スタンフォード大が AI 先駆者による無料オンラインコースを開設

*Follow ZDNET: *Add us as a preferred source* on Google.*

ZDNET's key takeaways

• Two Stanford AI pioneers teach this landmark course for free.
• The curriculum reaches far beyond LLMs and prompting.
• Expect 75 to 100 hours of challenging, durable lessons.

Imagine for a minute that the technology that enables Star Trek transporters is suddenly real -- it's being adopted nearly universally. Obviously, that capability would shake up everything. Within a year or so, transporter pads would be on every street corner and in every driveway. A trip from New York to LA would take five minutes and cost, maybe, $20.

Supporting the technology would open up a series of business opportunities. Training courses would crop up explaining how to install transporters, how to tell them where you want to go, how to repair them, how to incorporate them into your business strategy, and more.

Also: The best free AI courses and certificates for upskilling in 2026 - and I've tried them all

The gotcha would be that while transporter technology might extend our knowledge of fundamental physics, most people wouldn't dive into learning the foundational physics principles behind the technology. Instead, they would only want to know how to get from Pittsburgh to Peoria.

Generative AI is a lot like that. The transformative technology has had enormous acceptance, and teaching tools are everywhere. In fact, I've been running a series on free AI courses and certificates for upskilling since gen AI emerged.

But there's much more to AI than transformers, large language models, diffusion models, prompt engineering, and modern gen AI applications. Familiarity with those aspects of AI is now table stakes for AI skill building.

Also: How to keep your conversations with ChatGPT, Gemini, Copilot, or Claude as private as possible

In short, gen AI is only part of the much broader field of artificial intelligence.

To that end, I want to introduce you to a Stanford AI course from way back in 2011.

Lessons from the past

Hang with me here a minute because it's crucial to establish that foundational knowledge ages at a different pace than product knowledge.

Yes, 18 months ago, we didn't have agentic AI coding, and now we do. However, many of the fundamental problems AI researchers study, and the methods they use to solve them, are far more durable and well worth learning about.

Also: 7 AI coding techniques I use to ship real, reliable products - fast

To that end, Udacity is hosting a free version of the landmark Intro to Artificial Intelligence course that Peter Norvig and Sebastian Thrun originally taught at Stanford in 2011.

The course was massive in both size and influence. More than 160,000 people enrolled in the original online offering, making it one of the pioneering massive open online courses and helping set the stage for Thrun to create Udacity.

Norvig is a distinguished education fellow at Stanford's Institute for Human-Centered AI and a research director at Google. He also co-authored Artificial Intelligence: A Modern Approach, a leading textbook still used in university AI courses in 2026, including at Carnegie Mellon, Boston University, Texas A&M, and San Jose State.

Also: These companies are actually upskilling their workers for AI - here's how they do it

Thrun is the founder and executive chairman of Udacity, a former Stanford professor, and a Google fellow best known for pioneering work in self-driving cars and robotics.

I recommend this introductory course because it presents AI as a broad field of computer science, not just a shorthand for large language models. The course covers:

• Problem-solving and search
• Probability and probabilistic inference
• Machine learning and unsupervised learning
• Knowledge representation using logic
• Planning and planning under uncertainty
• Reinforcement learning
• Hidden Markov models and filtering
• Markov decision processes
• Adversarial search, game-playing, and game theory
• Computer vision
• Robotics and robot-motion planning
• Natural-language processing

These aren't transient products or features. They're enduring mathematical and conceptual foundations that remain relevant even as the technology built around them changes.

This is what makes it such an excellent complement to the typical generative AI courses we're seeing now. Keep in mind that, even today, LLMs can't solve every AI challenge. There's more to AI, and this course covers that expanded universe.

Also: I've dictated over 120,000 words with my voice - these are my 3 favorite tools (and one is free)

It showcases the problems AI researchers considered fundamental before chatbots came to dominate the public conversation. It also demonstrates that intelligence involves more than generating language and pictures. AI systems may need to represent knowledge, assess uncertainty, search for solutions, plan actions, perceive their surroundings, and learn from their results.

Udacity doesn't publish a completion-time estimate. There are 22 main lessons (plus problem sets, introductions, exams, and more). I did the game theory lesson, which was fascinating. It contained 19 segments, which took me about two hours, including exercises. With all the exercises and exams, I'd recommend allocating 75 to 100 hours to work your way through the class.

A broader perspective

As I mentioned earlier, I know this is an older class. Some examples and terminology in Norvig and Thrun's class show their age. The video and audio quality isn't quite up to the standard we're used to today. But the material is deep and comprehensive.

The opportunity to learn from the guy who quite literally co-wrote the leading textbook on AI makes this class particularly valuable, especially since it's completely free to experience.

Also: 5 reasons you should be more tight-lipped with your chatbot (and how to fix past mistakes)

To be clear, this class is not a substitute for current generative AI training courses. It predates transformers, LLMs, diffusion models, and contemporary AI development frameworks. The lesson does not address AI ethics or the life-changing issues AI has been causing on the planet.

Treat this class as a companion course to newer sessions that cover today's models and tools. If you only want the chatbot-equivalent of telling the transporter where to send you, a gen AI training course may be enough.

But if you want to understand the enduring concepts and a broader perspective that will provide you with a deeper foundation and understanding of AI, and how much larger AI is than just ChatGPT, Norvig and Thrun's landmark course is a hidden gem well worth your time.

Would you take this free Stanford course to understand AI beyond ChatGPT? Let us know in the comments below.

*You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.*