GPT-5 has gone out, the early updates come in, and they are not good.
Many chatgt fans take reddits and other communication platforms showing their frustrations and disappointments in the newest foundation model, issued on Thursday.
Fast display of Chatgt subsection (unchecked by Opelai) shows a subtle of GPT-5. As the model began out, subreddit filled the posting Posts GPT-5 “Disaster,” “and” the Great trash like a paid user. ”
In pain, ALTMAN and other members of the Openai group had reddit reddit to answer questions about GPT-5. On the papers before the videos, questions collected by expectation, by many operators who want to have Open-4o returns GPT-4 as another way of GPT-5.
OPENA says GPT-5 KNOW LONG – What is the data?
Which red sayings do you say with GPT-5
Many negative ideas say GPT-5 has no “Ubuntu” of GPT-4O, a cold lady, short answers. “GPT-4O had … warmth. It was wise, artistic, and wonderful itself, like talking to someone you found. said another red person. “Now? Everything happens … Seed.”
Another said, “GPT-5 does not have the context of Chatgt (Scic) from other AI bots. I honestly wish to return the estate of the estate or thing.”
Few Redditors also criticize the fact that Alepai has made the option to select different models, which makes us some users say they cancel their subscriptions. “I woke up this morning to find that ACCAi removes 8 models all night. No warnings. No option of One” I send to remove their account for the same reason.
As the manashable reported yesterday, GPT-5 includes various models in Openai on the one platform, and ChatGPT will now select the appropriate model based on user’s trial. Obviously, some users miss the old system and models.
Bright light speed
Logically, ACCAi is also painted criticism by having too many Model options; GPT-5 had to resolve this confusion by postponing previous models under GPT-5.
Sam Altman responds to criticism
When Altman and team goes into the videos, they face a portion of the requirements for returning GPT-4O.
“All right, we feel all in 4o,” Altman said in the middle of the Ama. “We appreciate the time to give us a reply (and love!)
Altman also faced an answer that GPT-5 seems to be a trash than it was, describing the “Autobitcher” that determines which GPT-5 version you can use is not working. “GPT-5 will appear to be wise begins today,” he said. ALTMAN also adds that Chatbot will make it clear that any image of the user responds. Opena will double the boundaries of Chatgpt users and users if the exempt is completed.
“As we said, we expected to ski as we pull out many things at a time. But it was more fuller than we expected!” Altman said to the Ama.
GPT-5 is an improvement, but not a single Exponential
Expected GPT-5 would not be more – and that would be a real problem with GPT-5.
Gary Marcus, an understanding and a historic scientist known as a Siyroscience
The last time of the abandon has issued the later two years ago for GPT-4. Since then, several rivals such as Google Gemini, Anthropic’s Claude, Grok’s Grok, Revesiek R1 is held on the benches, similar agentic features, and user’s integrity. For many, GPT-5 had the strength to strengthen or persist in Ocai as ai leader.
Through this mind, it cannot be prevented that some users can be disappointed, and many Chatgtters share good updates of GPT-5. The time may be diminished by this criticism as Openai make the development of the twins in GPT-5. The company also responded to historically in the user’s reply, with Altman’s most effective x.
“We are currently believing the best way to navigate the challenges of Ai DeploPlopment with a strong loop of a strict and careful response,” said the company’s machinery.
Exposed: Zift Davis, Parent’s company in Mashable’s, in April and filed the fight against Openai, allegedly Ziff Davis Copyrights in training and operating its Ai programs.
Update: Aug. 8, 2025, 3:20 PM EDT This story has been updated with SAM Altman’s answers from Reddit
