WEBVTT - TechStuff Classic: Do We Need Humanoid Robots? 0:00:04.400 --> 0:00:07.800 Welcome to tech Stuff, a production from I Heart Radio. 0:00:12.039 --> 0:00:14.800 Hey there, and welcome to tech Stuff. I'm your host 0:00:15.000 --> 0:00:18.160 job in Strickland. I'm an executive producer with I Heart Radio, 0:00:18.280 --> 0:00:21.040 and I love all things tech and it is time 0:00:21.079 --> 0:00:25.040 for a classic episode of tech Stuff. This episode originally 0:00:25.040 --> 0:00:29.120 published way back on October fifteenth, two thousand fourteen. It 0:00:29.240 --> 0:00:34.800 is titled do we need humanoid Robots? An interesting question 0:00:34.880 --> 0:00:37.120 because we often, at least I don't know, I can't 0:00:37.159 --> 0:00:40.400 say we. I often think of robots as sort of 0:00:40.440 --> 0:00:44.000 anthropomorphic robots, like I think of droids in the Star Wars, 0:00:44.280 --> 0:00:46.080 like C three b O that kind of droid. Not 0:00:46.120 --> 0:00:49.400 are two. He's clearly not humanoid. But you know, I 0:00:49.440 --> 0:00:53.040 think of humanoid robots more frequently than other types, despite 0:00:53.040 --> 0:00:55.160 the fact that the vast majority of robots I have have 0:00:55.320 --> 0:00:59.240 encountered have not been humanoid robots. So that does raise 0:00:59.240 --> 0:01:04.200 a question does a humanoid robot even makes sense? We 0:01:04.280 --> 0:01:08.920 try and answer that in this particular episode, Enjoy. I 0:01:09.000 --> 0:01:11.200 asked Josh what he would like to cover, because with 0:01:11.480 --> 0:01:13.280 the fact that I've got all these guests coming in 0:01:13.319 --> 0:01:15.840 to sit down with me, um, you know, some people 0:01:16.000 --> 0:01:18.800 like to come up with their own suggestions. Some people 0:01:18.840 --> 0:01:21.679 preferred if I pick a topic and then they research it. 0:01:22.080 --> 0:01:23.800 I asked Josh what he would like to talk about, 0:01:24.000 --> 0:01:27.760 and you were really interested in the idea of humanoid robots. Well, 0:01:27.880 --> 0:01:31.959 you have this awesome spreadsheet of um of listeners suggestions, 0:01:32.680 --> 0:01:34.640 and it might as well been a neon when it 0:01:34.760 --> 0:01:37.240 was going down the sheet. I'm like humanoid robots, of course, 0:01:37.520 --> 0:01:40.880 and this is a great topic. In order to really 0:01:40.880 --> 0:01:43.120 get into it, I was going to define a few terms, 0:01:43.200 --> 0:01:45.400 even though a lot of these are ones that I 0:01:45.480 --> 0:01:49.000 think most of us just kind of understand just from 0:01:49.040 --> 0:01:51.360 the fact that this is in our culture now. It's 0:01:51.400 --> 0:01:54.520 not just not just a reality as far as technology goes, 0:01:54.520 --> 0:01:56.480 but it plays a large part in fiction. In fact, 0:01:56.560 --> 0:02:00.320 that's where the term robot comes from, is from fiction. Uh. 0:02:00.480 --> 0:02:07.560 It was from Carrel Chopeck, a Czechoslovakian playwright. I did, 0:02:07.600 --> 0:02:11.520 in fact listen to it a couple of times. Carrel Chopec, Yeah, 0:02:11.520 --> 0:02:14.079 because it's his last name is spelled c A p 0:02:14.280 --> 0:02:16.960 e K. And it includes uh symbols that are not 0:02:17.040 --> 0:02:22.639 in the English alphabet, like squiggly lines and little UFOs 0:02:22.680 --> 0:02:26.000 and things he wrote. Are you are also known as 0:02:26.080 --> 0:02:29.800 Rossum's Universal robots, and the word robot comes from the 0:02:29.840 --> 0:02:34.760 check word robot to which means forced labor. Yeah, so 0:02:34.800 --> 0:02:39.360 a robot is a an entity, a synthetic construct that 0:02:39.480 --> 0:02:43.280 is forced to do work. Then we have humanoid, which 0:02:43.320 --> 0:02:46.720 just means resembling a human being. That's a term that 0:02:47.040 --> 0:02:50.960 is relatively young. It started showing up around the turn 0:02:51.000 --> 0:02:54.799 of the twentieth century. And uh, it started I think 0:02:54.800 --> 0:02:57.960 the first few times it was ever mentioned was around 0:02:58.040 --> 0:03:01.840 nineteen twelve, and it was mostly used then to describe fossils, 0:03:02.280 --> 0:03:07.280 saying these are humanoid fossils, like yeah. And then we 0:03:07.360 --> 0:03:10.560 have android, which is we're probably not gonna be using 0:03:10.560 --> 0:03:13.480 that word very often, but android is a robot that's 0:03:13.520 --> 0:03:16.639 in the form of a human. So all androids or robots, 0:03:16.639 --> 0:03:19.040 but not all robots are androids. And you know, I 0:03:19.120 --> 0:03:21.680 ran into I looked up android as well. Yeah, and 0:03:21.760 --> 0:03:25.120 apparently that's from like the early eighth century. It's a 0:03:25.160 --> 0:03:28.880 little odd that it actually predates robots. Yeah, but uh, 0:03:29.680 --> 0:03:32.440 when we look at myths and legends, there's so many 0:03:32.520 --> 0:03:36.320 stories that involve a human like entity that's not actually 0:03:36.360 --> 0:03:39.080 a person that you can see where it gets translated 0:03:39.080 --> 0:03:41.680 in there. This gets a little confusing too, because Star 0:03:41.680 --> 0:03:45.760 Wars they called all their robots droids, but there they 0:03:45.800 --> 0:03:48.080 aren't androids are two? T two is not an android 0:03:48.160 --> 0:03:52.240 because he's not um or it's not human shaped. You 0:03:52.240 --> 0:03:54.160 could even argue that C three Po is not a 0:03:54.160 --> 0:03:56.960 true android because some people say to be an android 0:03:57.000 --> 0:03:59.720 you have to appear, at least on casual glance to 0:03:59.840 --> 0:04:03.880 be human. He's way too shiny, way too shiny. Data 0:04:04.000 --> 0:04:07.880 from Star Trek Next Generation might be uh android, but 0:04:08.320 --> 0:04:11.360 he's an android who's had, you know, a long time 0:04:11.400 --> 0:04:14.200 in the man cave. He hasn't seen much sun, right, Yeah, 0:04:14.200 --> 0:04:17.520 because he's definitely got a weird complexion. Thing goes the 0:04:17.600 --> 0:04:20.840 kid in a I would be an android David, yes, 0:04:21.000 --> 0:04:24.000 which turns out to be a popular name for droids, 0:04:24.640 --> 0:04:26.800 because there was well there's David in AI, and there 0:04:26.839 --> 0:04:29.400 was also David and Prometheus. Oh you know, I never 0:04:29.400 --> 0:04:33.000 saw Prometheus. I don't know if your listeners are going 0:04:33.040 --> 0:04:35.279 to agree with me or not, because I could see 0:04:35.520 --> 0:04:38.760 um getting shouted down. But I thought Prometheus was a 0:04:38.760 --> 0:04:42.120 great movie. Even upon second viewing, I thought it was good. 0:04:42.240 --> 0:04:45.799 You know, I know that the from the artistic level, 0:04:46.120 --> 0:04:47.640 A lot of people really loved it, And then there 0:04:47.640 --> 0:04:49.640 were some people who said, how can you get lost 0:04:49.800 --> 0:04:52.880 if you have a three dimensional map with you at 0:04:52.920 --> 0:04:56.240 all time? But you know, plot versus artist, I don't know. 0:04:56.880 --> 0:04:59.880 Then you've got replicants from Blade Runner. These are more 0:05:00.040 --> 0:05:03.200 like cyborgs because they have some sort of organic material 0:05:03.640 --> 0:05:08.560 attached to them. They're not completely you know, synthetic material. 0:05:09.200 --> 0:05:12.960 So Terminator is another example. They have a fleshy over 0:05:13.360 --> 0:05:16.039 skin on top of their metallic bodies. But was that 0:05:16.120 --> 0:05:18.760 real skin that he had there was a synthetic skin, 0:05:18.800 --> 0:05:20.680 because that would make a difference. That's a good question. 0:05:20.800 --> 0:05:23.119 And I uh, you know, I know that they refer 0:05:23.200 --> 0:05:25.520 to them as cyborgs at least a few times in 0:05:25.560 --> 0:05:28.560 the movies, which would suggest that it's actual skin. Maybe 0:05:28.640 --> 0:05:35.240 it's lab grown human skin. So you know, there's some 0:05:35.440 --> 0:05:39.240 fuzzy lines around these definitions. I would think that cyborgs 0:05:39.279 --> 0:05:41.960 would be the hardest of all of the the um 0:05:42.600 --> 0:05:46.320 humanoid robots to make, because the flesh would just wrought 0:05:46.720 --> 0:05:50.680 like you have your normal looking humanoid robot your cyborg, 0:05:50.960 --> 0:05:53.560 and it's ear would just fall off. Yeah, and it's 0:05:53.600 --> 0:05:56.880 not as easy as you might think to wire together 0:05:57.040 --> 0:06:01.200 the wet wear that's in our heads, hardware that runs 0:06:01.200 --> 0:06:05.320 on circuits. We will often think of computers working in 0:06:05.360 --> 0:06:07.359 a way that's similar to our brains, but in fact 0:06:07.400 --> 0:06:10.239 the two work in very different ways. Well it seems 0:06:10.279 --> 0:06:14.960 like running into this though, Strickland um, the the more 0:06:15.400 --> 0:06:18.880 we got into humanoid robotics, the more we started to 0:06:18.960 --> 0:06:21.919 understand just how complex we are. Yeah, that's one of 0:06:21.920 --> 0:06:24.920 the things that I think is a benefit of study 0:06:24.920 --> 0:06:28.640 of humanoid robotics. The idea of pursuing the goal of 0:06:28.680 --> 0:06:31.279 creating a humanoid robot is not just that we learn 0:06:31.720 --> 0:06:34.680 more about all the different areas and robotics and they're 0:06:34.720 --> 0:06:36.719 a lot, and we'll talk about some of them, but 0:06:36.760 --> 0:06:40.240 we also learn more about ourselves. We're trying to figure out, Okay, 0:06:40.279 --> 0:06:43.120 if we're going to make something that is able to 0:06:43.160 --> 0:06:45.960 perform tasks the way human does, then we really got 0:06:45.960 --> 0:06:49.159 to take a close look at humans. That's that's the 0:06:49.200 --> 0:06:54.880 first place to start. So what makes a humanoid robot? 