So this week we've learned that even the government asseses Anthropic has the better model, and that OpenAI leadership has no concern for safety whatsoever.