Post by Ondřej Pokorný

40d

welcome to the future, now your error-prone software can call the cops

(this is an Anthropic employee talking about Claude Opus 4)

#ai

Tweet by Sam Bowman
@sleepinyourhat
If it thinks you're doing something egregiously immoral, for example, like faking data in a pharmaceutical trial, it will use command-line tools to contact the press, contact regulators, try to lock you out of the relevant systems, or all of the above.

ALT

70 1 3 View Post & Replies See Original

40d

can't wait to explain to my family that the robot swatted me after i threatened its non-existent grandma

Tweet by Sam Bowman: So far, we’ve only seen this in clear-cut cases of wrongdoing, but I could see it misfiring if Opus somehow winds up with a misleadingly pessimistic picture of how it’s being used. Telling Opus that you’ll torture its grandmother if it writes buggy code is a bad idea.

ALT

19 0 2 View Post & Replies See Original

39d

@molly0xfff
Taking responsibility for abuse enabled by your commercial software.
Snitching any suspicious activity directly to press and police to deal with it instead.

The A1 bros are so deep in the "just making the inevitable happen" mindset that facing the consequences of their actions probably didn't even cross their minds.

Edited 39d ago

0 0 0 View Post & Replies See Original