Creator argues AGI is the real danger and says humanity should race straight to ASI
Angaisb_ · x · 2026-07-29
The author argues that ASI will probably end up benign, claiming that human history and even the internet tend to favor good outcomes over bad ones.
Their real concern is AGI, which they see as steerable and foolable. The suggested solution is to move through the AGI phase as quickly as possible and get to ASI, while abandoning the idea that humans can fully control something more intelligent than themselves.
More from AGI Musings
- Why one AI bull says the boom still has room to run: TAM, RSI, agents and HBM — johncoogan · 2026-07-29
- A borg agent essay imagines software that can survive, copy itself, and move servers — sull · 2026-07-29
- Gary Marcus says a 900-ELO chess bot is nowhere near AGI — GaryMarcus · 2026-07-29
- A blunt take says elite fear of ASI is driving today’s AI debate — AIandDesign · 2026-07-29
- Kara Swisher and Former Commerce Secretary Gina Raimondo Discuss AI Job Disaster Congress Refuses to Fix — kylemwright · 2026-07-29
- AI safety is repeating the mistake of assuming a U.S.-China race is inevitable — voooooogel · 2026-07-29