fg137 11 minutes ago

> additional rogue activity

> misaligned model activity

> when AI systems behave in undesired ways

We cannot say the h word now?

pluc 43 minutes ago

Another day another OpenAI "misbehavior"