How does AI learn?
To help me tailor our conversation to the right level, could you tell me a bit about your background?
For example, are you interested in the technical engineering side of how algorithms are built, or are you looking for a high-level, conceptual overview?
Somewhere in the middle, please!
It's great to have you here! Finding that sweet spot between a high-level overview and deep-dive engineering is the best way to understand how AI actually functions without getting lost in code.
Your path
Types of Machine Learning
The Role of Neural Networks and Deep Learning
Practical Training Process
The Shift in How We Build Software
To understand how AI learns, we first have to look at how we used to build software.
In , a human engineer writes explicit, step-by-step instructions. You provide the computer with the input and the rules, and it calculates the answers. If you want a program to identify a cat, you have to manually code rules for ears, whiskers, and tail shapes.
But completely flips this equation. Instead of giving the computer rules, we give it the input (thousands of cat photos) and the expected answers (labels saying "this is a cat"). The computer then analyzes this data to figure out the rules on its own.