Re:Someone Asked an Autonomous AI to 'Destroy Humanity'
by hcta on Thursday April 13, 2023 @05:49PM
(#35562563)
I wouldn't guess the risk of this causing major damage is greater than >1%, but it still seems like an unnecessary risk. It is a finger searching for a button and there are buttons out there; e.g. if the agent stumbles on code that, intentionally or due to coding error, can cause denials of service or other havoc triggered by reading a twitter feed or scanning the public web. I guess the person responsible for this joke was too dumb to think of this?
[ Parent
| on HN ]
There are no comments on this comment yet.