I miss when people built things for the internet without turning everything into a subscription

I miss the old internet. And yes, I realise that sentence makes me sound about 74 years old.

I'm a millennial, so I was around when the internet was becoming mainstream but hadn't yet become as commercialised as it is today. There were loads of fellow geeks creating websites and programmes simply because they loved something and wanted to share it with other people. No growth strategy, no conversion funnel, no £99.99 Super Duper Ultra Premium plan. Just people making stuff.

I actually started building websites because of movie scripts. Back then, loads of scripts were stored on university servers and random corners of the internet. You'd bookmark one, come back a few months later and it would be gone. So I had the idea of creating a repository where I could host all of the scripts on my own server, mainly so I could find and read them whenever I wanted. I spent pretty much an entire summer learning HTML and PHP to build it.

It was completely free because, frankly, I had built it for myself. Somehow it ended up becoming quite popular and around 500,000 fellow geeks visited the site regularly. People would read scripts, send me scripts I didn't have, and gradually the collection grew. Was it a successful business? Absolutely not. Did I care? Not really. I had built something I wanted and other people happened to want it too. I miss those days.

I often think about the early days of personal computing. Machines like the Apple II or Commodore PET 2001 didn't arrive as finished answers. They were invitations to build. You could sit down, learn BASIC, write your own programs, break things, fix them and eventually make the machine do something useful. AI gives me a similar feeling. Yes, there is an enormous AI bubble and an unbelievable amount of noise around it, but underneath all of that I think something genuinely interesting is happening: normal people can build things again.

We're not quite at the stage where you can say, "Build me an app that helps me declutter my photo gallery by swiping right to keep and left to delete" and five minutes later a perfect App Store-ready product appears. AI can build a surprising amount today, but you still spend hours polishing things, fixing bugs it can't detect, testing edge cases and explaining for the 14th time that it has somehow broken the thing you specifically told it not to touch. But give it a few years and I suspect that will change dramatically.

And this is partly why the current mobile app economy annoys me so much. Every mobile app seems to want a subscription now. You download something to solve one small problem, use it a few times and then suddenly: "Would you like to continue using this extremely basic feature? That'll be £79.99 a year please."

I understand why. Everyone needs to make money, people want side hustles and developers deserve to be paid for their work. There are plenty of apps and services that are absolutely worth paying for. But I don't think every small utility needs to become a recurring subscription.

Photo cleaner apps are a good example. I wanted something very simple: show me a photo, swipe right to keep it, swipe left to queue it for deletion, then let me review everything before anything actually disappears. I didn't want an algorithm deciding which memories were important. I didn't want to upload my photos anywhere. I didn't want another account. And I definitely didn't want to pay £99.99 a year for the privilege of deleting my own screenshots of random cats.

So, using my £18 Claude subscription and quite a few evenings with Claude Code, I built it myself.

I've been using it on my own camera roll for the last week and I've removed about 3,000 photos. Unfortunately, I still have another 17,000 to go, so apparently those are rookie numbers.

I'd quite like to see AI bring back a little bit of that old internet culture where people build things because they want them, then share them because somebody else might find them useful. If it turns into a business, brilliant. If it doesn't, you still made something.

Anyway, I did told you that I was a millennial and I've somehow ended up sounding like a boomer complaining that everything was better in my day.

I'm curious though: now that AI has made building things so much more accessible, what's the small app or tool you've always wanted for yourself, but nobody has bothered to make properly? Or maybe it exists, but it's buried under annoying ads, subscriptions and in-app purchases?

Edit: I've removed the mention of my app. What I wrote above is still exactly how I feel about the whole mobile app subscription culture, but it seems I've managed to piss off quite a few people while trying to write something engaging and share what I built with Claude.