OpenAI CEO Addresses GPT-5 Rollout Challenges and User Feedback
OpenAI's Sam Altman discusses GPT-5's initial issues, user concerns, and steps to enhance performance and transparency.
OpenAI CEO Sam Altman recently addressed user concerns regarding the initial rollout of GPT-5 during a Reddit ask-me-anything session. Users reported that GPT-5 appeared less capable than its predecessor, GPT-4o. Altman attributed this to a malfunction in the real-time router system, which determines the appropriate model for each query. He assured users that the issue has been resolved and that GPT-5's performance should now be improved.
In response to requests from ChatGPT Plus subscribers to reinstate GPT-4o, Altman stated that OpenAI is evaluating the possibility of allowing Plus users to continue using the previous model. He emphasized the company's commitment to gathering data on the trade-offs involved.
Altman also addressed the presentation error during GPT-5's launch, where a bar chart inaccurately represented performance data. He acknowledged the mistake, referring to it as a "mega chart screwup," and noted that the correct charts were available in the published blog post.
To support users during the transition to GPT-5, Altman announced that ChatGPT Plus subscribers will have their rate limits doubled. This increase aims to provide users with more flexibility to explore and adapt to the new model without concerns about reaching usage caps.
Altman concluded by reaffirming OpenAI's dedication to addressing user feedback and enhancing the stability of GPT-5. He stated that the company will continue to work on improvements and maintain open communication with users.
Source
TechCrunchFact-checking
Fact-check the facts of the article using external sources and databases.