Sam Altman has been quoted as saying the rollout of GPT-5 was “screwed up” (see the article here: https://finance.yahoo.com/news/openai-ceo-sam-altman-concedes-171805241.html). We want to hear from you: have you tried using GPT-5, and if so, how has it compared with your experiences using earlier versions of OpenAI models, and with other model providers?
1 Like
Personally, I like it. However, I also wasn’t expecting much besides being more affordable and slightly better.
I don’t mind the slower response times, and hallucination has been reduced greatly!
Across all of these benchmarks, “GPT‑5 thinking” shows a sharp drop in hallucinations—about six times fewer than o3—marking a clear leap forward in producing consistently accurate long-form content.
Source: Open AI
I haven’t used GPT-5 in pickaxe a ton, but using ChatGPT, its pretty killer.
I’m happy to elaborate on anything!
2 Likes
I think for code generation, I’ve found it’s somewhat competitive with Claude but not consistently on par.
When it comes to sales and marketing copy, it seems to lose context in longer conversations and requires significant iteration to refine the output, even with well-structured prompts.
Its primary strengths right now are speed and brainstorming initial concepts.
2 Likes