Meta Launches New AI Model 'Muse Wonder'... Zuckerberg: "We've Caught Up with OpenAI"

Competition with Anthropic and OpenAI Heats Up
Zuckerberg Opposes Calls to Slow AI Development

Reuters-Yonhap News

Meta launched its new artificial intelligence (AI) model, Muse Wonder, on October 12 (local time), challenging established AI leaders Anthropic and OpenAI. After overhauling its AI business last year and committing substantial funding, Meta is now drawing attention as it seeks to strengthen its competitive position.

According to CNBC on October 12 (local time), Meta launched two new AI models that day: Muse Wonder and Muse Wonder Full. Meta said Muse Wonder outperformed the latest versions of OpenAI's ChatGPT and Anthropic's Claude in performance evaluations. The company said Muse Wonder can handle everyday tasks and coding, as well as complex tasks such as lease agreements, financial analysis, and health diagnoses.

Meta will soon also integrate Muse Wonder into Muse, its personal AI agent. The service will initially be available to paid subscribers. Muse Wonder is currently accessible through Meta's model API and Muse Code, its AI model specialized in coding.

Market reaction to the new product has been generally positive. Meta shares rose 3.38% from the previous session to close at $742.96 on the U.S. stock market that day. Meta plans to integrate Muse Wonder with its other services in the future to create a personal AI assistant capable of handling a broad range of tasks.

Muse Wonder is the result of an overhaul of Meta's AI business led by CEO Mark Zuckerberg. Since last year, Zuckerberg has pledged major investments and actively recruited researchers from competitors to help Meta make up ground in AI technology, where it had fallen relatively behind.

In an interview with the Financial Times (FT) that day, Zuckerberg said, "I think we've clearly caught up with the latest models from OpenAI and Anthropic." He added, "Getting to this stage required enormous computing resources and investment, as well as very rigorous scientific research conducted by the research team over a long period."

He also argued that calls recently raised in the AI development industry to slow the pace of AI development were unnecessary. "I have much greater confidence that AI is a technology that can be designed with standard safeguards and developed responsibly," he said. "I don't think AI is a human-like being that acts autonomously."

This content was produced with the assistance of AI translation services.