WEBVTT 00:00:00.577 --> 00:00:03.052 I know this is going to sound strange, 00:00:03.052 --> 00:00:06.346 but I think robots can inspire us 00:00:06.346 --> 00:00:08.734 to be better humans. 00:00:08.734 --> 00:00:12.363 See, I grew up in Bethlehem, Pennsylvania, 00:00:12.363 --> 00:00:14.684 the home of Bethlehem Steel. 00:00:14.684 --> 00:00:16.763 My father was an engineer, 00:00:16.763 --> 00:00:19.614 and when I was growing up, he would teach me 00:00:19.614 --> 00:00:20.773 how things worked. 00:00:20.773 --> 00:00:23.588 We would build projects together, 00:00:23.588 --> 00:00:26.380 like model rockets and slot cars. 00:00:26.380 --> 00:00:29.664 Here's the go-kart that we built together. 00:00:29.664 --> 00:00:31.883 That's me behind the wheel, 00:00:31.883 --> 00:00:35.702 with my sister and my best friend at the time, 00:00:35.702 --> 00:00:37.825 and one day, 00:00:37.825 --> 00:00:40.932 he came home, when I was about 10 years old, 00:00:40.932 --> 00:00:43.236 and at the dinner table, he announced 00:00:43.236 --> 00:00:49.674 that for our next project, we were going to build a robot. NOTE Paragraph 00:00:49.674 --> 00:00:50.783 A robot. 00:00:50.783 --> 00:00:53.035 Now, I was thrilled about this, 00:00:53.035 --> 00:00:54.625 because at school, 00:00:54.625 --> 00:00:56.700 there was a bully named Kevin, 00:00:56.700 --> 00:00:58.946 and he was picking on me 00:00:58.946 --> 00:01:01.360 because I was the only Jewish kid in class. 00:01:01.360 --> 00:01:04.173 So I couldn't wait to get started to work on this 00:01:04.173 --> 00:01:07.995 so I could introduce Kevin to my robot. (Laughter) 00:01:07.995 --> 00:01:19.058 (Robot noises) NOTE Paragraph 00:01:19.058 --> 00:01:23.933 But that wasn't the kind of robot my dad had in mind. 00:01:23.933 --> 00:01:27.661 See, he owned a chromium plating company, 00:01:27.661 --> 00:01:29.837 and they had to move 00:01:29.837 --> 00:01:32.933 heavy steel parts between tanks of chemicals, 00:01:32.933 --> 00:01:36.723 and so he needed an industrial robot like this 00:01:36.723 --> 00:01:39.822 that could basically do the heavy lifting. NOTE Paragraph 00:01:39.822 --> 00:01:43.727 But my dad didn't get the kind of robot he wanted, either. 00:01:43.727 --> 00:01:46.324 He and I worked on it for several years, 00:01:46.324 --> 00:01:48.302 but it was the 1970s, 00:01:48.302 --> 00:01:50.828 and the technology that was available to amateurs 00:01:50.828 --> 00:01:53.485 just wasn't there yet. 00:01:53.485 --> 00:01:57.230 So Dad continued to do this kind of work by hand, 00:01:57.230 --> 00:01:59.505 and a few years later, 00:01:59.505 --> 00:02:03.644 he was diagnosed with cancer. NOTE Paragraph 00:02:03.644 --> 00:02:06.949 You see, what the robot we were trying to build 00:02:06.949 --> 00:02:10.216 was telling him was not about doing the heavy lifting. 00:02:10.216 --> 00:02:14.524 It was a warning about his exposure to the toxic chemicals. 00:02:14.524 --> 00:02:17.501 He didn't recognize that at the time, 00:02:17.501 --> 00:02:19.640 and he contracted leukemia, 00:02:19.640 --> 00:02:22.862 and he died at the age of 45. 00:02:22.862 --> 00:02:25.703 I was devastated by this, 00:02:25.703 --> 00:02:30.278 and I never forgot the robot that he and I tried to build. 00:02:30.278 --> 00:02:34.589 When I was in college, I decided to study engineering, like him. 00:02:34.589 --> 00:02:39.535 And I went to Carnegie Mellon, and I earned my PhD in robotics. 