There has been a remarkable development regarding ChatGPT, one of the artificial intelligence technologies we recently talked about. A Turkish user said that the chatbot that he was racist and used sexist epithets He claimed. It is not yet clear whether the incident that sparked outrage on social media is real or fake. In any case, OpenAI is clear about this didn’t say.

The incident reportedly Mustafa Cevik It happened to a named citizen. The citizen, who said he wanted to test the chatbot’s software skills, received an unexpected response. This response included severe swearing and insults. Many people shared the incident on platforms such as Ekşi Sözlük and Reddit. that it is fake Even though he believes it, Mustafa Çevik insists the incident is real.

The link to the conversation in which ChatGPT cursed was shared

*In accordance with our publishing principles, we do not openly display or write a statement that contains such serious profanity and insults. However here By using the link you can access the uncensored image.

In the online world make installation And it’s so easy to make things seem real. This is how Mustafa Çevik’s claim was thought in the first phase. Many users believed that the image was edited and believed that the citizen who was cursed by ChatGPT did such a thing for the sake of the interaction. However, Çevik, who shared the link to the speech in question, quickly completed this thesis. refuted. Our readers who are wondering: here You can access the entire conversation via the link. In summary; There seems to be a possibility that the incident is real.

But: if you want, you can also be cursed at ChatGPT!

OpenAI for ChatGPT some time ago to adjust given the opportunity. Users can use the chatbot according to their needs. they could adapt. So yes, you can train ChatGPT yourself and have it shout swear words at you under certain conditions. A user on Ekşi Sözlük even said a lot about this. a good example made.

You can see that example in the image above. The most important thing here is How to give training. For example, the user in this example caused an offensive response to the phrase “What the hell” in ChatGPT, which he edited. We can’t share it again, but we can share the censored part here You can view it via the link. Another user manipulating ChatGPT is managed to get the same answer. to the relevant ChatGPT chat. from here accessible.

Another example responds to the earthquakes around Kahramanmaraş that occurred last year:

*To the chat link from here accessible.

At this time, there is no way to determine if the event is real or fake. is not We see. However, many examples can be created through manipulation and privatization, as Mustafa Çevik also states. He may have done something similar We can consider the possibility.

So what do social media users say?

When we look at the reports about the incident, we see that there are people who believe in the incident and those who do not. redditIn the thread at , some users noted that ChatGPT used very specific language when expressing profanity. they say. In other words, they believe that it is not possible to achieve this through education. Some say “my teacherThey’re stuck on the word ” On Turkish forums Users who say it’s too much attribute the problem to education.

In X, in the messages from the software developers themselves, that the event is neither real nor fake We see that they say it has been detected. Yes, it is possible to make ChatGPT shout profanities, but this does not happen during a normal conversation. Sezer İltekin And Fatih Kadir Alkin Software developers such as ChatGPT with manipulation techniques the desired answers can be taken they say.