On various social media platforms, however, ChatGPT users expressed frustration that GPT-5 continued to make up information and trip over simple math and spelling questions. Noah Giansiracusa, an associate professor of mathematics at Bentley University, said he felt the launch was “underwhelming.” While there were “some improvements,” he said, “they were much more marginal than I would’ve hoped.”
For months, OpenAI Chief Executive Officer Sam Altman has been hyping up the capabilities of GPT-5, setting up the launch as a seminal moment for the company. But in the first 24 hours after its release, the new model was met with mixed reviews.
In its announcement Thursday, OpenAI said GPT-5 was better at coding and reasoning through complex problems, and touted it as advanced enough to turn chatbot ChatGPT into a Ph.D.-level expert. Some with early access praised the model, with caveats. “It’s my new favourite model,” developer Simon Willison wrote in a blog post, calling it “competent” and “occasionally impressive.” He added: “It’s not a dramatic departure from what we’ve had before.”
