OpenAI’s GPT-5 Met With Mixed Reviews, Confusion in First Day

In a much-anticipated release, OpenAI's CEO Sam Altman has been promoting the capabilities of GPT-5, heralding it as a transformative advancement for the company. However, just 24 hours post-launch, the new AI model has sparked a range of reactions from users and experts alike. Announced on Thursday, OpenAI claimed that GPT-5 excels in coding and complex problem-solving, even suggesting it can elevate the ChatGPT experience to that of a Ph.D.-level expert. Early adopters, like developer Simon Willison, expressed enthusiasm, labeling it as their "new favorite model" and noting its "competence" and occasional impressiveness. Yet, he cautioned that it doesn’t represent a significant leap from previous versions. Conversely, many users took to social media to voice their disappointment. They reported ongoing issues with the model generating inaccurate information and struggling with straightforward math and spelling tasks. Noah Giansiracusa, a mathematics professor at Bentley University, described the launch as "underwhelming," citing only marginal improvements. The mixed reviews may partly stem from confusion surrounding the model's operational mechanics. Unlike its predecessors, GPT-5 employs an autoswitching feature that toggles between various levels of complexity based on user queries. While this is designed to optimize computing resources, it can lead to situations where users interact with a less advanced version of the model. For instance, in one test, GPT-5 incorrectly stated that the letter "b" appears three times in the word "blueberry," only to provide the correct answer when prompted to reconsider. In response to the feedback, Altman acknowledged that a technical issue had marred the initial performance of GPT-5, stating, "Yesterday, the autoswitcher broke and was out of commission for a chunk of the day, and the result was GPT-5 seemed way dumber." He assured users that improvements would be evident starting that day. The stakes surrounding this launch are substantial, as OpenAI aims to maintain its edge against rising AI competition from both U.S. and Chinese firms. The company is also striving to attract businesses and individual users to its premium services to finance its extensive investments in talent, technology, and infrastructure. Since the inception of the generative AI movement with ChatGPT, powered by the earlier GPT-3.5 model, OpenAI has rolled out increasingly sophisticated systems. However, as AI technology evolves, evaluating the relative performance of these models becomes increasingly complex. By midday Friday, GPT-5 had climbed to the top of various AI model leaderboards, although it lagged behind the latest version of Grok from Elon Musk's xAI on another benchmark. As nearly 700 million users engage with ChatGPT weekly, opinions on the model's performance are bound to differ. Notably, Ethan Mollick, a Wharton School professor, praised GPT-5 for its research capabilities and ability to assist novices with programming. He remarked, "GPT-5 just does stuff, often extraordinary stuff, sometimes weird stuff, sometimes very AI stuff, on its own. And that is what makes it so interesting." Meanwhile, during a recent Reddit AMA, Altman faced criticism from users seeking more transparency regarding which model they were interacting with. He committed to implementing changes to enhance user understanding and experience, acknowledging differing opinions on the writing quality between GPT-5 and its predecessor GPT-4.5, with many users expressing dissatisfaction with the new model. As more data and user experiences accumulate, the true impact of GPT-5 on personal and professional applications is yet to be fully realized.