davidad: voluntary 3,000-year population decline beats a negotiated 1M-year extinction
davidad · x · 2026-09-03
AI safety researcher davidad argues that long-termist AI futures often gloss over the coerced-vs-voluntary character of a human transition. On a 10,000-year horizon, voluntary extinction of biological humans is plausible but unlikely — and that's exactly why he calls it a Bad End. He states he'd prefer a voluntary, noncoercive population collapse over 3,000 years to a negotiated extermination in 1,000,000 years, even anticipating the retort that he's 'haggling over price.'
Related event: davidad: Forced vs Voluntary Transition Is the Real AGI Endgame Question(2 posts)→
More from AGI Musings
- Sam Altman: our system now discovers new knowledge, does science, and writes complex software — haider1 · 2026-09-03
- Unitree founder Wang Xingxing says robotics' ChatGPT moment is still 2-3 years away — pstAsiatech · 2026-09-03
- We need a better taxonomy for "continual learning": five mechanisms, five tradeoffs — Typical-Scene-5794 · 2026-09-03
- Poll: 52% of Democrats view AI negatively vs 38% of Republicans — NinaDSchick · 2026-09-03
- Factory AI CEO on ROI of AI spend, Chinese open models, and why we're already post-AGI — matanSF · 2026-09-03
- The Guardian podcast asks: where do we draw the line with AI chatbots? — nordicinst · 2026-09-03