Zuckerberg's Meta Joins AI Chatbot Race With Large Language Model Called 'LLaMA'

As ChatGPT contines to take over the world of technology, Meta (parent company of Facebook) has announced a research tool called LLaMa that will also rival Google's LaMDA - a family of neural language models.

In a Facebook post on Friday, Meta CEO Mark Zuckerberg said that language models "have shown a lot of promise in generating text, having conversations, summarizing written material and more complicated tasks like solving math theorems or predicting protein structures."

Unsplash

Meta joins AI chatbot race

For now, LLaMA is not in use across Meta services like Facebook and Instagram. According to Zuckerberg, the tech would be made available to researchers. Zuckerberg said that "Meta is committed to this open model of research."

Wondering what large language models are? Such AI systems learn and imbibe knowledge from large volumes of text available on the internet, based on which their responses get smarter and crispier. As is true for all AI algorithms, the more data such language models are fed, the better performing they are.

Also read: "Managers Managing Managers": Facebook Parent Meta CEO Mark Zuckerberg Hints At More Layoffs

Reuters

ChatGPT was built by OpenAI, a startup that has received backing from Microsoft. Recently, Microsoft said that it would incorporate the text tool into its Edge browser and Bing search.

Unsplash

Google's parent company Alphabet also has a language model of its own called LaMDA based on which it has built a ChatGPT-like tool called Bard. Meta has dabbled in large language models earlier with OPT-175B and the controversial Galactica that was quickly rolled back.

Also read: Facebook's New AI Bot Thinks Mark Zuckerberg Is A 'Manipulative Bad Person'

What do you think the new arena of language model wars? Let us know in the comments below. For more in the world of technology and science, keep reading Indiatimes.com.