Mark Zuckerberg (pictured) has mostly ignored provocations from Elon Musk
Meta and Microsoft released Llama 2, a large language model that serves as the foundation for artificial intelligence chatbots. AFP

KEY POINTS

  • Two of the world largest technology companies released a powerful AI tool to the public
  • Llama 2 is available for free for research and commercial use
  • The technology could rival OpenAI's ChatGPT

Two of the world's largest technology companies teamed up to build the foundation of an artificial intelligence language software that could rival industry leading ChatGPT. The pair is letting anyone who wants it access the technology for free.

On Tuesday, Microsoft Corp. (NASDAQ: MSFT) and Facebook-owner Meta Platforms Inc. (NASDAQ: META) announced the pair were working together to roll out so-called Llama 2. The camelid-inspired acronym stands for Large Language Model Meta AI.

In a Tuesday blog post, John Montgomery, Microsoft's corporate vice president for its Azure AI, said the leading technology companies "share a commitment to democratizing AI and its benefits" and are therefore taking an "open approach" with Llama 2.

In its own release, Meta said open access benefits everyone.

"Giving businesses, startups, entrepreneurs, and researchers access to tools developed at a scale that would be challenging to build themselves, backed by computing power they might not otherwise access, will open up a world of opportunities for them to experiment, innovate in exciting ways, and ultimately benefit from economically and socially," the company said in a Tuesday release. "We're now ready to open source ... Llama 2 and are making it available free of charge for research and commercial use."

Meta is due to release its second quarter of 2023 earnings on July 26. During its most recent earnings call, held on April 26, 2023, CEO Mark Zuckerberg touched on Llama 2 and other AI developments at the company.

"I think there's an opportunity to introduce AI agents to billions of people in ways that will be useful and meaningful," Zuckerberg said on the call, according to a Meta transcript. "We're exploring chat experiences in WhatsApp and Messenger, visual creation tools for posts in Facebook and Instagram and ads, and over time video and multi-modal experiences as well.

"I expect that these tools will be valuable for everyone from regular people to creators to businesses. For example, I expect that a lot of interest in AI agents for business messaging and customer support will come once we nail that experience."

In an email to International Business Times, Lauren Van Delden, a spokeswoman for Microsoft, said Llama 2 is a large language model designed by Meta to "enable developers and organizations to build generative AI-powered tools and experiences."

Large language models lie at the backbone of modern text generating AI software like OpenAI's popular ChatGPT. Language models enable computers to read and generate text.

ChatGPT, which is built on the multi-modal large language models GPT-3 and GPT-4 is, already turning heads around the world for its ability to disrupt academia and established businesses relying on writers and editors. The powerful technology is now drawing the scrutiny of regulators in the United States and the European Union.

Not to be outdone, Apple Inc. (Nasdaq: AAPL) is reportedly working on its own large language model AI tool to compete with ChatGPT and Alphabet Inc.'s (NASDAQ: GOOG) experimental Bard.

Amanda Felix, a spokeswoman for Meta, told IBT the company is not currently doing interviews. She declined to discuss the specifics of the technology and the partnership.

Llama 2, Van Delden said, is the successor to the Llama 1 test product released by Meta to the academic and science community on a limited basis. This trial yielded "rapid innovation and optimization" of the model allowing the second version to be released more widely.

"Meta is opening it up for commercial use on other platforms," Van Delden said. "Microsoft and Meta share a belief that AI can benefit everyone and we also believe in giving businesses and developers ... the ability to use and modify Llama 2 for their organizations."

Llama 2, like similar technologies, is open to various applications. Van Delden said it can be used for "content summarization, providing personalized customer service or internal reference experiences as well as a way to query large sets of data using natural language."

"Llama 2 is ideal for organizations within vertical industries such as healthcare, finance, and manufacturing that wish to have additional flexibility to personalize models on their hardware and for those looking for additional model options," Van Delden told IBT.

Hugging Face, an AI community, published a blog detailing the specifics of Llama 2 on Thursday. The community's blog post includes a demo version of the so-called Big Llama 2 Model.

The software is also being distributed through Amazon.com Inc. (NASDAQ: AMZN)'s SageMaker JumpStart.