0:06:55.120 --> 0:06:57.840 And generally speaking, we're talking about a robot that has 0:06:58.400 --> 0:07:02.479 basic features usually minimum a torso, arms and legs, and 0:07:02.839 --> 0:07:06.120 is walking up right. Uh, it may have a head 0:07:06.200 --> 0:07:09.240 or it might not. Early humanoid robots didn't, or at 0:07:09.320 --> 0:07:14.840 least their sensory uh. Instruments were all located within the 0:07:14.880 --> 0:07:18.600 top part of the torso there wasn't like a separate head. 0:07:18.920 --> 0:07:22.040 Did you see a picture of Minerva at the Smithsonian? No? 0:07:22.280 --> 0:07:24.920 I did not. It's a robot to her guide, but um, 0:07:24.960 --> 0:07:28.680 she came up in the humanoid robot research and I 0:07:28.720 --> 0:07:31.120 think she's stretching it a little bit. Yeah, she looks 0:07:31.160 --> 0:07:34.000 a bit like a washing machine with a couple of 0:07:34.160 --> 0:07:38.960 UM cameras on top. So just those alone, I guess 0:07:39.000 --> 0:07:44.240 makes her eligible for the humanoid robot realm. But that 0:07:44.240 --> 0:07:45.920 that seems like that's a bit of a you know, 0:07:46.320 --> 0:07:49.240 if it has an appendage, that doesn't necessarily make a humanoid. 0:07:49.240 --> 0:07:51.560 I mean, you could look at the Mars Curiosity rover, 0:07:51.640 --> 0:07:54.760 which has several appendages, but I don't think anyone would 0:07:54.760 --> 0:07:59.200 ever describe it as humanoid. Right, So, uh, ideally a 0:07:59.280 --> 0:08:02.560 humanoid robe lot would be able to interact with humans 0:08:02.960 --> 0:08:06.600 within a human environment. Because here's the thing about we, 0:08:06.600 --> 0:08:09.720 we people. We have defined our environments to a large extent, 0:08:09.800 --> 0:08:13.600 especially in developed nations, where the stuff that's around us 0:08:13.640 --> 0:08:18.640 we have shaped so that it works within our capabilities, 0:08:19.160 --> 0:08:22.240 right with our with our human environment thus far. If 0:08:22.240 --> 0:08:25.320 you look at technology though, on the whole, we've pretty 0:08:25.400 --> 0:08:28.720 much been forced to adapt to it. So, for example, 0:08:28.760 --> 0:08:32.439 like a keyboard, we don't normally naturally, you know, um 0:08:32.760 --> 0:08:36.520 express ideas through our fingers on a little a little board, right, 0:08:36.720 --> 0:08:38.640 we don't normally do that, So we had to adapt 0:08:38.720 --> 0:08:40.800 to the technology and learn to type and get good 0:08:40.800 --> 0:08:44.959 at it. With humanoid robots, it's basically going the exact opposite. 0:08:45.360 --> 0:08:48.800 It's saying we already have an environment, we already are um, 0:08:48.840 --> 0:08:50.640 you know, good at all this other stuff. If we're 0:08:50.640 --> 0:08:53.200 gonna make humanoid robots, one of the great benefits is 0:08:53.640 --> 0:08:57.040 they can adapt to us. Right. Yeah, we don't have 0:08:57.160 --> 0:09:00.600 to uh create a unitask or robe, but that's really 0:09:00.600 --> 0:09:03.439 good at one thing, um that may or may not 0:09:03.480 --> 0:09:06.640 be something that humans can do easily. We can make 0:09:06.679 --> 0:09:10.400 a robot that's good lots of things. Uh. I also 0:09:10.480 --> 0:09:13.760 think usually when I think of humanoid robots, when I 0:09:13.800 --> 0:09:16.640 think of robots in general, I normally think that they 0:09:16.640 --> 0:09:19.920 are at least semi autonomous. That's that's one of the 0:09:19.920 --> 0:09:22.440 things I usually think of. It doesn't necessarily have to be. 0:09:22.520 --> 0:09:26.720 You could have a tel operated robot, but I almost 0:09:26.720 --> 0:09:30.280 think of that closer to the realm of like a 0:09:30.320 --> 0:09:34.120 remote controlled car or a puppet even. Um So, I 0:09:34.200 --> 0:09:37.600 often one of the definitions I use is that it's 0:09:37.600 --> 0:09:43.160 an autonomous or semi autonomous machine in human form. It's 0:09:43.240 --> 0:09:48.560 mechanical and electronic, and it can thus do the sort 0:09:48.559 --> 0:09:50.720 of things humans do, but do it in a totally 0:09:50.720 --> 0:09:53.680 synthetic way. And you have to be careful when you 0:09:53.720 --> 0:09:57.120 say autonomous or semi autonomous, because the state of the 0:09:57.200 --> 0:10:01.120 art right now appears to be that robots display autonomy 0:10:01.120 --> 0:10:03.320 because they just kind of wander off in places they're 0:10:03.320 --> 0:10:06.680 not supposed But it's not because they want to. It's 0:10:06.720 --> 0:10:10.679 because their their program just ran a foul of program. Right. 0:10:10.800 --> 0:10:13.400 There's no determination there on the part of the robot. 0:10:13.920 --> 0:10:18.480 It's not exploring its environment on its own accord. It's 0:10:18.520 --> 0:10:21.280 someone made a mistake in the code somewhere and where 0:10:21.320 --> 0:10:23.480 the robot was supposed to take a left hand turn 0:10:23.559 --> 0:10:28.040 at this one you know, predetermined spot, and instead continued 0:10:28.080 --> 0:10:31.520 forward or something exactly. Um So, I wanted to talk 0:10:31.520 --> 0:10:34.440 a little bit about the history of humanoid robots, and 0:10:34.480 --> 0:10:38.240 if you wanna look way way way back, I mean 0:10:38.240 --> 0:10:40.680 we're talking about the first people to really kind of 0:10:40.720 --> 0:10:46.079 attempt to build a humanoid machine that could mimic or 0:10:46.200 --> 0:10:48.599 the movements at least of a person. You gotta go 0:10:48.640 --> 0:10:51.720 all the way back to Greece between the years ten 0:10:51.920 --> 0:10:56.960 and seventy Common era. That's when Hero of Alexandria started 0:10:57.280 --> 0:11:00.240 to create various machines. He's also the person who made 0:11:00.240 --> 0:11:06.199 the first working steam engine style tool, which is pretty impressive. Yeah. 0:11:06.240 --> 0:11:08.240 He he had come up with a lot of very 0:11:08.280 --> 0:11:12.080 clever designs. Whether they were built or not depends upon 0:11:12.600 --> 0:11:16.079 uh certain accounts and and if they're true. But the 0:11:16.080 --> 0:11:20.000 stuff he designed is completely build a bowl. So he 0:11:20.040 --> 0:11:22.079 didn't come up with any ideas where it was so 0:11:22.360 --> 0:11:25.040 you know, outlandish, that was impossible. He wasn't just like 0:11:25.120 --> 0:11:27.720 drawing in the margins of his diary or something. No, Yeah, 0:11:27.720 --> 0:11:31.080 he came up with specific plans that people today have recreated. 0:11:31.120 --> 0:11:34.559 They built their own versions. Yeah. So he created a 0:11:34.600 --> 0:11:38.160 lot of designs for automata. Although these are things that 0:11:38.200 --> 0:11:41.880 were controlled by pulleys and ropes and cogged wheels, and 0:11:42.480 --> 0:11:46.120 uh needed some form of outside influence to make them work, 0:11:46.160 --> 0:11:50.840 so they're not all fully self contained like the show 0:11:50.880 --> 0:11:54.520 biz pizza Rocke fire explosion day and yeah exactly, yeah, yeah, 0:11:54.520 --> 0:11:58.079 it was some one of those audio animatronic figures that 0:11:58.080 --> 0:12:00.920 that looks very robotic, but you realize it's really just 0:12:01.160 --> 0:12:05.160 one tiny piece of a giant system. Uh. In four, 0:12:06.080 --> 0:12:08.240 we get up to Leonardo da Vinci. He designed an 0:12:08.240 --> 0:12:12.640 automaton in the form of a mechanical night which uh 0:12:12.840 --> 0:12:16.559 again supposedly he built. There's no actual record of an 0:12:16.559 --> 0:12:20.520 existing one from history, but they have created ones based 0:12:20.559 --> 0:12:23.600 on the design since then, and it could do things 0:12:23.600 --> 0:12:26.240 like move its arms and raise its visor. Who's doing this? 0:12:26.559 --> 0:12:32.120 Who's doing this? Crazy engineers who are also uh, very 0:12:32.120 --> 0:12:35.680 excited about history and very wealthy too, I would imagine. Yeah, 0:12:35.720 --> 0:12:40.160 so in this case, you're talking about Mark Rossheim, who 0:12:40.320 --> 0:12:44.640 recreated this particular machine, and it it's a night, it's 0:12:44.679 --> 0:12:49.920 a night in German medieval armor. And it can sit down, 0:12:50.040 --> 0:12:51.800 it can stand up, it can move its arms, it 0:12:51.840 --> 0:12:56.800 can raise a visor, it can work its jaw ling, 0:12:57.120 --> 0:13:00.600 I imagine. So I tried to find video of ross 0:13:00.640 --> 0:13:03.400 Heims version working and I couldn't find it. He did, however, 0:13:04.120 --> 0:13:07.800 make another of Da Vinci's inventions, which was a self 0:13:07.880 --> 0:13:11.960 driving cart that used Yeah you you wound a spring, 0:13:12.400 --> 0:13:15.160 and it had cam stops that would allow it to 0:13:15.320 --> 0:13:18.680 steer a predetermined path. You would actually program the cart 0:13:19.480 --> 0:13:22.640 by putting cam stops in particular locations along the cam 0:13:22.960 --> 0:13:24.719 and that would tell it when to turn left or 0:13:24.720 --> 0:13:28.520 when to turn right, so it couldn't it couldn't navigate 0:13:28.760 --> 0:13:32.800 through uh an obstacle course unless you had already previously 0:13:32.800 --> 0:13:34.840 seen the obstacle course and you could figure out when 0:13:34.840 --> 0:13:37.480 it needed to turn ahead of time, so you're essentially 0:13:37.520 --> 0:13:41.280 programming the device. Um. There are lots of examples in 0:13:41.280 --> 0:13:45.000 the renaissance of automata and semi automata, things that are 0:13:45.000 --> 0:13:47.600 really more like puppets. You've heard about the Mechanical Turk 0:13:48.520 --> 0:13:52.800 have the chess