AI Safety Researcher Audits X-Risk Bibliography With ChatGPT, Paper by Paper
dhadfieldmenell · x · 2026-09-23
AI safety researcher David Hadfield-Menell responded to the debate over whether peer-reviewed literature supports AI x-risk claims by having ChatGPT analyze a bibliography paper by paper. His nuanced conclusion:
- The bibliography does not support the claim of "a large peer-reviewed literature showing AI x-risk is high"
- It does support "a substantial literature establishing technical premises of serious x-risk arguments, plus a smaller literature assembling those premises into a full argument"
He also flags a confounder: given the field's history, explicit existential-risk discussion was likely grounds for rejection, so papers show argument steps while the connective work happened elsewhere.
More from AGI Musings
- David Krueger: 'Pause AI' means stopping more powerful general-purpose AI, which doesn't exist yet — DavidSKrueger · 2026-09-24
- The case that AI safety means ASI acting in humanity's interest, not human control — basedjensen · 2026-09-24
- Jensen Huang fires back: 'Nobody's building more compute than the people asking to slow down' — Hesamation · 2026-09-24
- Should We Start Labeling AI Deniers as Conspiracy Theorists? — Safe-Bar-6300 · 2026-09-24
- Gebru and Bender: Don't be fooled by this summer of AI hype, says MIT Tech Review op-ed — MilagrosMiceli · 2026-09-24
- Lufthansa's blunt customer reply sparks joke: pre-AI, businesses said 'the customer is always right' — nikitabier · 2026-09-24