Human versus Machine: Artificial intelligence through the lens of human cognition

This course will explore the promise and limits of artificial intelligence (AI) through the lens of human cognition. Amid whispers of robots one day taking over the world, it is tempting to imagine that AI is (or soon will be) all-powerful. But few of us understand how AI works, which may lead us to overestimate its current (and even its future) capabilities. As it turns out, intelligence is complicated to build, and while computers outperform humans in many ways, they also fail to replicate key features of human intelligence at least for now. We will take a conceptual, non-technical approach (think: reading essays, not writing code). Drawing upon readings from philosophy of science, computer science, and cognitive psychology, we will examine the organizing principles of AI versus human intelligence, and the capabilities and limitations that follow. Computers vastly outperform humans in tasks that require large amounts of computational power (for example, solving complex mathematical equations). However, you may be surprised to learn the ways in which humans outperform computers. What is it about the human brain that allows us to understand and appreciate humor, sarcasm, and art? How do we manage to drive a car without hitting pedestrians? Is it only a matter of time before computers catch up to these abilities? Or are there differences of kind (rather than degree) that distinguish human intelligence from AI? Will robots always be constrained to the tasks that humans program them to do? Or could they, one day, take over the world? By the end of this course, you will be able to discuss the current capabilities, future potential, and fundamental limitations of AI. You may also arrive at a newfound appreciation for human intelligence, and for the power of your own brain.