> AI is everywhere

Learn how AI actually works.

Explore how neural networks learn through hands-on, interactive lessons. Follow the numbers, see the math, and discover how AI works. No experience required. Free for all learners.

Start Learning How AI Learns
live_preview
  1. 01forward
    propagation
  2. 02back
    propagation
  3. 03weight & bias
    update

How it works

See every connection

A neural network is built from nodes and connections. See every node's value and every connection's weight.

Trace the numbers

Follow each number as it moves through the network. Every step along the way uses elementary math.

Watch the network learn

See the network learn from its mistakes and get better at making predictions.

Three units. Three stages of AI learning.

01 Forward Propagation

Make a prediction

See how a network turns its inputs into an answer.

Start here 02 Back Propagation

Find the error

Trace the error back to what caused it.

Start here 03 Weight Update

Update the model

Watch the network change its weights to get it right.

Start here

Neural networks students will explore

A network with one input and one output
Weight Update

Delivery Time Predictor

It gets smarter with every delivery.
A network with two inputs and two outputs
Weight Update

Fruit Sorter

Apple or orange? Teach it to tell.
A network with three inputs and two outputs
Examples

Spam Filter

Help it catch the junk.
A network with three inputs and three outputs
Weight Update

Drink Identifier

Juice, coffee or soda?
A network with two inputs, a hidden layer of two, and one output
Forward Propagation

Hallway Light Switch

A puzzle only a hidden layer can crack.
A network with four inputs, three hidden layers of six, and four outputs
Back Propagation

Weather Forecaster

Your biggest network yet.

Ready to learn how AI learns?

Start Learning How AI Learns