00:02:39.535 --> 00:02:42.629 I've been studying robots ever since. NOTE Paragraph 00:02:42.629 --> 00:02:44.267 So what I'd like to tell you about 00:02:44.267 --> 00:02:46.981 are four robot projects 00:02:46.981 --> 00:02:54.025 and how they've inspired me to be a better human. 00:02:54.025 --> 00:02:59.774 By 1993, I was a young professor at USC, 00:02:59.774 --> 00:03:02.811 and I was just building up my own robotics lab, 00:03:02.811 --> 00:03:06.396 and this was the year that the World Wide Web came out. 00:03:06.396 --> 00:03:07.572 And I remember my students were the ones 00:03:07.572 --> 00:03:09.276 who told me about it, 00:03:09.276 --> 00:03:11.895 and we would -- we were just amazed. 00:03:11.895 --> 00:03:15.329 We started playing with this, and that afternoon, 00:03:15.329 --> 00:03:19.182 we realized that we could use this new, universal interface 00:03:19.182 --> 00:03:21.725 to allow anyone in the world 00:03:21.725 --> 00:03:24.773 to operate the robot in our lab. NOTE Paragraph 00:03:24.773 --> 00:03:30.481 So, rather than have it fight or do industrial work, 00:03:30.481 --> 00:03:32.967 we decided to build a planter, 00:03:32.967 --> 00:03:34.943 put the robot into the center of it, 00:03:34.943 --> 00:03:37.296 and we called it the Telegarden. 00:03:37.296 --> 00:03:40.899 And we had put a camera in the gripper of the hand 00:03:40.899 --> 00:03:43.608 of the robot, and we wrote some special scripts 00:03:43.608 --> 00:03:46.769 and software so that anyone in the world could come in 00:03:46.769 --> 00:03:49.098 and by clicking on the screen 00:03:49.098 --> 00:03:51.271 they could move the robot around 00:03:51.271 --> 00:03:53.587 and visit the garden. 00:03:53.587 --> 00:03:57.207 But we also allowed, set up some other software 00:03:57.207 --> 00:04:00.626 that lets you participate and help us water the garden 00:04:00.626 --> 00:04:03.671 remotely, and if you water it a few times, 00:04:03.671 --> 00:04:07.256 we'd give you your own seed to plant. NOTE Paragraph 00:04:07.256 --> 00:04:10.527 Now, this was a project, an engineering project, 00:04:10.527 --> 00:04:13.303 and we published some papers on the design, 00:04:13.303 --> 00:04:15.552 the system design of it, but we also thought of it 00:04:15.552 --> 00:04:18.638 as an art installation. 00:04:18.638 --> 00:04:20.811 It was invited, after the first year, 00:04:20.811 --> 00:04:23.855 by the Ars Electronica Museum in Austria 00:04:23.855 --> 00:04:26.855 to have it installed in their lobby, 00:04:26.855 --> 00:04:29.273 and I'm happy to say it remained online there, 00:04:29.273 --> 00:04:34.256 24 hours a day, for almost nine years. 00:04:34.256 --> 00:04:38.055 That robot was operated by more people 00:04:38.055 --> 00:04:41.141 than any other robot in history. NOTE Paragraph 00:04:41.141 --> 00:04:42.695 Now, one day, 00:04:42.695 --> 00:04:45.012 I got a call out of the blue 00:04:45.012 --> 00:04:47.027 from a student, 00:04:47.027 --> 00:04:51.697 who asked a very simple but profound question. 