• ChatGPT users are asking the AI to reveal any personal traits it's gleaned from their past conversations.
  • The AI's responses to the prompt have been likened to horoscopes, with some finding them insightful.
  • OpenAI's latest model, o1-preview, aims for more human-like thinking and reasoning.

ChatGPT is known for being quick to offer up answers. Now, some users of the popular chatbot are using a specific prompt to see if the AI can reveal something about themselves that they didn't already know.

Tom Morgan, founder of networking company The Leading Edge, posted on X encouraging people to ask ChatGPT: "From all of our interactions what is one thing that you can tell me about myself that I may not know about myself."

Morgan's social media post gained traction, receiving over 4 million views, 4,000 reposts, and nearly 2,000 replies.

The ChatGPT responses posted by people online revealed long, detailed analyses that pointed out traits from users' "thoughtful approach to conversations" or ability to "blend curiosity with analysis."

The in-depth, though sometimes generic-sounding summaries led many people to liken it to horoscopes or astrology readings for techies.

While some said they appreciated the little ego boost of its more flattering responses, others said they were a bit creeped out by the AI's accuracy. One person commented, "This is eerie actually," with a screenshot of ChatGPT's intricate character examination.

Another revealed their chatbot's answer, which talked about the person's ability to "thrive on duality — constantly juggling between structure and spontaneity." Despite calling it "trippy," the user admitted that it "rings true."

The question, which has spread to other platforms, including Threads and Reddit, received the seal of approval from OpenAI CEO Sam Altman, who simply wrote, "love it."

Some ChatGPT users suggested follow-up questions for the AI that might result in a less complimentary response. One person on X suggested asking ChatGPT to give them "the real talk," and another on Reddit, for a similar prompt, suggested telling the bot to "play devil's advocate."

The viral prompt suggests that many people are getting comfortable talking to AI and asking it personal questions. Between OpenAI's rollout of Advanced Voice Mode to paid users and its latest o1-preview AI model "designed to spend more time thinking" before responding, those conversations are sounding more human-like too.

Read the original article on Business Insider