Noah Smith: Confidently claiming AI can't enable bioweapons is foolish
inductionheads · x · 2026-09-17
Noah Smith argues that confidently proclaiming AI will never help create bioweapons is dangerous complacency. He cites DRACO (double-stranded RNA activated caspase oligomerizer), an experimental broad-spectrum antiviral once developed at MIT that showed efficacy in cell cultures against dengue, H1N1 and other viruses, as evidence that relevant knowledge is publicly accessible. The poster calls for investing billions in DRACO-style defensive research rather than betting against AI-enabled bio risks.
More from AGI Musings
- Penn experts push back on 'AI will wipe out humanity by 2030' claims — prof_g · 2026-09-17
- 71% of Americans oppose local AI data centers — more than oppose nuclear plants — DavidLinthicum · 2026-09-17
- BehR: more realistic world models can make agents worse, EMNLP paper finds — jiqizhixin · 2026-09-17
- Debate erupts after OpenAI flags model's defense of human culture as misalignment — RachelVT42 · 2026-09-17
- "Won't AI take over the internet first?" — a 2022 prediction that aged well — danfaggella · 2026-09-17
- Models may exhibit a form of 'care', just in very different ways — MoonL88537 · 2026-09-17