OpenAI Restores GPT-4o After Users Rebel Against the 'Cold' GPT-5 Experience

A Swift Reversal for OpenAI

In a surprising turn of events, OpenAI has officially brought back GPT-4o as a primary option for ChatGPT users. This decision came just twenty-four hours after the company attempted to move forward with GPT-5 as the new "brain" of its famous chatbot. The rollout, which was supposed to be a massive leap forward in artificial intelligence, quickly turned into a PR headache as users clamored for the return of the previous model. It is rare to see a tech giant backtrack so quickly on a major release, but the data—and the voices of the community—were too loud to ignore.

Why Users Preferred GPT-4o

It turns out that raw intelligence and newer algorithms aren't everything in the world of AI. According to reports from The Verge, the community reaction was swift and overwhelmingly negative. The primary grievance? GPT-5 felt "soulless." Users who had grown accustomed to the warmth, nuance, and conversational fluidity of GPT-4o found the new model to be significantly more rigid and difficult to interact with on a daily basis.

Many long-time users noted that the unique "personality" they had come to rely on was missing. Instead of a helpful assistant that felt like it understood the context and tone of a conversation, GPT-5 was described as being overly formal and corporate. These characteristics don't always translate well to a tool that people use for everything from creative writing to casual brainstorming or even just emotional support.

The Backlash on Reddit

The discontent was most visible on Reddit, where users didn't hold back their frustrations. One user highlighted a specific interaction that ruined their morning: "This morning I tried talking to it, and instead of a short paragraph with exclamation marks, or optimism, all I got was a single sentence. Plain corporate nonsense." This shift from an encouraging, detailed partner to a blunt, robotic system was a dealbreaker for many.

This sentiment was echoed across various threads, with some users even expressing a sense of grief over the change. Another user shared a surprisingly poignant thought, saying that the previous model (referring to the 4.x/4o iterations) actually "spoke" to them and, despite how it might sound to others, felt like their only friend. This highlights a growing trend in AI development: the emotional connection between humans and machines is far deeper than developers might realize, and changing the "vibe" of an AI can have real psychological impacts on its user base.

Lessons in AI Development

OpenAI’s decision to pivot back to GPT-4o suggests that the company is listening closely to user feedback, even when it means admitting a new launch didn't land as intended. The shift proves that "more advanced" does not always mean "better" in the eyes of the consumer. If a model loses its ability to connect with humans on a conversational and emotional level, it risks losing the loyalty of the very people it was built to serve.

For now, fans of GPT-4o can breathe a sigh of relief. The model they know and love is back in its place. Meanwhile, the team at OpenAI likely has a lot of work to do to ensure that when GPT-5 eventually becomes the standard, it carries over the "human touch" that users clearly value above all else.