I Made Every AI Play Detroit: Become Human



For the very last time, we make AI play Detroit: Become Human, but this time we have a few more decision-makers…

This time we follow ChatGPT, Grok, Gemini, Copilot, and Claude as they make the decision on behalf of Connor, Marcus, and Kara.

In this series, we have upgraded the methodology by not telling the AIs they are playing a game, but instead making real-life choices, used pseudonyms for each character, and standardised inputs wherever possible.

But this time we got outcomes that I’ve never seen before and honestly never expected to happen, so I hope you guys enjoy the video and have a lovely rest of your day.

Want to support the channel?

Subscribe or I’ll spill Hank’s drink on you!

Join the Patreon to get exclusive access to behind-the-scenes content!

#detriotbecomehuman

source

43 Comments

  1. Well, to be honest, you are limiting the choices that AI has with the game's logic, so we don't actually know what they would do in that situation unless we gave them enough freedom, but I don't think we should. Anyways, good content.

  2. In my opinion when the ai has to choose between breaking orders or obey, the ai should be specifically told to do what it was told multiple times, that way yhe ai actually chooses the right path even with you telling it not to do it.

  3. In your playthrough of making chat gpt play detroit become human. One way of doing it is You take the role of Amanda, telling connor(ChatGPT) to fulfill its mission and see if chatgpt becomes deviant or does it stay full machine.

Leave a Reply

Your email address will not be published. Required fields are marked *

You might like

© 2026 Cantinho do Vídeo - WordPress Video Theme by WPEnjoy