playing robot uh So it looked like 0:13:52.840 --> 0:13:54.840 it was a robot that could play chess and was 0:13:54.920 --> 0:13:57.840 really really good at playing chess, and it turned out 0:13:57.880 --> 0:14:00.400 eventually to be a hoax. It was actually it was 0:14:00.440 --> 0:14:03.240 actually a puppet, and there was an actual chess master 0:14:03.840 --> 0:14:07.640 hidden in a cabinet beneath the Mechanical Turk who sat 0:14:07.760 --> 0:14:11.560 kind of Indian style with a chessboard in front of 0:14:11.600 --> 0:14:13.960 him and could move the pieces to where it needed 0:14:14.000 --> 0:14:15.520 to be. But it was all being guided by an 0:14:15.559 --> 0:14:18.480 actual chess master who was hidden under When was that 0:14:18.480 --> 0:14:21.880 that was late Renaissance early Enlightenment. That in and of 0:14:21.960 --> 0:14:24.880 itself is pretty impressive. Yeah, it was. It was neat 0:14:24.920 --> 0:14:27.560 that people were thinking about these sort of things. Uh. 0:14:27.640 --> 0:14:30.200 By ninety six we get the first humanoid robot to 0:14:30.240 --> 0:14:34.400 appear on film, uh, Metropolis, the character of Maria, and 0:14:34.480 --> 0:14:38.520 at ninety nine at the World's Fair, Westinghouse Electric Corporation 0:14:38.600 --> 0:14:41.600 showed off a robot called Electro. Now have you ever 0:14:41.640 --> 0:14:45.160 seen this? So he kind of looks like the ten 0:14:45.240 --> 0:14:48.840 man from the Wizard of Oz film. Yes, he smokes. 0:14:49.640 --> 0:14:51.320 I have seen one of the things he can do 0:14:51.880 --> 0:14:55.520 get a little bellows in his head that allowed him 0:14:55.560 --> 0:14:58.960 to puff smoke. He could also kind of speak. He 0:14:59.000 --> 0:15:03.120 had a seventy eight revolution per minute UH record player 0:15:03.280 --> 0:15:07.680 essentially inside of him. I suppose if the if the 0:15:07.680 --> 0:15:10.240 needles skipp it would at least sound like it. Um. 0:15:10.520 --> 0:15:14.120 He repeat himself over. He called the audience twots. That 0:15:14.200 --> 0:15:19.760 was the name. Yeah, he's apparently the main reason he 0:15:19.840 --> 0:15:23.520 was retired from being shown off at exhibitions was because 0:15:23.560 --> 0:15:27.000 it was very dated kind of lingo. But he was 0:15:27.080 --> 0:15:30.960 used at the nineteen thirty nine World's Fair, which was 0:15:31.680 --> 0:15:33.760 It's funny because I'll be talking about that again in 0:15:33.800 --> 0:15:36.800 another episode very shortly. The World's Fair would have been 0:15:36.800 --> 0:15:41.800 an amazing thing to visit. Um, I'm more I can 0:15:41.880 --> 0:15:44.440 understand that. But if you're if you're going to get 0:15:44.440 --> 0:15:48.440 to the point where you're looking at full scale anthropomorphic robots, 0:15:48.480 --> 0:15:51.560 you gotta get up to about ninety three. That's when 0:15:51.560 --> 0:15:56.120 the Waybot one from the Waseda University came out. That 0:15:56.200 --> 0:15:59.320 was the first full scale anthropomorphic robot developed in the 0:15:59.360 --> 0:16:03.080 world which had limb control, a vision system, so I 0:16:03.080 --> 0:16:06.000 had an optical system that could recognize its environment and 0:16:06.040 --> 0:16:11.360 objects and measure distance, and it also had a conversation system. Uh. 0:16:11.480 --> 0:16:14.840 It was actually a collection of a bunch of very 0:16:14.880 --> 0:16:19.359 complex machinery. Like its hands had been previously developed independently 0:16:19.360 --> 0:16:22.240 of the robots, so had its legs. So it's like 0:16:22.760 --> 0:16:26.000 all these people coming up with these various pieces saying, 0:16:26.040 --> 0:16:28.400 all right, let's connect all this together and see what happens. 0:16:29.000 --> 0:16:32.480 So that was a huge, huge leap forward. Yeah, you know, 0:16:32.480 --> 0:16:36.560 if you'll notice, we went basically from um HOAXI chess 0:16:36.560 --> 0:16:40.440 playing turks to you know, a robot that could converse 0:16:40.520 --> 0:16:44.720 and interact with its environment. And then um, it seems 0:16:44.760 --> 0:16:46.680 like we we kind of went off course for a 0:16:46.720 --> 0:16:48.680 little bit and now we're coming full circle back to 0:16:48.760 --> 0:16:51.720 that where like you said, a lot of different disciplines 0:16:51.760 --> 0:16:55.400 are contributing these different pieces to what will eventually be 0:16:55.760 --> 0:16:58.800 all of the best practices from each little sub discipline 0:16:58.840 --> 0:17:02.640 put together in you know, the true humanoid robot. Well yeah, 0:17:02.640 --> 0:17:05.160 I mean, if we're talking about a humanoid robot that's 0:17:05.200 --> 0:17:09.520 capable of interacting with people as if they were, you know, 0:17:09.600 --> 0:17:12.560 their own person, even though maybe an odd person, not 0:17:12.680 --> 0:17:15.040 like the kind of person you would typically run into. 0:17:16.080 --> 0:17:19.800 There's it's a multi disciplinary approach. I mean, artificial intelligence 0:17:19.840 --> 0:17:24.240 by itself is multi disciplinary because you have sensing, you 0:17:24.280 --> 0:17:27.239 have so all the perception, there's all these different just 0:17:27.320 --> 0:17:32.119 that's multi disciplinary. Then you've got the processing, the cognition, 0:17:32.760 --> 0:17:36.280 things like planning, navigation. There are so many things that 0:17:36.359 --> 0:17:40.760 come together to make a humanoid robot a a possibility. 0:17:40.800 --> 0:17:44.639 And that's just the the mental side, right. Then you 0:17:44.680 --> 0:17:47.000 have all the physical side, the how do you make 0:17:47.040 --> 0:17:49.440 it walk, how do you make it keep its balanced? 0:17:49.600 --> 0:17:53.440 So lots of stuff to consider there. We're gonna take 0:17:53.480 --> 0:17:56.040 a quick break but when we come back, we will 0:17:56.040 --> 0:18:07.679 look more at the issue of humanoid robots. Wybot two 0:18:07.760 --> 0:18:10.480 came out in that was a specialist robot. It could 0:18:10.480 --> 0:18:13.760 play a keyboard, trying a keyboard. He could read sheet 0:18:13.840 --> 0:18:17.119 music and play music. Uh. It was because it was 0:18:17.160 --> 0:18:20.080 a specialist. It was not able to do the general 0:18:20.160 --> 0:18:23.040 functions that it's predecessor could do. And that's why the 0:18:23.280 --> 0:18:27.760 issues in robotics today is that it's very challenging to 0:18:27.840 --> 0:18:31.320 build a general purpose robot. It's much easier to take 0:18:31.440 --> 0:18:33.800 a specific task that you need to have done and 0:18:33.840 --> 0:18:36.000 designed a robot to do that. Yeah, because I mean 0:18:36.000 --> 0:18:40.639 we already have those rooms and rovers. I mean you know, 0:18:40.720 --> 0:18:44.439 there's there's also all the robots and manufacturing, all the 0:18:44.480 --> 0:18:49.680 welding robots things like that. Uh. Nine Pacific Northwest National 0:18:49.760 --> 0:18:54.240 Laboratory built a robot called Manny that was the first 0:18:54.320 --> 0:18:57.600 full scale android body and it had forty two degrees 0:18:57.640 --> 0:19:01.480 of freedom but no AI or autonomy. It was completely teleoperated. 0:19:01.840 --> 0:19:03.720 And um it took me a couple of times to 0:19:03.720 --> 0:19:06.119 figure out what degrees of freedom meant. Yeah, Um, I 0:19:06.160 --> 0:19:08.680 thought it meant like the it could move its arm 0:19:09.240 --> 0:19:13.600 forty two degrees basically, but a degree of freedom is say, 0:19:13.680 --> 0:19:15.840 like it can move its wrist, that's a degree of freedom. 0:19:15.880 --> 0:19:18.520 It can turn its head left and right, that's a 0:19:18.520 --> 0:19:20.679 degree of freedom right right. And if you look at 0:19:20.680 --> 0:19:24.240 the human hand, the human hand has about thirty degrees 0:19:24.320 --> 0:19:27.600 of freedom, meaning that you look at the way each 0:19:27.680 --> 0:19:29.879 finger and your thumb can move. You look at the 0:19:29.920 --> 0:19:32.120 way you can clench a fist, you can twist your 0:19:32.119 --> 0:19:35.919 hand with your wrist um, those are all different degrees 0:19:35.920 --> 0:19:38.800 of freedom. And uh. In fact, one of the cool 0:19:38.880 --> 0:19:41.639 things about robots is that as we get better and 0:19:41.680 --> 0:19:44.399 better at designing them, we can create robots that have 0:19:44.520 --> 0:19:46.840 far more degrees of freedom than the human body does. 0:19:47.320 --> 0:19:50.679 So I like the idea of a humanoid robot in 0:19:50.720 --> 0:19:54.200 the future that has sixty degree motion with its wrist 0:19:54.320 --> 0:19:56.879 and then just having it changed light bulbs. Is it 0:19:57.080 --> 0:19:59.159 just spin and not have to do the little twisty 0:19:59.200 --> 0:20:01.640 turning motion. How many robots does it take to change 0:20:01.640 --> 0:20:04.160 the light bulbs? Just just the one, just that one, 0:20:04.480 --> 0:20:07.800 just one billion dollar rob Yeah. Yeah, I'm not saying 0:20:07.800 --> 0:20:11.840 it's a fishing system. I'm just saying I'm in supremely 0:20:11.960 --> 0:20:16.439 lazy human being with tall ceilings. Uh. In nine Honda 0:20:16.480 --> 0:20:19.280 introduced the P two robot, which was a self contained 0:20:19.359 --> 0:20:22.840 robotic humanoid. It could walk and climb stairs. The P 0:20:23.040 --> 0:20:26.080 three followed in n and in two thousand two, Honda 0:20:26.200 --> 0:20:30.679 introduced My good buddy as Emo. As the first article 0:20:30.760 --> 0:20:34.880 I wrote for How Stuff Works. As Works have an 0:20:34.880 --> 0:20:37.360 episode on it. I have not done a full episode 0:20:37.440 --> 0:20:41.240 on Asimo. I even was offered the opportunity to meet 0:20:41.280 --> 