Skip to main content

A place to understand AI

Understand AI.One layer at a time.

Connect ideas, code, and practice into understanding.
Follow your curiosity. Build at your own pace.

Start learning Open learning plan

Move your cursor. Follow your curiosity.

IDEAS BECOME CLEARER IN CODE

Read it. Run it. Make it click.

A small example can unlock a big idea.
Start with the pattern. See how a model finds it.

learn.py

Python / NumPy

1import numpy as np
2
3# A small dataset. A pattern to discover.
4x = np.array([1., 2., 3., 4.])
5y = 2 * x + 1
6w, b = 0., 0.
7
8for step in range(200):
9    error = (w * x + b) - y
10    w -= 0.05 * (2 * error * x).mean()
11    b -= 0.05 * (2 * error).mean()
12
13print(f"w={w:.3f}, b={b:.3f}")

Expected output

$ python learn.py

w=2.005, b=0.986

The model approaches y = 2x + 1.

ONE EXAMPLE, THREE CONNECTIONS

01

Find the patternUnderstand the relationship in the data.

02

Follow the gradientWatch small updates reduce the error.

03

Build your intuitionConnect the equation to working code.

Explore gradient descent

BUILD YOUR KNOWLEDGE MAP

Choose your next layer.

Four paths. A connected way to learn.

Artificial intelligence

Explore reasoning, agents, and the infrastructure that runs AI.

Foundations · Agents · AI infrastructure

Machine learning

Find patterns in data. Understand how models learn and generalize.

Data · Models · Evaluation

Deep learning

Go inside neural networks, from a forward pass to modern architectures.

Networks · Training · Architecture

Reinforcement learning

Learn how agents make decisions through actions and feedback.

Environments · Policies · Rewards

You don’t have to learn it all at once.Understand an idea. Run the code. Solve a problem.

Try an exercise