Abu Dhabi-developed AI Language Model for Arabic Unveiled

A new Abu Dhabi-developed artificial intelligence (AI) language model for Arabic, called Jais, has been unveiled. Developed by Inception, a unit of Abu Dhabi AI company G42, Mohammed bin Zayed University of Artificial Intelligence, and Silicon Valley-based Cerebras Systems, Jais aims to bring the Arabic language into the mainstream of AI technology.

Jais is an open-source bilingual Arabic-English model that is more accurate than other existing language models for Arabic. It captures the linguistic nuances of various Arabic dialects and can comprehend language, context, and cultural references, making it more accurate and contextually relevant than its counterparts.

The model is expected to be useful in generative use cases such as generating responses to questions, generating documents, translations, emails, and providing advice and recommendations, according to Andrew Jackson, the CEO of Inception.

Jais has been developed for government use and is targeted towards sectors such as finance, energy, climate, and healthcare. It is available for download on Hugging Face, a machine learning platform.

Several public and private organizations in the UAE, including the Ministry of Foreign Affairs, the Ministry of Industry and Advanced Technology, and ADNOC, have signed on as launch partners for Jais.

The language model is trained on the Condor Galaxy, which is considered the world’s largest AI supercomputer. It uses billions of Arabic and English tokens to process and generate language.

Arabic is one of the most widely spoken languages worldwide and is the official language in 22 countries. However, its online presence is minimal, with only about 1% of Arabic content available online.

In an effort to boost Arabic content, Inception is spearheading an initiative to collect more Arabic data from offline sources and exploring ways to synthesize and translate English to Arabic more effectively.

The availability of language models like Jais will enhance AI capabilities and assist developers in bringing new models to the market more quickly.

Subscribe Google News Channel