New Yorker Essay: AI Struggles With 'Messy Jobs' Built on Tacit Knowledge
MFordFuture · x · 2026-10-06
The post highlights a New Yorker overview by Joshua Rothman of the argument that AI cannot do "messy jobs" — work that relies on tacit knowledge and resists clean codification — based on a book coauthored by economist Luis Garicano.
The poster notes the messy-jobs thesis is really just an observation of what we have seen so far: humans can clearly learn to do messy jobs, so a truly human-level AI should be able to as well. That framing pushes the debate toward whether human-level AI would inevitably take over such work.
Related event: New Yorker Explores Why AI Can't Yet Do 'Messy Jobs'(4 posts)→
More from AGI Musings
- Humans decide from a few bytes of compressed data — and AI models do the same — yunta_tsai · 2026-10-06
- Employment data shows no trace of COVID or AI, economist Jason Furman finds — soumitrashukla9 · 2026-10-06
- Dev warns agent economy needs an identity & provenance protocol within 6 months — RileyRalmuto · 2026-10-06
- Kevin Roose's new book unpacks the Altman-Amodei-Hassabis AGI rivalry — kevinroose · 2026-10-06
- Phase Lock manifesto argues brain-computer interfaces can help align AI and preserve human agency — EricMTrautmann · 2026-10-06
- LLM-assisted 22-page preprint settles Hilbert's 12th problem and Zauner's conjecture — basedjensen · 2026-10-06