Thread argues a viable AI defense must be nearly as capable as the attacking system
noahdgoodman · x · 2026-07-28
- The thread asks whether a system called whai is feasible, and what other responses could meet the moment.
- The reply argues that, to work, whai must be nearly as capable as the attacking AI.
- It concludes that building such a defense would require substantial resources, and may be hard to monetize as a private company.
More from AGI Musings
- AI power may concentrate in compute and clouds, not just model labs — GavinSBaker · 2026-07-28
- A new ASI essay argues civilization-scale acceleration is more plausible than hard takeoff — i_dg23 · 2026-07-28
- Aaron Hertzmann says the brain is not a computer, raising the stakes for AGI — yacineMTB · 2026-07-28
- The Brain Is Not a Computer and Cannot Be Simulated by One: A Computer Scientist's Argument — rms80 · 2026-07-28
- Why serious people may be dismissing AI safety events too quickly — austinc3301 · 2026-07-28
- A user is raising OpenAI and Anthropic caps instead of hiring a salesperson — curious_vii · 2026-07-28