Sam Altman addresses GPT-5 concerns, promises it'll be smarter now
What's the story
OpenAI's CEO Sam Altman has responded to user concerns about the performance of its latest AI model, GPT-5, and requests to bring back its previous model, GPT-4o. During a recent Reddit AMA session, he promised that "GPT-5 will seem smarter starting today." The clarification comes after users noticed a decline in performance compared to the previous model, GPT-4o.
Technical glitch
Initial poor performance due to technical glitch
Altman explained that the reason for GPT-5's initial poor performance was a technical glitch with its router. The issue prevented the model from functioning as intended at launch, making it seem less capable than its predecessor. "Yesterday, we had a sev and the autoswitcher was out of commission for a chunk of the day," he said.
Improvement measures
Model decision boundary improvements on the way
Altman also revealed that OpenAI is working on improving the decision boundary of GPT-5. This change will help users get the right model for their queries more often. "We will make it more transparent about which model is answering a given query," he promised, assuring users of better performance in the future.
User feedback
Bringing back GPT-4o for Plus subscribers
Responding to user requests, Altman said OpenAI is bringing back GPT-4o for Plus subscribers. *We will let Plus users choose to continue to use 4o. We will watch usage as we think about how long to offer legacy models for." he wrote on X. He also promised to double the rate limits for Plus users as they complete the rollout of GPT-5, giving them more room to explore and adapt this new model without running out of monthly prompts.
Twitter Post
Take a look at Altman's post
GPT-5 rollout updates:
— Sam Altman (@sama) August 8, 2025
*We are going to double GPT-5 rate limits for ChatGPT Plus users as we finish rollout.
*We will let Plus users choose to continue to use 4o. We will watch usage as we think about how long to offer legacy models for.
*GPT-5 will seem smarter starting…
Presentation blunder
Altman addresses 'chart crime' incident during GPT-5 launch
Altman also addressed the 'chart crime' incident, where an inaccurate chart was presented during the live presentation of GPT-5. He recently called it a "mega chart screwup" on X. Despite the error, he assured users that the company is working on fixing issues that have been most concerning to them.