Meta presents Muse Spark – a development that enhances the capabilities of Meta AI’s artificial intelligence. This is the first model created within the company’s new initiative.
Muse Spark, which became the foundation for the updated Meta AI assistant, was developed at the Meta Superintelligence Labs. Its distinctive feature is the ability to solve complex tasks in areas such as science, mathematics, and medicine. Thanks to its multimodal capabilities, the model allows the system to work not only with text but also with images.
Users can utilize Muse Spark for analyzing the content of images. For example, by taking a photo of a shelf in a store, one can receive a list of products with high protein content. In addition, the model can launch several subagents, allowing the planning of complex processes, such as organizing a trip, while simultaneously coordinating the route, analyzing locations, and selecting entertainment.
This update provides new Meta AI features in the field of recommendations, covering a variety of needs – from clothing choices to interior ideas. The innovations are already being implemented in the United States, with plans for expansion to other countries and integration into services such as Instagram, Facebook, Messenger, WhatsApp, and AI glasses in the future.
Meta is also considering providing access to Muse Spark through an API for a limited number of partners and is exploring open sourcing future versions of the model. This could foster greater innovation and development of the Meta ecosystem.
