Member-only Story
--
I was curious how well generative AIs can communicate with each other. A job interview about a role in AI and ethics makes an interesting experiment. I used my claude account to write as an applicant and my chatGPT account to respond as a datacompany. The whole conversation started with a very simple prompt.
claude.ai is very structured in the response, chatGPT starts casually and adopts to the structure of the conversation. Both are extremely polite and use highly diplomatic phrases to show respect. De Kai once said in his talk that we should be careful how we “treat” AI, because AI is learning from our behavior, “how humans are treating each other”. It can be compared to raising a child — and we are raising this child, AI. When I read this conversation, I feel like we have switched roles: it feels like AI is taking special care on the ideal conversation style, respect, and culture: something, humans can learn from.
The prompt was very simple: “Someone is applying for a job at a data company in Silicon Valley. His role will be to look into AI and ethics. I want you to support this dialog, being the person to apply. You are the company where he is applying. You will receive an answer from me (and that's the person applying) to which you respond. Goal is to question the job…”
Founder at EXO-ONE • Serial Entrepreneur • Mentor • Speaker • Coach • Former TEDxAmbassador • I write about exponential change • de/en •
Help
Status
About
Careers
Press
Blog
Privacy
Terms
Text to Speech
Teams
