Elon Musk is working on an artificial intelligence project for Twitter

Last month, the head of the social network, along with hundreds of researchers, signed a plea to take a break from research on the latest generation of artificial intelligence.

By Le Figaro with AFP

published On 04/12/2023 at 03:16to update On 04/12/2023 at 08:10

Elon Musk is investing in artificial intelligence (AI) within Twitter, according to an article published on Tuesday, April 11. from the insideWhile the billionaire signed last month An invitation to take a break from next-generation AI research.

According to anonymous sources fromfrom the inside, The leading company recently purchased about 10,000 graphics processors, and the Twitter artificial intelligence project will include a language model. These models, trained from large amounts of data, are able to generate all kinds of scripts and answer user questions via interfaces such as ChatGPT (OpenAI) or Bard (Google). Elon MuskTwitter, which has laid off thousands of Twitter employees since it bought the platform in October, has also hired artificial intelligence specialists, according to Twitter. from the inside. The article mentions hiring Igor Babushkin and Manuel Kroes, both of whom went through DeepMind, the AI ​​subsidiary of Alphabet (Google’s parent company).

When will software be ‘generally smarter than humans’?

specialized site the information He had already indicated at the end of February that the head of Tesla, SpaceX and Twitter was trying to team up to compete with OpenAI, which had created a buzz since last year with the success of ChatGPT. Asked by AFP, Igor Babushkin did not immediately reply and the press’s Twitter address automatically sent a typical poop expression.

From Microsoft to Snapchat, all the tech giants and many startups are investing in generative AI. But the adoption of this technology and its very rapid progress also raises many concerns, especially when Sam Altman, head of OpenAI, talks about the coming of the upcoming artificial intelligence called “General”when the programs “Smarter than humans in general”.

Is it desirable to evolve non-human spirits that could eventually outnumber us, deceive us, and take our place? »

Excerpt from the appeal signed by Elon Musk

At the end of March, hundreds of academics, presidents, and personalities, including Elon Musk, Steve Wozniak (co-founder of Apple), and Imad Mostaque (head of Stability AI), called for a six-month pause in the search for AI. More powerful than GPT-4, OpenAI’s latest language model. They mention the various risks associated with this technology, and ask themselves in these terms: “Is it desirable to develop non-human spirits that might eventually outnumber, outpace, and supplant us?”

Elon Musk co-founded OpenAI in 2015, before leaving the company in 2018. The California startup has been funded primarily by Microsoft since 2019. He has since criticized the company, particularly as he believes in a tweet last December that it leads to IA being “awake” (a term for a fringe of the American left), ie “to lie”.