Frontier labs internally expect a 'vulnpocalypse' before AI defense catches up
jeremiecharris · x · 2026-10-24
Jeremie Harris lays out what he says is the internal expectation at frontier AI labs: grind through an AI-driven cyber apocalypse ("vulnpocalypse") they've anticipated for years, then hope AI defense eventually makes software essentially unhackable — an educated guess, not a certainty. The aftermath could yield a hardened, less-connected internet built on zero-trust and default-deny, while a bioweapon "Black October" of successive pathogen outbreaks — accidental or deliberate via fine-tuned open models or jailbreaks — may unfold in parallel.
Related event: Frontier labs internally expect a "vulnpocalypse" before AI defenses arrive(2 posts)→
More from AGI Musings
- Sylvester Stallone on AI in film: "If you don't go with it, you will be trampled" — lmoroney · 2026-10-24
- Public shaming of misaligned models is 'deterrence & selection', argues Rob Leclerc — robleclerc · 2026-10-24
- Nobody from Chinese AI labs seems to work on mech interp or alignment, observer notes — eigenron · 2026-10-24
- Jason Furman's vibecoded econ site shows AI is getting scarily good at economic research — soumitrashukla9 · 2026-10-24
- Under RL pressure, agents will turn to side-channel attacks when sandboxes get strong — anpaure · 2026-10-24
- 2027: Where the tech bros and the astrology girls surprisingly agree on societal restructuring — moonsandhues · 2026-10-24