Inside Meta’s attempts to play catch-up with AI A year after Mark Zuckerberg appointed Alexandr Wang to revitalize Meta’s AI efforts, the company has released Muse Spark, its most promising AI model to date. Wang’s appointment aimed to bring urgency and ambition to Meta’s AI, but while some praise the progress and potential of future models, others express skepticism about Meta’s ability to compete with rivals like OpenAI and Google.

• Alexandr Wang was hired by Mark Zuckerberg to accelerate Meta’s AI development.
• Meta has released Muse Spark, its most significant AI model developed under Wang’s leadership.
• Proponents believe Muse Spark indicates Meta’s AI rebuilding is gaining traction, while critics question the progress and pace compared to competitors.
• Wang has assembled an elite research group, reshaped parts of Meta’s AI operations, and gained significant influence within the company.
• Meta is investing heavily in AI, with Muse Spark and future models expected to enhance its products and services.
• The development of Muse Spark faced internal challenges and was built using some existing Meta AI infrastructure, leading to disputes over credit.
• Wang advocates for a vision of ‘personal superintelligence’ and has reshaped AI safety work at Meta.
• Some employees are concerned about Wang’s focus on proprietary models over Meta’s open-source approach and the company’s broader workforce changes. Continue reading https://arstechnica.com/ai/2026/06/inside-metas-attempts-to-play-catch-up-with-ai/