On today’s show we are going to take a closer look at AI tools that are making headlines. A couple of weeks ago I put out an episode on the OpenAI framework and the software ChatGPT which uses that framework as the underlying AI engine. In that episode I gave some live examples of questions and answers that I put to the software.
In that episode, I concluded that the results were underwhelming and no threat to us humans.
It turns out that my conclusions missed the mark in that episode. Nothing I said was misleading. But where I missed the mark was by asking the software some very simple questions.
If you ask an unsophisticated question, then you are going to get an unsophisticated answer. I suppose humans would respond in the same way. Ask a stupid question and you will get a stupid answer. Ask a better question and you’re likely to get a better answer.
Host: Victor Menasce