Developers have shared mixed experiences with OpenAI’s GPT-5. Jenny Wang, an engineer and creator of styling agent Alta, noted its improved ability to handle complex coding tasks in one go, outperforming previous models like GPT-3 and GPT-4. For her company’s press page, GPT-5 accurately generated design-specific code in a single attempt, despite hallucinating URLs. Another developer praised GPT-5’s capability to tackle deep technical problems, noting its effectiveness in providing project recommendations. Some enterprise partners, including Notion, confirmed its fast and thorough performance. However, critical feedback emerged quickly; Kieran Klassen stated that GPT-5’s coding capabilities felt outdated, reminiscent of older models. Amir Salihefendić deemed it “underwhelming” in coding tasks. Additionally, while some found it verbose, adjustments can be made for less redundancy. Overall, the sentiments reflect a balance of innovation and disappointment, as developers assess GPT-5’s position in the evolving AI landscape.
Source link
Share
Read more