AI21 Labs Launches Language Model, One of the Largest to Date

The Tel Aviv-based artificial intelligence (AI) development company AI21 Labs has announced the launch of what may be one of the largest AI models of its kind. AI21 Labs’ Jurassic-1 Jumbo Language Model works as an application programming interface (API) which is now available for developers and researchers in beta via the company’s recently launched AI21 Studio.

“You shouldn’t have to be an AI researcher working at a big tech company to do this stuff,” said Yoav Shoham, the company’s co-founder and co-CEO. “Now anyone — publishers, students, artists, businesspeople, researchers — can build language-based applications that rival those being dreamed up in big AI labs.”

AI21 Labs was initially founded in 2017, by Shoham and his colleagues Ori Goshen and Amnon Shashua. Jurassic-1 was launched through the AI21 Studio Aug. 11, and according to a press release from the company, is available for free to anyone who is interested in using or exploring the program. AI21 Labs has been working on developing Jurassic-1 since the company was founded in 2017, and the model has cultivated a vocabulary of 250,000 lexical items. 

Venture Beat’s Kyle Wiggers noted that the model could even rival OpenAI’s GPT-3 model — GPT-3 is currently recognized as the largest AI model of its kind, however AI21 Labs says its Jurassic-1 model is “slightly larger” than GPT-3, with more than 178 billion computational parameters. Additionally, GPT-3’s vocabulary is about one-fifth the size of Jurassic-1’s, at just around 50,000 lexical items. 

Both language models are effective in generating written language, such as recipes, emails, website layouts, and code — however, AI21 Labs stresses the fact that Jurassic-1 is much more accessible to the public than other language models with similar capabilities. Unlike other language models like GPT-3, Jurassic-1 is available for free to anybody who signs up for an AI21 Studio account.

“Our mission at AI21 Labs is to fundamentally reimagine the way humans read and write by introducing machines as thought partners, and the only way we can achieve this is if we take on this challenge together,” reads the company’s announcement.

Andrew Warner
Andrew Warner is a writer from Sacramento. He received his B.A. in linguistics and English from UCLA and is currently working toward an M.A. in applied linguistics at Columbia University. His writing has been published in Language Magazine, Sactown Magazine, and The Takeout.


Weekly Digest

Subscribe to stay updated

MultiLingual Media LLC