Are AI models getting safer over time, or is it just bullshit?

3 points| homanp | 3 months ago |superagent.sh