OpenAI Chief Executive Officer Sam Altman has acknowledged that last year’s launch of the GPT-5 language model fell short of expectations and alienated many of ChatGPT’s “hundreds of millions” of users. Speaking at a private dinner with reporters in San Francisco, Altman said, “I think we totally screwed up some things on the rollout,” marking his most candid assessment to date of the product’s reception. Altman said the company is accelerating work on GPT-6 to regain momentum but warned that industry-wide shortages of high-performance graphics-processing units could delay the model’s debut. He offered no timetable for release, adding only that the next iteration would arrive “faster” than its predecessor if sufficient compute capacity can be secured.
At a private meal with reporters in San Francisco, Sam Altman admitted that GPT-5's launch upset many of ChatGPT's hundreds of millions of users. “I think we totally screwed up some things on the rollout,” he said #MacroEdge
OPENAI 首席执行官山姆·奥特曼承认 GPT-5 是一次失误,押注 GPT-6。
Wasn’t it just yesterday he said that GPT-6 was going to be: “Basically GPT-5, but as personalized chat bots” After so long hyping 5 maybe it’s time to admit LLMs have plateaued and we need other approaches to reach AGI https://t.co/yhPQmc27Qn