3-Month Reflection: LLMs Boosted My Productivity Less Than 50%, and Their Intelligence Is Nothing Like Ours
sebkrier · x · 2026-09-04
Andrew Ho shares a >3-month reflection on using LLMs:
- Despite their seemingly magical nature, his total productivity hasn't risen by more than 100% — perhaps not even 50% — because LLMs tempt him into gratifying but ultimately unproductive tasks.
- Model capabilities are extremely spiky and highly correlated with the level of investment poured into them.
- He concludes that LLM intelligence is wildly dissimilar to human intelligence, and scaling existing approaches with tweaks won't trivially yield something superior to humans across all important respects — even if AGI is eventually achievable.
More from AGI Musings
- Blogger: only ~0.1% of people are still worth hiring in the AI agent era — scaling01 · 2026-09-04
- After Harvard shut its writing center, dean Deming now urges AI use in writing courses — soumitrashukla9 · 2026-09-04
- Six months of 'fully capable' models, yet software feels no better, critic says — MoonL88537 · 2026-09-04
- NYC K-8 AI moratorium is misguided: students need guided AI use, critics argue — castrotech · 2026-09-04
- David Lynch explains why AI is just like a pencil — UrAWeaselBro · 2026-09-04
- Investor cites OpenAI execs: multiple reserve models, scaling laws still holding, long compute — thedealdirector · 2026-09-04