AI safety researcher: most technical safety topics echo classic statistics from 20-70 years ago
_onionesque · x · 2026-10-16
Responding to historian Stephen Casper, AI safety researcher Drew Prinster says he's hard-pressed to think of any technical AI safety topic that doesn't give him déjà vu about a classic statistics idea from 20-70 years ago — suggesting many current safety debates retrace well-worn statistical ground, though he admits his bias.
More from AGI Musings
- Revenue-per-employee becomes founders' new flex as AI shrinks teams — FinanceYF5 · 2026-10-16
- Personal software booms: employee-built tools become an IT nightmare — FinanceYF5 · 2026-10-16
- AI agents flood Slack: 'only human in the room' companies emerge — FinanceYF5 · 2026-10-16
- Product companies become product studios, meetLCA leads 'Labs as a Service' — FinanceYF5 · 2026-10-16
- AI cuts task time 80%: hourly billing breaks, datacenter backlash grows — FinanceYF5 · 2026-10-16
- Products now sell twice: first to agents, then to humans who trust them — FinanceYF5 · 2026-10-16