Meta’s highest-paid worker Alexandr Wang has reportedly signaled how the corporate plans to compete with rivals akin to OpenAI, Anthropic and Google within the AI race. In a current interview, Alexandr Wang stated that the longer term Meta AI fashions will stand out due to their health-related capabilities. Talking on the Bloomberg Tech convention in San Francisco, Wang stated well being will play a significant function as Meta develops bigger and extra superior AI fashions. “Well being is an space that we view as actually important as we scale these fashions out to billions,” he stated, giving a glimpse into Meta’s AI technique because the social media large goals to distinguish its fashions from choices by ChatGPT, Claude and Google’s Gemini. For these unaware, Wang joined Meta in 2025 following the corporate’s $14 billion funding in Scale AI. He now leads Meta Superintelligence Labs (MSL), the unit overseeing the corporate’s next-generation AI efforts.
Meta bets on health-focused AI
In accordance with Alexandr Wang, well being capabilities have emerged as one of many strongest areas for Meta’s newest AI mannequin, Muse Spark. The mannequin was launched in April and marked the primary main AI launch from Meta Superintelligence Labs since CEO Mark Zuckerberg started overhauling the corporate’s AI division.Wang acknowledged that Muse Spark is “not on the tier of the main frontier fashions” akin to Anthropic’s Claude and OpenAI’s ChatGPT. Nevertheless, he added that the mannequin carried out “higher, frankly, than we anticipated internally.”He stated Meta expects future variations to develop into more and more aggressive and that well being will stay a key focus space. The corporate can be exploring methods to combine these capabilities into client merchandise akin to Fb, Instagram and WhatsApp.Well being has develop into one of the vital carefully watched alternatives within the AI trade. Many customers Wang additionally revealed that Muse Spark raised biological-risk considerations throughout growth, although he didn’t present particulars. He stated Meta addressed these dangers earlier than releasing the mannequin. The findings additionally influenced the corporate’s choice to not launch Muse Spark as an open-source mannequin, which might have made its core expertise broadly obtainable to builders.