In the field of artificial intelligence (AI), tasks that are hypothesized to require artificial general intelligence to solve are informally known as AI-complete or AI-hard.[1] Calling a problem AI-complete reflects the belief that it cannot be solved by a simple specific algorithm.
Expert systems, that were popular in the 1980s, were able to solve very simple and/or restricted versions of AI-complete problems, but never in their full generality. When AI researchers attempted to "scale up" their systems to handle more complicated, real-world situations, the programs tended to become excessively brittle without commonsense knowledge or a rudimentary understanding of the situation: they would fail as unexpected circumstances outside of its original problem context would begin to appear. When human beings are dealing with new situations in the world, they are helped by their awareness of the general context: they know what the things around them are, why they are there, what they are likely to do and so on. They can recognize unusual situations and adjust accordingly. Expert systems lacked this adaptability and were brittle when facing new situations.[8]
DeepMind published a work in May 2022 in which they trained a single model to do several things at the same time. The model, named Gato, can "play Atari, caption images, chat, stack blocks with a real robot arm and much more, deciding based on its context whether to output text, joint torques, button presses, or other tokens."[9] Similarly, some tasks once considered to be AI-complete, like machine translation,[10] are among the capabilities of large language models.[11]
AI-complete problems
AI-complete problems have been hypothesized to include:
Computational complexity theory deals with the relative computational difficulty of computable functions. By definition, it does not cover problems whose solution is unknown or has not been characterized formally. Since many AI problems have no formalization yet, conventional complexity theory does not enable to formally define AI-completeness.
Research
Roman Yampolskiy[20]
suggests that a problem is AI-Complete if it has two properties:
It is in the set of AI problems (Human Oracle-solvable).
Any AI problem can be converted into by some polynomial time algorithm.
On the other hand, a problem is AI-Hard if and only if there is an AI-Complete problem that is polynomial time Turing-reducible to . This also gives as a consequence the existence of AI-Easy problems, that are solvable in polynomial time by a deterministic Turing machine with an oracle for some problem.
Yampolskiy[21] has also hypothesized that the Turing Test is a defining feature of AI-completeness.
Groppe and Jain[22] classify problems which require artificial general intelligence to reach human-level machine performance as AI-complete, while only restricted versions of AI-complete problems can be solved by the current AI systems. For Šekrst,[23] getting a polynomial solution to AI-complete problems would not necessarily be equal to solving the issue of artificial general intelligence, while emphasizing the lack of computational complexity research being the limiting factor towards achieving artificial general intelligence.
For Kwee-Bintoro and Velez,[24] solving AI-complete problems would have strong repercussions on the society.
^Shapiro, Stuart C. (1992). Artificial IntelligenceArchived 2016-02-01 at the Wayback Machine In Stuart C. Shapiro (Ed.), Encyclopedia of Artificial Intelligence (Second Edition, pp. 54–57). New York: John Wiley. (Section 4 is on "AI-Complete Tasks".)
^Mueller, Erik T. (1987, March). Daydreaming and Computation (Technical Report CSD-870017)Archived 2020-10-30 at the Wayback Machine PhD dissertation, University of California, Los Angeles. ("Daydreaming is but one more AI-complete problem: if we could solve anyone artificial intelligence problem, we could solve all the others", p. 302)
^Yampolskiy, Roman (2013), "Turing Test as a Defining Feature of AI-Completeness", Artificial Intelligence, Evolutionary Computing and Metaheuristics, Studies in Computational Intelligence, vol. 427, pp. 3–17, doi:10.1007/978-3-642-29694-9_1, ISBN978-3-642-29693-2
^Groppe, Sven; Jain, Sarika (2024), "The Way Forward with AI-Complete Problems", New Generation Computing, 42: 1–5, doi:10.1007/s00354-024-00251-8
^Bintoro, Ted; Velez, Noah (2022), "AI-Complete: What it Means to Be Human in an Increasingly Computerized World", Bridging Human Intelligence and Artificial Intelligence, Educational Communications and Technology: Issues and Innovations, Cham: Springer, pp. 257–274, doi:10.1007/978-3-030-84729-6_18, ISBN978-3-030-84728-9