Abstract
In this essay, I consider the important role of heuristics in research on artificial intelligence and cognitive systems. After clarifying the different senses of this term, I recount how views on heuristics have changed since their introduction, leading many in the AI community to see them in a very different light than intended originally by the field’s founders. In addition, I present four claims about how heuristic methods and structures can influence high-level cognition and how these postulates differ from mainstream views on the topic. In closing, I propose some actions that cognitive systems researchers can take to redress the situation and restore heuristics, in the original meaning, to their rightful place in the computational study of intelligence.
