Cognitive Science: An Introduction/Kinds of Intelligences

Cognitive science is about intelligent agents in general, not just human ones. As such it's useful to think about what kinds of agents there are in the broadest possible terms. Philosopher Daniel Dennett offers a useful classification scheme for this, with each class named after a relevant scientist.[1]

Darwinian CreaturesEdit

Some agents are created (or born) with every competency they will ever have. They cannot learn. We will call these "Darwinian creatures."

Skinnerian CreaturesEdit

In addition to the competencies they are created/born with, Skinnerian creatures (named after the behaviourist and learning scientist B.F. Skinner) can learn. You can see the chapter on learning for more information about what learning is, but for purposes of this discussion you can think of learning in terms of reinforcement learning (as its known in AI) or operant conditioning (as it's known in psychology): making behaviours more or less likely according to feedback from the environment.

Popperian CreaturesEdit

Popperian creatures have all the skills of Skinnerian creatures, but can also consider hypothetical situations and make decisions based on the outcome in their imagination. For example, you might imagine that hot sauce on ice cream would taste bad, so you choose not to eat it. You learned, in this case, from a simulation in your imagination, rather than directly from experience with the environment. This class is named after Karl Popper, a philosopher of science.

Gregorian CreaturesEdit

The most complex intelligence class (that we know of so far), is that of the Gregorian creatures, who have various thinking tools, from math to capitalism to computer programming. These tools are shared culturally, and in artifacts outside of the creatures themselves. This class of creatures is named after Richard Gregory, a British psychologist.

  1. Dennett, D. C. (2017). From bacteria to Bach and back: The evolution of minds. New York: WW Norton & Company.