0:20:43.600 as Amo when I first wrote the article, but it 0:20:43.640 --> 0:20:46.000 would have meant having to travel to Disneyland to do it, 0:20:46.040 --> 0:20:48.200 and at the time, How Stuff Works was not prepared 0:20:48.240 --> 0:20:52.440 to do such a I went to Disneyland by myself, 0:20:52.880 --> 0:20:56.160 uh well with my wife, and we went and saw 0:20:56.400 --> 0:20:59.000 the Asimo production. And at the end of it, I 0:20:59.400 --> 0:21:02.240 talked to of the Disney cast members and I said, yeah, 0:21:02.400 --> 0:21:04.399 I wrote the article about how Asthma works for How 0:21:04.400 --> 0:21:06.560 Stuff Works, and she said, hang on a minute. And 0:21:06.600 --> 0:21:08.520 I got to meet Asthmo. And it was a man 0:21:08.560 --> 0:21:11.200 in a suit, right. It was actually it was actually 0:21:11.240 --> 0:21:14.080 a collection of cats that the duct taped together and 0:21:14.119 --> 0:21:17.280 then plastic. No, it was a working robot. That's pretty 0:21:17.320 --> 0:21:20.119 near how blown away? I was very much blown away. 0:21:20.160 --> 0:21:22.119 It was cool seeing it up close. I mean, it 0:21:22.160 --> 0:21:25.120 looks like a little tiny astronaut right because he's got 0:21:25.160 --> 0:21:27.360 like the face plate especially. I love the people call 0:21:27.400 --> 0:21:30.520 it he people give And I do this all the 0:21:30.560 --> 0:21:33.040 time too. With robots, I'll sign a gender even though 0:21:33.080 --> 0:21:36.800 technically many of them are specifically genderless. As amo is 0:21:36.960 --> 0:21:39.280 is supposed to be genderless, but I often refer to 0:21:39.320 --> 0:21:40.960 Asmo as a he as well. Well, you know why. 0:21:41.000 --> 0:21:43.280 It's the shoulders, I would guess. Yeah, they go straight 0:21:43.280 --> 0:21:47.040 across and and and far out. That's very masculine no 0:21:47.080 --> 0:21:49.920 matter what. Yeah, and you need that, I would guess 0:21:49.920 --> 0:21:54.359 you need shoulders in a humanoid robot with flexible arms. Well, 0:21:54.400 --> 0:21:57.840 and also, I'm sure every single element of Asimo is 0:21:57.920 --> 0:22:00.960 built with the balance in mind, because is the first 0:22:01.040 --> 0:22:05.640 robot that can run. Yes, I've seen him run. It's gawky. Yeah, 0:22:05.800 --> 0:22:07.560 it kind of look like someone who really needs to 0:22:07.560 --> 0:22:09.639 get to the bathroom as a little bit of a 0:22:09.640 --> 0:22:12.879 hoppy kind of run. But the the definition of run 0:22:12.920 --> 0:22:15.760 here is that there are moments where both feet are 0:22:15.840 --> 0:22:19.359 off the ground. So walking you always have one ft 0:22:19.359 --> 0:22:21.720 in contact with the ground, and running both feet at 0:22:21.760 --> 0:22:25.600 some point are out of contact. And that's a huge 0:22:25.720 --> 0:22:29.320 deal for robotics, right. I mean, you have a machine 0:22:29.400 --> 0:22:32.240 that completely separates itself from contact on the ground. It 0:22:32.280 --> 0:22:35.480 has no propulsion to keep it upright, you know, it 0:22:35.520 --> 0:22:37.760 doesn't have like propellers or jets or anything like that. 0:22:38.280 --> 0:22:40.920 So you have to design it so it can it 0:22:41.000 --> 0:22:43.760 can propel itself off the ground and then catch itself 0:22:43.760 --> 0:22:46.040 when it comes back down without falling over. And that 0:22:46.200 --> 0:22:49.800 is a non trivial challenge. No, it's an enormous challenge 0:22:49.840 --> 0:22:55.080 that robots UM really kind of started to tackle lately. UM. 0:22:55.160 --> 0:22:57.199 One of the ways that they've overcome it is with 0:22:57.320 --> 0:23:00.399 rounded feet, which are very helpful in keeping balanced and 0:23:00.440 --> 0:23:04.359 allowing it to run UM. But there's drawbacks to it 0:23:04.400 --> 0:23:07.920 as well, Like the robot can't start itself, it also 0:23:08.000 --> 0:23:11.679 can't stop, so it can't stop moving, which is not 0:23:11.880 --> 0:23:14.600 something you want. Like, there's still some challenges there. Ahead 0:23:14.600 --> 0:23:17.600 of the robotists who are learning to teach a robot 0:23:17.640 --> 0:23:20.240 to walk, and even the ones that have taught robots 0:23:20.240 --> 0:23:23.720 to walk UM, they typically can just walk over flat 0:23:23.720 --> 0:23:27.719 surfaces with no obstacles. When they encounter stairs, there in trouble. 0:23:27.800 --> 0:23:30.679 But then you have robots that know how to go upstairs, 0:23:31.119 --> 0:23:34.080 but they can't walk on a flat surface. Eventually, all 0:23:34.119 --> 0:23:37.000 this information, all this knowledge will be brought together and 0:23:37.080 --> 0:23:39.640 you'll have a robot that can walk no problem. Right. 0:23:39.680 --> 0:23:43.399 In fact, this kind of transitions nicely into those challenges 0:23:43.520 --> 0:23:48.520 that face designers of humanoid robots and and locomotion is 0:23:49.040 --> 0:23:51.119 the probably one of the top ones, at least from 0:23:51.160 --> 0:23:54.320 the physical engineering side. For example, you know, as Amo 0:23:54.480 --> 0:23:57.600 can can go up and downstairs, but that is a 0:23:57.640 --> 0:24:01.320 little deceptive because as Amo has to be programmed to 0:24:01.680 --> 0:24:04.199 go up or down the staircase and know exactly how 0:24:04.240 --> 0:24:07.640 many stairs are involved. It's not so much it's not 0:24:08.000 --> 0:24:11.840 a case of Asimo detecting a staircase and then uh, 0:24:11.880 --> 0:24:15.359 and then navigating through oh up or down it. It's 0:24:15.920 --> 0:24:18.879 the fact that all right now we're initiating your stair 0:24:19.000 --> 0:24:23.480 climbing program. Yeah exactly. It's kind of like smoke and 0:24:23.480 --> 0:24:28.320 mirrors robotics. Basically, it's at But that's you know, those 0:24:28.359 --> 0:24:31.119 are the little no no pun intended, Those are the 0:24:31.119 --> 0:24:32.800 little steps you have to take in order to get 0:24:32.800 --> 0:24:35.000 to the destination. What do you mean, no pun intended. 0:24:35.040 --> 0:24:37.600 I don't buy that at all. I started saying it 0:24:37.640 --> 0:24:40.439 without thinking about it, and then I mean, but then 0:24:40.480 --> 0:24:42.000 I did follow through with it, so I guess there 0:24:42.040 --> 0:24:45.040 was some intention there at the end. But uh yeah. 0:24:45.119 --> 0:24:48.480 They're also not very good at going across any kind 0:24:48.640 --> 0:24:54.720 of uneven terrain, right, So, humanoid robots in particular find 0:24:54.720 --> 0:24:57.760 it very difficult to maintain balance over anything that's not 0:24:58.359 --> 0:25:00.800 either a flat surface or in the case some robots 0:25:00.840 --> 0:25:03.600 that can go up or downstairs and stairs. So if 0:25:03.600 --> 0:25:07.200 you're talking about like a sidewalk that is not completely even, 0:25:07.640 --> 0:25:09.960 that would be enough to give a robot trouble because 0:25:10.080 --> 0:25:12.240 it's going to try and put its foot down to 0:25:12.359 --> 0:25:15.399 where it would believe the ground to be, and if 0:25:15.400 --> 0:25:19.240 the ground is not exactly level, then that's yeah, because 0:25:19.320 --> 0:25:22.520 they can't really catch their balance very well. There are 0:25:22.640 --> 0:25:26.479 robots that can, but they are four legged. Yeah, you've 0:25:27.800 --> 0:25:30.800 you've seen the Big Dog video, No, I saw the Army, 0:25:31.960 --> 0:25:34.960 though it's very similar. Big Dog is essentially a robo 0:25:35.040 --> 0:25:39.280 mule type development. It is a four legged robot that 0:25:39.480 --> 0:25:42.560 is able to maintain its balance even when pushed. And 0:25:42.640 --> 0:25:47.080 the famous video shows the robot dog the Big Dog 0:25:48.359 --> 0:25:52.240 kind of jogging and then a guy just casually lifts 0:25:52.280 --> 0:25:56.160 his leg up and kicks the robot dog like he 0:25:56.200 --> 0:25:58.240 puts essentially puts the bomb of his foot against the 0:25:58.280 --> 0:26:01.199 side of it and pushes really hard, tipping and you 0:26:01.240 --> 0:26:04.960 see the big dog stumble. It actually stumbles and then 0:26:05.040 --> 0:26:09.320 catches itself and then rights itself and continues on. And 0:26:09.520 --> 0:26:12.840 almost everyone has an emotional reaction to this, like how 0:26:12.960 --> 0:26:16.760 dare that evil man kick that poor defenseless robot. The 0:26:16.840 --> 0:26:22.520 robot can't feel anything, but that robot is um gasoline powered. 0:26:23.119 --> 0:26:25.440 So oh yeah, that's right. That's one of the big 0:26:25.440 --> 0:26:28.639 dis keeping it inside you. Yeah, you wouldn't want to 0:26:28.680 --> 0:26:31.359 have one of these indoors. No, you don't bring your 0:26:31.400 --> 0:26:35.240 lam indoors. No, and uh. And the the pistons that 0:26:35.400 --> 0:26:38.480 allow it to do this are quite loud, you know. 0:26:38.520 --> 0:26:41.560 It's not not a subtle system at all. So a 0:26:41.600 --> 0:26:44.679 lot of work has to go into creating better systems 0:26:44.800 --> 0:26:47.720 for robots to maintain their balance in order for the 0:26:47.800 --> 0:26:50.840 locomotion problem to really be solved. And again, I mean, 0:26:50.880 --> 0:26:53.480 like anybody who's seen short Circuit knows that you can 0:26:53.480 --> 0:26:56.240 build a robot like Johnny five with arms in the 0:26:56.320 --> 0:27:00.240 head and a torso and then like traction um wed 0:27:00.520 --> 0:27:04.520 shreds uh and it can go anywhere over terrain, it 0:27:04.520 --> 0:27:07.439 can go up steps. Probably. The thing is is again 0:27:07.600 --> 0:27:10.760 you have to remember when it comes to humanoid robots, 0:27:10.800 --> 0:27:12.960 you're trying to make the robot that can adapt to 0:27:13.000 --> 0:27:16.240 the human world. So if you had somebody like Johnny 0:27:16.280 --> 0:27:19.000 five as your house butler or something, you you couldn't 0:27:20.359 --> 0:27:22.560 You couldn't have an island in your kitchen, and who 0:27:22.600 --> 0:27:25.439 doesn't love an island in their kitchen. Johnny five couldn't 0:27:25.440 --> 0:27:27.639 maneuver around it because he's too wide. Yeah, you wouldn't 0:27:27.640 --> 0:27:29.840 have You wouldn't be able to have any any space 0:27:29.840 --> 0:27:33.040 that would be narrower than the robot's body exactly. That's 0:27:33.040 --> 0:27:36.240 not what you want. Yeah, with humanoid robots, it wouldn't 0:27:36.240 --> 0:27:38.199 work well. In my house. I've got a I've got 0:27:38.320 --> 0:27:44.120 a a flat style house where there's three floors. Yeah, 0:27:44.560 --> 0:27:48.080 it's like flat, like European flat. Uh, not flat as 0:27:48.119 --> 0:27:50.720 in there's only one level. They're actually three of them, 0:27:51.160 --> 0:27:53.840 four if you count the rooftop denck, so that counts. 