OpenAI GPT-5 Receives Mixed Reactions Amid Technical Quirks

OpenAI's latest model, GPT-5, has sparked polarized responses from developers and testers. While some praised its ability to handle complex prompts and generate structured code, others highlighted persistent flaws, including factual inaccuracies and mathematical errors. Early adopters like Simon Willison noted the model's competence but questioned whether it represented a significant leap from GPT-4.

Criticism intensified on social media, with users pointing out spelling mistakes and inconsistent performance. Noah Giansiracusa, a mathematics professor, labeled the release 'underwhelming,' citing marginal improvements over its predecessor. The model's dynamic architecture—which switches between sizes to optimize processing—further confused users, as some interactions revealed unexpected limitations.

OpenAI faced backlash on platforms like Reddit and X, where users expressed frustration over opacity in model attribution. Despite the controversy, a subset of testers reported adapting to GPT-5's quirks, though skepticism about its reliability persists.