The short answer

Artificial intelligence is the broad goal — systems that solve problems typically associated with human reasoning. Machine learning is the main technique used to reach that goal today. All modern machine learning is AI. Not all AI is machine learning.

AI

AI is the umbrella. It includes:

  • Rule-based systems (expert systems, symbolic AI)
  • Search and planning
  • Machine learning
  • Neural networks and deep learning
  • Generative models

Historically, AI meant rules written by hand. That works only when the rules are known and stable — chess openings, tax computation, grammar checking. It breaks when the problem is fuzzy.

Machine learning

Machine learning takes a different approach: give the system many examples of inputs and correct outputs and let it discover the mapping. Three broad kinds:

  • Supervised learning — labelled examples. Spam vs not-spam, image class, sales forecast.
  • Unsupervised learning — no labels. Grouping customers, detecting anomalies, learning representations.
  • Reinforcement learning — an agent learns by acting and receiving rewards. Used in games, robotics and some recommendation systems.

Deep learning and modern AI

Deep learning is a subset of machine learning using large neural networks. It dominates image, audio and language tasks, and it is what powers ChatGPT, Claude and Gemini. When people say "AI" today, they usually mean systems built with deep learning.

Why the distinction matters

  • If you want to build models, learn machine learning fundamentals — data, features, evaluation.
  • If you want to apply AI to real problems, focus on describing tasks well and choosing the right tool.
  • Job descriptions often use "AI" and "ML" loosely. Read the actual responsibilities.

Where to go next