Nick Bostrom on a16z: Superintelligence Could End Aging, but Delay Has Costs Too

a16z · youtube · 2026-11-19

In an a16z podcast episode, philosopher Nick Bostrom argues the hard challenge isn't racing toward or stopping superintelligence, but preserving the ability to slow down when safety lags without freezing technological progress. He favors "pacing the frontier" over indefinite pauses, notes alignment may get harder as models become more situationally aware, and contends safely developed superintelligence could radically extend lifespans and tackle disease and poverty—so delay carries its own risks. The conversation spans Deep Utopia, governance, post-human life, the simulation argument, and whether AI could surpass humans at philosophy.

Related event: Bostrom: Delaying Superintelligence Has Its Own Costs(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →