How to read imperfect AI-job-impact studies: update Bayesian-style as signals pile up
danielrock · x · 2026-09-21
Researchers discuss how to interpret studies on AI hurting early-career workers. Alex Olegimas argues no paper has gold-standard methodology, so read them Bayesian-style: if six papers from different angles show the same thing, you should update. Daniel Rock agrees, noting academic causality norms shape the hedged wording — it's enough to say exposed early-career workers are having a tougher time, causality unclear, and wait for better evidence.
More from AGI Musings
- OpenAI Researcher: Limited-Agency AI Is a Defensible Path, Not a Nonsense Distinction — jachiam0 · 2026-09-21
- Zvi: Should Your AI Lawyer Refuse You When You're Being Evil? — TheZvi · 2026-09-21
- Why China lets charisma-free nerds reach the top: the Liang Wenfeng effect — teortaxesTex · 2026-09-21
- tszzl Says the Research-vs-Consumer Agent Distinction Is Nonsensical: Science Needs Broadly Deployed Agents — tszzl · 2026-09-21
- Thomas Dietterich: human supervision must tighten for non-routine agent steps in volatile settings — tdietterich · 2026-09-21
- Dietterich vs Ambrosio: why isn't anyone working on continual learning? — tdietterich · 2026-09-21