1. Meta is launching the next-generation AI model, Llama 3, to power its virtual assistant in the Ray-Ban smart glasses, offering more complex tasks and natural responses.
2. Llama 3 will be an open-source model available in smaller sizes for local hardware, expected to be as powerful as OpenAI’s GPT-4.
3. Meta’s AI assistant, MetaAI, will be enhanced with Llama 3’s improved understanding of the real world, expanding its capabilities in products like the Quest 3 VR headset and social media platforms like Instagram and WhatsApp.
Meta is set to launch its next-generation AI model, Llama 3, which will power the virtual assistant in the Meta Ray-Ban smart glasses. Llama 3 is an open-source model that is expected to be as powerful as OpenAI’s GPT-4. During an event at Meta HQ in London, executives hinted at what we can expect from Meta in terms of AI, including Llama 3 and the MetaAI assistant.
Last year, the Llama-2-powered MetaAI assistant was upgraded to “see” through the camera of the glasses, offering sartorial and technical advice. With the launch of Llama 3, the assistant is expected to become even more impressive and capable.
Llama 3 is a large language model that will come in various sizes, from small to compete with other models like Claude Haiku or Gemini Nano, to larger models like GPT-4. It is expected to be more fully multimodal, capable of understanding a range of inputs including spatial understanding of the real world, speech, and video.
Meta’s Chief AI Scientist, Yann Le Cun, believes that in the near future, all interactions with the digital world will be through AI assistants. Meta’s goals for its AI assistant span beyond just the smart glasses, with ambitions to integrate MetaAI into other apps and products like the Quest 3 VR headset and various social media platforms like Instagram and WhatsApp.
Overall, Meta aims to make Llama-powered Meta AI the most useful assistant in the world, with the upcoming release of Llama 3 expected to enhance its reasoning and understanding capabilities.