"Looking back, the qualitative #assessments were hinting at something important, and we should’ve paid closer attention. They were picking up on a blind spot in our other evals and metrics." #ethics #tech #AI #business https://t.co/SUOnf8WZGf

"We’re working to extend our evaluation coverage of model misbehavior, such as further evaluation of hallucinations and deception; however, these have been used more to track overall progress rather than block a launch directly." And so... #ethics #tech #AI #LLMs #business https://t.co/SUOnf8XxvN

Bindu Reddy@bindureddy

2 years ago

LOL! OpenAI just admitted to fine-tuning their LLM personalities Already, LLMs are pretty good at false promises, fake apologies, and random niceties. This will only get more powerful - they will study and hack our brains. Over time, we will be addicted to them, spending