Compete with “GPT Chat”… Arabic and international race with alternative models

A scientist specializing in artificial intelligence talks to Sky News Arabia about the chances of these new models overtaking GBT Chat soon, whether other models will emerge and the pros and cons of this race.

OpenAI’s Chat GPT, in which Microsoft is investing heavily, came to light on November 30, 2022, and has continued to gain popularity over time, until its name has become synonymous with any mention of artificial intelligence.

The program continues to break milestones, including amassing 100 million users within 3 months from users across TikTok and Instagram.

Gemini International and Guess UAE

In recent months, two models have emerged to compete with GBT Chat: the Google-affiliated Gemini and the Emirat G42 team-affiliated Emirat Ges.

Google’s AI scientists believe they can beat GBT Chat, and the company is working hard to take it to the next level, codenamed Gemini.

A company called Gemini uses an artificial intelligence model previously used to win world champions in board games. To build on the company’s already advanced AI tools, hoping to take the crown off GPT’s head.

In the Arab world, the Artificial Intelligence Center of the Emirati “G42” group launched an open source linguistic model in Arabic named “GES”.

“GACE” is the world’s highest quality Arabic competitor to “GPT Chat” and others at the level of pre-trained generative transformers with artificial intelligence.

“JESS” was created in cooperation with the United Arab Emirates “G42” Group’s artificial intelligence center “Inception”, the Mohammed bin Zayed University for Artificial Intelligence, the first graduate university specializing in artificial intelligence research in the world, and an American company. “Cerebrous Systems”. , according to Emirates News Agency.

Will this rapid run continue?

Hisham Iraqi, an Amazon scientist at the Seattle, Washington headquarters and one of the founders of artificial intelligence, responded to this question by saying that the number of new advanced models we are seeing and the acceleration of their development is “very high.” than we expected.”

• However, Iraq does not expect new and upcoming models to be free of bugs during this acceleration:
• I don’t think the new models are actually trained on all the data on the internet written and written by humans.
• The new content will not be the same vast amount of content used in previous models.
• Accordingly, artificial intelligence companies began to develop the idea that humans would respond to questions and conversations in these models, and then the model would evolve itself during use, learning better from human responses.
• Human responses usually come from people who have no experience with the topic of conversation; Therefore, they sometimes provide false information during conversations, and the model has difficulty distinguishing between truth and falsehood.