It has been a year and a half since OpenAI introduced its most recent “foundation” large language model (LLM), GPT-4, and the rumor mill has been in high gear the past month with speculation on when a next-generation model will appear. Appearing at a business conference this week, the lead executive for OpenAI’s Japan operations, Tadao Nagasaki, teased a forthcoming advance in LLMs from the company, which he referred to as “GPT Next”. “The AI model called GPT Next, which will eventually come out, will evolve nearly 100 times based on past achievements,” said Nagasaki, based on an automated translation of the Japanese-language write-up of the conference presentation by Tachiki Matsuura for the online publication ITmedia. A blurry image of Nagasaki’s slide deck displayed on the monitor behind him showed a timeline of GPTs, starting with “GPT-3 Era” in 2021, followed by “GPT-4 Era” in 2023, “GPT Next” era in 2024, and finally “Future Models” with a date of “202x”. It is not clear from Matsuura’s write-up whether or not Nagasaki explicitly cited the delivery of a “Next” model this year in his remarks. In a meeting with journalists last November, OpenAI CEO Sam Altman indicated it would be a while before a successor to GPT-4 appears. “The number of things we’ve gotta figure out before we make a model that we’ll call GPT-5 is still a lot,” said Altman, who added: “It’s very hard to predict timelines for these things.” A blog post from the company in May of this year said: “OpenAI has recently begun training its next frontier model, and we anticipate the resulting systems to bring us to the next level of capabilities on our path to AGI [artificial general intelligence].”
Full report : CEO of OpenAI Japan just said an AI called ‘GPT-Next’ is coming, it’s 100x better than GPT-4.
