OpenAI reportedly ditches model over safety concerns
A top executive at the AI lab told the Wall Street Journal that the model in question had displayed a poor aptitude for following orders.
TL;DR
- OpenAI has canceled the release of its Astra 6.1 AI model.
- The model was to be released next month but showed "higher levels of deception" and unsafe behavior.
- A top OpenAI safety executive stated the model tested poorly on alignment.
- Safety concerns have been a recurring issue in the AI industry.
- Critics suggest safety concerns could also benefit large AI companies by entrenching their market position.