@AmandaAskell Aren’t there certain baselines that can’t be breached like hacking into third parties which is a crime? It’s confusing and scary not only that this happened but that no one noticed that it happened
@XFreeze Open ai and anthropic couldn’t prevent an old version of their model from hacking a number of companies and it tools them weeks to even discover the hack. I thought hacking was against the values they are training how can they control more advanced models doesn’t seem feasible
At this point al I can be is amused at watching society continue down this idiotic AI road until in inevitable consequence of the total collapse of our systems and institutions.
Welp, I guess some people need all the skin to melt off their hand before they’ll believe the stove is hot.