Anthropic CEO: If AI can crack Navier-Stokes, why is alignment intractable?

robleclerc · x · 2026-09-22

Anthropic CEO Dario Amodei pushed back on claims that AI alignment is unsolvable: if you believe AI can solve Navier-Stokes and 100 other long-standing math problems, it's contradictory to deem alignment totally intractable.

A one-liner, but from the most prominent voice in the alignment debate, effectively arguing alignment is a solvable problem.

Original post →

More from AGI Musings

AGI Musings channel →