AI-complete

From Lesswrongwiki
Jump to: navigation, search
Smallwikipedialogo.png
Wikipedia has an article about

A problem is considered AI-complete or AI-hard if solving it is equivalent to creating Artificial General Intelligence.

For example, natural language processing (or machine translation) is often considered AI-complete because understanding arbitrary language constructs seems to require broad general knowledge. The term was coined by the computer scientist Fanya Montalvo as an analogy with NP-complete, a class of problems in complexity theory. Problems labeled AI-complete like graceful degradation or computer vision tend to be framed at human-level intelligence; there may be many problems that AIs can solve that humans cannot. While mathematical formalizations of the class have been attempted, the term is usually used to communicate the qualitative difficulty of a problem.

Roman V. Yampolskiy proposed a classification in AI-Complete, AI-Hard and AI-Easy problems:

  • A problem C 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 C by some polynomial time algorithm.
  • A problem H is AI-Hard if and only if there is an AI-Complete problem C that is polynomial time Turing-reducible to H.
  • A problem X is AI-Easy if and only if there exists some AI problem Y such that X is polynomial-time Turing reducible to Y. This means that given an oracle for Y, there exists an algorithm that solves X in polynomial time.

References

See also