March 8, 2026
Hype vs. Help
Ask HN: Why there are no actual studies that show AI is more productive?
AI’s big productivity promise — commenters ask “Where are the apps and proof”
TLDR: A viral thread asks why, despite massive buzz, there aren’t clear studies or obvious signs that AI makes people more productive. Commenters split: skeptics point to the missing flood of new apps and lack of hard data, while others argue it’s too early—like electricity, big productivity gains take time.
An Ask HN post lit the fuse: if AI is such a miracle worker, where’s the hard proof? The original poster waves the “110% adoption in five minutes” meme and points to Andrej Karpathy’s interview noting even the PC boom didn’t visibly bump GDP. Translation: we’ve had time—so where are the receipts?
The comments went full courtroom drama. One camp slammed the hype: “If AI boosts productivity, why isn’t the App Store bursting?” anovikov asks, saying mobile apps are perfect grunt work and AI should’ve flooded the charts by now. Lionga calls out “snake oil,” claiming the few studies show perceived gains but little real-world lift. austin-cheney plays the grown-up: some folks want evidence before burning cash, and that’s not irrational.
On the other side, defenders say chill: vjk800 reminds everyone that electricity took 50 years to transform industry; AI tools are barely two years old and only recently usable. flawn adds engineering reality: AI can spit out prototypes, but keeping systems reliable as requirements change is the real productivity test—and that’s still a human art. Between the “show me the apps” crowd and the “give it time” team, the thread turned into a meme-laced tug-of-war over whether AI is a rocket ship… or just a really loud leaf blower.
Key Points
- •The post asks why there are few rigorous studies showing clear AI-driven productivity gains.
- •It notes AI adoption has been rapid and argues that, with existing internet infrastructure, outcomes should be visible within a few years.
- •It cites Andrey Karpathy’s remark that computers did not produce a discernible jump in GDP, suggesting macro metrics may lag.
- •The author argues many users underutilize tools (e.g., Excel), limiting realized productivity gains.
- •The post contends that significant AI product successes require time for maturation, user feedback, reliability, and marketing; product–market fit is not automatic.