What are some introductory videos about AI safety?

Why would really smart AI be dangerous?

Why would an AI want to do bad things?

(Learn more about instrumental convergence.)

Why would a smart AI do such random/dumb-seeming things?

(Learn more about the orthogonality thesis.)

Why couldn't we just fix or "unplug" a misbehaving AI?

(Learn more about corrigibility.)

See a longer YouTube playlist here for more!