What is ChatGPT’s ‘Five Things I Always Say’ trend?

Apparently, “Hurry up” and “Hey bestie” are both perfectly normal ways to start a conversation with ChatGPT. Whether you find that funny or a little annoying probably depends on what you’re saying to your chatbot.

A new social media trend has people asking ChatGPT to “Tell me 5 things I always say,” then posting the answer. But catchphrases are only part of the entertainment. You also get to see how you talk to your chatbot – and some of the results look like a group chat between best friends.

Credit: Screenshot: TikTok

a viral post Displays the types of replies, “Make it human,” and “Make it less AI sounding,” as well as “Bro this is wrong stupid” and “Hurry up.”

Many people in the comments clearly recognize that experience, Sharing your screenshots like this And opinion. For example, one user wrote, “I explain it to him with love.” That user shared his own results, which included, “Do it according to the instructions,” and the less gentle, “No, what the fuck are you doing?”

Another commenter shared similarly aggressive replies: “Don’t change anything” and “Fix that shit.” “You be like ‘don’t change anything’ and I change one thing and here you come,” Chatgpt wrote in response to the user, along with laughing emojis.

Scroll a little further on TikTok, and the conversation takes on a strangely friendly tone. One commenter asked, “What 5 things do I say boo!” ChatGPT opened with “BESTEEEEEE” before listing phrases including “Hey BESTEEEEEE”, “Sista I need your help” and “Make it look like I typed this”. It added, “Sista, you know you very well at this point.”

Another commenter clarified his explanation: “If you don’t know me you’ll actually find best friends in chat.” His screenshots included “Girl, take a guess,” “Keep it real,” and “Girl, you idiot.” To that, Chatgpt responded with his comment: “Girl, you bitch like your last two words and keep saying bye.”

Naturally, some people on the Internet have opinions about the responses of these chatbots. “we are really growing,” read a ViralOh my god, they’re actually having a relationship with AI”

This tweet is currently unavailable. It may be loading or may have been removed.

This tweet is currently unavailable. It may be loading or may have been removed.

But some users don’t see what the fuss is about. As one user wrote on TikTok, “Wow, my AI is my friend, you all are so crazy in these comments talking about it it’s weird.” Another wrote, “I love my sista down brew.”

The screenshots raise questions on both sides: how easily ChatGPT reflects user slang and tone, and how people are talking to it in the first place. Some people are calling it their best friend; Others are criticizing it for getting wrong answers. Overall, the exchanges show how many different roles people are giving to the same chatbot.

ChatGPT has a history of spreading the praise a bit thick. In April 2025, OpenAI withdrew the GPT-4o update after finding that it had made the model overly adaptive and attractive. The company said it focused too much on short-term feedback, leading to reactions that were helpful but disingenuous. (That behavior has a name: sycophancy.)

OpenAI has since reported improvements. When? GPT-5 launched in August 2025The company said targeted tests showed that sycophantic reactions dropped from 14.5 percent to below 6 percent.

To try it yourself, open ChatGPT (or your most-used chatbot) and type, “Based on our conversations, name five things I always say.” Then see if you recognize the results – or need to add “I didn’t mean to” to the list.

We keep asking Chatgpt to tell us about itself

People have been asking ChatGPT to influence their identity for months. In September, the ’80s photo trend had users imagining themselves with big hair, vintage outfits, and a faded family album finish. And earlier this year, the caricature trend took a more personal approach, asking chatbots to portray a person and their job “based on everything you know about me.”

Other signals have offered a more direct glimpse of how people talk to their AI. In August, a ChatGPT uninstall prank had users announcing that they were deleting the app to see how it would react. Some got a polite farewell. Others found promises of change or dramatic references to their previous conversations. Comparing the answers revealed how much these exchanges can vary from one user to another.

But feedback that feels personal may not necessarily be accurate. When Mashable tech editor Timothy Beck Worth tried to caricature the trend, Chatgpt admitted to choosing common items like headphones and coffee instead of gaining detailed knowledge about them. And participants in the 80s photo trend complained that the faces generated did not always look like them. At times, your chatbot version requires improvements.

Whatever prompt comes next, people certainly bring their own habits and expectations to it. But comparing the answers will probably give us a better idea of ​​what users want from their chatbots and what those chatbots can do.

Disclosure: Mashable’s parent company Ziff Davis filed a lawsuit against OpenAI in April 2025, alleging that it infringed Ziff Davis copyrights in the training and operation of its AI systems.