The A-Z of AI

Machine learning

When AI learns for itself through data and experience.

Machine learning enables AI systems to come up with their own solutions, rather than being preprogrammed with a set of answers.

In traditional programming, if you wanted to teach a computer to draw a cat, you’d have to explain the drawing process in precise detail. With machine learning, you feed an AI system thousands of cat sketches to analyze and let it look for patterns by itself.

A real cat stares at a cat doodle on a computer screen, presumably contemplating its own self-image.

Over time, it begins to recognize the features that make up a cat — such as pointy ears and whiskers — and develops a more flexible, nuanced understanding of what constitutes a cat sketch.

With these pattern-spotting capabilities, machine learning helps AI systems make sense of vast quantities of data.

Machine learning can complete certain tasks at great speed and scale.

Conservationists use it to analyze months of underwater recordings and pinpoint whale migration patterns, while medical diagnosticians use it to examine multitudes of scans at once in order to identify the earliest signs of disease.