GPT 5 was launched as a unified system featuring an automatic model router intended to select the best variant — be it speedy or deeply thoughtful — according to the query’s complexity. But users quickly found the new experience lacking. Users revolted on Reddit, flooding it with complaints and saying GPT 5 doesn’t have a personality, emphasizing that its responses are shorter, colder, less creative, and even erroneous.
One user even mourned the loss of GPT 4o, marking it “The day they killed GPT 4o, it felt like watching a friend die.” Another called GPT 5 a “corporate beige zombie.” These responses highlighted just how far people had connected with their AI versions — not just for performance, but for companionship.
Also Read: OpenAI CEO surprising predictions could blow your mind
The backlash wasn’t just noisy — it was swift. OpenAI responded within days by reinstating GPT 4o for paid users, restoring access via settings to “show legacy models.” CEO Sam Altman took to X (formerly Twitter) to admit that GPT 5’s router had “broken” briefly, resulting in “way dumber” performance that fell short of expectations. The company also promised clearer signaling of which model is answering and hopes to bring back GPT 4.1 for Plus users if demand merits it.
GPT-5 rollout updates:
— Sam Altman (@sama) August 8, 2025
*We are going to double GPT-5 rate limits for ChatGPT Plus users as we finish rollout.
*We will let Plus users choose to continue to use 4o. We will watch usage as we think about how long to offer legacy models for.
*GPT-5 will seem smarter starting…
In addition, Altman pledged performance improvements, increased usage limits (especially for Plus users), and a warmer, more engaging AI personality in future updates.
This saga is a powerful reminder: AI isn’t just code — it’s company. When people engage with these systems daily, design changes need empathy, not just efficiency, to avoid alienating the very community they serve.