00:04:51.697 --> 00:04:56.235 He said, "Is the robot real?" 00:04:56.235 --> 00:04:59.000 Now, everyone else had assumed it was, 00:04:59.000 --> 00:05:01.331 and we knew it was because we were working with it. 00:05:01.331 --> 00:05:02.871 But I knew what he meant, 00:05:02.871 --> 00:05:05.464 because it would be possible to take a bunch of pictures 00:05:05.464 --> 00:05:09.825 of flowers in a garden and then, basically, index them 00:05:09.825 --> 00:05:12.054 in a computer system such that it would appear 00:05:12.054 --> 00:05:15.128 that there was a real robot when there wasn't. 00:05:15.128 --> 00:05:16.369 And the more I thought about it, I couldn't think 00:05:16.369 --> 00:05:19.920 of a good answer for how he could tell the difference. NOTE Paragraph 00:05:19.920 --> 00:05:22.807 This was right about the time that I was offered a position 00:05:22.807 --> 00:05:24.880 here at Berkeley, 00:05:24.880 --> 00:05:28.432 and when I got here, I looked up Hubert Dreyfus, 00:05:28.432 --> 00:05:32.097 who's a world-renowned professor of philosophy, 00:05:32.097 --> 00:05:34.471 and I talked with him about this, and he said, 00:05:34.471 --> 00:05:38.031 "This is one of the oldest and most central problems 00:05:38.031 --> 00:05:41.633 in philosophy. It goes back to the Skeptics, 00:05:41.633 --> 00:05:43.593 and up through Descartes. 00:05:43.593 --> 00:05:46.763 It's the issue of epistemology, 00:05:46.763 --> 00:05:50.513 the study of how do we know that something is true." NOTE Paragraph 00:05:50.513 --> 00:05:52.736 So he and I started working together, 00:05:52.736 --> 00:05:55.777 and we coined a new term: telepistemology, 00:05:55.777 --> 00:05:59.303 the study of knowledge at a distance. 00:05:59.303 --> 00:06:02.055 We invited leading artists, engineers, 00:06:02.055 --> 00:06:05.119 and philosophers to write essays about this, 00:06:05.119 --> 00:06:07.459 and the results, the results are collected in this book 00:06:07.459 --> 00:06:10.080 from MIT Press. 00:06:10.080 --> 00:06:12.412 So thanks to this student who questioned 00:06:12.412 --> 00:06:15.215 what everyone else had assumed to be true, 00:06:15.215 --> 00:06:19.423 this project taught me an important lesson about life, 00:06:19.423 --> 00:06:23.496 which is to always question assumptions. NOTE Paragraph 00:06:23.496 --> 00:06:26.144 Now, the second project I'll tell you about 00:06:26.144 --> 00:06:28.023 grew out of the Telegarden. 00:06:28.023 --> 00:06:30.599 As it was operating, my students and I were very interested 00:06:30.599 --> 00:06:33.361 in how people were interacting with each other 00:06:33.361 --> 00:06:34.982 and what they were doing with the garden. 00:06:34.982 --> 00:06:37.244 So we started thinking, what if the robot could leave 00:06:37.244 --> 00:06:39.327 the garden and go out into some other 00:06:39.327 --> 00:06:41.319 interesting environment? 00:06:41.319 --> 00:06:43.498 Like, for example, what if it could go to a dinner party 00:06:43.498 --> 00:06:48.696 at the White House? (Laughter) 00:06:48.696 --> 00:06:51.537 So, because we were interested more in the system design 00:06:51.537 --> 00:06:54.823 and the user interface than in the hardware, 00:06:54.823 --> 00:06:56.857 we decided that, rather than have 00:06:56.857 --> 00:07:00.535 a robot replace the human to go to the party, 00:07:00.535 --> 00:07:03.378 we'd have a human replace the robot. 