In the field of [[artificial intelligence]], the most difficult problems are informally known as '''AI-complete''' or '''AI-hard''', implying that the difficulty of these computational problems is equivalent to solving the central artificial intelligence problem—making computers as intelligent as people, or [[strong AI]].

The term was coined by [[Fanya Montalvo]] by analogy with [[NP-complete]] and [[NP-hard]] in [[complexity theory]], which formally describes the most famous class of difficult problems. {{Harv|Mallery|1988}} Early uses of the term are in Erik Mueller's 1987 Ph.D. dissertation and in Eric Raymond's 1991 jargon file.

To call a problem AI-complete reflects an attitude that it won't be solved by a simple algorithm, such as those used in [[ELIZA]]. Such problems are hypothesised to include:

*[[Computer vision]] (and subproblems such as [[object recognition]])
*[[Natural language understanding]] (and subproblems such as [[text mining]] and [[machine translation]])
*Dealing with unexpected circumstances while solving any real world problem, whether it's [[robotic mapping|navigation]] or [[automated planning and scheduling|planning]] or even the kind of [[reasoning]] done by [[expert system]]s.

For example, consider a straight-forward, limited and specific task: [[machine translation]]. To translate accurately, a machine must be able to [[understand]] the text. It must be able to follow the author's argument, so it must have some ability to [[artificial intelligence#Deduction, reasoning, problem solving|reason]]. It must have extensive [[commonsense knowledge|world knowledge]] so that it knows what is being discussed — it must at least be familiar with all the same commonsense facts that the average human translator knows. Some of this knowledge is in the form of facts that can be explicitly represented, but some knowledge is unconscious and closely tied to the human body: for example, the machine may need to understand how an ocean makes one ''feel'' to accurately translate a specific metaphor in the text. It must also model the authors' goals, intentions, and emotional states to accurately reproduce them in a new language. In short, the machine is required to have wide variety of human intellectual skills, including [[artificial intelligence#Deduction, reasoning, problem solving|reason]], [[commonsense knowledge]] and the intuitions that underly [[robotics|motion and manipulation]], [[machine perception|perception]], and [[artificial intelligence#Social intelligence|social intelligence]]. [[Machine translation]], therefore, is believed to be AI-complete: it may require [[strong AI]] to be done as well as humans can do it.

AI systems can solve very simple restricted versions of AI-complete problems, but never in their full generality. When AI researchers attempt to "scale up" their systems to handle more complicated, real world situations, the programs tend to become excessively [[brittle (software)|brittle]] without [[commonsense knowledge]] or a rudimentary understanding of the situation: they fail as unexpected circumstances begin to appear. When human beings are dealing with the world, they are helped immensely by the fact that they know what to expect: they know what all 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. A machine without [[strong AI]] has no other skills to fall back on. {{Harv|Lenat|Guha|1989|pp=1-5}}

==References==
* Engels, Robert & Bremdal, Bernt (2000, July 28). [http://www.ontoknowledge.org/countd/countdown.cgi?del5.pdf ''Information Extraction: State-of-the-Art Report''].
* {{Citation | last=Lenat | first=Douglas | last2=Guha | first2=R. V.| year = 1989 | title = Building Large Knowledge-Based Systems | publisher = Addison-Wesley| author-link=Douglas Lenat }}
* {{Citation| last=Mallery | first=John C. |year=1988 | url=http://citeseer.ist.psu.edu/mallery88thinking.html | contribution=Thinking About Foreign Policy: Finding an Appropriate Role for Artificially Intelligent Computers | title=The 1988 Annual Meeting of the International Studies Association. | location=St. Louis, MO }}.
* Mueller, Erik T. (1987, March). [ftp://ftp.cs.ucla.edu/tech-report/198_-reports/870017.pdf ''Daydreaming and Computation'' (Technical Report CSD-870017)] Ph.D. dissertation, University of California, Los Angeles. ("Daydreaming is but one more ''AI-complete'' problem: if we could solve any one artificial intelligence problem, we could solve all the others", p. 302)
* Raymond, Eric S. (1991, March 22). [http://catb.org/esr/jargon/oldversions/jarg282.txt Jargon File Version 2.8.1] (Definition of "AI-complete" first added to jargon file.)
* Shapiro, Stuart C. (1992). [http://www.cse.buffalo.edu/~shapiro/Papers/ai.ps Artificial Intelligence] 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".)

[[Category:Artificial intelligence]]
[[fr:IA-complet]]
[[ja:AI完全]]
[[ru:AI-полный]]