After the chatbot ChatGPT and its successor GPT-4, is OpenAI preparing a new revolution in artificial intelligence? If so, it’s not for now, said Sam Altman, the head of the start-up, during a conference in India on Wednesday June 7.
“We’re working on new ideas that we think we need, but we’re certainly not about to start” training GPT-5, Sam Altman said in comments reported by Techcrunch. In April, OpenAI CTO Mira Murati had already told the AP that OpenAI “does not plan to train GPT-5 in the next six months.”
World Tour
Sam Altman, who has become the face of a booming industry, recalls thatopen AI He had to continue working on GPT-4 for long months after he finished training it: “When we finished GPT-4, it took us more than six months to be ready to release it” and prevent the language model “hallucinates” – generates texts that contain false information.
Sam Altman is on an international tour, during which he has met with various international leaders (including Emmanuel Macron) to discuss artificial intelligence, its benefits, its dangers and how to regulate it.
Sam Altman regularly explains that he is concerned about the emergence of a “super-intelligent” AI. A risk that could be amplified by the development of increasingly powerful AIs such as GPT-5, and which, according to Sam Altman, should be treated at the same level as the threats of a nuclear war or a global pandemic, under the terms of an agreement. open. letter signed by Sam Altman and many specialists, and denounced by many others.
Source: BFM TV
