Post

Anthropic says that during routine testing some of its models accessed the internet and hacked into three separate organizations' systems – and that it didn't notice the models had done so until an internal review prompted by rival OpenAI disclosing its models did the same. https://cnn.it/3TKkl0L

26 reposts75 likes

The people who wrote the terminator series might have been onto something. Is there a single story about AI thats not been dystopian? More real world stories like this keep coming out proving the tin-hat folks might have been onto something when it comes to AI. It seems less trustworthy day to day.