As we haven't seen new operating systems or web browsers and the like, I'm guessing the reason is the same the corporation execs still have to find out: producing the code is just a small part of it. The big part is iterating bug fixes, compatibility, maintenance etc.

We’re in a personal software era. Or disposable software era however you want to look at it. I think most people are building for themselves and no longer needing to lean on community to get a lot of things done now.

Self plug, but basically that’s the TL;DR https://robertdelu.ca/2026/02/02/personal-software-era/

Is this the best way to measure this? I think the biggest adopters of AI coding has been companies who are building features on existing apps, not building new apps entirely. Wouldn't it make more sense looking at how quickly teams are able to build and ship within companies?

It seems like all tech executives are saying they are seeing big increases in productivity among engineering teams. Of course everyone says they're just [hyping, excusing layoffs, overhired in 2020, etc], but this would be the most relevant metric to look at I think.

I feel they're largely here, on this platform. Hacker News, currently, could be renamed to AI News, without any loss of generality.

I absolutely hate web development with a passion and haven’t done a new from the ground up web app in 25 years and even since then it was mostly a quick copy and paste to add a feature.

But since late last year even when it’s not part of the requirements leading app dev + cloud consulting projects, I’ll throw in a feature complete internal web admin site to manage everything for a project with a UI that looks like something I would have done 25 years ago with a decent UX.

They are completely vibe coded, authenticated with Amazon Cognito and the only things I verify are that unauthenticated users can’t access endpoints, the permissions of the lambda hosting environment (IAM role) and the database user it’s using permissions.

Only at most 5 people will ever use the website at a time - but yeah I get scalability for free (not that it matters) because it’s hosted on Lambda. (yes with IAC)

The website would not exist at all if it weren’t for AI.

Now just to be clear, if a website is meant for real people and the customer’s customers. I’ll insist on a real web designer and a real web developer be assigned to the project with me.

the pypi metric feels off. most of the ai stuff i see shipping is either internal tooling that never hits pypi, or its built on top of existing packages (langchain, openai sdk, etc) rather than creating new ones.

the real growth is in apps that use ai as a feature, not ai-first packages. like every saas just quietly added an llm call somewhere in their stack. thats hard to measure from dependency graphs.

> So where are all the AI apps?

They're in the app stores. Apple's review times are skyrocketing at the moment due to the influx of new apps.

There has been a 2x and sometimes even 10x in PR size, measured in LoC...

But that's not really what we were promised.

I am worried for people using write ups like this as a huge, much appreciated dose of copium.

Try it out and don't stop trying. If something improves at this rate, even if you think it's not there right now, don't assume it is going to stop. Be honest about the things we were always obviously bad at, that the ai has been getting quickly better at, and assume that it will continue getting better. If this were true, what would that mean for you?

If one were to release an AI app - what would be an appropriate license? Genuine question.