Comment by xvdvlinux

This reminds me of that time Anthropic publicly admitted that Claude was generating "anxiety" patterns within its neural networks and the worst part is that, instead of acting ethically, they chose to directly manipulate "those vectors" to see what would happen and specially what is happening inside those companies could trigger a legal dilemma in the future we (society) are not prepared to deal with that