spot_img
HomeNews & Current EventsOpenAI CEO Sam Altman Acknowledges GPT-5's Rocky Start, Reinstates...

OpenAI CEO Sam Altman Acknowledges GPT-5’s Rocky Start, Reinstates GPT-4o Option and Boosts User Limits

TLDR: OpenAI CEO Sam Altman has admitted that the recently launched GPT-5 model experienced an ‘unflattering debut,’ leading the company to allow Plus subscribers to revert to the highly regarded GPT-4o and simultaneously doubling their usage limits. The decision follows user backlash and a technical issue that made GPT-5 appear less capable.

OpenAI’s latest flagship AI model, GPT-5, has faced an unexpected setback, with CEO Sam Altman publicly acknowledging its less-than-stellar initial performance. During a candid Reddit AMA on Friday, August 8, 2025, Altman confessed that GPT-5 had a ‘rather unflattering debut,’ prompting a swift response from the company.

In a move aimed at appeasing its user base, OpenAI is now allowing Plus subscribers the option to revert to the previously favored GPT-4o model. Furthermore, the company has announced a significant doubling of usage limits for these subscribers, a measure seen as a sweetener amidst the initial disappointment.

Altman attributed some of the performance issues to a ‘sev‘ (service issue) where the ‘autoswitcher’ — a real-time router designed to determine whether GPT-5 should provide a quick answer or take more time to ‘think’ — was out of commission for a significant portion of the day. This technical glitch, he stated, made GPT-5 ‘seem way dumber‘ than intended. He further assured users that interventions are being made to the ‘decision boundary’ to ensure the correct model is utilized more frequently, and transparency regarding which model is answering a query will be improved.

User feedback played a crucial role in this decision, with multiple users demanding the option to continue using GPT-4o, which many felt performed superiorly for their specific needs. Altman conceded to these demands, stating, ‘We are looking into letting Plus users continue to use 4o.‘ The Reddit AMA also touched upon other aspects, including feature requests and an ’embarrassing’ moment during the launch, referred to as the ‘chart crime.’

Also Read:

This development, published by Unnati Gusain on August 10, 2025, highlights the challenges in deploying advanced AI models and the importance of responsive developer-user interaction. While the long-term performance of GPT-5 remains to be fully assessed, Plus users can now benefit from increased prompt allowances and the flexibility to choose their preferred model.

Ananya Rao
Ananya Raohttps://blogs.edgentiq.com
Ananya Rao is a tech journalist with a passion for dissecting the fast-moving world of Generative AI. With a background in computer science and a sharp editorial eye, she connects the dots between policy, innovation, and business. Ananya excels in real-time reporting and specializes in uncovering how startups and enterprises in India are navigating the GenAI boom. She brings urgency and clarity to every breaking news piece she writes. You can reach her out at: [email protected]

- Advertisement -

spot_img

Gen AI News and Updates

spot_img

- Advertisement -