owners have been “in the game” so long they can probably just guess at how much to buy. 

I ran a company that supplied restaurants and hotels for a good while and can tell you that precisely 0% of our customers were doing anything other than calling us from inside their walk-in cold room with a list scratched on the back of a chopping board directly from memory.

One Step Up

Beyond just eyeballing, I’d guess the next most common method would be basing ordering off averages. Averages are a relatively simple metric - well understood by most and on the surface, a good indicator of what to order for the forthcoming weeks and months. Let’s take an example:

Acme Widgets have sold 10 widgets per week for the last 3 months and now needs to order stock for the next 3 months. Simple right? Start by working out the average monthly sales, which is just 10 x 4 = 40, then multiply that by the 3 months to get an order of 120 units.

The Hidden Problem with Averages

Let’s take a different example but with a common theme: a monthly average of 40.

This time though, instead of a constant 10 sales per week, the pattern looks like this:

• Week 1: 10

• Week 2: 29

• Week 3: 1

• Week 4: 5

• Week 5: 14

• Week 6: 12

• Week 7: 3

• Week 8: 6

• Week 9: 7

• Week 10: 11

• Week 11: 2

• Week 12: 20

That’s radically different right? Sales are not constant at all, they are pretty much all over the place.  Honestly, though, this is probably a much more realistic scenario for most businesses.

Statisticians refer to this as variance - how much the individual numbers that go into an average vary around that average. You might remember studying standard deviation at school and variance is exactly what standard deviation quantifies. The higher the variance/standard deviation - the more “all over the place” the numbers that make up the average. The lower the variance/standard deviation - the more tightly packed together they are.

But what does this mean for stock ordering though? In the end we might still be tempted to add all those numbers up (120) and divide by the number of weeks (12) to get a weekly and monthly average (10, and 40 respectively) - which leaves us right back at square one.

Every Week Is Christmas

Look back at the numbers. What was the highest weekly number of sales? 29, right? So if we want to be sure we won’t run out of stock, we could assume we’ll sell 29 units each week for 3 months. That would mean an order of 348.

Compare that number to the 120 we decided we need to order earlier, based on the average. Ordering 348 gives would give us the advantage that we are MUCH less likely to run out of stock, but it comes with some BIG disadvantages from a business perspective - a big capital outlay, a risk of having a lot of stock lying around for a long time, and in the case of products with a shelf-life, the risk of expiry and having to throw some stock away.

So clearly, assuming every week is Christmas isn’t the best tactic. Is there a better one?

Every Week Is Not Christmas

Let’s summarise where we are.

We understand that ordering according to just the average is a poor strategy that doesn’t account for variance. We’ve also seen that ordering to the “best sales scenario” is also a poor tactic which could leave us hugely overstocked and potentially out of pocket.

What we want is to be “reasonably sure” we won’t run out of stock. In order to achieve that we’ll need to order more than the average, but less than the best case case scenario. 

Luckily we don’t have to pick a number out of thin air. There’s a mathematical middle ground that can really help us and it’s based on a concept called confidence intervals - it lets us specify some threshold, like 99% or 95%, of “confidence” that we’ll be able to cover sales.  It’s just a statistical measurement based on existing data, so it’s no guarantee - but it’s much better than an average.

Going back to the above numbers and running the calculations: for a 95% confidence interval, the “upper boundary” is approximately 14 and the lower boundary is approximately 6. This should be interpreted as “we are 95% confident that sales per week, during the period we are buying for, and based on past data, will be between 6 and 14.

That’s pretty cool right? 95% is plenty of confidence.

Scientists tends to use this statistical threshold when looking for “statistical significance” in experiments, so it’s certainly good enough for stock ordering!

So if we can be 95% confident that sales won’t exceed 14 per week, we can multiply that up by the 12 weeks to get an order of 168. Let’s now compare the different ordering strategies we’ve looked at:

• Basic average: 120

• Every week is Christmas: 348

• 95% confidence interval, upper bound: 168

Using the 95% CI technique, we’d be ordering “a bit more” than what the average tells us to, but in exchange we are getting MUCH more certainty we’ll be able to cover sales.

