Chaos and Comedy in the GPT-5 Rollout

During a lively AMA session on Reddit, OpenAI's CEO Sam Altman faced a barrage of questions about the much-anticipated GPT-5 and concerns about its predecessor, GPT-4o. Fans were craving clarity on the recently launched model and even begged for the return of the adored 4o.

Introducing the New Features—or the 'Chart Crime'?

Among the fresh features of GPT-5 is a real-time router designed to optimize responses by determining the best model to handle specific prompts. This system, however, faced hiccups at launch, leading many users to voice frustrations over GPT-5's performance compared to 4o.

A Technical Glitch? Altman Responds!

Altman revealed that a significant technical issue caused the initial rollout to falter. "GPT-5 will appear smarter starting today. We experienced a server incident that temporarily disabled the autoswitcher, resulting in a notably underperforming model. We're also making adjustments to the decision-making process to ensure better model selection going forward," he clarified.

The Fans Demand GPT-4o Back!

The outcry for GPT-4o was so substantial that Altman assured users he would explore the possibility of allowing Plus subscribers to revert back to the classic model. He wrote, "We are looking into allowing Plus users to continue using 4o as we gather more data on the tradeoffs." This would help users smoothly transition or maintain their preferred experience.

More Prompts, Less Worry for Plus Subscribers!

In a bid to enhance user experience, Altman also announced a doubling of rate limits for Plus users as the rollout proceeds. This update would empower users to explore the new capabilities of GPT-5 without the anxiety of exhausting their prompt limits.

The Infamous 'Chart Crime': A Hilarious Blunder

Naturally, amidst all the questions, the infamous 'chart crime' came up. Altman's team presented a perplexing chart during the live demo, leading to a wave of humorous memes and commentary across social media. While the published blog post contained accurate visuals, the initial mistake left many giggling over the thought of using GPT for corporate presentations.

A Promise for Improvement

Despite the blunders, Altman concluded the AMA with a reassuring message: "We will continue to work tirelessly to stabilize these issues and listen to your feedback." His commitment to refining the experience left fans hopeful for the future of GPT technology.