News

The goal of AI is to create intelligent machines, thus prompting the question: what is intelligence? Unfortunately, intelligence is a concept that has resisted simple definition. According to the psychologist Robert Sternberg, there are almost as many definitions of intelligence as there are experts who have tried to define it. In lieu of attempting to provide a single, unified account, I discuss a number of aspects of the concept of intelligence that have played important roles in the history of AI. These aspects include concepts such as goals, computation, learning, embodiment, and abstraction. This is mostly a descriptive talk, but I do hope to persuade the audience of two modest claims. First, that intelligence is in fact multi-faceted in an interesting way (or, equivalently, any unified account will leave something important out). Second, that this fact is relevant for the engineering challenge of building intelligent machines.