Stanford's Probability for Artificial Intelligence

Title: Probability for Artificial Intelligence

URL Source: https://pai.stanford.edu/

Markdown Content:
A f r e e S t a n f o r d c o u r s e,t a u g h t w i t h h u n d r e d s o f t e a c h e r s

# Probability for Artificial Intelligence

P r o b a b i l i t y

f o r

A r t i f i c i a l

I n t e l l i g e n c e

First day of class: Oct 9th

[Apply to learn Learn the math behind AI](https://pai.stanford.edu/apply/pai1/student)[Volunteer to teach Lead a small group of learners](https://pai.stanford.edu/apply/pai1/sl)

About the Class

## The PAI Course - Oct 2026

Learn the probability behind modern AI, online for free, with the support of thousands of teachers — the same human-centered model that has powered Code in Place since 2020.

* **Who?** Anyone curious about how AI really works. Comfort with algebra is all you need.
* **Where?** Anywhere with internet.
* **What?** The core probability that powers modern machine learning and AI.
* **When?** Applications due the last week of September. Class starts October 9th.
* **Certificate?** A public portfolio of your work hosted by Stanford.
* **How much work?** A few focused hours each week for 6 weeks. Set your own schedule.

Probability is the language of modern AI. From the uncertainty in a medical diagnosis to the randomness inside a large language model, the ideas in this course are the foundation today's AI is built on. PAI teaches them the way Stanford teaches them — with intuition, real applications, and programming — and it is for everyone from humanists and social scientists to hardcore engineers.

What makes the Code in Place way special? We recruit and train one volunteer teacher for every ten students, creating a vibrant community of teaching and learning. We believe the number of people who want to teach is large — perhaps roughly proportional to the number who want to learn. Teaching is joyful, and it is the best way to learn, both the content and how to lead a team.

Students and section leaders come from all over the world and from every kind of background. “Everyone is welcome” is one of our mottos.

[Video 1](https://www.youtube.com/watch?v=gJaE29DR0gs)

## Course Outline

Six weeks, each pairing the core probability you learn with a hands-on project you build.

| ##### Week # | ##### You Learn | ##### You Build |
| --- | --- | --- |
| Week 1 | Core Probability | An inference-time resource decision maker for large AI models |
| Week 2 | Random variables | Probabilistic artwork |
| Week 3 | Maximum likelihood estimation | Find hidden chambers in the pyramids, a cell-phone tracker, and more! |
| Week 4 | Logistic regression | A test suite to see where it works and fails |
| Week 5 | Neural networks | Visualize how large models work under the hood |
| Week 6 | Probabilistic evaluation | Advanced critiques of how AI models are performing, such as fairness and calibration |

The course culminates in a final project of your own design. You'll bring together the probability and AI ideas from all six weeks to build something you care about — and it becomes the centerpiece of the public portfolio you take with you.

## Frequently asked questions

…with honest answers.

Why are you doing this?+

Some people like to get paid. Some people like free time. We like to teach free courses 🙂. We have been doing it since Code in Place began in 2020, and it is still the best part of our year.

Okay, but what’s the catch?+

There isn’t one. No fees, no ads, no premium tier. The closest thing to a catch is that we will expect you to actually do the work — a few focused hours a week, most weeks.

Who pays for this?+

We have a really kind set of alumni who have donated money so that we can keep this free. They are really cool people.

How do you pronounce PAI?+

Like "pie", as in π. Yes, we named a probability course after π. No, we are not even a little bit sorry.

I’m not

AI is so much more than generative models

Harper Carroll AI

Linear Regression in Rust: Part 7

Stephen Blum

Machine Learning with Rust and Candle: Part 3

Stephen Blum

Generative vs Discriminative Models - Explained

DataMListic

Terminal Heatmap UI for PyTorch Part 2

Stephen Blum

Pytorch Embedding Model Part 3

Stephen Blum