There are plenty of misnomers in science and mathematics. Atoms aren’t indivisible. Hubble’s constant isn’t a constant. And in 9/10 cases, X’s theorem was usually first discovered by someone else (in 4/10 cases, it was discovered by Gauss). Another bad piece of terminology, according to some: “artificial intelligence” or AI.

Given that we don’t have a good definition of human intelligence, the term “artificial intelligence” is inherently vague. Because AI sounds cool, people use the term quite liberally. Logistic regression in Excel? AI! But it’s unclear what qualifies as “intelligent enough”. As AI systems become more capable, we seem to raise the bar. Previously, calculators and spell checkers were considered artificial intelligence.

When speaking of artificial general intelligence, people usually specify exactly what they mean. For example, people might refer to powerful/strong AI, AI with expert-level science and engineering capabilities or Process for Automating Scientific and Technological Advancement (PASTA). But sometimes it wouldn’t hurt being more specific when speaking about narrow forms of AI either. Indeed, consider the following thought experiment from AI Snake Oil:

Imagine an alternate universe in which people don’t have words for different forms of transportation, only the collective noun “vehicle.” They use that word to refer to cars, buses, bikes, spacecraft, and all other ways of getting from place A to place B. Conversations in this world are confusing. There are furious debates about whether or not vehicles are “environmentally friendly,” but (even though no one realizes it) one side of the debate is talking about bikes and the other side about trucks. There is a breakthrough in rocketry, but when the media focuses on how vehicles have gotten faster, people call their car dealer (oops, vehicle dealer) to ask when faster models will be available. Meanwhile, fraudsters have capitalized on the fact that consumers don’t know what to believe when it comes to vehicle technology, so scams are rampant in the vehicles sector.

Now replace the word “vehicles” with “artificial intelligence,” and we have a pretty good description of the world we live in.

So it’s helpful having vocabulary for the different kinds of AI. For example, you might differentiate between generative, predictive and conversational AI. Another distinction is between large language models and reinforcement learning agents.

Not to be a pedant, but sometimes these distinctions matter a lot. For example, agentic AI is much more likely to pose a safety risk than non-agentic AI.

However, it’s annoying adding a long descriptor before the word “AI” in everyday conversation. Furthermore, many AI systems fall under multiple categories; for example, Claude is both a language model and a reinforcement learning agent. Finally, we need a generic term for referring to the capability of a machine to simulate intelligent behaviour, and the term “artificial intelligence” does the job.

Interestingly, this choice of term was a conscious decision. AI emerged as its own field of study after a workshop in 1956. In their workshop proposal, the organisers introduced the term “artificial intelligence”. They were also considering the name “automata studies”. Maybe more appropriate, but pretty lame.