The latest Frontier Artificial Intelligence (AI) model of Chatgpt users Openai is not happy with GPT-5. The new model was released on Thursday as the integration of the GPT-series and the O-series. The San Francisco -based company said they wanted to make a model that could do everything in itself, so users had to not use model picker (dropdown menu that shows many available models). Openai also retired old models with the launch of GPT-5. However, the move can be left behind as many users have started complaining about removing GPT-4O.
Chatgpt user GPT-4o wants back
Many users have started posting about GPT-5 getting worse than GPT-4o on social media platforms. A reddit user complained that the latest AI model reactions were small and insufficient and without personality.
Similarly, at an OOSC Mi Etho (AMA) session on Reddit with OpenAII CEO Sam Altman, the user U/Naddybear said, “I have long been using GPT-4o and have built a very specific dynamic and workflow with it.[..]I know that GPT-5 is designed to be strong for complex arguments, coding and professional tasks, but we don’t all need a Pro coding model. Some of us rely on 4Os for creative collaboration, emotional nuances, rollplays and other long-form, high-content interactions. “In one answer, the CEO promised to fix the issue.
Hours later, Altman posted on X (formerly known as Twitter) and announced several changes that users would meet within the next few days. He said that plus subscribers would soon get double the rate limit for GPT -5 (some users complained about killing the rate limit within an hour).
Additionally, Openai’s CEO stated that the plus subscriber will continue to use the GPT-4O model. However, the company will monitor the use to determine how long the model should be made available. He also explained why people would have found GPT -5 ineffective.
According to Altman, “Autoswitcher” or real -time router that switchs to a different capacity based on Quarry (argument, convertible, creative, etc.), “was out of commission.” As a result, GPT-5 cannot perform according to its capacity. The CEO believes that this is the reason why the AI model felt Dumber.
Openai is now making some changes that the model-switching works will ensure that users are getting the correct model more often. Altman also said that Chatgpt would show transparently which model is being used to generate a response from users to eliminate any confusion.
Openai CEO says the company is running at loss, but it is fine
During an interview with CNBC after the launch of GPT-5, Altman highlighted that the company was currently running at a loss despite hitting billions of dollars annually. However, he was allegedly not worried about it. He said, “As long as we are getting better and better on this different curve of the model, I think it is the rational thing that it is just ready to run the loss.”
Last year, Openai reportedly produced $ 3.7 billion in revenue, but lost $ 5 billion. A major part of the company’s operational expenses leads to powering and maintaining the in-house server and paying for the third-party providers to pay for servers. This may be the reason behind deciding to retire old models of high cost company.
However, it appears that each of these old AI models was trained to do something different, and GPT -5 is still unable to catch it. It puts Openai in a catch -22 position, as the company will continue to suffer heavy losses by continuing the old model, but takes them away from the user base in the backlash.
