Meta Launches New Language Model Amid AI Push

As the newest business to enter the AI race, Meta Platforms Inc. announced on Friday that it was introducing a new huge language model powered by artificial intelligence (AI) that was targeted at the research community.

With the release of Microsoft-backed OpenAI’s ChatGPT late last year, the fight for supremacy in the AI technology market, which had previously been in the background, officially began. This move pushed tech giants from Alphabet Inc. to China’s Baidu Ltd. to develop their own products.

According to a blog post by Meta, its LLaMA, or Long Language Model Meta AI, will be made accessible under a non-commercial license to researchers and organizations connected to the public sector, academia, and civil society.

For customers to modify the model and use it for research-related use cases, the company will make the underlying code available. The model, whose computing requirements, according to Meta, are “much less,” is trained on 20 languages with a concentration on those using Latin and Cyrillic alphabets.

Gil Luria, the senior software analyst at D.A. Davidson, said that Meta’s statement today “seems to be a step in proving their generative AI capabilities so they may apply them to their products in the future.”

Although Meta has less expertise with this new type of AI application, it is crucial for the future of their company.

AI has emerged as an attractive area for investment in the tech sector, which has experienced significant layoffs and a decrease in experimental bets as a result of its halting development.

 Meanwhile, increasingly well-liked apps like search are fusing Microsoft Corp.’s, Baidu’s, and Alphabet’s Google’s potent AI language engines.

In May of last year, Meta launched the huge language model OPT-175B, which was likewise intended for researchers and served as the foundation for a new version of its chatbot BlenderBot.

Later, it unveiled a program called Galactica that, according to the company, could write academic papers and resolve mathematical issues, but its demonstration was eventually taken down because it kept producing stuff with an authoritative tone.