Security vet: AI finds real vulns, but demo videos hide how autonomous it really is
dyn___ · x · 2026-11-19
julianor clarifies he doesn't dismiss AI-assisted vulnerability research: experienced teams are using AI to find real vulnerabilities and build working PoC exploits, and a demo video can reflect weeks of skilled guidance. His skepticism targets one specific claim — that an inexperienced person can autonomously develop exploits — arguing most viewers never see the environment, assumptions, and limits behind the demos, so people end up arguing about different things.
More from AGI Musings
- On CNN to discuss Anthropic's IPO, extinction risk was literally the first question — GarrisonLovely · 2026-11-19
- Ben Recht revisits his 2019 AI reading list: the critiques predate GPT-2 and still hold — beenwrekt · 2026-11-19
- Old clips show SV wrote off Anthropic 3 years ago — a lesson in the illusion of understanding — c_valenzuelab · 2026-11-19
- Nick Bostrom on a16z Podcast: The Real Challenge Is Keeping the Ability to Slow Down — a16z Podcast · 2026-11-19
- Why AI won't kill SaaS: I pay for more niche tools than ever — doooyle · 2026-11-19
- AI researcher joins Star Trek 60th anniversary panel on hopeful AI futures — mdredze · 2026-11-19