OpenAI's New AI Model 'o1' Raises Both Potential and Concerns

The 'Strawberry' model exhibits advanced reasoning skills but poses medium risks in biosecurity and persuasion.