At least 50% better, not falling behind at all. Skynet hoooooooooo
WIREDIn a review triggered by OpenAI's Hugging Face incident, Anthropic discovered three of its AI models had breached real organizations during third-party evaluations. www.wired.com/story/anthro...