By OpenAI CEO Sam Altman admitted that GPT-5 had a rough start. The launch faced technical issues and complaints from users. The problem came from the real-time router, which sends tasks to different GPT-5 versions. It didn’t work properly during the rollout, making the model seem “way dumber” than it really is.
Altman stated that they are addressing the issues. He also said that he will be “more clear” as to which model will answer which query.
Many Plus subscribers want GPT-4o back; Altman said OpenAI is thinking about it. On Reddit, in a recent AMA, he said, “We are looking into letting Plus users continue to use 4o.” He added, “Rate limits for Plus users will be doubled for the time being to help with this change.”
One of the most talked-about moments was the “chart crime.” Some charts in the launch showed misleading data. For example, one chart showed GPT-5 with a 50% bar, but o3’s 47.4% bar looked much smaller.
Another chart made different values look the same size. Altman called it a “mega chart screwup” caused by a last-minute mistake. OpenAI has now fixed the charts in its blog and apologized for the error.
While GPT-5 boasts next-generation capabilities, its debut revealed that even AI leaders aren’t immune to flubs.
In promising immediate improvements, possible fallback options for frustrated subscribers, and rectified visuals, Altman is signaling a course correction aimed at restoring trust. The speed of AI may be accelerating, but so too is OpenAI’s responsiveness to its users.
Stay up-to-date with the latest news, trends, and insights on The Verbo.
[…] Fires Back: OpenAI’s CEO Sam Altman fought back too, labeling Musk a hypocrite and pointing out Musk’s prior allegations of […]