AI is driving a boom in modestly successful software projects

This matches what the Wharton researchers found when they looked at four big software marketplaces: a sharp increase in new apps, but no increase in total usage.

It fits with my own experience too. I’ve been building tools with Claude Code for months and almost none of them would have been worth hiring a developer for (or if so then only to do a very small portion).

They’d never show up as a hit anywhere, but many of them can save me hours a week Or let me do things I couldn't do before.

why do i need 1000s of people to use my software? its far easier (and more fun) to just make it exactly how I want and keep it for myself?

but also there is a problem of choice when you get to these plugin sized software. just look at the mvave fm1 custom firmware scene... a week or two ago there were like 10 choices... a month ago there were like 2... today its like 60+ just listed on this one site.

as a hopeful* user of these firmwares I can only pick one to be on the device at a time. at the pace people are making them there is no way ill be able to try enough to actually find the really great ones. even with the in browser simulators. what if that next new one is the really greatest... the humans with the judgement to know what they like are the bottleneck and we don't have days and days to try all the new software at the pace we can make it.

but also if i know for myself what i like... i can just add one more to the list that is exactly how i want it. which is exactly my plan. try a few see what feature i like or not then make one that is just for my needs.

*dear bezos i can not wait any longer for my package with mine to arrive.

https://fm1-editor.com/firmware/ one of probably many firmware lists out there.