Meta Debuts Llama 3, Latest AI Model for Training Chatbots

Meta Debuts Llama 3, Latest AI Model for Training Chatbots
Artist during a Pop Up for musician and artist organised by Meta, at its India headquarters in Gurugram, India on Monday, June 20, 2022. Photographer: T. Narayan/Bloomberg

Facebook parent company Meta Platforms Inc. debuted a new version of its powerful Llama artificial intelligence model, its latest effort to keep pace with similar technology from companies like OpenAI and Alphabet Inc.’s Google. 

Llama 3, unveiled Thursday, is an upgrade from an AI model that Meta released last summer. Chief Product Officer Chris Cox said that model, Llama 2, has been downloaded 170 billion times. Llama 2 was billed as open source, though its license included some restrictions such as requiring companies with more than 700 million users to ask for permission to use it.

Meta, the world’s largest social media company, is using Llama 3 to run its own in-app AI assistant, called MetaAI, which exists across several products, including Facebook, WhatsApp and Meta’s Ray-Ban smart glasses. The MetaAI assistant will be updated for users on Thursday as well, the company said.

Many major tech companies are racing to develop AI products and services, spending billions on high tech chips and employee resources to develop large language models and other AI-focused products. Meta is using AI to power its content feeds and recommendations, but also to improve ad targeting and its virtual reality headsets.

Llama 3 is “industry leading on a number of benchmarks for models of this size,” Cox said. “We’re no longer playing catch-up to have a model that’s best-in-class.”

The company is also working on several iterations of Llama 3 that will be rolled out as updates, Cox added.


Comments - Please in order to comment.

Please peer review 3 community comments before your comment can be posted