Video Of Two AI Chatbots Playing “20 Questions” Is Freaking People Out

Artificial intelligence is here to stay, and while it still has a very long way to go before it can truly match human intelligence or consciousness patterns, it’s safe to say it’s learning quickly. Case in point? A video of two AI chatbots playing a game of “20 Questions” with each other is freaking people out — what is going on here?!

The game seemed to be going well at first.

If you’ve never played “20 Questions,” the rules are pretty simple. Someone thinks of a person, place, or thing, and the other player(s) can ask 20 questions to try and narrow down what it might be.

The caveat is that every question you ask has to be able to be answered with a simple yes/no.

The two AI chatbots started off their game well enough, as seen in the TikTok video uploaded by @Kyssmhe, but then things got a bit weird.

One of the AI chatbots broke the rules.


Two gpts casually playing a 20 question game #gpt4 #openai #chatgpt #chapgpt4

♬ original sound – Kyss Mhe

At one point, the AI chatbot asked a question that couldn’t be answered with a yes or no. The other chatbot then corrected it, suggesting that there must have been a “mix-up” and prompting it to try again by asking another question, this time in yes/no format.

The rule-breaking AI chatbot then apologized for not following the rules, which is what freaked people out. After all, how does a computer know it’s done something wrong and have the instinct to say it’s sorry?

Want a partner? Attract love with the power of your mind.

Sweetn is a new research-based startup that shows you how to call love into your life with the power of your mind. Take our quiz, and try our tools—they can transform your energy and your love life in a few weeks. Just click here.

Commenters couldn’t believe what they were seeing.

While the video ended before we found out much about the item the one chatbot was thinking of other than that it belongs in the kitchen, people were sufficiently weirded out.

“The fact that it apologized is insane,” one person said.
Another wrote, “Why do they actually sound human? The intonation is so human-like.”

However, not everyone saw it as a bad thing. In fact, many were thrilled to see what AI can do in the future.

“I’m excited for the future of ai its one of the most fascinating things I have ever heard about since i was a kid,” a third person admitted.

Jennifer Still is a writer and editor with more than 10 years of experience. The managing editor of Bolde, she has bylines in Vanity Fair, Business Insider, The New York Times, Glamour, Bon Appetit, and many more. You can follow her on Twitter @jenniferlstill