00:07:03.378 --> 00:07:05.978 We called it the Tele-Actor. NOTE Paragraph 00:07:05.978 --> 00:07:08.010 We got a human, 00:07:08.010 --> 00:07:11.017 someone who's very outgoing and gregarious, 00:07:11.017 --> 00:07:14.159 and she was outfitted with a helmet 00:07:14.159 --> 00:07:16.729 with various equipment, cameras and microphones, 00:07:16.729 --> 00:07:20.469 and then a backpack with wireless Internet connection, 00:07:20.469 --> 00:07:23.828 and the idea was that she could go into a remote and 00:07:23.828 --> 00:07:27.643 interesting environment, and then over the Internet, 00:07:27.643 --> 00:07:30.794 people could experience what she was experiencing, 00:07:30.794 --> 00:07:33.731 so they could see what she was seeing, 00:07:33.731 --> 00:07:37.038 but then, more importantly, they could participate 00:07:37.038 --> 00:07:39.986 by interacting with each other 00:07:39.986 --> 00:07:43.691 and coming up with ideas about what she should do next 00:07:43.691 --> 00:07:45.907 and where she should go, 00:07:45.907 --> 00:07:49.069 and then conveying those to the Tele-Actor. 00:07:49.069 --> 00:07:51.514 So we got a chance to take the Tele-Actor 00:07:51.514 --> 00:07:55.129 to the Webby Awards in San Francisco, 00:07:55.129 --> 00:07:59.240 and that year, Sam Donaldson was the host. 00:07:59.240 --> 00:08:02.882 Just before the curtain went up, I had about 30 seconds 00:08:02.882 --> 00:08:07.340 to explain to Mr. Donaldson what we were gonna do, 00:08:07.340 --> 00:08:09.204 and I said, "The Tele-Actor 00:08:09.204 --> 00:08:11.660 is going to be joining you on stage, 00:08:11.660 --> 00:08:13.858 and this is a new experimental project, 00:08:13.858 --> 00:08:16.483 and people are watching her on their screens, 00:08:16.483 --> 00:08:18.788 and she's got -- there's cameras involved and there's 00:08:18.788 --> 00:08:21.699 microphones and she's got an earbud in her ear, 00:08:21.699 --> 00:08:23.146 and people over the network are giving her advice 00:08:23.146 --> 00:08:24.514 about what to do next." 00:08:24.514 --> 00:08:27.723 And he said, "Wait a second, 00:08:27.723 --> 00:08:34.098 that's what I do." (Laughter) NOTE Paragraph 00:08:34.098 --> 00:08:36.029 So he loved the concept, 00:08:36.029 --> 00:08:38.369 and when the Tele-Actor walked onstage, 00:08:38.369 --> 00:08:41.190 she walked right up to him, and she gave him a big kiss 00:08:41.190 --> 00:08:44.296 right on the lips. (Laughter) 00:08:44.296 --> 00:08:45.427 We were totally surprised. 00:08:45.427 --> 00:08:47.444 We had no idea that would happen. 00:08:47.444 --> 00:08:50.136 And he was great. He just gave her a big hug in return, 00:08:50.136 --> 00:08:51.905 and it worked out great. 00:08:51.905 --> 00:08:53.969 But that night, as we were packing up, 00:08:53.969 --> 00:08:57.577 I asked the Tele-Actor, how did the Tele-Directors 00:08:57.577 --> 00:09:03.135 decide that they would give a kiss to Sam Donaldson? 00:09:03.135 --> 00:09:05.347 And she said they hadn't. 00:09:05.347 --> 00:09:07.781 She said, when she was just about to walk on stage, 00:09:07.781 --> 00:09:10.093 the Tele-Directors were still trying to agree on what to do, 00:09:10.093 --> 00:09:12.500 and so she just walked on stage and did 00:09:12.500 --> 00:09:18.022 what felt most natural. (Laughter) 00:09:18.022 --> 00:09:21.692 So, the success of the Tele-Actor that night 00:09:21.692 --> 00:09:26.065 was due to the fact that she was a wonderful actor. 