 I've always wondered why it was so hard for me to learn to dance. Yet, I can learn new sports pretty readily. Was it all about strengthening my muscles? Or was there something about my brain that made one skill easier to learn than another? So I sought to figure this out. If we could better understand this, we could perhaps help people learn more quickly and to a higher level of proficiency. For example, a child learning math, a stroke patient learning to walk again, or me learning to dance like Fred Astaire. It is well documented that the brain changes during learning. With practice, the neurons that make up our brains can change the way in which they are wired up, and this in turn changes the patterns of activity that are produced by the brain. However, it is not well understood how these changes support our improvements in sports, dancing, or other skills. Consider the simple task of moving a computer mouse with your hand, an everyday skill that we often take for granted. Your brain sends signals down the spinal cord to move your hand. For thousands of patients worldwide, this pathway is cut off, and as a result, these patients are no longer able to move their hand. So to help these patients, what we can do is to feed their brain activity into a computer, and the computer can interpret whether they would like to move the cursor to the left or to the right. In other words, the person can move the cursor just by thinking this so-called brain-machine interface is being actively tested in human patients. But that's not what I'm going to be talking about today. I'm going to talk about how moving a cursor with the brain can teach us about learning to dance, to play sports, or any other skill in everyday life. So to build a brain-machine interface, we implant a grid of electrodes that are smaller than your fingernail. We can get the detailed activity of tens of neurons, and we feed this brain activity into a computer. We have built the system such that when the person thinks of moving left, the cursor moves to the left. Similarly, when the person thinks of moving right, the cursor moves to the right. And when the person gets good at controlling the cursor with their thoughts, we can then make it harder. So now, when the person thinks of moving left, the cursor moves instead to the right. And with practice, we find that the user gets better at controlling the cursor. In other words, the person is learning to perform this new task. It turns out that something as simple as moving a cursor with your thoughts can serve as a platform for studying learning. In particular, my colleagues and I have asked what types of manipulations are easier or harder for the brain to learn. What you find easier or harder to learn depends on how your brain is wired up, which depends on previous experience. In other words, perhaps the reason why it's easier for me to learn to play tennis than to learn to dance is because my brain is already wired up to play basketball, which shares some common traits with tennis. Our findings apply not only to physical skills, but also to cognitive skills, such as learning math or learning a foreign language. Our work is likely to have two major areas of impact. First, our work on brain machine interfaces and our work to understand how the brain works is likely to lead to better treatment options for patients around the world. Secondly, our work on learning may lead to new ways to teach our children new skills, whether it be learning a new sport or learning math. And maybe one day I can learn to dance like Fred Astaire. The question I would like to leave the audience with is, what can a brain and machine do together that neither could do alone?