"Humans were never very good at research either": AI verification argument resurfaces
RexDouglass · x · 2026-11-07
A widely shared observation from @natalienkhalil: just as we're learning humans were never very good at writing secure code, we'll soon realize humans were also never very good at research—every scientific claim rests on evidence and reasoning, and humans make mistakes in both, often.
The quip extends the lesson from AI code security to scientific practice itself, feeding into debates over whether AI should take a bigger role in verifying research.
More from AGI Musings
- Roger Martin on AI's Core Problem: 'It Organizes and Recycles' — RogerLMartin · 2026-11-07
- Reorganization effect can reverse technology effect on AI-era wages, paper shows — soumitrashukla9 · 2026-11-07
- Unitree built its own automation lines — and cheap Chinese labor means it barely needs them — teortaxesTex · 2026-11-07
- Cybersecurity problems are organizational dysfunction in disguise, and AI won't fix them — nptacek · 2026-11-07
- AI makes you faster only at what you already know how to do — feregri_no · 2026-11-07
- The slicing problem: a clever reductio against computational functionalism about consciousness — JacquesThibs · 2026-11-07