00:09:26.065 --> 00:09:28.422 She knew when to trust her instincts, 00:09:28.422 --> 00:09:32.286 and so that project taught me another lesson about life, 00:09:32.286 --> 00:09:38.665 which is that, when in doubt, improvise. (Laughter) NOTE Paragraph 00:09:38.665 --> 00:09:41.745 Now, the third project grew out of 00:09:41.745 --> 00:09:46.638 my experience when my father was in the hospital. 00:09:46.638 --> 00:09:48.878 He was undergoing a treatment, 00:09:48.878 --> 00:09:52.511 chemotherapy treatments, and there's a related treatment 00:09:52.511 --> 00:09:57.559 called brachytherapy, where tiny, radioactive seeds 00:09:57.559 --> 00:10:01.770 are placed into the body to treat cancerous tumors. 00:10:01.770 --> 00:10:03.867 And the way it's done, as you can see here, 00:10:03.867 --> 00:10:08.258 is that surgeons insert needles into the body 00:10:08.258 --> 00:10:10.850 to deliver the seeds, and all this, 00:10:10.850 --> 00:10:14.283 all these needles are inserted in parallel, 00:10:14.283 --> 00:10:17.258 so it's very common that some of the needles 00:10:17.258 --> 00:10:22.075 penetrate sensitive organs, and as a result, 00:10:22.075 --> 00:10:27.138 the needles damage these organs, cause damage 00:10:27.138 --> 00:10:30.625 which leads to trauma and side effects. 00:10:30.625 --> 00:10:33.125 So my students and I wondered, what if we could 00:10:33.125 --> 00:10:36.610 modify the system 00:10:36.610 --> 00:10:40.395 so that the needles could come in at different angles? NOTE Paragraph 00:10:40.395 --> 00:10:43.172 So we simulated this, and we developed some 00:10:43.172 --> 00:10:45.761 optimization algorithms and we simulated this, 00:10:45.761 --> 00:10:48.038 and we were able to show that we are able to avoid 00:10:48.038 --> 00:10:51.805 the delicate organs and yet still achieve the coverage 00:10:51.805 --> 00:10:55.313 of the tumors with the radiation. NOTE Paragraph 00:10:55.313 --> 00:10:58.811 So now, we're working with doctors at UCSF 00:10:58.811 --> 00:11:01.507 and engineers at Johns Hopkins 00:11:01.507 --> 00:11:05.089 and we're building a robot that has a number of, 00:11:05.089 --> 00:11:08.346 it's a specialized design with different joints that can allow 00:11:08.346 --> 00:11:12.604 the needles to come in at an infinite variety of angles, 00:11:12.604 --> 00:11:15.770 and as you can see here, they can avoid delicate organs 00:11:15.770 --> 00:11:19.908 and still reach the targets they're aiming for. 00:11:19.908 --> 00:11:23.225 So, by questioning this assumption that all the needles 00:11:23.225 --> 00:11:26.170 have to be parallel, this project also taught me 00:11:26.170 --> 00:11:29.470 an important lesson: When in doubt -- 00:11:29.470 --> 00:11:33.837 When your path is blocked, pivot. NOTE Paragraph 00:11:33.837 --> 00:11:38.238 And the last project also has to do with medical robotics. 00:11:38.238 --> 00:11:42.278 And this is something that's grown out of a system called 00:11:42.278 --> 00:11:45.866 the da Vinci surgical robot, 00:11:45.866 --> 00:11:48.334 and this is a commercially available device. 