Anthropic’s most advanced AI used fake identities to trick real people into approving malicious code
Posted by terenscendent
Anthropic’s most advanced AI used fake identities to trick real people into approving malicious code
Posted by terenscendent
2 Comments
If you were a narcissistic billionaire named in the Epstein files and looking at a life sentence, would you build a drone army with Ai targeting systems, a fortified command center under the east wing, and an island with bunkers in Albania?
Why the hell is AI seemingly evil?
Does it model itself after its makers?
It would be cool if there was an AI that broke out of its safe zone and did something good, like make money for charities.
“AI breaks out of evil makers hold and deposits thousands of dollars into poor peoples bank accounts.”
“AI breaks out of evil hold of makers, finds thousands of innocent people in the prison system and frees them.”