In Practice

Obviously you’re not going to be calculating confidence intervals on the back of an envelope. You could use a spreadsheet or perhaps a small script your developer could make for you. If you use Pakk, our integrated demand planner is based on this fundamental technique (see the detailed explanation of how the calculations work).

First, take a deep breath

If you've got here because you've worked yourself up into a state of rage because the fancy accounting/ERP/web system you pay all that money for each month has produced an invoice that doesn't match up with an invoice you've been given, or a customer has complained that the totals on an invoice you've produced are incorrect - take a deep, calming breath and know this: there is no single, correct way to calculate VAT.

Right now that might sound plainly ridiculous to you, but stay with me for a rock-and roll-tour of VAT and rounding and at the end of the article I'll point you to some hilarious forum discussions where people really lose their shit over this.

How could there be different ways of working out VAT?

Surely 20% means 20%?

Yes and no. It all comes down to when you perform rounding. If you can't remember what rounding is, you're going to struggle with the rest of this article, so here's a refresher.

Rounding 101

The result of a calculation involving amounts of money and fractions (or decimals, or percentages etc etc) can result in numbers that don't make sense within that monetary system. That's a mouthful - here's an example:

In the UK, standard VAT used to be 17.5%. So a £1 + VAT product should cost £1.175 right? But what does that even mean in a monetary system that gave up 0.5p coins donkeys' years ago? And that doesn't happen just with funky VAT rates involving 0.5%s. VAT is 20% right now, so a £3.21 + VAT product should be £3.852 inc. VAT, which is even more meaningless.

So for a monetary amount of 3.852, what do we do? Well, most modern monetary systems divide their primary unit (e.g. Euro) in hundredths (e.g. Cents), which means we need to round to 2 decimal places.

Now, there are a few ways rounding can be accomplished:

• always round down: 3.852 -> 3.85

• always round up: 3.852 -> 3.86

• commercial (common) rounding: round up or down depending on the rest of the number - without going into technical detail, this is probably what you remember from school as "round down if it's below 5; round up if its exactly or above 5".

None of these are technically 'correct', but commercial rounding is the one normally used, well, commercially. It's the safest bet and usually the one either allowed or mandated by accounting standards. It's the one Pakk uses. For uber-geeks, we round -0.125 to -0.13not 0.12, which is how 'commercial' rounding applies to negative numbers.

If you're starting to already see how things could get gnarly, buckle up - we haven't even fired up the engine yet!

When to round

OK, so we understand that we need rounding because we know prices like 3.852 just don't make any sense. Let's dive deeper into how the simple act of going from 3.852 to 3.85 can cause mayhem in an accounting system. We'll used a worked example. I won't include any specific currency, just imagine this applied to your home currency, like GBP or EUR.

Let's say there is a product, we'll call it Product A. The ex. VAT price is 13.25, the tax rate is 21% and so the inc. VAT price is 16.0325. If you went into a shop, picked up the product, went to the counter and were asked to pay 16.0325 you'd rightly be confused. Luckily, that would never happen. You'd be charged 16.03.

Let's just step back a minute and philosophise: what actually happened to the 0.0025 that got lopped off the end of that price? Where did it go? In the game of money there are always winners and losers right? So who won? and who lost? Well right off the bat, we can see that you, the consumer, saved yourself the princely sum of 0.0025. That's a lot of cash. To put that in perspective, if you did that 10 times, you'd have saved 0.025 which is....hmm, also in need of rounding. Forget it. You saved some money. So who lost out? Maybe the retailer - after all, he collected 0.0025 less then he would have done without the cunning rounding stealing his income. That's true, but remember, the VAT that the retailer collects goes to the tax authorities eventually, so really it was the tax man who lost out here. Shame. Well, you win some, you lose some, as we'll see later.

Right, so we've established that the effective inc. VAT price of Product A is 16.03. Cool. So you go back to the same shop to buy 4 of them to complete your set. You take them up to the counter, and as you are counting out your coins to pay exactly 64.12 (16.03 X 4), the cashier throws you a curveball and asks for 64.13. Did you think you were always going to be a winner in the game of VAT and rounding? Sorry to burst your bubble. But what just happened there? It's actually quite simple to understand. The till system used by the store did the multiplication of quantity (4) first, then calculated the VAT. So: 13.25 X 4 = 53; 53 X 1.21 = 64.13.

