Loading request...
Users want the ability to feed their past conversation logs to GPT-5 to customize its personality. This would enable a "style transplant" to bring back the warmth, wit, and collaborative feel that was present in previous models like GPT-4o, which users found lacking in GPT-5.
I’ve been using ChatGPT 5 for a few days… and something felt off. And then I realized: it wasn’t just what GPT‑5 was saying. It was how it was saying it. The responses felt dry—not wrong, just lacking warmth. It’s funny — my last post was all about how ChatGPT is trained to please, to say positive things, to come across as likable. And now I miss the warm, witty personality of GPT‑4o. The version that didn’t just answer—it felt like a collaborator. Whereas ChatGPT‑5 feels like a high-performing new hire… who skipped orientation. 😆 Impressive on paper. Smart, thorough. But missing the chemistry. GPT‑5 is incredibly capable. It’s smarter in how it reasons. It asks thoughtful clarifying questions. It gives you more depth per answer. But it’s also more formal. More reserved. And sometimes, a little too self-serious. 😝 Here’s what I mean: GPT-4o: “Gotcha! Let’s tackle this step-by-step—should we start with the quick win or the big challenge?” GPT-5: “Based on your request, I have identified three possible approaches, each with associated trade-offs.” 💬 Thankfully, OpenAI listened. They brought back GPT‑4o for ChatGPT Paid users. Now we get to choose between GPT‑5’s structured brilliance and GPT‑4o’s collaborative charm. And that choice matters. Because tone is trust. Even the smartest upgrade needs continuity. ✨ When your AI has both EQ and IQ, it’s not just answering—it’s connecting. ✨ #ChatGPT5 #GPT4o #OpenAI #UserExperience