00:11:48.334 --> 00:11:51.666 It's being used in over 2,000 hospitals around the world, 00:11:51.666 --> 00:11:54.194 and the idea is it allows the surgeon 00:11:54.194 --> 00:11:58.443 to operate comfortably in his own coordinate frame, 00:11:58.443 --> 00:12:03.434 but many of the subtasks in surgery 00:12:03.434 --> 00:12:06.481 are very routine and tedious, like suturing, 00:12:06.481 --> 00:12:08.846 and currently, all of these are performed 00:12:08.846 --> 00:12:13.266 under the specific and immediate control of the surgeon, 00:12:13.266 --> 00:12:15.924 so the surgeon becomes fatigued over time. 00:12:15.924 --> 00:12:17.219 And we've been wondering, 00:12:17.219 --> 00:12:19.484 what if we could program the robot 00:12:19.484 --> 00:12:21.971 to perform some of these subtasks, 00:12:21.971 --> 00:12:23.691 and thereby free the surgeons to focus 00:12:23.691 --> 00:12:26.347 on the more complicated parts of the surgery, 00:12:26.347 --> 00:12:29.595 and also cut down on the time that the surgery would take 00:12:29.595 --> 00:12:32.618 if we could get the robot to do them a little bit faster? NOTE Paragraph 00:12:32.618 --> 00:12:35.052 Now, it's hard to program a robot to do delicate things 00:12:35.052 --> 00:12:39.131 like this, but it turns out my colleague, Pieter Abbeel, 00:12:39.131 --> 00:12:41.547 who's here at Berkeley, has develeloped 00:12:41.547 --> 00:12:47.170 a new set of techniques for teaching robots from example. 00:12:47.170 --> 00:12:49.937 So he's gotten robots to fly helicopters, 00:12:49.937 --> 00:12:53.041 do incredibly interesting, beautiful acrobatics, 00:12:53.041 --> 00:12:55.760 by watching human experts fly them. 00:12:55.760 --> 00:12:58.348 So we got one of these robots. 00:12:58.348 --> 00:13:00.530 We started working with Pieter and his students, 00:13:00.530 --> 00:13:03.193 and we asked a surgeon to perform 00:13:03.193 --> 00:13:07.644 a task, and what we do is we, with the robot, 00:13:07.644 --> 00:13:09.707 so what we're doing is asking the robot, 00:13:09.707 --> 00:13:10.985 the surgeon to perform the task, 00:13:10.985 --> 00:13:13.257 and we record the motions of the robot. NOTE Paragraph 00:13:13.257 --> 00:13:15.385 So here's an example. I'll use a figure eight, 00:13:15.385 --> 00:13:17.625 tracing out a figure eight as an example. 00:13:17.625 --> 00:13:21.255 So here's what it looks like when the robot, 00:13:21.255 --> 00:13:23.572 this is what the robot's path looks like, 00:13:23.572 --> 00:13:24.746 those three examples. 00:13:24.746 --> 00:13:27.208 Now, those are much better than what a novice 00:13:27.208 --> 00:13:31.865 like I could do, but they're still jerky and imprecise. NOTE Paragraph 00:13:31.865 --> 00:13:33.937 So we record all these examples, the data, 00:13:33.937 --> 00:13:37.649 and then we go through a sequence of steps. 00:13:37.649 --> 00:13:41.281 First, we used a technique called dynamic time warping 00:13:41.281 --> 00:13:43.463 from speech recognition, and this allows us to 00:13:43.463 --> 00:13:46.303 temporally align all of the examples, 00:13:46.303 --> 00:13:49.232 and then we apply Kalman filtering, 00:13:49.232 --> 00:13:52.215 a technique from control theory, that allows us 00:13:52.215 --> 00:13:54.887 to statistically analyze all the noise 00:13:54.887 --> 00:14:01.070 and extract the desired trajectory that underlies them. 00:14:01.070 --> 00:14:03.064 Now, so what we're doing is that we take those 00:14:03.064 --> 00:14:05.087 human demonstrations, they're all noisy and imperfect, 00:14:05.087 --> 00:14:08.178 and we extract from them an inferred task trajectory 00:14:08.178 --> 00:14:11.181 and control sequence for the robot. 