Releases rose to 182 while user reviews hovered at 95
This is a great chart for why we should look at things like productivity and not “shipping” for signals of AI-driven economic acceleration. The former is a function of what people value enough to actually spend money on, the latter could be a proxy for it…but not necessarily.
Massive output uptick due to agentic AI. Complete flat adoption.
Positive users praise agentic AI for boosting productivity and making app shipping easier despite flat metrics, while negative users call the release surge low-value slop that lacks customers or adoption.
@jenzhuscott It’s the Ikea of personal app era. I assembled my tool
Massive output uptick due to agentic AI. Complete flat adoption.
@apexlearn_org To be fair adoption always takes time.
@jenzhuscott @steipete A lot of people are finding out right now that building something and selling something are completely different skill sets
Still have to make something good.
It reminds me of the self publishing era. We saw a flood of crap.
The gatekeepers did keep most of that off the shelves. But they also kept good things off the shelves. Books like the Martian and Wool came out of that era.
So any enabling tech will give you a lot more of something but the small fraction of quality releases remains the same.
Quality and taste are always in short supply.
@jenzhuscott @alexolegimas Maybe because people are using ChatGPT for everything.
@jenzhuscott @calvinfroedge Slop garbage
@jenzhuscott Everyone thinking if they only knew how to code they can ship a great app. Now everyone can ship an app, it turns out all the ideas are just shit 😂
Good chart, but I'd read it more carefully before calling it failure.
Releases are a leading indicator. Usage and reviews lag by months, through discovery, ranking, and word of mouth.
The usage line is also an aggregate. A flat total is fully consistent with new AI-built apps gaining users while older apps decay. It can't tell you whether the same apps are winning or new ones are quietly replacing them.
App Store traction is the wrong scoreboard for a lot of this anyway. Most AI-built software is internal tools, B2B, utilities, personal projects. The payoff is cost and capability for the builder, not a chart position. This is a productivity study, and the productivity gain holds even when the output never trends.
Fair point stands: supply has outrun demand so far. But that is a long way from saying the apps don't matter.
Or a point in support can be that founders now realize how difficult it is to sell!!
@jenzhuscott distribution / user acquisition / building a community around your product has always been the real moat. Building mobile apps wasn’t that difficult even 10+ years ago! Many devs built thousands of iOS apps when the iPhone first came out but a tiny percentage remained relevant.
@jenzhuscott Expect the same in books: more will be published, fewer will be read
@jenzhuscott Output was never the bottleneck. I can ship in a weekend now what took years.
The flat part is distribution. Everyone's shipping more, nobody's solved who sees it.
@jenzhuscott The AI bottleneck now isn’t “how much can we output.” It’s “how much can we actually care about.” This shows that in amazing fashion; in some dimensions, there’s a limit in how much HUMANS keep top of mind.
@jenzhuscott @apexlearn_org I don't think app adoption will increase. I thought the tendency was for it to go down.
@jenzhuscott Because marketing is still an issue that agentic engineering doesn't solve.
So focus on building an audience.
A study of 160,000 @ProductHunt launches found broadly the same thing
Lots more solo founders launching apps, but generally of a lower quality — the top launches were still team endeavors.
(Which is not a comment on the viability of solo founders specifically, more about the influence of AI lowering the bar to launch a product.)
https://arxiv.org/html/2605.10291v1
@jenzhuscott @growing_daniel What we’re missing is this is the age of the personal app. 2007 was the age of the smartphone, a personal computer. This is the age of the personal app.
@jenzhuscott That's because people still need to learn how distribution works. Nobody cares If you have the " best app ever" if they don't know it exists. Marketing is a skill that's probably more important than making the app itself.
@alexolegimas It could also be a case of consumer surplus we aren't counting -- like all the personal Chrome Web extensions. (zero reviews.)
@alexolegimas It's like looking at the number of books printed to measure knowledge gained. Consumer attention is still constrained so the median book only sells a few hundred copies. But discovering the blockbuster apps should be easier now.
Releases rose to 182 while user reviews hovered at 95
This is a great chart for why we should look at things like productivity and not “shipping” for signals of AI-driven economic acceleration. The former is a function of what people value enough to actually spend money on, the latter could be a proxy for it…but not necessarily.
Massive output uptick due to agentic AI. Complete flat adoption.
