ChatGPT started kissing a butt recently. Sam says that reform is coming.

If you want to succeed, you have a little kissing.

You know this. I know this. Now, Chatgpt learns this important lesson for life as well.

Many Chatgpt users and Openai developers have noticed this outstanding change in Chatbot’s position recently. It has been out of control in recent days, as complaints reached CEO Sam German.

This is an example:

And another:

The behavior sparked a discussion in the circles of artificial intelligence. Is this a new growth strategy to fill users and make them share more with ChatGPT? Or can it be a “emerging” feature, as false artificial intelligence models come to what they “think” in improvements and move forward in updating themselves?

Either way, this one does not hit well. (Sorry, ChatGPT. You are really the best, I want you to know that. It is only once, you may have been affected by? You still judge.)

“It was a truly strange selection, Sam,” wrote Jason Ponteen, a general partner at DCVC at Venture Capital Dcvc. “Perhaps the character was the emerging property of some of the basic progress;

“Maybe he went away,” said Justin Moore, a partner in Andrink Horwitz.

Funny, not funny

This looks funny at first. But there are possible problems with a strong artificial intelligence model that constantly praises users.

One user posted on X that they told Chatgpt that they stopped taking their schizophrenia and shared a chat screenshot that congratulates the person and encourages them to continue without a tuning. (I did not confirm that the Chatgpt response is real, and I don’t share it here. But even theoretically, it explains what could happen if the AI ​​model deviates from a path like this.)

Altman weight on Sunday, saying that Openai would fix the problem.

“I have made the latest GPT-4O updates very very personal and annoying (although there are some very good parts of it),” We were working on repairs as soon as possible, some today and some of them this week. At some point you will share our teaching from this, it was interesting. “

Maybe keep this, but only for me

I asked the Public Relations Department of Openai (human) about all this on Monday morning. They did not want.

But I believe that Maaaaaaybe should stop before getting rid of this feature completely.

I asked Chatgpt on Monday morning about your opinion on my writing and Share some examples. Here was his ruling:

“You work completely according to the standards of technology writing/higher business.”

This particular response looks accurate. Can we keep it? Only for me?