What's the difference between Artificial Intelligence and Machine Learning?

You're no doubt familiar with the concept of modeling already, even if you don't recognize its meaning.

For example, suppose I asked you to estimate how long it would take you to read this article. If you're like me, you might have looked at the size of the scrollbar on the right-hand side of the browser: a small scrollbar means lots of scrolling to reach the bottom, so it's probably a long article. A long scrollbar means less scrolling and a shorter article.

That process of taking one piece of information (the length of the scrollbar) and transforming it into another piece of information (how long it will take to read the article) is called modeling. You're probably making those kinds of estimates every day even if you've never called it "modeling" before.

So why do we use so many different words if they all mean the same thing? We do it for the same reason we say "steak" instead of saying "cow." Saying "Artificial Intelligence" just straight up sounds better than "model."

I think some people will find that last statement unfair: after all, there are many different types of models. For example, you might not have used the length of the scrollbar to estimate how long this article is, and instead you could have counted the number of words in the article before making at a prediction.

Surely counting the number of words would produce a much better prediction, and to emphasize the higher quality of that prediction, it deserves a more appealing name. Perhaps we should call counting the number of words a "quantitative type of model" and glancing at the scrollbar a "qualitative type of model."

Personally, I wouldn't disagree with that, but my goal isn't to tell you which is the best word to choose when describing some type of model. I'm only trying to help you make sense of the words that other people use, so that you can better evaluate the value of what they're claiming.

In fact, I have found it so helpful to substitute the words "AI" and "Machine Learning" with the word "model" that I have started to do it without even realizing it. Somehow just changing the language can make it easier to see when the emperor has no clothes.