A rules-based symbolic AI program developed by Newell, Simon, and Shaw in 1957. Intended to handle a wide variety of problems and synthesise the problem-solving process. It could solve certain well-defined logic puzzles but fell short of real-world problems.