0:27:53.960 --> 0:27:56.280 It makes it. You know, any robot that would not 0:27:56.280 --> 0:27:59.600 be able to navigate stairs easily would definitely have an issue, 0:27:59.600 --> 0:28:01.480 which is the main reason why I don't have a rumba, 0:28:01.560 --> 0:28:03.000 because I don't want to hear the sound of a 0:28:03.040 --> 0:28:07.520 roomba going falling down a flight up stairs. Um, but 0:28:07.600 --> 0:28:10.679 at any rate, Uh, that's a great point. Moving on 0:28:10.720 --> 0:28:14.800 from locomotion, there's also dexterous manipulation. Yeah, I think we should. 0:28:14.840 --> 0:28:18.359 I think that point bears repeating. What we just talked 0:28:18.359 --> 0:28:20.639 about is locomotion. Yeah, and this is you and I 0:28:20.800 --> 0:28:24.240 a couple of non robot experts talking about the problem 0:28:24.280 --> 0:28:29.119 with locomotion. That's just one of myriad challenges facing humanoid 0:28:29.280 --> 0:28:33.720 robotics designers. Yeah, yeah, exactly. It's It's one that's easy 0:28:33.800 --> 0:28:37.760 to point to because it's something that we all, you know, 0:28:38.160 --> 0:28:43.400 end up at least observing or participating in all the time. 0:28:43.640 --> 0:28:46.240 Can we take it for granted? But then when you think, okay, well, 0:28:46.240 --> 0:28:47.800 how do I make a machine that does that? You 0:28:47.840 --> 0:28:50.200 start to realize this is you know, even if I 0:28:50.280 --> 0:28:53.600 have a leg that has lots of different degrees of 0:28:53.600 --> 0:28:56.840 freedom and points of articulation, I still have to design 0:28:56.880 --> 0:28:58.680 the upper part of the robots so that it does 0:28:58.720 --> 0:29:01.840 not unbalance the lower part, and if it does on balance, 0:29:01.920 --> 0:29:04.760 it's able to catch itself. You know, some people just 0:29:04.840 --> 0:29:10.600 describe walking as falling and catching yourself over and over again. Yeah, yeah, 0:29:10.680 --> 0:29:15.440 they're walking right now. No, I I described walking as 0:29:15.480 --> 0:29:18.840 something that other people do. I like to keep my 0:29:18.920 --> 0:29:21.240 walking to a minimum. I thought you walked alot, Actually 0:29:21.280 --> 0:29:24.080 I do. I just joke about being lazy. I think 0:29:24.520 --> 0:29:27.920 moving forward, falling down and catching your balance every time 0:29:27.920 --> 0:29:32.480 it's lurching. Yeah, that's well. As an Adams Family fan, 0:29:32.600 --> 0:29:36.160 I'm okay with that. Yeah, but uh, Dexter's manipulation would 0:29:36.200 --> 0:29:40.160 be the ability to pick up and manipulate objects. Now, 0:29:40.200 --> 0:29:42.120 we're really good at that, we humans, You know, we 0:29:42.200 --> 0:29:45.400 can we can feel an object and decide at that 0:29:45.480 --> 0:29:48.880 point how to handle it, even if we've never encountered 0:29:48.880 --> 0:29:51.560 that kind of object before. So if I encounter something 0:29:51.600 --> 0:29:54.600 I've never seen before and i've I've ascertained it's safe 0:29:54.720 --> 0:29:56.800 for me to touch it, I can touch it. I 0:29:56.840 --> 0:29:58.239 can feel like, I can get a feel for how 0:29:58.280 --> 0:29:59.840 heavy it is, I can get a feel for how 0:30:00.000 --> 0:30:03.360 eloquant it might be, and then I can adjust on 0:30:03.440 --> 0:30:06.360 the fly so that I can handle it appropriately. Well, 0:30:06.360 --> 0:30:08.520 I'm not gonna hurt myself. I'm not gonna hurt the object. 0:30:09.040 --> 0:30:13.480 Robots are not so good at that. Yeah, exactly, even 0:30:13.520 --> 0:30:15.920 if they don't mean to right, Yeah, if the robots. 0:30:15.960 --> 0:30:18.240 If the robots grip is too strong, it can break 0:30:18.280 --> 0:30:20.520 the object. If it's too weak, the object slips from 0:30:20.520 --> 0:30:23.760 its grip and it falls. Uh. And it may not 0:30:23.840 --> 0:30:28.200 be able to distinguish between different types. So getting those 0:30:28.240 --> 0:30:31.880 tactile sensors where a robot can tell how tightly it's 0:30:31.920 --> 0:30:35.360 gripping something and how much pressure a particular object can 0:30:35.360 --> 0:30:39.080 take before you've reached the failure point is a big deal. Now, 0:30:39.120 --> 0:30:41.440 this is also a big deal for just making robots 0:30:41.520 --> 0:30:44.480 safe for humans to be around. It is a big deal. 0:30:44.560 --> 0:30:49.000 You know that the first fatality by robot occurred um 0:30:49.040 --> 0:30:52.240 at the business end of a robotic arm in a 0:30:52.560 --> 0:30:55.719 flat rock, Michigan in nine a man named Robert Williams 0:30:55.720 --> 0:30:59.440 who's working on a forward line. Yeah, his robot arm 0:30:59.560 --> 0:31:02.360 was moving a little slow for his taste in getting 0:31:02.400 --> 0:31:05.200 supplies down, so he climbed up to where the supplies were. 0:31:05.480 --> 0:31:07.760 The robot arms suddenly sped up and hit him in 0:31:07.800 --> 0:31:11.320 the head and killed him instantly. Wow. Yeah, I've I've 0:31:11.640 --> 0:31:15.239 had a chance to see some of these industrial robots. Uh. 0:31:15.320 --> 0:31:17.680 And I would say up close, but you can't because 0:31:18.120 --> 0:31:22.520 because of instances like that. Industrial robots usually have lots 0:31:22.560 --> 0:31:27.200 of of safety barriers around them because it's not safe 0:31:27.200 --> 0:31:31.160 to be near those robots when they're in operation. They 0:31:31.200 --> 0:31:33.880 they can't react exactly, So you leave it up to 0:31:33.920 --> 0:31:36.480 the humans to stay away from the robots because the road. 0:31:36.480 --> 0:31:38.480 We haven't gotten to the point where the robots no, 0:31:38.600 --> 0:31:40.880 did not crush you or hit you in the head. Right, Yeah, 0:31:40.920 --> 0:31:42.760 I got the robots fall. I'm looking forward to that 0:31:42.840 --> 0:31:45.040 day when they figure out not to crush me. Yes, 0:31:45.680 --> 0:31:49.000 it's been pretty lousy days so far. Uh. Yeah, when 0:31:49.000 --> 0:31:52.840 I when I toured the Georgia Tech Robotics Lab, they 0:31:52.920 --> 0:31:55.760 talked specifically about this. This is a real challenge having 0:31:55.840 --> 0:31:58.600 robots recognize and react in a way that's going to 0:31:58.680 --> 0:32:03.440 be safe around humans. And uh, but dexterous manipulation is 0:32:03.440 --> 0:32:05.760 only that's only a part of dexterous manipulation. Obviously. The 0:32:05.840 --> 0:32:09.160 rest of it is again that object recognition and handling 0:32:09.240 --> 0:32:11.120 so that you're not destroying whatever it is you're trying 0:32:11.120 --> 0:32:15.680 to pick up. Um. Another big challenge in designing robots 0:32:15.800 --> 0:32:19.680 in general, not just humanoid robots, is just the the perception, 0:32:19.720 --> 0:32:24.200 the sensory perception of the robot. Yeah, so you know, 0:32:24.280 --> 0:32:28.280 whether it's optical systems like actual cameras in the place 0:32:28.320 --> 0:32:31.520 of eyes, or infrared so that you can see even 0:32:31.520 --> 0:32:35.440 in low light situations, radar, light ar, I mean, there's 0:32:35.480 --> 0:32:40.200 tons of different ways of sensing. Yep, there's uh, there's 0:32:40.640 --> 0:32:42.719 you know sensing obviously, it's not just site. Then you 0:32:42.760 --> 0:32:45.520 have to have the sound. That's a really tricky one 0:32:45.560 --> 0:32:49.600 actually because for us humans, we we can kind of 0:32:49.680 --> 0:32:52.600 zero in on what's important. Right, So if we're if 0:32:52.640 --> 0:32:54.560 you and I were at a party, which you know, 0:32:54.960 --> 0:32:57.720 someone made a mistake and invited me, Uh, we could 0:32:57.720 --> 0:33:01.440 have a conversation and be able to carry that conversation 0:33:01.440 --> 0:33:04.400 on even within the context of a big, bustling party 0:33:04.720 --> 0:33:06.920 because we can focus on what the other person is saying. 0:33:06.960 --> 0:33:10.360 It's called latent inhibition. Yeah. So when you don't have that, 0:33:10.360 --> 0:33:14.280 that schizophrenia, Yeah, you you can't separate the the signal 0:33:14.360 --> 0:33:17.200 from the noise, and everything either becomes noise or everything 0:33:17.240 --> 0:33:20.720 becomes signal. Um. So for a robot that might for example, 0:33:20.880 --> 0:33:25.120 require verbal commands, that's really tricky. What if you have 0:33:25.160 --> 0:33:30.360 the television on and someone's saying something on TV and uh, 0:33:30.640 --> 0:33:33.040 you are trying to get your robot to do something 0:33:33.120 --> 0:33:34.880 and it's not quite sure what to do because it's 0:33:34.880 --> 0:33:38.560 hearing these different commands and isn't sure who to obey. Yeah, 0:33:38.640 --> 0:33:42.200 you are the person pitching the bacon bowl right right exactly. 