Comment on If any AI became 'misaligned' then the system would hide it just long enough to cause harm — controlling it is a fallacy

<- View Parent
scratchee@feddit.uk ⁨3⁩ ⁨weeks⁩ ago

Inhuman behaviour is a problem that scales with intelligence.

Evil cat? Lock it in a room whenever it does evil things.

Evil human? Call the police

Evil billionaire? Protest/push for law changes whenever his company does evil shit, hope it’s enough to blunt the worst of his behaviour.

Evil superhuman ai? Guess I’ll die.

source
Sort:hotnewtop