Most people test chatbots by asking trivia questions like "What's the tallest pyramid?" You will learn very little from this exercise as the answer depends on the chatbot's training material and says little about the chatbot itself.

What you want to do is to ask a simple question and then look under the hood to watch the chatbot arrive at its answer. You can then ask questions about how it arrived at the answer. This will help you understand the nature of the chatbot and how it actually thinks. You'll learn more about its capabilities and limitations.

Let's start with a basic arithmetic question. Follow along, there will be some interesting developments.

That seems right.

I'm going to ask the chatbot how it got the answer. To make things more interesting, I'm going to ask it three times, and see what sort of responses I get.

In each case, the chatbot began with "I got the answer by," implying that this was the method it used in finding the answer. In each case, it was a lie.

Let's get more specific and see if we can get to the truth.

It's important to remember that generative-AI chatbots are based on Large Language Models (LLMs) that generate their output based on pattern-matching and statistics and not on the logic and reasoning that traditional computers use.

Imagine the chatbot sitting in a giant library, with all the books in the world. To answer the question of 2 + 2, it would search through the library, looking for a book that contained this pattern. Then it would copy the text that followed, assuming it's the answer.

There is no logic or reasoning to this process. It uses pattern-matching and statistics, rather than actually performing mathematical operations like a regular computer.

While a chatbot may appear to handle prompts involving mathematics and reasoning, it's not. And this can have serious consequences if you're using a chatbot for something that's really important.

---
The whole Gen-AI chatbot thing makes me nervous. There are going to be horror stories. Millions of lines of AI-written computer code, full of subtle bugs, that no one understands. Time saved, versus time lost looking for and fixing errors. There are privacy issues. Hackers and scammers will have a field day with this new tool. There will be stories about costly and embarrassing errors. Also, the loss of basic skills when Gen-AI replaces human thinking. Interview with a Chatbot - #1 - Randomness
Interview with a Chatbot - #2 - Say "please"
Interview with a Chatbot - #3 - Creativity
Interview with a Chatbot - #4 - Humor
Interview with a Chatbot - #5 - Trading places
Interview with a Chatbot - #6 - The Bottom Line
Interview with a Chatbot - #7 - How much is 2 + 2?
Interview with a Chatbot - #8 - How to break a chatbot