In order to compete with Google’s AI project Bard and Microsoft-backed OpenAI’s ChatGPT in the race to build generative AI, Meta (formerly Facebook) is said to be preparing to issue a commercial version of their AI model.
The Financial Times cited sources when reporting that Meta will make use of new software that was driven by massive language models (LLMs), which can produce text, graphics, and code.
The internet giant earlier this year provided its own language model, known as LLaMA, to academics and researchers, but the new version is reportedly more accessible and customizable by businesses.
Since Meta’s LLMs are “open-source,” the details of the new model will be made available to the public.
According to the article, the new AI model should be released shortly.
This contrasts with rivals like OpenAI, whose most recent model, GPT-4, is a so-called “black box,” which means that the data and code used to create the model were inaccessible to outsiders.
Yann LeCun, vice president and chief AI scientist at Meta, was reported as saying: “The competitive landscape of AI is going to completely change in the coming months, in the coming weeks maybe. There will be open source platforms that are actually as good as the ones that are not.”
According to the source, Meta’s founder and CEO Mark Zuckerberg and other company leaders have hinted at a drive to develop a variety of AI chatbots for customers, marketers, and companies on Facebook, Instagram, and WhatsApp that are powered by the company’s LLMs.
Elon Musk, a tech millionaire, has also started a brand-new AI business called xAI with the stated goal of “understanding the true nature of the universe.”
Members of the team under Musk’s leadership had previously worked at OpenAI, Google Research, Microsoft Research, and Google’s DeepMind, among other well-known AI companies.