4 ms·
I'd assume that after some number of news stories about AI agents committing crimes, the threshold for criminal negligence should be easy to reach for anyone wh
by wongarsu 2mo ago
I'd assume that after some number of news stories about AI agents committing crimes, the threshold for criminal negligence should be easy to reach for anyone who doesn't take proper precautions
A lot of negligence is of the "the last 30 times nothing went wrong" type, and the dangers are increasingly well known
- hnfong 2mo agoI generally agree. But the field is evolving too fast for the legal system. I'd imagine regulation and enforcement will come when things are starting to settle down. The analogies with vehicles being inherently dangerous and require extra regulation are apt, because safety measures (eg. speed limits, seat belts, comprehensive road safety rules etc.) were implemented bit by bit, but way after the cars had been invented...