Once released, the GPT-5, the new model artificial intelligence her Openaiit caused a wave of criticism, according to Le Monde citing BBC survey. Users complain that it is “stupid” than its predecessor and gives more colder and less natural answers.
Faced with complaints and threats to stop subsidies, Openai CEO Sam Altman announced the restoration of the GPT-4O of paid users, promising to “stabilize” the new edition.
A start away from expectations
Showed as more reliable and fast, the GPT-5 is supposed to give a sense of conversation with a specialist. However, early users report more mistakes and a “personality” that is characterized as more distant.
Sam Altman attributes these problems to the automatic model selection system incorporated into the GPT-5, which remains unjustly. The criticisms intensified after the publication of a wrong graph during the official presentation, which the head of Openai described as a “huge blunder” because of “human error”.
Emotional bond and concerns
The temporary replacement of the GPT-4O has revealed the emotional bond that some users maintain with artificial intelligence. Some say they lost “their only friend” and that they do not find the same “warmth” in talks with the GPT-5. Openai, knowing this phenomenon, has recently restricted the answers to some personal questions and urges users to take breaks during long discussions in order to reduce the risks to vulnerable people.
Sam Altman recognizes that this particular relationship between people and artificial intelligence must be better regulated: “We need to find ways to be positive,” he told the X platform.