AI Fundamentals
What is intelligence — and what happens when you build it into a machine? This course starts at the beginning. Not with code or tools, but with the question of what intelligence actually is: the ability to recognise differences, similarities, and identities, and to perceive, pose, and resolve problems. Humans do this naturally. AI does it computationally. Understanding the connection between the two is the foundation for everything else. From there, you'll learn how machines learn from data — the algorithms, datasets, and patterns that make it work. You'll see the three main approaches to machine learning (supervised, unsupervised, and reinforcement learning) and understand why each exists. You'll discover how neural networks and deep learning triggered the AI revolution we're living through. You'll explore how AI processes the world — understanding language through natural language processing and interpreting images through computer vision. Then you'll meet the technology behind the current boom: generative AI, large language models, and the chatbots that have put AI in everyone's hands. The course closes where it must: with the real-world applications of AI, the biases it can inherit, and the ethical responsibility that comes with building systems that make decisions affecting people's lives. No prior technical knowledge is needed. Just curiosity about how the most transformative technology of our time actually works.
Section 1: What Is Intelligence?
Every intelligent act — whether performed by a human, an animal, or a machine — comes back to three operations: recognising what's different, what's similar, and what's the same. Intelligence is also the ability to perceive that a problem exists and work toward solving it. These aren't abstract philosophical ideas — they're the practical foundation that makes AI possible. Before you can understand artificial intelligence, you need a clear picture of what intelligence itself means.
Section 2: How Machines Learn
Traditional software follows rules that humans write. Machine learning inverts this — you show the system examples and it discovers the rules itself. The quality of the data, the choice of algorithm, and the ability to find meaningful patterns in noise are what determine whether an AI system works well or fails. Understanding this machinery is understanding how modern AI is actually built.
Section 3: Three Ways Machines Learn
Not all learning looks the same. Sometimes you have labelled examples and a clear right answer. Sometimes you have raw data and no labels at all. Sometimes the only way to learn is through trial and error. Machine learning has three main approaches — supervised, unsupervised, and reinforcement learning — and understanding when to use each one is as important as understanding how they work.
Section 4: The Deep Learning Revolution
The ideas behind neural networks have existed for decades. What changed around 2012 was a convergence: massive datasets from the internet, powerful GPUs that could process them, and algorithmic breakthroughs that made deep networks trainable. The result was a revolution — AI systems that could suddenly see, hear, read, and generate at levels that surprised even their creators. Understanding neural networks and deep learning means understanding why AI is where it is today.
Section 5: AI That Understands the World
Language and vision are two of the richest ways humans experience and communicate about the world. Natural language processing gives machines the ability to read, write, translate, and converse. Computer vision gives them the ability to interpret images, recognise objects, and navigate physical spaces. Together, they represent AI's growing ability to process the world as humans do — through words and pictures.
Section 6: AI That Creates
Until recently, AI was mostly about analysis — classifying, predicting, recognising. Generative AI changed that. Now AI systems can write essays, generate images, compose music, and produce code that works. Large language models predict the next word in a sequence, and this simple mechanism — applied at massive scale — produces systems of remarkable capability. But these systems also have a fundamental flaw: they can generate confident, plausible content that is completely wrong.
Section 7: AI in the Real World
AI isn't just chatbots and research papers. It's the technology behind the recommendations you see, the searches you run, the fraud detection on your bank account, and the quality control in the factories that make your products. AI is reshaping every industry — from healthcare to agriculture, finance to education. Understanding where AI is already at work helps you see both the opportunities and the limits of the technology.
Section 8: Building AI Responsibly
The power to build systems that make decisions at scale comes with responsibility. AI systems learn from human data — and human data contains human biases. Without care, AI can scale discrimination faster than any human institution. Fairness, transparency, accountability, and privacy aren't constraints on AI — they're design requirements. This section confronts the risks, the principles, and the choices that determine whether AI serves everyone or just some.