A factory can use a vision model to flag possible surface defects in photographs. A planning system can search for a sequence of actions that satisfies production constraints. Both are examples of artificial intelligence, though they perform different tasks and use different methods.
Machine learning learns patterns from data and is one part of the field. Generative AI produces outputs such as text or images. AI also includes approaches that reason over written rules or search a defined space of possibilities. People sometimes call those older approaches traditional AI; rule-based expert systems are a classic example. The label does not specify one architecture or level of autonomy.
This matters when assessing a product claim. Successful defect detection does not establish conversational ability or artificial general intelligence. Ask what inputs the system receives, what decisions it makes and how its outcomes are evaluated. “AI-powered” alone does not tell a team whether the system meets its task requirements.
Sources
- Poole and Mackworth: What is artificial intelligence? — Frames AI around computational agents, intelligent behavior, tasks and environments.
Go deeper
- UC Berkeley CS188: Introduction to artificial intelligence course
Explore search, planning, uncertainty and learning as distinct parts of the field.