LlamaIndex founder: AI agents boost my productivity, but far from superhuman
lateinteraction · x · 2026-09-13
Jerry Liu (LlamaIndex founder) offers a candid first-person take on agent productivity
- Models are "extremely good by past AI standards" — good enough to be daily drivers for much of his work — but "superhuman at any coherently broad skill" he cares about, they are not.
- His experience is jagged: overall productivity is up, but a considerable fraction of the gain is gamification. Agents make work feel like play, letting him persist on tasks he'd have found too unstimulating and spread judgment across more exploratory, less essential frontiers.
- The most important threads of work may not benefit — or may even get hurt. He's happier working in agent software, but says the reality is "orders of magnitude off" from the hype on X.
Related event: Practitioners say frontier models are useful but far from superhuman(4 posts)→
More from AGI Musings
- Reddit users speculate Musk, Amodei and Altman know of an undisclosed AI incident behind slowdown calls — Traditional-Chip8339 · 2026-09-13
- tszzl predicts open-source AI will be banned after a major disaster, wants monitored APIs — mimi10v3 · 2026-09-13
- Commentators argue frontier labs fear open-source AI killing revenue, not extinction — whurley · 2026-09-13
- UK GDP beats forecasts at 0.4% in July, with programming and info services adding 0.12pp — krishnan · 2026-09-13
- 'In Weeks to Months': Building Apps Will Be Like Driving a Car, Professor Argues — deepakns · 2026-09-13
- Billion-dollar company walked back all-in AI coding after 6 months, admits no real productivity gains — jmclondon97 · 2026-09-13