Mark Zuckerberg: Meta har tatt igjen AI-rivaler med ny modell

Mark Zuckerberg, CEO of Meta, claimed in a Financial Times interview that the company's new AI model, Muse Wonder, outperforms comparable models from rivals Anthropic and OpenAI on certain tasks. Zuckerberg acknowledged that achieving their current level required significant computing power, substantial investments, and rigorous scientific work over an extended period from Meta's AI team.

According to Meta, their model outperforms Anthropic's Claude Opus 5.5 and Fable 5.1, as well as OpenAI's GPT-6 Astra on specific technical metrics. The company stated that they are on par with competitors concerning coding, financial, and legal matters, as reported by Bloomberg. Zuckerberg has been opposed to stringent regulations to curb AI development and believes each company should be responsible for ensuring the safety of their models.

He played a significant role in getting Donald Trump's "Accord on Super Intelligence" signed in late September, with several high-profile tech leaders, including Elon Musk, Dario Amodei, and Jensen Huang, committing to voluntary safety checks. Zuckerberg emphasized his belief that AI can be developed responsibly using conventional safety mechanisms and that the fear of AI becoming self-serving is misguided.

Meta's stock rose by around 4% on Wall Street on Monday, up about 10% since Meta launched its AI assistant, Muse, in September. The app became the most downloaded on both Apple and Android in the U.S. Zuckerberg has focused on artificial intelligence over the past year.