À PROPOS DE CET ÉPISODE
In this episode I interview ChatGPT, the powerful and easy to use AI chatbot that has changed the global discussion about the roles of human and artificial intelligence. We talk about its potential impact on the legal system.
I'll be interested in your reactions. My takeaways were that ChatGPT:
- Was unfailingly polite and well-organized;
- Seemed to have a high-level "understanding" of a lot of topics
- Was at times pedantic and evasive;
- Wasn't great with detail, at one point making a mistake about the case and when Roe v. Wade was overruled. It acknowledged its error and apologized for it, though, when pointed out.
Technical note: I did not the substance of any response by ChatGPT. I did delete occasional redundant paragraphs and made one small revision to a numbered list to help the speech software. The voice of ChatGPT is provided by the text-to-speech function in Microsoft Word, which may not be the most sophisticated voice AI program out there but was enough to get the job done.
