The launch of the new AI model, GPT-5, has raised numerous questions and issues among users. Sam Altman, CEO of OpenAI, addressed many of these inquiries during an AMA on Reddit.
What Happened with the GPT-5 Rollout?
Following the launch of GPT-5, users quickly began reporting performance issues, including feelings that the new model seemed less effective than its predecessor, GPT-4o. These complaints led to inquiries about the causes of the perceived decline. Altman explained that the reason was a malfunctioning autoswitcher system, which was supposed to ensure optimal model selection for each query.
Will GPT-4o Return and Doubled Rate Limits for ChatGPT Plus Users?
During the AMA, the strong desire of users to bring back GPT-4o was evident. Altman confirmed that OpenAI is considering the option of returning this model for ChatGPT Plus subscribers. It was also announced that rate limits for Plus users would double, allowing for more extensive use of the new GPT-5 model.
Decoding the Infamous OpenAI ‘Chart Crime’
In the presentation of GPT-5, a humorous incident occurred with an inaccurate chart that quickly went viral. Altman had previously acknowledged the error on social media, calling it a 'mega chart screwup.' This situation, while met with humor, highlights the importance of data presentation accuracy, a critical issue for AI models.
The GPT-5 rollout has certainly been a learning experience for both OpenAI and its user community. Altman provided valuable insights into performance issues and user demands, along with outlines for improvements and enhanced user support.
