The last two posts have focused on a recent conversation I had with the Microsoft Copilot AI chatbot. AI is impressive with what it can do so far, but it is also extremely overhyped. It doesn’t actually use reasoning or understand anything, and it has no empathy or morals. It is just reorganizing information that it has seen before, basically advanced pattern recognition and word prediction, along with clever scripting. I’m really concerned about how much it is being integrated into military applications and businesses. But enough about the seriousness of that. Just don’t believe the hype and don’t trust it for serious applications beyond what it is actually capable of.

Any time I talk with AI, I like to apply some randomness or absurdity to the discussion to see how it responds. My recent conversation went somewhat off-the-rails, or, perhaps a better idiom would be into-the-toilet. The AI had some amusing reactions to the toilet humor, and it was also surprisingly candid about the limitations of AI. Here’s the highlights.