General Problem Solver

The General Problem Solver (GPS ) is a technology developed by Herbert Simon and Allen Newell from 1957 software for realizing a general problem-solving method in the context of the then incipient research on artificial intelligence. The software was described by Simon and Newell in the article GPS, A Program That Simulates Human Thought. In this experiment, which was finally viewed as a failure, flowed both cognitive scientific aspects as well as methods for mathematical formalization of problems and problem solving. The software and the developed theoretical framework had a significant influence on the development of cognitive psychology and artificial intelligence. The failure of the general problem-solving approach eventually led to the development of expert systems that should go to a narrower field of knowledge to better results.


GPS based on the principle of problem reduction. In this case, a problem is decomposed into sub-problems that can be solved by dissolving the obtained individual sub-problems. This method is compared with the so-called problem transformation, the problem is transformed into another problem whose solution is already known or can be found more easily. Due to the use of problem reduction, the GPS the group of state-based problem solver can assign: A problem here is formalized through a space of possible states between which transitions are brought about by means of operations. The use of targeted operations in the event of GPS just to problem reduction.

A number of authors have pointed out later that this approach was generally not, as had been supposed in the initial euphoria and due to the overbearing name. This was the comment about McDermott in a famous article Artificial Intelligence Meets Natural Stupidity? not without irony that the program probably instead of GPS better LFGNS (Local Feature Guided Network Searcher ) should have said. In fact, GPS could only be applied to well-defined problems such as proving theorems of logic and simple geometry, word puzzles or chess.


Simon and Newell give an example of the solution of the logical problem of the transformation of L1 = R * ( P ⇒ Q) into L2 = (Q \ / P) * R ( Newell & Simon, 1972, p 420). This example provides an insight into the operation of the GPS:

  • Goal 1: Transform L1 into L0
  • Goal 2: Reduce difference in between L1 and L0
  • Goal 3: Apply R1 to L1
  • Goal 4: Transform into L1 condition (R1 )
  • Produce L2 ( P ⇒ Q) * R
  • Goal 5: Transform L2 into L0
  • Goal 6: Reduce difference in between left ( L2) and left ( L0)
  • Goal 7: Apply R5 to left (L2 )
  • Goal 8: Transform left ( L2) into condition (R5 )
  • Goal 9: Reduce difference in between left ( L2) and condition (R5 )
  • Rejected: No Easier than Goal 6
  • Goal 10: Apply R6 to left (L2 )
  • Goal 11: Transform left ( L2) into condition (R5 )
  • Produce L3: (P \ / Q) * R
  • Goal 12: Transform into L3 L0
  • Goal 13: Reduce difference in between left ( L3) and left ( L0)
  • Goal 14: Apply R1 to left (L3 )
  • Goal 15: Transform left ( L3) into condition (R1 )
  • Produce L4: (Q \ / P) * R
  • Goal 16: Transform into L0 L4
  • Identical, QED