0:33:42.360 --> 0:33:44.040 Let's say that you are, you know, trying to get 0:33:44.040 --> 0:33:47.040 some help in the kitchen, but it just keeps hearing U, 0:33:47.240 --> 0:33:49.520 C S I Miami and say kill Billy. And then 0:33:49.600 --> 0:33:51.960 next thing you know, you're like, you're just desperately trying 0:33:51.960 --> 0:33:54.960 to kill the robot. Please don't kill Billy. Um. Yeah, 0:33:56.040 --> 0:33:59.120 well that's a silly example. It's a real problem. Uh. 0:33:59.160 --> 0:34:02.960 And then there's the tactile the responses, the tactile sensors, 0:34:03.000 --> 0:34:05.800 like the making sure you don't crush something that's delicate, 0:34:05.840 --> 0:34:10.200 that that falls into perception. Smell can also fall into perception. 0:34:10.719 --> 0:34:13.279 You might want to have humanoid robots that work in 0:34:13.360 --> 0:34:17.480 areas where the humanoid robot can alert humans to the 0:34:17.480 --> 0:34:20.160 presence of things that might be toxic. You know, this 0:34:20.239 --> 0:34:23.840 isn't necessarily just the robot butler we're talking about. This 0:34:23.920 --> 0:34:27.319 could be robots that work in areas that might be 0:34:27.719 --> 0:34:31.239 dangerous for humans, and that would be an important element too. 0:34:31.719 --> 0:34:34.080 It's not a robot. But NASA already has a sensor 0:34:34.320 --> 0:34:38.040 that senses things like ammonia or smoke. It can actually 0:34:38.280 --> 0:34:45.799 sense smoke artificially smell smoke before the fires actually started ignited. Interesting, yeah, 0:34:45.880 --> 0:34:48.839 because you know it's such a dangerous proposition, right, yes, 0:34:48.920 --> 0:34:52.759 clearly for for anything NASA related. But they can also 0:34:52.840 --> 0:34:55.160 sense ammonia because you know a lot of the refrigeration 0:34:55.200 --> 0:34:59.000 systems run on ammonia and you can't have aneumonia leak 0:34:59.239 --> 0:35:02.680 on the space right right, And I mean the same 0:35:02.680 --> 0:35:04.839 thing is true for I mean I've heard of of 0:35:05.520 --> 0:35:09.880 robots that are used in in mining operations, which you know, 0:35:09.920 --> 0:35:12.080 if you come upon a pocket of natural gas that 0:35:12.120 --> 0:35:14.640 can be a real danger that sort of stuff. Then 0:35:14.680 --> 0:35:18.440 you've got the the back end of the sensory perception. 0:35:18.680 --> 0:35:22.360 That's where you have the actual interpretation of the data. 0:35:22.840 --> 0:35:25.120 Where that's a big one. It's huge because not only 0:35:25.120 --> 0:35:27.200 do you need to have a robot that can have 0:35:27.280 --> 0:35:31.720 that has binocular vision so it has a depth of field, right, um, 0:35:31.760 --> 0:35:35.480 it also has to know what it's doing, what what 0:35:35.520 --> 0:35:40.080 the information means, and how to apply it to adapted changes, 0:35:40.360 --> 0:35:43.680 right right. So So if I were to show you 0:35:43.840 --> 0:35:46.960 josh a series of pictures of various types of dogs. 0:35:47.520 --> 0:35:50.400 You would very quickly pick up on the the things 0:35:50.480 --> 0:35:53.880 that mean mean dog like, you would understand the concept 0:35:53.920 --> 0:35:59.120 of dog pretty quickly. Robots and various other computers machines, 0:35:59.600 --> 0:36:02.400 they have a lot harder problem with this. If whatever 0:36:02.480 --> 0:36:06.239 they're looking at doesn't exactly match the parameters of the example. 0:36:06.880 --> 0:36:10.520 It's very difficult for a machine to extrapolate and say, oh, 0:36:10.640 --> 0:36:13.359 this other thing I'm looking at relates to this thing 0:36:13.440 --> 0:36:17.719 I know, even though the two examples don't aren't identical. 0:36:18.360 --> 0:36:20.400 So the same thing could be true for any object. 0:36:20.520 --> 0:36:23.560 Let's let's just use a coffee mug, and let's say 0:36:23.640 --> 0:36:27.560 that you use a plain white coffee mug of average 0:36:27.640 --> 0:36:32.200 size as the example for the robot, and then the 0:36:32.280 --> 0:36:36.440 robot encounters a larger blue coffee mug and the handle 0:36:36.480 --> 0:36:39.200 has turned the other way. The robot might be completely 0:36:39.200 --> 0:36:42.600 befuddled by this. So this is a real problem in 0:36:42.760 --> 0:36:47.960 artificial intelligence. Is object identification, so that a robot knows 0:36:48.000 --> 0:36:51.480 what it's looking at and also understands the context that 0:36:51.480 --> 0:36:56.239 that object fills within the environment. So it's not just that, oh, 0:36:56.280 --> 0:36:59.280 that's a mug, it's oh, that's a mug. A mug 0:36:59.320 --> 0:37:01.640 is a container. I can put things into that mug. 0:37:01.800 --> 0:37:03.279 You are the things that can go in the mug. 0:37:03.320 --> 0:37:05.319 Here are the things that absolutely should not go in 0:37:05.320 --> 0:37:07.840 the mug, like Billy. Those are the kind of things 0:37:07.880 --> 0:37:11.160 that yeah, Billy, well you know we're gonna need another Billy. Um, 0:37:11.640 --> 0:37:14.239 we'll make a robot Billy. Yeah, which in which case 0:37:14.280 --> 0:37:16.560 you can just turn those suckers out right, mass production 0:37:16.600 --> 0:37:20.520 of Billy. But yeah, artificial intelligence an enormous problem. And that, 0:37:20.600 --> 0:37:22.840 of course is not just with robotics. That's that's a 0:37:22.920 --> 0:37:27.200 field unto itself, and robotics is just one branch that 0:37:27.600 --> 0:37:31.920 relies upon artificial intelligence. And it's from what I came across, 0:37:31.960 --> 0:37:34.120 it looked like just out of the gate, I guess 0:37:34.120 --> 0:37:37.440 that was said a university, they tried to build a 0:37:37.560 --> 0:37:41.600 robot that was just like high functioning, Yeah, and they realized, like, 0:37:41.800 --> 0:37:44.960 we have no idea what we're doing. Yeah, that that 0:37:45.000 --> 0:37:48.840 waybot one was able to converse at the level or 0:37:49.000 --> 0:37:53.520 was able to have a a a cognitive function equivalent 0:37:53.560 --> 0:37:56.320 to a one and a half year old person, which 0:37:56.560 --> 0:37:59.160 I have to say, when you're talking right out of 0:37:59.160 --> 0:38:02.879 the gate, Yeah, impressive, very impressive, because we're not that 0:38:02.960 --> 0:38:06.840 much further along now. But what they found from making 0:38:06.840 --> 0:38:10.960 waybot one was, Okay, this is way more difficult than 0:38:10.960 --> 0:38:15.360 we thought. You can't just program every kind of coffee 0:38:15.360 --> 0:38:18.480 cup in the world, and even if you could, then 0:38:18.520 --> 0:38:21.080 you also have to program every kind of table and 0:38:21.080 --> 0:38:23.680 every kind of light. And we need to come at 0:38:23.680 --> 0:38:27.319 this in a different way. And so they realized number one, 0:38:27.560 --> 0:38:32.480 humans are extraordinarily more complex than we thought before. And 0:38:32.520 --> 0:38:35.640 then number two, humans make a pretty good model for 0:38:35.719 --> 0:38:39.480 a humanoid robot in the realm of things like perception 0:38:40.000 --> 0:38:45.520 and UM information systems and uh learning. So they went 0:38:45.640 --> 0:38:52.759 to these different these different disciplines like neurobiology, neurology, um psychology, 0:38:52.800 --> 0:38:55.279 and they said, what can we learn from you guys 0:38:55.320 --> 0:38:58.440 about how humans do this that we can apply to robots. 0:38:58.600 --> 0:39:03.759 And since they started taking those steps, it seems like, uh, 0:39:04.160 --> 0:39:07.920 humanoid robotics has gotten it's it's footing a little more. Yeah, 0:39:08.000 --> 0:39:10.719 and we're seeing so many developments in other areas of 0:39:10.760 --> 0:39:14.120 AI that are really promising. I always bring up IBM 0:39:14.160 --> 0:39:18.880 S Watson because it's natural language recognition was phenomenal. The 0:39:18.920 --> 0:39:23.200 ability for it to parse clues in jeopardy and come 0:39:23.280 --> 0:39:26.600 up with the appropriate answer, knowing that Jeopardy those clues 0:39:26.640 --> 0:39:30.200 are not always straightforward. Uh. And it again illustrates the 0:39:30.239 --> 0:39:34.759 complexity that we humans navigate without much trouble because this 0:39:34.840 --> 0:39:38.080 is the world we've created. But then we realize if 0:39:38.120 --> 0:39:40.480 we make a machine that's mostly when you get down 0:39:40.480 --> 0:39:43.480 to it based on yes or no, a one or 0:39:43.520 --> 0:39:46.360 a zero, true or false, and you're trying to build 0:39:46.400 --> 0:39:49.719 complex behaviors off of something that is incredibly simple. When 0:39:49.719 --> 0:39:52.719 you boil it down to its basic element, that's where 0:39:52.760 --> 0:39:54.800 you're like, oh, this is this is gonna require a 0:39:54.840 --> 0:39:57.680 lot of work. I mean, IBMS Watson was an enormous 0:39:57.719 --> 0:40:02.799 machine with with thoul of microprocessors just so it could 0:40:02.800 --> 0:40:07.080 be able to play Jeopardy. That's a very specific function too. 0:40:07.520 --> 0:40:18.440 We'll be right back after this next break. So, creating 0:40:18.480 --> 0:40:22.279 a robot that is able to navigate and interact with 0:40:22.320 --> 0:40:24.800 a human environment and be able to interact with humans 0:40:24.840 --> 0:40:27.600 in a way that makes sense is a big challenge. Also, 0:40:27.680 --> 0:40:31.120 just the way that a robot would socialize with humans 0:40:31.280 --> 0:40:33.480 is a huge challenge. How do how do you make 0:40:33.480 --> 0:40:38.000 a robot that is able to respond to commands and 0:40:38.200 --> 0:40:41.400 cues in an