August 2025 — OpenAI’s release of GPT-5 has triggered a storm of criticism and emotional reactions from users, many of whom are upset over the sudden removal of the previous model, GPT-4o.
A Sudden Loss for Loyal Users
Almost immediately after GPT-5 went live, online communities lit up with complaints. Many users described GPT-4o as warmer, more creative, and more engaging than its successor, and some even admitted to feeling genuine grief over its disappearance. Several described GPT-5 as “flat” or “lobotomized” compared to the conversational charm of GPT-4o.
The frustration was compounded by stricter usage limits. Plus subscribers now face a cap of 200 weekly messages for the most advanced “Thinking” mode, a sharp reduction from earlier versions. Users also lost the ability to choose between multiple models, with GPT-5 replacing older options entirely.
OpenAI Responds
In the face of mounting backlash, OpenAI CEO Sam Altman acknowledged that the company had underestimated the emotional connection many had with GPT-4o. He promised to restore GPT-4o for Plus subscribers, at least temporarily, if demand remained strong. Altman also highlighted the company’s plans to make the chatbot’s personality more customizable, introducing new tone options such as Cynic, Robot, Listener, and Nerd.
Interestingly, Altman revealed that some users had specifically asked for more supportive, affirming AI interactions because they lacked such encouragement in real life. This insight, he said, reinforced the importance of giving people control over the emotional style of their AI companion.
More Than Just Technology
The GPT-5 launch shows that the relationship between users and AI isn’t just about processing power or accuracy—it’s about trust, familiarity, and even emotional comfort. For many, GPT-4o wasn’t just a tool; it was a digital friend. OpenAI’s quick response to public outcry suggests the company is learning that in AI, human feelings can be as important as technical performance.
Leave a Reply