The Jigsaw Doesn't Make You a Carpenter

Yesterday morning I had a discussion with my colleague Yannis at work. We spoke about AI and where the world is going, what’s going to happen to jobs, what’s going to happen to apps and the digital world. And one thing that came to my mind is a realization that I’ve had over the past 20 years.

I’ve been in this industry since around 2007 or 2008, full time. But from the very early 90s, as a kid, I was already playing with computers. Very early on, even at the very early stage of my high school, I was fixing computers and making my first money.

And the core realization that I’ve had, working with clients, working with people, fixing their computers, is that we often forget that people actually very much dislike computers. I would say that throughout my career, a majority of people don’t want to fix their bugs. Even if they can. They don’t want to be cleaning their computers. Even if they knew that they can open it up and vacuum clean it. People don’t want to be changing their thermal paste. They don’t want to be unscrewing coolers. And they certainly don’t want to be building their own applications and maintaining their own applications.

It was never about capability

When I think about AI, which I still treat as some sort of a miracle and in that sense don’t treat it as a hype, I do believe there is a loud narrative around this idea that people will all be building their apps. That is simply false in my experience working with computers and people all these years.

Not because people are not capable of doing it, by vibe coding or whatever else. It has nothing to do with capabilities. It has to do with the fact that that’s not what people want to do.

Most people don’t want to be sitting at home looking at some screen or text editor, deploying, building, and debugging apps. Let’s even imagine we are at a place where you can just sit and perfectly debug everything with no knowledge of programming. Most people still prefer other activities. Many that I know can barely sit for two hours in front of a computer unless they’re playing video games, let alone eight hours a day staring at a screen writing some gibberish.

Saying that everyone will be a programmer because AI exists is like saying everyone is a carpenter because jigsaws exist. The tool is there. That does not mean people want to use it.

99.9% are consumers

Think about any sort of creation on this planet. Music, art, programming, digital art, building games, building applications. It’s not even 0.1% that are the builders. I would say 99.9% of people on this earth are consumers. They’re not builders. This is also what Yannis found very intriguing as an argument.

We are not all building apps. We are not all producing music. We are not all producing our food. We are not producing our own cheese. A very, very tiny portion of the planet is doing the actual work, enabling people to have tools, enabling people to have products, creating hardware for people to use and consume. The rest of us are doing things we prefer, whatever our hobbies and interests are.

I’ve seen this personally too. Over the years, whenever I tried to build something on the side, a digital startup or just a project after work, I would try to find colleagues who are willing to devote some extra time to build things together. It has been incredibly, incredibly hard. In any company where I worked, at least here in the EU, finding people who have that spirit was almost impossible. Even the programmers, the people who could have built things after work, would much rather go home and do something else than sit down and do more of what they already do all day. Even among builders by profession, the ones willing to put in extra work beyond their job were almost impossible to find.

This argument that everybody is going to be builders just doesn’t hold.

And then you have to maintain it

Even if people were more incentivized to build their own apps, the amount of time they would be spending to maintain them, to upgrade them, to keep dependencies up to date, to do all the maintenance, would be incredibly inefficient. It is way, way cheaper to pay for existing services that are often so cheap that it would make absolutely no sense to build your own.

If we’re talking about building individual tools for yourself or for a couple of people, it is way cheaper to buy a $5/month app than to build your own, host it, maintain it, ensure that it’s secure, that it doesn’t get hacked. And the assumption that you just build something and you leave it there forever and it works is just silly, because we know that the most expensive part of software is maintenance, upgrades, and extensions.

I do believe AI is an amazing tool. For experienced engineers it already makes you faster and lets you take on more. But I don’t think that just because the tool exists, everyone will be a programmer now. It has never been about the capability or accessibility, as we often like to call it. It’s always been about the fact that for most people, there is nothing fun about staring at the screen and building to-do apps or CRM platforms, or even gaining deep knowledge in these domains so they can actually do that.