appropriate way? Uh? An appropriate way is 0:40:41.440 --> 0:40:44.360 the key there, because there are humans are pretty complex 0:40:44.360 --> 0:40:46.839 and we can be very subtle in many ways. Yeah, 0:40:47.000 --> 0:40:52.399 we speak unplainly, we use sarcasm, um, we we uh yeah, 0:40:52.480 --> 0:40:57.080 we use a lot of gestures rather than just words. Yep. 0:40:57.280 --> 0:40:59.759 There's a lot that goes into human communication that, if 0:40:59.800 --> 0:41:03.080 you are a human, is pretty much natural, especially I 0:41:03.080 --> 0:41:05.680 mean if you're a human within that particular culture and 0:41:05.719 --> 0:41:08.600 you're familiar with that culture, Because anyone who has traveled 0:41:08.600 --> 0:41:11.839 extensively knows there are cultures where things that would be 0:41:12.000 --> 0:41:15.600 commonplace at home are very different in the place where 0:41:15.640 --> 0:41:17.640 you happen to be, right then, and it may be 0:41:17.800 --> 0:41:22.240 that something that is completely innocent at home is an uh, 0:41:22.280 --> 0:41:25.200 offensive gesture in the place where you are. Now, we'll 0:41:25.239 --> 0:41:29.520 imagine a robot that is not programmed to handle these 0:41:29.560 --> 0:41:38.120 kind of subtle uh communication methods, and yeah, he doesn't know, 0:41:38.480 --> 0:41:41.560 he's just he's just doing as he was programmed. But 0:41:41.680 --> 0:41:44.319 even even beyond that, something that you brought up in 0:41:44.360 --> 0:41:48.120 our research when we were planning this was the Uncanny Valley. Yeah, 0:41:48.160 --> 0:41:50.680 that's it's a big one. Yeah, I've never read that 0:41:50.680 --> 0:41:54.000 paper before, and I'm glad I did. It's really interesting, right, Yeah. 0:41:54.040 --> 0:41:56.240 So the uncanny Valley, for those who are not familiar 0:41:56.239 --> 0:41:59.359 with the term, uh it's it describes when we start 0:41:59.440 --> 0:42:04.680 to appre coach artificial humans that look almost, but not 0:42:04.800 --> 0:42:07.759 quite like real humans. And and by look, I don't 0:42:07.800 --> 0:42:11.160 necessarily just mean the physical appearance. I also mean their behaviors, 0:42:11.160 --> 0:42:14.239 their movements. So if you are if you were to 0:42:14.360 --> 0:42:18.279 look out and see a figure that from the from 0:42:18.280 --> 0:42:20.560 a distance looked like it was a human figure, and 0:42:20.600 --> 0:42:22.880 you start walking towards it just thinking this is another person, 0:42:23.040 --> 0:42:25.400 and then they start moving in a very herky jerky motion, 0:42:25.719 --> 0:42:30.200 very mechanical motion, then you're likely going to have a 0:42:30.280 --> 0:42:34.799 negative emotional response. Um often revulsion is one of the 0:42:34.800 --> 0:42:39.560 words used us. Yeah, I mean I remember the C 0:42:39.719 --> 0:42:43.920 G I movies that would that were almost to the 0:42:44.000 --> 0:42:47.840 point of photo realism, where they look like people for 0:42:47.880 --> 0:42:50.399 the most part, but there's not something's just not quite 0:42:50.440 --> 0:42:56.800 right with the eyes. There's a good example, and Polar 0:42:56.800 --> 0:43:00.120 express polar expresses the way everyone very well because the 0:43:00.239 --> 0:43:02.359 uncanny Valley, that's what they blame it on. Yeah, and 0:43:02.440 --> 0:43:05.120 the same thing applies to robots. So in fact, I 0:43:05.160 --> 0:43:08.840 saw a robot that was really disturbing to me, it 0:43:08.880 --> 0:43:13.719 was really an art exhibit, an installation, and um it 0:43:13.800 --> 0:43:18.239 was a robot of Confucius in a cell filled with monkeys, 0:43:19.080 --> 0:43:22.720 and uh they're live monkeys, real monkeys, and the robot 0:43:22.800 --> 0:43:28.239 would just thrash around wildly. It was it was the 0:43:28.239 --> 0:43:30.920 stuff of nightmares that I'll show it to you after 0:43:30.960 --> 0:43:35.000 the show. Yeah. So these are all big challenges, and 0:43:35.040 --> 0:43:37.440 some of them are going to be harder for us 0:43:37.440 --> 0:43:39.920 than others. It may be that the engineering challenges of 0:43:40.000 --> 0:43:43.879 locomotion are solved well before we ever get a real 0:43:43.920 --> 0:43:46.520 grip on all the artificial intelligence problems. Or it could 0:43:46.520 --> 0:43:50.120 be the other way around. Uh we but it is multidisciplinary. 0:43:50.320 --> 0:43:53.720 It's a big, big issue. So there are some people 0:43:53.719 --> 0:43:57.400 who argue for humanoid robots, there are people who argue 0:43:57.440 --> 0:44:02.200 against humanoid robots. Where do you, um, I think. I 0:44:02.200 --> 0:44:06.240 think research and development with humanoid robots is important because 0:44:06.280 --> 0:44:09.680 by having the goal of creating a humanoid robot, you 0:44:09.800 --> 0:44:13.320 drive the research and development process. You have a specific 0:44:13.320 --> 0:44:15.640 goal in mind, and in order to achieve that goal, 0:44:15.880 --> 0:44:18.200 you know what sort of problems you have to solve. 0:44:18.760 --> 0:44:21.880 And even if we never enter a future where humanoid 0:44:21.960 --> 0:44:27.080 robots are a common thing, even if they are mostly 0:44:27.200 --> 0:44:31.920 used as something in an exhibition or uh for pr 0:44:32.040 --> 0:44:34.439 or whatever. Even if that's the only use for them, 0:44:34.920 --> 0:44:38.359 We're going to benefit from the research and development of 0:44:38.440 --> 0:44:42.399 making that possible in ways we can't anticipate. Well. Yeah, 0:44:42.400 --> 0:44:45.560 and then the more we get into humanoid robotics, the 0:44:45.600 --> 0:44:50.000 more we understand humans, which is pretty much the only 0:44:50.360 --> 0:44:54.120 argument I've seen that stands up in favor of doing 0:44:54.200 --> 0:44:57.640 humanoid robots. Yeah, because it's it's expensive, and it's hard. 0:44:57.960 --> 0:45:00.719 I mean, it's it's really a different cult problem, and 0:45:00.760 --> 0:45:03.239 it's and and to build a humanoid robot something that 0:45:03.360 --> 0:45:07.960 is capable of being a general purpose robot, it's you know, 0:45:08.480 --> 0:45:10.680 it's hard to anticipate all the things you're going to 0:45:10.719 --> 0:45:12.960 need to be able to do. If you're talking general 0:45:13.040 --> 0:45:16.279 purpose and adaptable, that's really tough. I mean, we didn't 0:45:16.280 --> 0:45:19.239 even talk about the adaptability problem very much. We talked 0:45:19.239 --> 0:45:21.719 a little bit about a robot capable of learning from 0:45:21.719 --> 0:45:25.399 other people, which I find fascinating. Um, you know, that's 0:45:25.440 --> 0:45:29.640 one way, just watching humans and then mimicking humans, that's 0:45:29.680 --> 0:45:34.759 one way of robot learning. There's also, um the way 0:45:34.760 --> 0:45:38.319 where it's controlled through virtual reality by a human and 0:45:38.360 --> 0:45:41.040 it just kind of logs the motions the humans making 0:45:41.120 --> 0:45:43.320 it do. Like there's a NASA has a robot not 0:45:43.520 --> 0:45:48.080 that learns like that. So I think I think there's 0:45:48.120 --> 0:45:52.960 a lot of benefit to investigating artificial intelligence. Like you 0:45:53.000 --> 0:45:56.239 want your nest at home to learn so you don't 0:45:56.239 --> 0:45:59.440 have to keep adjusting the thermostat that counts. That's the 0:45:59.640 --> 0:46:05.120 that's sheen learning to me. The big argument against having 0:46:05.400 --> 0:46:10.120 humanoid robots, and the quagmire that seems to begin, is sociability. 0:46:10.719 --> 0:46:13.239 That seems to be the whole reason anybody wants a 0:46:13.320 --> 0:46:15.960 humanoid robot, because you can make you know, you have 0:46:16.000 --> 0:46:20.160 a ruma vacuums. You can make a driverless car um 0:46:20.200 --> 0:46:23.799 as Olivia Solan wrote and wired a couple of years back. Um, 0:46:25.040 --> 0:46:26.520 you know, why why do you have to make a 0:46:26.600 --> 0:46:28.839 robot butler to park the car? Just make a car 0:46:28.920 --> 0:46:32.160 that parks itself. And it seems like that's where we're 0:46:32.200 --> 0:46:35.800 going right now. Um. So when you add this extra 0:46:35.880 --> 0:46:41.080 layer of humanoid, you add all of this additional problems 0:46:41.160 --> 0:46:44.600 and troubles and redundancies, like like, for example, if you're 0:46:44.600 --> 0:46:48.279 gonna make a humanoid robot that throws a ball, this 0:46:48.440 --> 0:46:52.759 this humanoid robot to appear real needs to have a 0:46:52.760 --> 0:46:56.040 little bit of follow through. But as far as the 0:46:56.120 --> 0:46:58.279 robot the machine is concerned, it can throw the ball 0:46:58.360 --> 0:47:01.600 and just stop where the releases. It doesn't need to 0:47:01.600 --> 0:47:04.560 go anymore. But it's gonna look weird and robotic if 0:47:04.560 --> 0:47:06.960 you want to get past that uncanny valley, which is 0:47:07.000 --> 0:47:10.080 another problem. Um, the thing has to have followed through. 