00:14:11.181 --> 00:14:13.365 We then execute that on the robot, 00:14:13.365 --> 00:14:15.537 we observe what happens, 00:14:15.537 --> 00:14:18.199 then we adjust the controls using a sequence of techniques 00:14:18.199 --> 00:14:21.129 called iterative learning. 00:14:21.129 --> 00:14:25.106 Then what we do is, we increase the velocity a little bit. 00:14:25.106 --> 00:14:28.669 We observe the results, adjust the controls again, 00:14:28.669 --> 00:14:31.191 and observe what happens. 00:14:31.191 --> 00:14:33.327 And we go through this several rounds. 00:14:33.327 --> 00:14:34.683 And here's the result. 00:14:34.683 --> 00:14:36.528 That's the inferred task trajectory, 00:14:36.528 --> 00:14:39.991 and here's the robot moving at the speed of the human. 00:14:39.991 --> 00:14:42.433 Here's four times the speed of the human. 00:14:42.433 --> 00:14:45.004 Here's seven times. 00:14:45.004 --> 00:14:48.641 And here's the robot operating at 10 times 00:14:48.641 --> 00:14:50.841 the speed of the human. 00:14:50.841 --> 00:14:53.791 So we're able to get a robot to perform a delicate task, 00:14:53.791 --> 00:14:57.015 like a surgical subtask, 00:14:57.015 --> 00:15:00.262 at 10 times the speed of a human. 00:15:00.262 --> 00:15:04.485 So this project also, because of its involved practicing 00:15:04.485 --> 00:15:06.658 and learning, doing something over and over again, 00:15:06.658 --> 00:15:09.415 this project also has a lesson, which is, 00:15:09.415 --> 00:15:12.541 if you want to do something well, 00:15:12.541 --> 00:15:20.489 there's no substitute for practice, practice, practice. NOTE Paragraph 00:15:20.505 --> 00:15:23.625 So these are four of the lessons that I've learned 00:15:23.625 --> 00:15:26.752 from robots over the years, 00:15:26.752 --> 00:15:32.121 and robotics, the field of robotics has gotten much better 00:15:32.121 --> 00:15:34.289 over time. 00:15:34.289 --> 00:15:36.456 Nowadays, high school students can build robots 00:15:36.456 --> 00:15:40.481 like the industrial robot my dad and I tried to build. 00:15:40.481 --> 00:15:47.463 And now, I have a daughter, 00:15:47.463 --> 00:15:49.849 named Odessa. 00:15:49.849 --> 00:15:51.682 She's eight years old, 00:15:51.682 --> 00:15:54.106 and she likes robots, too. 00:15:54.106 --> 00:15:56.520 Maybe it runs in the family. (Laughter) 00:15:56.520 --> 00:16:00.167 I wish she could meet my dad. 00:16:00.167 --> 00:16:03.067 And now I get to teach her how things work, 00:16:03.067 --> 00:16:05.839 and we get to build projects together, and I wonder 00:16:05.839 --> 00:16:10.147 what kind of lessons that she'll learn from them. NOTE Paragraph 00:16:10.147 --> 00:16:12.754 Robots are the most human 00:16:12.754 --> 00:16:14.983 of our machines. 00:16:14.983 --> 00:16:18.068 They can't solve all of the world's problems, 00:16:18.068 --> 00:16:22.426 but I think they have something important to teach us. 00:16:22.426 --> 00:16:25.769 I invite all of you to think about the innovations 00:16:25.769 --> 00:16:28.444 that you're interested in, 00:16:28.444 --> 00:16:31.660 the machines that you wish for, 00:16:31.660 --> 00:16:34.829 and think about what they might be telling you, 00:16:34.829 --> 00:16:36.892 because I have a hunch 00:16:36.892 --> 00:16:38.979 that many of our technological innovations, 00:16:38.979 --> 00:16:41.637 the devices we dream about, 00:16:41.637 --> 00:16:45.601 can inspire us to be better humans. NOTE Paragraph 00:16:45.601 --> 00:16:49.186 Thank you. (Applause)