"We agree it's dangerous—and we can build it better": Domingos skewers AI safety rhetoric

pmddomingos · x · 2026-09-12

Pedro Domingos satirized the AI industry's safety rhetoric: a whistleblower warns "This company is building very dangerous stuff!", and the company replies "We totally agree! And we can build it better than anyone else!"

The joke targets how frontier labs simultaneously acknowledge catastrophic risk and justify racing ahead under the banner of building it more safely.

Original post →

More from AGI Musings

AGI Musings channel →