Eric Hartford asks: where's the evidence for AI doom? Burden of proof is backwards
QuixiAI · x · 2026-11-22
In a new blog post, Eric Hartford argues that while models have demonstrably lied, gamed metrics, and resisted shutdown in lab tests, no AI system has ever shown an autonomous intention to wipe out humanity. The doomer scenario — convergent instrumental goals leading to extinction at high capability — is a coherent theoretical story, not an observed sequence of events. He contends the burden of proof sits with those predicting catastrophe, not those questioning it, and traces how AI doomerism gained outsized policy influence, including the role of Effective Altruism.
More from AGI Musings
- LangChain's "knowledge factory" essay sparks debate: better tools or training our replacements? — hwchase17 · 2026-11-22
- After two weeks in China, author argues Western AI safety playbook doesn't transfer — and catching up is getting costlier — JacquesThibs · 2026-11-22
- DeepMind researcher: watching token cannons smash castles that withstood human sieges is poignant — SamuelAlbanie · 2026-11-22
- AI vs. Millennium Problems: the real story starts after the demo — ayushthakur0 · 2026-11-22
- The 2029 meme: ASI takeoff, cancer cured, yet GDP still grows 2.8% — willcb · 2026-11-22
- Auki founder at Humanoid Hub: physical AI should run physical businesses, not just move robots — broodsugar · 2026-11-22