 I'm an engineer, turned psychologist. And I was an undergraduate, I got interested in control systems. And I realized that the brain is the most sophisticated control system in the known universe. Just consider a simple reaching movement. You've got 37 muscles in your arm, 17 in your hand, not to mention 600 muscles in your body. Your brain needs to learn to control that. It has to deal with the real physics, the real equations of motion, of how to move the limb. And we've yet to crack the code on how we do this. Robots can be made to do very specific tasks very well, but they lack the flexibility to do common everyday tasks that we take for granted. Why do they fail so miserably? Well, they maybe lack our remarkable ability to learn and also use our semantic knowledge about the world to contextualize the task. As an engineer, you can write out nice adaptive neural network control systems that can refine action execution over time, such as making movements smoother, but it lacks a psychological aspect of motor control. How do we come up with novel cognitive strategies to solve a motor task, such as opening a wine bottle with your shoe? Useful trick tonight if you lose your corkscrew. Okay, it works. A great example comes from the sport of high jumping. At the turn of the century, people were going over the bar using the scissors technique, basically hurtling sideways. By the end of the century, everybody was going over backwards. This is Dick Fosbury doing the Fosbury flop. How did this change in strategy happen? It really led to a breakthrough in how many feet someone can clear the bar. I went to look at the world records and they actually look quite gradual. Like a nice little progression of heights that you could achieve. But if you actually look closer, you see that there's four different strategies that are used. Each time a new strategy was invented in an arrow here, it gave way to a new wave of record breaking. Just consider the last two of this era, the parallel straddle and the Fosbury flop. As you can see, they're diametrically opposed to one another. With the parallel straddle, you're going over with your chest first. With the Fosbury flop, you're going over with your back first. It wouldn't be the gradual refinement of action execution, of practicing the parallel straddle that would ever have given rise to the Fosbury flop. You need a whole new strategy. How does this come about? Well, the classic theory says that it starts off in a cognitive stage. You think about what the strategy is explicitly, and that then gives way to an associative learning process that you then slowly get better at. And much of research is focused on the latter phase. But all the action, 80% of your learning is happening in this cognitive stage. How can we get at that? In my lab, we try to do this. And while we love this study of high jumping, it's a bit unpractical. So what we have people do is make reaching movements while they hold on to robotic manipulantum and move in a virtual environment, and this robot knocks them off their path, and they have to learn these new forces that are put on their limb. And as you can see here, they can learn overall in purple, but we can actually measure two dissociable learning processes. In blue are strategies that they use in the beginning a lot, and it kind of peters out a little bit towards the end, and then a slow gradual rise, and that's something implicit that they don't know about accumulates over time. But both are kind of always operating, and they're there at the end of learning. And this is very different from the classic theory that posits a unitary process of strategies into skill. And we actually know that these are dissociable things in the brain by looking at people with neurological damage. This is an individual that has lost her hippocampus due to viral encephalitis. She cannot remember anything for more than about five minutes. So if she met you, she'd quickly forget your name, who you were, or ever having a memory of meeting you before. But she can learn new motor skills. She can learn something like drawing a figure while looking through a mirror and retain that across days. Okay, and very different from an explicit thing that she would know like your name. The problem is that she actually learns much slower than anybody else. She always does much worse. And we think this is because she can't use strategies. She can't remember them, and she can't employ them. So she's stuck with just action refinement. She can't come up with a new way to solve this trick. So it got us to think, can we actually get her to learn a motor test as a one-shot strategy that's not very apparent? This is just a little cute parlor trick that I'd like to show you called Jacob's Revenge. And the goal is to get the ball to the neck of the bottle. The problem when you try to tilt it is that dowel gets in the way. It blocks the ball. And you have to actually spin the bottle to get it to work. You use centrifugal force to move that ball down to the neck of the bottle. So what you're going to see here is a video. I've already explained to her how to do this. You don't hear the audio, but she's telling me back how to solve the task. I'm doing it for her. I'm going to then spin the bottle. She knows how to solve it. She just told me how to do it. And then we're going to have her perform it. But again, she can't remember anything more than about five minutes. She learns in that permanent present. This is just 15 minutes later after doing a different task, distracting her. And now I ask her again if she can solve this task. And she says that she can't. She doesn't know what to do. She looks at the bottle, and she tries to solve it in that direct way. Kind of tipping, trying to get the ball down. She's lost the memory of how to spin. So in this one-shot strategy task, motor skill task, she's unable to do it. She's committed at just following that kind of orange trace of doing this action refinement. She can't use strategies over time, that blue trace, that allow us to achieve high levels of performance. And an analogy back to the Fosbury flop is that she would be committed to doing the scissors technique, the direct approach to getting over that high jump bar. She'd never come up with Fosbury flop. And the strategies are an integral part of our learning process. It isn't just through practice and refinement that we can ever achieve top levels of performance. Thanks.