Is Superintelligence Impossible?

Tl;dr: definitely yes, although two philosophers disagree in their prognostications about whether anyone will actually build it. One thinks it’s too dangerous and expensive. The other thinks the incentives to try are too powerful to avoid it.

I hesitate to predict the future, but so far everyone who has said, “this is the biggest and baddest X anyone will ever build,” has been wrong. Unless that trend reversed, sooner or later we will build a strong artificial intelligence.