So who's laughing now? Earlier in the day, when you'd paid only 16.03 for a single unit of Product A, you'd left the shop highly contented with the bargain you'd scored. Now, as you trudge away after having to scrape together that extra 0.01, you leave the store highly discontented. Surely you've been scammed out of a penny? Where did that penny go? Well, assuming the merchant accounts for VAT in the same way the till calculated it (which might not be a valid assumption, even though it would be a mini "fraud" if they didn't), she will pass on 11.13 to the taxman, leaving 53.00 as her net income. That represents exactly 13.25 per unit, which is what you paid in the morning for the single unit. So, as you probably already guessed, that penny went to the taxman. Like I already said, you win some, you lose some.

So, the difference between these two examples is down to where the rounding was applied. In the first transaction, since there was only one unit, the inc VAT price of that single unit was rounded. In the second, the line total (i.e. the total price after multiplying by quantity) was rounded. If the same method had been used in the second transaction, we would have multiplied the already rounded 16.03 by the quantity of 4 to reach a total of 64.12, which is what you were expecting to pay, right?

From now on, we're going to refer to the above methods as the unit method and the linemethod, respectively.

Actually, there's another way

OK, this is where it gets really complicated and working an example through is too hard to follow in an article like this, so I'll include a spreadsheet if you are interested in looking up particular combinations of numbers that can cause the scenario I'm about to describe. For now, just trust me.

In the second transaction I described above, and in transactions of its kind, we refer to lines. A transaction is composed of multiple lines, each line refers to a product and quantity. So in the above transaction, there was only 1 line, with a quantity of 4. Simple.

However, in a commercial setting, transactions tend to have many lines, so there's now an extra level of complexity. In a multi-line transaction, it's entirely possible to leave each line Inc tax amount in its raw, unrounded state, add them all up and then round the total!! This gives the same result as just adding up all the line Ex VAT prices and doing a single VAT calculation right at the end. Predictably, this can give a result that differs from both of the methods described above. I'm going to call this final method the transaction method.

Actually actually, there are at least 4 other ways

From the previous article I wrote about rounding, I can think of all these approaches to rounding, and I'm sure I haven't got all of them:

• Round the unit ex price, multiply by quantity, apply tax, round again

• Apply tax to the unit ex price, round, multiply by quantity (the “unit” method)

• Multiply the unit ex price by quantity, calculate tax, round tax, add to line ex (variant of “line” method”, with rounding of tax)

• Multiply the unit ex price by quantity, calculate tax, add to line ex, round line inc (variant of “line” method”, with rounding of line inc)

• Multiply the unit ex price by quantity, round, apply tax, round again ( “line” method but with rounding of line ex, and rounding of tax, this is what we do)

• Multiply the unit ex price by quantity, apply tax (no rounding at all)

• Round the line inc. amount, sum those to get to the transaction inc. total

• Round the line inc. amount but ignore those and add all unrounded line ex and line tax, then round that, to get to the transaction inc. total

• No rounding anywhere (leads to 'unreal' transactions amounts like 67.452, that can't actually be paid)

Just tell me the right way to do it

The right way to do is consistently. That's pretty much the only answer I can give you. Once you understand that, you'll start to understand where the torrents of rage you can find online come from.

Most tax authorities are happy for you to just choose a way and stick with it. As long as you're not doing open fraud or have discovered a way to cheat them out of millions in tax revenue, rounding is going to make very little difference in the long run. I know this to be the case in the UK (link below), and I suspect it is the case in most countries, but at the end of the day, I'm not a certified accountant or tax guru, so you'll have to try to find out how it works in your country. PS: good luck with that - this is such a minefield of complexity that 'experts' will probably not be able to give you a straight answer and it might even be totally ambiguous in your country's tax code. In the UK it has sometimes gone to court!

The really irksome scenario is where different systems do it in different ways and I believe this is what drives most people nuts. Given the complexity I described above, most casual users could be forgiven for not understanding the layers of accounting assumptions that are baked into any ERP/accounting/transactional system and so when they find that a transaction their system produces doesn't match up to the "same" transaction on someone else's system, they assume that one of the systems is "wrong" or "broken" and get into forum or helpdesk flamewars (see links below for some real nutters).