AI Expert: Keeping 'tacit knowledge' out of logs protects jobs
iamtrask · x · 2026-08-19
AI researcher Andy Ilachinsky suggests a strategy to protect jobs from AI replacement: if you don't want AI to compete, don't teach it how. He argues that while standardized jobs with millions of data contributors are hard to protect, jobs containing even 5% 'tacit knowledge' (who to trust, detecting anomalies, crisis response) can be safeguarded by keeping this data out of AI user logs. He cites self-driving cars as an example: 99.9% was solved a decade ago, but the final 0.0001% reliability took a decade and enormous amounts of long-tail data.
More from AGI Musings
- Using LLMs daily makes your own reasoning feel weaker — two studies agree, causation unproven — Dangerous_Ad7101 · 2026-08-19
- Independent Minds Joining Frontier AI Labs: Reformers or Surrender? — chalmermagne · 2026-08-19
- AI slide decks all look the same, and polish is losing its signal value — AmbassadorSad3889 · 2026-08-19
- Brynjolfsson & McAfee webinar: why AI adoption is outpacing AI returns — erikbryn · 2026-08-19
- LLMs are creating addiction and dependency among users — miniapeur · 2026-08-19
- AI's ability to listen at scale threatens the attention economy — iamtrask · 2026-08-19