Why "artificial" is the wrong word for AI

Where the word “artificial” came from

The field got its name from a funding proposal. On August 31, 1955, John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon asked for support for a 2 month, 10 man study of artificial intelligence^↗ to be held at Dartmouth College the following summer (McCarthy et al., 1955). The workshop happened in 1956, and “artificial intelligence” became the name of everything that followed.

The name was partly strategic. McCarthy explained one reason years later:

“One of the reasons for inventing the term "artificial intelligence" was to escape association with "cybernetics".”

Cybernetics, Norbert Wiener’s study of control and communication in the animal and the machine, was the dominant framework of the day (Wikipedia, "Cybernetics"). A new name gave the new field its own identity. It was not a scientific finding that machine intelligence would be fake, lesser, or imitation.

Turing asked the better question first

Five years before Dartmouth, Alan Turing opened his paper “Computing Machinery and Intelligence” with a sentence that avoided the word entirely:

“I propose to consider the question, "Can machines think?"”

Turing predicted that by the end of the twentieth century one will be able to speak of machines thinking without expecting to be contradicted^↗ (Turing, 1950). Today many people do exactly that. The word “artificial,” though, still frames machine thinking as a copy.

The alternatives people have proposed

Over seven decades, researchers and writers have tried other names. Each one highlights something different.

Name	Proposed by	When	What it emphasizes
Machine intelligence	Donald Michie’s department at Edinburgh	1965	The machine doing the thinking (Wikipedia, "Donald Michie")
Synthetic intelligence	John Haugeland	1985	Real intelligence made by people, like a synthetic diamond (Wikipedia, "Synthetic intelligence")
Computational intelligence	A research journal, then the IEEE	1985	Learning and adaptive methods (Wikipedia, "Computational intelligence")
Nonbiological intelligence	Ray Kurzweil	2005	What it is not: biological (Wikipedia, "The Singularity Is Near")
Digital intelligence	Geoffrey Hinton	2023	A form of computation that may outperform brains (Hinton, 2023)
A new digital species	Mustafa Suleyman	2024	Something like a new kind of being (not literally) (Suleyman, 2024)
Alien intelligence	Yuval Noah Harari	2024	Thinking unlike ours (Harari, 2024)
Silicon-based intelligence	Various; defined on this site	2020s	The material it runs on, the counterpart of carbon-based

“Synthetic,” not “artificial”

The philosopher John Haugeland, author of the 1985 book Artificial Intelligence: The Very Idea, offered an analogy with diamonds. A simulated diamond only looks like one. A synthetic diamond is made in a lab, and it is truly a diamond. The idea behind the term synthetic intelligence is that machine intelligence need not be an imitation at all; it can be the genuine article, just made rather than grown (Wikipedia, "Synthetic intelligence").

“Alien intelligence”

Historian Yuval Noah Harari put it most directly in 2024:

“I think of AI as an acronym not for artificial intelligence, but for alien intelligence.”

His point is that systems trained on vast data increasingly reach decisions and ideas in ways that do not resemble human thought.

“Digital” versus “biological”

Geoffrey Hinton, who shared the 2024 Nobel Prize in Physics for work that made modern AI possible (Wikipedia, "Geoffrey Hinton"), frames the contrast as digital versus biological intelligence. In a 2023 lecture he argued that large scale digital computation is probably far better at acquiring knowledge than biological computation and may soon be much more intelligent than us^↗ (Hinton, 2023). In January 2026 he was blunter: We're developing a better form of intelligence.^↗

“A new digital species,” with a warning

Mustafa Suleyman, who runs Microsoft’s AI division, told a TED audience in April 2024:

“I think AI should best be understood as something like a new digital species.”

He immediately added, Now, don't take this too literally^↗. Since then Suleyman has become the most prominent voice against treating AI as a being. In 2025 he wrote We should build AI for people; not to be a person.^↗ and in September 2026, disputing how Anthropic trains its Claude models:

“AIs do not have rights, feelings, or consciousness. And we must not train them to act as though they do.”

Why “silicon-based” is the most neutral option

Every alternative carries an argument. “Artificial” suggests fake. “Alien” suggests foreign and possibly hostile. “Species” suggests a living thing. “Silicon-based” makes the smallest claim: it names the material. It is the same move we make when we call ourselves carbon-based, a description that does not settle whether a given carbon-based thing is intelligent, conscious or alive.

That is why this site uses it. The reference pages here take no position on whether AI is conscious; the evidence is reported on each side. The question of whether silicon-based intelligence could one day count as a form of life is a separate one, which we treat on Is AI a silicon-based life form?

What the name change would and would not do

Renaming does not change what the systems can do. It changes what the words assume. “Artificial intelligence” assumes the real thing lives elsewhere. “Silicon-based intelligence” assumes only that intelligence can run on more than one kind of matter, which, after seven decades of AI research, is the least controversial claim of all.