An AI navigation system automatically calculates the fastest route to a destination.

Question:
Why can a navigation algorithm repeatedly recommend a road even though this makes the road worse?

Solution

The algorithm evaluates each trip individually.

If a road currently appears slightly faster than others,

it may be recommended to many drivers at once.

But this creates additional traffic.

The road becomes slow only after many cars have already chosen it.

This is sometimes called a “self-reinforcing effect.”

The recommendation changes the reality it was trying to measure.

Result:
A navigation system can optimize routes —
but when many people follow the same recommendation,
the advice itself can create a new traffic problem.

The Navigation That Always Finds the Shortest Route

Solution