March 24, 2026
Shhh… the apps are hiding
So where are all the AI apps?
On our laptops, not the charts — and please stop calling it “AI”
TLDR: Data shows no flood of new public apps, but AI-related projects that do exist are updating far faster than others. Commenters say the real boom is in unbranded, personal tools—people want useful features, not “AI” labels—fueling a quiet surge outside the charts that could reshape how software gets built.
The charts say there’s no explosion of public AI apps, but the comment section is screaming, “You’re looking in the wrong place.” The analysis shows no big boom in new public packages and only modest release upticks—except for AI-focused projects, which are shipping updates like crazy. Cue the drama: users say the real action isn’t on PyPI or app stores, it’s on personal machines and scrappy demos.
One commenter flexed that with a chatbot co-pilot, they’ve cranked out dozens of personal apps in a year. Another roasted executives for slapping “AI” stickers on everything: “People don’t want ‘AI’; they want magic that works.” The crowd’s verdict? Don’t brand it as AI, just make it useful. One zinger summed it up: nobody cares if it’s AI or “goblins” under the hood—just ship the feature.
Meanwhile, the nerdy plot twist: the only place the charts do pop is for AI-related packages, especially popular ones, which are updated way more often. Translation for non-tech folks: public app counts aren’t soaring, but the AI projects that exist are iterating at breakneck speed. The running joke? “Where are the apps?” Answer: “On Show HN,” says one wry commenter, and apparently, on everyone’s laptop, quietly running the world.
Key Points
- •Counting new packages shows no clear post-ChatGPT surge in genuine package creation; observed spikes are linked to spam/malware.
- •Analysis of the 15,000 most downloaded PyPI packages (December 2025) shows higher first-year release frequency in recent cohorts, continuing a trend that began around 2019.
- •The increased first-year release frequency may be influenced by CI tools like GitHub Actions, not solely AI.
- •Release frequency declines as packages age across all cohorts, a pattern unchanged in the post-ChatGPT period.
- •AI-related packages show markedly higher initial release frequencies (e.g., median ~20 in 2023), and popular AI packages reach ~21–26 median releases, suggesting the effect isn’t just popularity.