GPT-5 has been launched, and the earliest feedback is not positive.
A number of ChatGPT fans have expressed their discontent and frustration on Reddit and other social media sites concerning OpenAI’s new foundation model that was introduced on Thursday.
A quick glance at the ChatGPT subreddit (unaffiliated with OpenAI) showcases severe criticisms of GPT-5. Since the model’s debut, the subreddit has been inundated with messages branding GPT-5 as a “disaster,” “terrible,” and the “worst piece of garbage even as a paying user.”
Curiously, Altman and other members of the OpenAI team are set for a Reddit AMA at 2:00 p.m. ET today to respond to inquiries regarding GPT-5. Questions have already started piling up in anticipation of the event. Numerous users are calling on OpenAI to bring back GPT-4o as an option in place of GPT-5.
OpenAI asserts GPT-5 hallucinates less — what does the evidence show?
Reddit users’ views on GPT-5
Many of the early unfavorable perceptions indicate that GPT-5 misses the “personality” of GPT-4o, with remarks on its more distant, terse responses. “GPT-4o had this… warmth. It was clever, imaginative, and surprisingly relatable, like interacting with someone who understood you. It didn’t simply provide answers; it felt as though it listened,” one redditor commented. “Now? Everything seems so… clinical.”
Another user remarked, “GPT-5 lacks the spirit and character that distinguished Chatgpt (sic) from other AI bots. I genuinely hope they reintroduce 4o as a legacy model or something similar.”
Several redditors also condemned the elimination of the ability to select different models, prompting some users to cancel