(San Francisco) Meta (Facebook, Instagram) ensures that Meta AI, its generative artificial intelligence (AI) assistant, is now more efficient thanks to the new version of its Llama 3 language model, unveiled Thursday to compete with others tech giants.
“We believe that Meta AI is now the smartest artificial intelligence assistant you can use freely,” said Mark Zuckerberg, the boss of the world leader in social networks, in a video on Instagram.
Technology companies have been engaged for more than a year in a frantic race to develop and deploy generative AI (production of texts, images, and other content, upon simple request in everyday language).
In the lead, OpenAI, which launched this new technological wave with ChatGPT, its main investor Microsoft, and its competitor Google, are competing with conversational robots (“chatbots”) and other virtual assistants to help humans do online research , to create content or even educate their children.
Thanks to Llama 3, Meta AI is supposed to provide better answers to user queries and generate images more quickly (“in real time, it’s pretty crazy!” commented Mark Zuckerberg). He will have access to recent information via Google and Bing, Microsoft’s search engine.
The assistant presented for the first time in September will also be easier to use on all applications (Facebook, Instagram, WhatsApp, Messenger), on a dedicated website (meta.ai) and with Ray-Ban Meta connected glasses .
The Californian group has had to invest massively to catch up with its neighbors in this area, and has adopted a different strategy: its language model is “open source” (free access to the programming code) for companies and researchers.
This free access makes it possible, according to Meta, to democratize this powerful technology and also to reduce the risks associated with its use.
The group presented two first versions of LLama 3 on Thursday, one with 8 billion parameters and the other with 70 billion parameters. “Parameters” are the variables in an AI model and allow them to be compared in terms of size and capabilities.
“Our goal in the near future is to make Llama 3 multilingual and multimodal, capable of integrating more context and […] better in reasoning,” the company detailed in an online press release.