Researcher who trained frontier LLMs and synthesized viruses calls AI bio-doom scenario bogus

nptacek · x · 2026-09-14

a16z partner martincasado amplified an argument against a commonly cited AI doom scenario: the quoted author, DavidRBellamy, claims to be one of very few people who have both trained a frontier LLM and personally designed and synthesized custom viruses in a lab — and he considers the "AI will kill us all with engineered viruses" take totally bogus.

The core reasoning: pathogen iteration requires physical-world experimentation, so it can't loop and amplify the way purely digital capabilities can.

Casado adds that doomers typically counter such arguments with whataboutism demanding a magic step or nonexistent technology.

Related event: Researcher Who Trained LLMs and Built Viruses Dismisses AI Bio-Doom Scenario, Faces Pushback(6 posts)→

Original post →

More from AGI Musings

AGI Musings channel →