OpenAI is trying to make ChatGPT less of a charmer

If you recently noticed that ChatGPT is showing its overly appeasing personality or trying hard to respond as a human, you aren’t alone. This has been happening since OpenAI pushed a few updates to GPT-4o. Well, the good news is that OpenAI is aware of this and is working on fixes.

OpenAI will roll out fixes in batches to tone down ChatGPT’s personality

OpenAI CEO, Sam Altman, recently informed that the team is aware of ChatGPT’s “sycophant-y and annoying” personality. He further confirmed that the team is working on a fix to tone it down a little.

Today, in a post on X, Altman noted, “[The] last couple of GPT-4o updates have made the personality too sycophant-y and annoying (even though there are some very good parts of it), and we are working on fixes asap, some today and some this week. [At] some point will share our learnings from this, it’s been interesting.”

When an X user commented on his post asking whether ChatGPT could return to its old personality, the AI tech giant’s CEO replied, “Eventually, we clearly need to be able to offer multiple options.” While everyone awaits the fix, some genius Reddit users have taken this matter into their hands. 

Prompts that can tone down ChatGPT’s personality

As spotted by TechRadar, a bunch of Redditors shared different prompts to tone down the AI chatbot’s personality. The outlet found two specific prompts that can be useful to alter ChatGPT’s personality.

The first prompt is by a Reddit user, TedHoliday, which removes the flattery and unnecessary comments from ChatGPT. Please note that there is strong language used in that prompt, so proceed accordingly. You can check the prompt here.

All you have to do is write down this prompt in ChatGPT, and you should be good to go. If you don’t want ChatGPT to have any sort of personality, Reddit user MrJaxendale has an incredible prompt for you to try. 

The post OpenAI is trying to make ChatGPT less of a charmer appeared first on Android Headlines.