Elements of AI is one of those rare free courses that genuinely earned its reputation. It was built by the University of Helsinki together with MinnaLearn, and it set out to demystify artificial intelligence for ordinary people rather than to train engineers. That goal shapes everything about it. The first course, Introduction to AI, needs no maths and no coding at all, and it walks you through what AI actually means, how machines make decisions, what a neural network is in principle, and where the real limits and risks sit.
The writing is calm and clear, the quizzes make you think rather than just recall, and the worked examples are well chosen. The second course, Building AI, goes a little deeper and introduces some light Python, but even there the emphasis stays on ideas over implementation. What I like most is the honesty of it. It does not oversell what AI can do, and it spends real time on ethics, bias and the things these systems get wrong, which is more than a lot of paid courses manage.
The catch is timing. Much of the material was written before the generative AI and large language model boom, so if you came specifically to understand ChatGPT or to build with modern tools, you will find this course pointed at an earlier era of the field. It is still correct and still valuable for the fundamentals, but it is showing its age in places. Treat it as the conceptual grounding, then go and learn the current tools somewhere more recent.