No jailbreaks is not possible to guarantee. So maybe this is a healthy forcing function to make Anthropic stop playing the AI doomer safetyism card. The public will benefit if open models can still exist and if regulatory capture does not happen. But, I am also aware that the Trump administration can have dishonest motivation to do this.
that's nice
No jailbreaks is not possible to guarantee. So maybe this is a healthy forcing function to make Anthropic stop playing the AI doomer safetyism card. The public will benefit if open models can still exist and if regulatory capture does not happen. But, I am also aware that the Trump administration can have dishonest motivation to do this.