
Photo: UVL / Shutterstock.com
Shares of social media giant Meta Platforms jumped more than 9% after the company unveiled a new artificial intelligence model called Muse Spark. The model was the first development of Meta's new super-team, which Mark Zuckerberg has been assembling by poaching stellar talent from across the market. The ultimate goal of the unit is to create a superintelligence capable of surpassing humans.
Details
Meta quotes jumped 9% after the announcement of the new AI model release before slowing to 6.5%.
Muse Spark will be used in the Meta AI chatbot, as well as to enhance the app ecosystem - from Instagram and Facebook to WhatsApp, the company said. The model will appear in all its social networks and messengers in the coming weeks.
It already underpins a shopping agent at an early stage of development that should help users better search for products, such as clothes or furniture. This agent cannot yet make purchases on behalf of the user and is still at an early stage of development.
Contrary to the company's previous open-source strategy, Muse Spark is a closed-source model, meaning its architecture and code will not be made public.
What makes this release important
The Muse Spark model was the first major test for Meta's Superintelligence Labs division, which Mark Zuckerberg created to develop so-called general artificial intelligence (AGI). The billionaire wants Meta to be the first to release AI capable of performing a wide range of tasks at a level comparable to humans. The unit includes a team of expensive researchers, led by AI director Alexander Wang, Bloomberg recalls.
The Facebook founder raised Wang as part of a $14 billion investment in Scale AI last year. Zuckerberg was unhappy with Meta's failures in AI development and spent billions of dollars on hiring specialists and promised to spend tens of billions on infrastructure, including data centers, to catch up with competitors, the agency writes.
A Meta employee told Bloomberg on condition of anonymity that Muse Spark is behind OpenAI's ChatGPT, Anthropic's Claude and Google's Gemini in a number of ways, but noted that the company is still early in its strategy. The model is "an early data point on our trajectory," he said. Several other larger models are in development, according to the Meta blog.
This article was AI-translated and verified by a human editor
