In contrast, a heuristic is an approach to problem solving that may not be fully specified or may not guarantee correct or optimal results, especially in problem domains where there is no well-defined correct or optimal result. Using human characteristics as descriptors of machines in metaphorical ways was already practiced by Alan Turing with terms such as 'memory', 'search' and 'stimulus'.

More advanced algorithms can perform automated deductions (referred to as automated reasoning) and use mathematical and logical tests to divert the code execution through various routes (referred to as automated decision-making). Algorithms are used as specifications for performing calculations and data processing. In mathematics and computer science, an algorithm ( / ˈ æ l ɡ ə r ɪ ð əm/ ( listen)) is a finite sequence of rigorous instructions, typically used to solve a class of specific problems or to perform a computation.

Ada Lovelace's diagram from 'note G', the first published computer algorithm

Blank flowchart templates for word