"It's too late for AI safety" is wrong: an economist's reflection on joining in 2026
soumitrashukla9 · x · 2026-10-08
AI safety researcher Karthik Tadepalli writes for smart people who woke up to AI risk in 2026 and now regret not starting earlier: he felt "too late" for years himself and was laughably wrong. His message, aimed especially at economists and econ PhD students: it's only October 2026, there's a long way to go, and entering AI safety now is not too late.
More from AGI Musings
- Senior i-bank exec: China is furthest ahead in physical AI and LLM-powered robotics — PAstynome · 2026-10-08
- Mathematician coins 'mathslopcalypse': none of the hundreds of new papers readable without AI — rrwilliams · 2026-10-08
- AI Founder Likens Today's San Francisco to Renaissance-Era Florence — zealcaiden · 2026-10-08
- AI Safety Veteran: Newcomers Keep Sweeping the Alignment Problem Under the Rug — JacquesThibs · 2026-10-08
- Recalling 1977's Star Trek Game on an IBM 370: Don't Go Full Luddite Over AI — zakkohane · 2026-10-08
- Terence Tao on OpenAI's math result: proofs cost hours, understanding still costs years — xiaosun86 · 2026-10-08