What's the actual difference between AI, machine learning, and deep learning?

Think of them as nesting dolls. Artificial intelligence is the biggest doll — it's the whole idea of machines doing smart things. Machine learning is a smaller doll inside it, and deep learning is an even smaller one inside that. AI is the broad goal: getting computers to mimic human intelligence in some way. Machine learning is the method we use most often to get there. Instead of programming exact rules, you feed a machine learning system tons of examples, and it figures out the patterns on its own. Deep learning is a specific, powerful type of machine learning that uses very complex networks inspired by the brain's structure. When you hear about an AI that can describe what's in a photo, that's deep learning at work. A simpler machine learning model might just tell you if the photo is of a cat or a dog. An old-school AI, without any learning, would need a programmer to hand-code every rule about whiskers and ear shapes. The confusion is understandable since people use these terms loosely. When Apple's director of machine learning resigned recently, the news wasn't about a robot uprising — it was about a leader in the specific field of algorithms that learn from data. I've found a helpful way to keep it straight: if it's learning from data, it's machine learning. If it uses a neural network with many layers, it's deep learning. If it's just a cleverly programmed set of if-then rules, like a chess engine from the 90s, it's AI without the learning part. The Chaostron, a fascinating 1961 device, was an early attempt at a learning machine, but it was purely hardware-based and couldn't adapt its own rules — a far cry from the software that learns patterns today.