I had high school curriculum night last night. In the geometry class the teacher referenced a quote from Euclid (the ancient Greek mathematician): "There is no royal road to geometry."

The story goes that Ptolemy I, king of Egypt, was working through Euclid's Elements and asked if there was a shorter way. Euclid's reply became famous. Kings had roads built just for them, smooth and fast and closed to everyone else. Geometry had no such road. Even a king had to walk it.

It made me think of AI.

AI is the royal road everyone is hoping for. Type a prompt, get an essay. Type another, get a strategy. It feels like a shortcut that skips the work and lands you at the answer.

Here is what a language model does. It predicts the most likely next word, over and over, based on everything it has read. Left to its defaults, it gives you the middle of the distribution. The average. The most common take on your question, written fluently.

Average is useful. It is also what everyone else gets when they type the same question. If a thousand people ask for a market overview, a thousand people receive roughly the same overview, nicely polished.

Differentiated insight does not come from average.

A better essay, a sharper investment thesis, a more creative idea will not come from AI. The person who wrestled with the problem builds true understanding.

Each step rests on the one before it.

So if AI looks like the "royal road" how do you use it in a way that builds true understanding?

Do these things:

1. Think first. Write your own answer before you open the chat. Ten minutes, rough, ugly, whatever it takes. This is the struggle, and it is the part that makes the idea yours.
2. Ask for the average on purpose. "What is the most common answer to this question?" Now you know what everyone else will say. Go somewhere else.
3. Make it attack you. Paste your draft and ask for the three strongest objections. You do the rebuilding. If you're doing creative work, ask for the edge cases or the least probable answers.

In all three, you do the thinking and AI sharpens it. Order matters. Start with the model and you inherit its average. Start with your own mind and the model becomes a sparring partner.

Tools can be wonderful guides. A guide who has walked the route before can save you from wrong turns. But you still have to do the walking. The walk is where the learning comes.

So here is my question for you this week: what is one problem you will struggle with for ten minutes before you ask AI?

Alex