24 February 2025
Artificial intelligence, or AI, is the capacity of computers to perform tasks that were previously thought impossible for machines, that is to mimic human thought processing either for specific tasks or in general. The term appears in the 1950s. More recently, the term generative artificial intelligence has come in to use to refer to AI that is “trained” on large sets of data and used to produce intelligible text and images in response to prompts by users.
As is the case with many technical concepts, artificial intelligence first appears in science fiction, only the initial appearances there were in a somewhat different sense, that of robot’s mental capabilities. Edmond Hamilton writes in his 1951 novella Moon of the Forgotten:
Grag, the towering manlike giant who bore in his metal frame the strength of an army and an artificial intelligence equal to the human, rumbled a question in his deep booming voice. But Curt Newton only vaguely heard him.
The term artificial intelligence as we know it today in the real world dates to at least 1955 when a group of computer scientists, John McCarthy, Marvin. L. Minsky, Nathaniel Rochester, and Claude E. Shannon proposed a research project to be conducted at Dartmouth College:
We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. An attempt will be made to find how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves.
By 1958, artificial intelligence was starting to appear in general audience publications. On 16 April 1958 Lansing, Michigan’s State Journal had an article on technical predictions for the future and included this prediction for the year 2000, one that is rather accurate and only off by about twenty-five years:
“And if you think some of those things are out of this world,” Mr. Davis told the writers, “your grandchildren will enjoy artificial intelligence machines which will do things people do now—write letters, do bookkeeping, translate languages, file and retrieve information, teach students individually, plan and operate factories, cook, serve meals, clean houses, drive automobiles, and fly airplanes.”
The abbreviation A.I. appears by 1963, when Marvin Lee Minsky used it to shorten the title of MIT’s Artificial Intelligence Project Memo 61.
And we see generative artificial intelligence by the turn of the twenty-first century, when it appears in a 2001 technical report published by the US Air Force Research Laboratory:
The core component of the Suggestions Module/Server is a generative artificial intelligence (AI) planner, developed by Carnegie Mellon University (CMU) called Prodigy. Prodigy [9] is a multi-strategy planning and learning architecture that can solve planning problems in a number of different ways. Simply stated, Prodigy searches for a sequence of actions that transform an initial state (the current state) into a final state (the goal state).
That’s the history of the terms. Whether or not AI has fulfilled its promises is a mixed bag. Some AIs created for specific tasks perform as well, if not better, than humans do. AIs trained to examine medical imagery for signs of cancer, for instance, have been shown to perform better than human radiologists. And AIs powering self-driving cars perform remarkably well, although they may not yet be safe enough for city streets. But the generative AIs, at best, merely produce mediocre and inaccurate text and images and show no signs of improving because they are not truly “intelligent.” They possess no actual knowledge of the real world, nor can they, as the scientists in 1955 hoped, “form abstractions and concepts.” They merely output what their training data tells them is the statistically most likely response. They can form a grammatically correct sentence, but they cannot form an insightful, original, or stylistically innovative one.
Sources:
Hamilton, Edmond. “Moon of the Forgotten.” Startling Stories, 22.3, January 1951, 118–34 at 121/1. Archive.org.
Historical Dictionary of Science Fiction, 17 November 2024, s.v. artificial intelligence, n.; 16 December 2020, s.v. AI, n.
McCarthy, John., Marvin. L. Minsky, Nathaniel Rochester, and Claude E. Shannon. “A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence” (31 August 1955). AI Magazine, 27.4, Winter 2006, 12–14 at 12. DOI: 10.1609/aimag.v27i4.1904.
Minsky, M. L. A.I. Memo 61: Mathscope Part 1. A Proposal for Mathematical Manipulation-Display System (typescript). Cambridge: Massachusetts Institute of Technology, 18 November 1963.
Mulvehill, Alice M., Clinton Hyde, and Dave Rager. Joint Assistant for Deployment and Execution (JADE). Air Force Research Laboratory, AFRL-IF-RS-TR-2001-171, August 2001, 8. Defense Technical Information Center (DTIC).
“New Cures Predicted.” State Journal (Lansing, Michigan), 16 April 1958, 64/2. ProQuest Historical Newspapers.
Oxford English Dictionary, third edition, December 2023, s.v. artificial intelligence, n., AI, n.2., generative artificial intelligence, n.
Image credit: Mike MacKenzie, 2018. Wikimedia Commons. Licensed under a Creative Commons Attribution 2.0 Generic license.