A bilingual Arabic and English chatbot developed in the UAE, Jais Chat, has been unveiled by Core42, a unit of Abu Dhabi's artificial intelligence and cloud company, G42.
The app, available to download free of charge on iPhone devices, is proficient in both Arabic and English. According to Core42, the app is engineered with an Arabic-centric model for efficient processing of Arabic text.
The interface of Jais Chat is familiar to those who have used OpenAI's ChatGPT or Microsoft's CoPilot.
Users can use a variety of prompts to get answers, gather information, and find solutions that might elude conventional search engines.
Jais Chat is based on G42's large language model for Arabic, Jais, which was released last year in collaboration with Mohamed bin Zayed University of Artificial Intelligence and Silicon Valley-based Cerebras Systems.
“With its Arabic-first approach, Jais reshapes how bilingual individuals interact with technology, and with Jais Chat for mobile, we are delivering the advancement in our mission to democratise AI access for the world,” said Andrew Jackson, executive vice president and chief AI officer at Core42.
On a more specific level, Jais Chat uses "Jais 30B", which, according to developers, is trained on a data set made up of a total of 1.63 trillion tokens, 475 billion of which are Arabic tokens, with the remaining comprising of English and code tokens.
In the world of AI, tokens are basic units used for calculating text, and higher token counts yield better and more comprehensive results.
“As a result, Jais Chat delivers unparalleled results in Arabic language processing and accuracy while delivering uncompromising English language processing, rivalling top-performing English language models of similar size,” Core42 said.
Core42 said that future iterations of Jais Chat would enable document processing, voice conversation capabilities, and enterprise support with subscription models for businesses seeking customised functionalities.
The name of the app, Jais, refers to the UAE's highest peak in Ras Al Khaimah.
Although Arabic is one of the most widespread languages in the world and is spoken by about 400 million people, during the initial growth of AI developments and large language models it wasn't at the forefront of AI offerings, with English proving to be most prevalent.
The complexity and diversified Arabic dialects, coupled with various language nuances, also posed a challenge for engineers and programmers trying to perfect machine learning technologies.
That changed last year with the introduction of various Arabic LLMs, including Jais, along with other advancements and efforts made to include Arabic.
“Since the inception of Jais in August of 2023, the response has been overwhelmingly positive,” said Mr Jackson. “With the recent launch of Jais 30B, we've witnessed a significant enhancement in its performance metrics compared to its predecessor,” he said.
Core42 describes Jais 30B as “the world's most performant Arabic large language model”.
