As a former kid, I remember challenging the idea of learning math, due to the calculator existing. The counter argument was always, “you’re not always going to have a calculator in your pocket”. It turns out that was dead wrong.

However, I’m still glad I learned math and wish I had learned even more.

A calculator is useless without knowing what to type in, as well as having a rough idea of what kind of answer to expect incase it’s typed in wrong. The counter argument against, “but I have a calculator”, isn’t, “you won’t always have a calculator”. The real counter argument is that knowing how to do the math enables the use of the calculator, and the more math a person knows, the more useful that calculator becomes. A mathematician or physicist can do things with a calculator that I don’t even know are possible.

When a person doesn’t know what they don’t know, even AI can become nerfed. To get the most out of the AI, they need to know that what is being asked is possible, how to ask it in an intelligent way, and how to understand and make use of the result. All of this requires a base of foundational knowledge. The larger that base, the further the AI can be pushed while also maintaining understanding and control.