Understanding AI (1/4): 10 Terms to shine in meetings

A broad field of techniques that let machines imitate human capabilities: learning, reasoning, perception, language understanding, and decision-making.

A logical set of instructions to solve a problem. In AI, algorithms analyze data and produce predictions—like “recipes” that learn from data.

A program that perceives its environment, decides, and acts to reach a goal, improving through experience (e.g., a robot avoiding obstacles).

A branch of AI where computers learn from data without being explicitly programmed for every situation. Models find patterns to predict or decide.

Supervised learning uses labeled examples (the “right answer”). Unsupervised learning discovers structure on its own (e.g., customer clustering).

A subfield of ML using multi-layer neural networks to extract complex features from images, audio, or text—powering speech recognition and many modern assistants.

Inspired by the brain: connected artificial neurons organized in layers. More layers let models learn increasingly abstract patterns.

A model is the result of learning from data. Input is what it receives; output is the prediction produced (e.g., image → “cat”).

Training: learning from large amounts of data. Inference: using what was learned to answer new questions—like studying first, then applying knowledge.