So where are all the AI apps?

I can't really think of a polite way to phrase this, but I'm not surprised throwaway mobile apps do benefit, while relatively mature python packages do not. That matches my estimation of how much programming skill you can reasonable extract from the current LLMs.

Really the one thing that conclusively has changed is that the 'ask it on stackoverflow' has become 'ask it an LLM'. Around 95% of the stackoverflow questions can be answered by an LLM with access to the documentation, not sure what will happen to the other 5%. I don't think stackoverflow will survive a 20-fold reduction in size, if only because their stance on not allowing repeat questions means that exponential growth was the main thing preventing them from becoming stale.

> I'm not surprised throwaway mobile apps do benefit, while relatively mature python packages do not.

Right.

I don't think you even need cynicism or whatever you felt you were having impolite thoughts about:

I'd expect the top mature libraries to be the most resistant to AI tool use for various reasons. They already have established processes, they don't accept drive-by PR spam, the developers working on them might be the least likely to be early adopters, and -- perhaps most importantly -- the todo list of those projects might need the most human comms, like directional planning rather than the sort of yolo feature impl you can do in a one-man greenfield.

All to further bury signals you might find elsewhere in broader ecosystems.

I’m not a developer by trade. I’ve screwed around with some programming classes when I was in school, and have written some widely used but highly specific scripts related to my work, but I’ve never been a capital-D developer.

In the last few months, Gemini (and I) have written for highly personal, very niche apps that are perfect for my needs, but I would never dream of releasing. Things like cataloguing and searching my departed mom‘s recipe cards, or a text message based budget tracker for my wife and I to share.

These things would never be released or available as of source or commercial applications in the way that I wanted them, and it took me less time to have them built with AI then it would have taken me to Research existing alternatives and adapt my workflow/use case to fit whatever I found.

So yeah, there are more apps but I would venture to say you’ll never see most of them…

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

But that's not really what we were promised.

> if your goal is to solve some specific problem for yourself, your friends/family, community or your team, then the "last step" you mention - the one that "takes majority of time and effort" - is entirely unnecessary, irrelevant, and a waste of time.

To a point, but I think this overstates it by quite a bit. At the moment I'm weighing some tradeoffs around this myself. I'm currently making an app for a niche interest of mine. I have a few acquaintances who would find it useful as well but I'm not sure if I want to take that on. If I keep the project for personal use I can make a lot of simplifying decisions like just running it on my own machine and using the CLI for certain steps.

To deploy this to for non-tech users I need to figure out a whole deployment approach, make the UI more polished, and worry more about bugs and uptime. It sucks to get invested in some software that then constantly starts breaking or crashing. GenAI will help with this somewhat, but certainly won't drop the extra coding time cost down to zero.