What counts as artificial intelligence?
The honest answer: a moving boundary. John McCarthy, who coined the term for the 1956 Dartmouth workshop, defined it as the science and engineering of making intelligent machines (Stanford’s McCarthy pages, read 12 October 2026). In practice the label covers rule-based systems, machine learning that finds patterns in data, deep learning that stacks neural networks, and the current center of gravity: LLMs and the generative AI built on them. Once a capability becomes routine, we quietly stop calling it AI, which is why the definition keeps moving.
What is the difference between narrow AI and general AI?
Everything deployed today is narrow: superb within a scope, helpless outside it. A model that drafts contracts cannot drive a truck. Artificial general intelligence, a system matching human flexibility across domains, remains a research goal and a forecasting debate, not a product. Every AI decision a company makes today is a decision about narrow systems, however broad their language sounds.
How do AI, machine learning and generative AI relate?
Nested families. Machine learning is the subset of AI that learns from data instead of explicit rules; deep learning is machine learning with many-layered networks; generative AI is deep learning that produces new content rather than classifications. The newest layer is action: an AI agent wraps a model in a loop that plans, calls tools and checks results, the approach covered under agentic AI. The 2026 shift is from AI that answers to AI that acts.
What should a buyer actually take from the definition?
That “AI” in a product sheet tells you almost nothing. Ask narrower questions: what does it learn from, what does it do autonomously, what happens when it is wrong, and who is accountable then. Evaluate the specific capability against your specific process, not the label against the hype, and the rest of this glossary exists to make those narrower questions precise.