The article with GPT-5: Openaai Wärmmtchmtcht Chatgpt first appeared at the online magazine Basic Thinking. You can start the day well every morning via our newsletter update.
With GPT-5, Openai released its latest AI model in early August 2025. It should be faster, more intelligent and useful than its predecessors. But although Openai boss Sam Altman GPT-5 had fully announced as a AI revolution, disillusionment quickly spread for many users. Now the company wants to counteract.
Background: Chatgpt-5 is more impersonal
- GPT-5 should decide for itself which task it uses which AI model. It was different under GPT-4. But with the publication of the new version, Openai initially stamped its old models. This has caused frustration for many users, there GPT-5 responds much impersonal and scarce. So Openai has GPT-4O released.
- According to a Study of the University of Oxford Are sPrach models that provide warmer and more empathetic answers, also more prone to errors. Models that personal answers would tend to flatter, which some users find uncomfortable. The researchers warn that properties such as empathy can affect other important skills.
- With GPT-5, Openai wanted to Minimize hallucinations as well as flatteryaround that To improve the language model. However, many users criticize that the new version is too impersonal. The company therefore wants to counteract and has announcedTo make GPT-5 warmer and friendlier.
Classification: GPT-5 provides frustration
It is not the first time that Openai Chatgpt has improved. At the end of April 2025, the company was forced to withdraw GPT-4O in the meantime because the Ki simply behaved too nicely And to almost everything yes and Amen said – with partially devastating consequences.
The O3 and O4-Mini models also pointed out one significantly higher error rate as their predecessor models. The reason: hallucinations. However, even Openaai could not explain why this was the case. Instead, the company with GPT-5 again caused frustration for many users.
The reason: Openai has The focus of his new model not only focused on performance and effectiveness And neglected the emotional quality of the interaction, but also took users the opportunity to select between different models themselves.
The announced changes in terms of tone and friendliness could in turn pull deterioration in other areas.
Voices
- Above all via reddit leave Numerous users run free their displeasure: “You have completely ruined chatt. There are such short answers and makes some of the most basic things wrong.” Another added: “It definitely does not have the same atmosphere as 4o. It is in a way better organized, but I think that the answers are shorter and scarcer in the tone.”
- Openai boss Sam Altman showed himself insightful: “We are working on an update of the personality of GPT-5, which should feel warmer than the current personality, but (for most users) not as annoying as GPT-4O. However, a knowledge from the past few days is that we really need a world in which the personality of the model can be more adapted to the individual user.”
- Openai Complemented only a few days later: “The changes are subtle, but chatt should now appear more accessible. You will notice small, authentic details such as ‘good question’ or ‘great beginning’ that are not flattering. Internal tests show no increase in submissiveness compared to the previous GPT-5 personality. The changes can take up to a day, further updates will follow shortly.”
Outlook: Openai canceled a smaller variant of Chatgpt-5
With GPT-5, Openai focused primarily on companies. Above all, this decision should be economically motivated. To address both corporate customers and private users, the company could return to a sliced model.
Because the upgrade to GPT-5 shows how sensitive usage habits affect the acceptance of AI. Even small changes in tone or availability can endanger the bond.
Openai is therefore doing well to react flexibly in order not to endanger user trust and market shares. Soon Should for example A smaller variant of GPT-5 appearwhich should also be economically motivated, since the operation of large models causes enormous costs.
At the same time, Openai increases the individualization of GPT-5 by introducing four new and already announced personalities. Chatgpt should then be able to answer as a cynic, robot, listener or nerd.
Also interesting:
- First German AI factory: independence from the USA and China?
- Goes to the doctor, diagnosis: algorithm disorders
- Fewer entry-level jobs: AI hype could be expensive to get to the company
- Anyone who has AI as a friend no longer needs enemies
The article with GPT-5: Openaai worse-minded Chatgpt first appeared on Basic Thinking. Follow us too Google News and Flipboard Or subscribe to our update newsletter.
As a Tech Industry expert, I am excited about the advancements that come with GPT-5 from OpenAI. With this new version, the chatbot technology has been significantly improved, making conversations more natural and engaging. The ability to have more realistic and coherent conversations with a chatbot opens up a world of possibilities for customer service, virtual assistants, and even entertainment purposes.
However, it is important to also consider the potential ethical implications of this technology. As chatbots become more sophisticated and indistinguishable from humans, there is a risk of misinformation and manipulation. It will be crucial for companies and developers to implement safeguards and guidelines to ensure that GPT-5 is used responsibly and ethically.
Overall, I believe that GPT-5 represents a significant step forward in artificial intelligence and chatbot technology, and I am excited to see how it will be integrated into various industries in the future.
Credits