If it seems that the chatzipt is lying on a bit thick flattering, you are not just the person who noticed.
CEO of OpenEni, the company responsible for ChatzPT, has recently said Sam Altman X posts The latest batch made in its GPT -4O model has satisfied it with a lot of people. He also added that the company is working on personality tweets. (Publish: GEF Davis, the main CNET, in April, filed a lawsuit against the OpenAI, alleged that it had violated the GEF Davis Copyright in training and operating its AI systems.)
“GPT-4O update has made the personality very psychopant-y and annoying (though it has some very good part), and we are working on some today and some fixes this week,” he said-forgot the word psychopantic.
He also added that the “one -point” company will share what has been learned from updates. “It was interesting,” said Ultman.
In recent weeks, OpenAI has Roll out It is a handful of subtle changes in the way users react to the ability to guide the conversation, how to listen to and how to follow the instructions, to work more cooperatively and dial its emoji use.
‘Losing customers is not an option’
Changes are part of the greater effort to make the generator AI more intuitive and conversation in natural language, as it becomes a larger part of everyday life.
Recest Hayden, chief analyst of the Market Research Farm ABI Research, says companies are focusing on improving the user’s experience to steer their competitive edge, so it is understood that the personality of the chatabot is in the equation.
“A annoying experience must stop using customer and initiative and it needs to be picked to make sure that it remains near the chatboat in the market,” he said. “It remains as a market that is bleeding in cash, and losing customers is not an option, even a strong first-match facility for any company like Openai.”