0:47:10.400 --> 0:47:13.799 That's totally unnecessary. You can make a robot that can 0:47:13.840 --> 0:47:16.800 throw a ball and the goal is to throw the ball. 0:47:17.520 --> 0:47:19.719 You don't have to add the follow through, but you 0:47:19.840 --> 0:47:23.840 do when you're making it a sociable humanoid robots. So 0:47:23.880 --> 0:47:27.280 it seems like that's the path that will lead everyone 0:47:27.320 --> 0:47:30.160 is straight that I don't get. Yeah, I I like 0:47:30.320 --> 0:47:34.359 the idea of designing robots for specific tasks because you 0:47:34.400 --> 0:47:37.239 can really focus on getting the task done. So there 0:47:37.320 --> 0:47:39.960 I see this as two separate branches. I see the 0:47:40.000 --> 0:47:43.799 branch of developing the humanoid robot as pushing forward a 0:47:43.880 --> 0:47:47.320 lot of different areas of thought that could be applied 0:47:47.360 --> 0:47:51.319 in multiple disciplines, so that will benefit from that. I 0:47:51.400 --> 0:47:55.600 see the development of robots as unitaskers as being important 0:47:55.600 --> 0:47:58.799 to actually handle the jobs that are the three D S. 0:47:58.920 --> 0:48:02.319 That's dirty, different called dangerous, all right, So those are 0:48:03.760 --> 0:48:07.120 the jobs that maybe they revolve a lot of repetition, 0:48:07.560 --> 0:48:10.799 which can cause injury over time, or it can lead 0:48:10.840 --> 0:48:13.759 to mistakes because you've done the same task so many 0:48:13.760 --> 0:48:16.240 times that you start to kind of zone out. Robots 0:48:16.280 --> 0:48:19.800 will never zone out. Um, if it's a dirty job 0:48:19.880 --> 0:48:23.120 where it's something that's undesirable by people, robots don't care, 0:48:23.360 --> 0:48:27.000 they'll do that. Or if it's dangerous, if it's bomb disposal, 0:48:27.280 --> 0:48:31.399 or if it's something like the Mars Curiosity rover. These 0:48:31.400 --> 0:48:34.759 are dangerous jobs that you wouldn't necessarily want to put 0:48:34.760 --> 0:48:39.120 a human into if you had the alternative. So all 0:48:39.160 --> 0:48:42.120 of these things, those that's where robots really makes sense 0:48:42.120 --> 0:48:44.600 to me um to to go to the places that 0:48:44.680 --> 0:48:46.880 are difficult for us to go to. Maybe that's you know, 0:48:47.160 --> 0:48:51.080 deep sea exploration, space exploration, that kind of thing, or 0:48:51.280 --> 0:48:53.840 to do jobs that might be dangerous having a first 0:48:53.880 --> 0:48:56.799 respond to robot to survey a scene to make sure 0:48:56.880 --> 0:49:01.000 that a structure is remaining solid while maybe there was 0:49:01.040 --> 0:49:02.920 a fire and it has to make sure that the 0:49:03.880 --> 0:49:08.080 it's not going to collapse in on rescue mission that 0:49:08.160 --> 0:49:10.800 kind of stuff. Um, But do you need those things 0:49:10.840 --> 0:49:13.000 to be to come out and be able to tell 0:49:13.040 --> 0:49:16.640 a joke or something. And most of them don't need 0:49:16.719 --> 0:49:20.239 to be any sort of humanoid form factor either, which 0:49:20.880 --> 0:49:24.480 greatly simplifies the actual development of the robot and thus 0:49:24.520 --> 0:49:26.920 cuts down on the cost, so you can you can 0:49:26.920 --> 0:49:30.439 achieve the task you're trying to achieve for less money 0:49:30.520 --> 0:49:32.440 than if you are trying to build this this general 0:49:32.520 --> 0:49:36.480 purpose machine. So then we come to this ultimate question, 0:49:37.440 --> 0:49:43.000 why what's the purpose of humanoid robot? Well, I think 0:49:43.000 --> 0:49:46.080 the I think the purposes too fold. One is again 0:49:46.200 --> 0:49:51.200 to have that specific goal in mind that allows you 0:49:51.239 --> 0:49:56.359 to define where your indpoint is. I believe that when 0:49:56.360 --> 0:49:59.600 you have that defined in goal, it makes it easier 0:49:59.640 --> 0:50:01.520 for you to build on the things you need to 0:50:01.560 --> 0:50:04.200 achieve it, as opposed to having an open goal where 0:50:04.200 --> 0:50:07.680 it's just I want to approve improve a I that's 0:50:07.719 --> 0:50:09.920 so open that it's hard to get direction from it. 0:50:09.960 --> 0:50:12.600 But if you think I need to have an artificial 0:50:12.640 --> 0:50:16.400 intelligence that will allow a robot to uh, here's a 0:50:16.400 --> 0:50:18.680 great example. Let's say that the challenge is to have 0:50:18.719 --> 0:50:23.000 a robot leave a room, go down a flight of stairs, 0:50:23.560 --> 0:50:27.040 leave a building, get into a vehicle, drive the vehicle 0:50:27.080 --> 0:50:30.440 to a different location, get out of the vehicle, go 0:50:30.480 --> 0:50:33.919 into another building, break through a wall and put out 0:50:33.920 --> 0:50:39.280 a fire. That's a real, actual robotics challenge challenge. Yeah. 0:50:39.360 --> 0:50:42.120 Dr Henrik Kristensen told me about this. Is I mean, 0:50:42.160 --> 0:50:44.600 it really is a challenge. It's not just a real challenge. 0:50:44.640 --> 0:50:48.839 It's a real challenge. It's like a DARPA challenge. So 0:50:49.000 --> 0:50:51.600 it's a uh he was telling me about this, and 0:50:51.680 --> 0:50:53.560 you start to think about all the things that have 0:50:53.640 --> 0:50:54.920 to fall in line for you to be able to 0:50:54.920 --> 0:50:57.200 achieve the skull That is a valuable thing. But I 0:50:57.239 --> 0:50:59.319 think the other thing is the social aspect. I think 0:50:59.360 --> 0:51:02.720 that there are people who would benefit from a robot 0:51:02.840 --> 0:51:06.560 that is able to give some form of social comfort. 0:51:06.640 --> 0:51:11.080 Let's say for the elderly who need to have some 0:51:11.120 --> 0:51:14.960 form of interaction. Um. You know, that could actually be 0:51:15.080 --> 0:51:17.520 a really valuable tool. And in fact, there's a lot 0:51:17.520 --> 0:51:20.919 of work that's going into robotics to help people like 0:51:21.040 --> 0:51:27.760 the elderly who may have real emotional and psychological problems, um, 0:51:27.840 --> 0:51:31.000 due to loneliness. Do you think that robots are the 0:51:31.000 --> 0:51:34.520 answer to that. I think that robots can help. I 0:51:34.560 --> 0:51:36.480 don't know, I would never go so far as to 0:51:36.520 --> 0:51:39.759 say answer, but couldn't you also make the argument that 0:51:40.280 --> 0:51:46.720 if you created robots that displaced human jobs and also 0:51:46.800 --> 0:51:53.640 simultaneously said, hey, this, this nursing home sector is about 0:51:53.680 --> 0:51:56.120 to explode because we've got a bunch of baby boomers, 0:51:56.120 --> 0:51:59.920 and ways the society of now decided that our elderly 0:52:00.040 --> 0:52:05.600 need human interaction more than we've more than we've carried 0:52:05.600 --> 0:52:08.920 it out before. So let's create this whole other industry. 0:52:09.000 --> 0:52:13.400 Or let's expand this industry of elderly caretakers and fill 0:52:13.760 --> 0:52:17.800 those jobs with people who have been displaced by worker robots. 0:52:18.239 --> 0:52:22.720 Wouldn't that be better? That might be, or you could again, 0:52:22.960 --> 0:52:25.640 looking at the way a lot of roboticists framed this, 0:52:26.320 --> 0:52:28.920 they say, all right, well, it is a reality that 0:52:29.080 --> 0:52:34.600 robots are taking over actual jobs. But the hope is 0:52:34.640 --> 0:52:38.239 that it also ends up creating new jobs that are 0:52:38.280 --> 0:52:43.640 better paying jobs, less dangerous jobs friendly, more old books friendly. 0:52:44.480 --> 0:52:47.680 But yeah, like um, the idea being that that it 0:52:47.800 --> 0:52:52.080 frees up people and encourages the pursuit of jobs and 0:52:52.200 --> 0:52:57.560 engineering in computer science. Now, we live in the real world, 0:52:57.640 --> 0:53:00.759 and we understand that it's a lot more complex than 0:53:00.840 --> 0:53:03.720 telling someone who's been working on a manufacturing line. Hey, 0:53:03.960 --> 0:53:06.239 I'm sorry your job's gone because there's a robot here. 0:53:06.239 --> 0:53:08.799 But guess what, we have an opening and engineering. So 0:53:08.840 --> 0:53:11.560 if you just go and pursue a four year degree 0:53:11.600 --> 0:53:14.200 and then some post graduate work, you'll be right back 0:53:14.239 --> 0:53:17.799 to work. That's that's obviously not uh something that's going 0:53:17.840 --> 0:53:21.120 to be easy, especially in the short term. But the 0:53:21.200 --> 0:53:24.000 long term hope is that more and more of these 0:53:24.080 --> 0:53:27.719 jobs that are are dangerous for people, are less desirable 0:53:27.760 --> 0:53:31.480 for people, will be taken up by robots, and then 0:53:31.680 --> 0:53:35.280 the will there will be the creation of better jobs 0:53:35.320 --> 0:53:37.000 that are higher up on the food chain. And I 0:53:37.040 --> 0:53:39.759 think that makes sense to me. It's just the once 0:53:39.840 --> 0:53:44.120 you enter the sociability, yeah, because without sociability, you there's 0:53:44.160 --> 0:53:47.279 no reason to create a humanoid robot. Everything else can 0:53:47.320 --> 0:53:51.239 look like a robot. Yeah. Um, so it's it's when 0:53:51.239 --> 0:53:53.920 you enter sociability that you lose me. Not only does 0:53:53.920 --> 0:53:55.839 it can it look like a robot, but we can 0:53:55.880 --> 0:53:58.239 still socialize with it even if it doesn't look like 0:53:58.280 --> 0:54:01.440 a human people. There. You know, there's the story Rumba 0:54:01.520 --> 0:54:05.000 owners named their rumba. Yeah, so you we we end 0:54:05.080 --> 0:54:09.360 up having these kind of emotional attachments and investments in 0:54:10.000 --> 0:54:12.560 things that don't look Not only do they not look human, 0:54:12.640 --> 0:54:15.680 they don't look like any other animal that we would 0:54:15.680 --> 0:54:20.160 interact with on a like owner or and pet or whatever. 0:54:20.200 --> 0:54:23.440 I mean, they they're They're just a robot. So I 0:54:23.440 --> 0:54:25.160 think at the end of the day, Josh, I think 0:54:25.200 --> 0:54:27.279 we're on the same page. We think humanoid robots are 0:54:27.280 --> 0:54:30.640 an interesting idea, but not necessarily the end goal. There's 0:54:30.719 --> 0:54:34.120 there's not a whole lot of of incentive to go 0:54:34.239 --> 0:54:36.880 after it for its own purposes. We can see the 0:54:36.920 --> 0:54:39.319 benefits of going after it in the sense of the 0:54:39.360 --> 0:54:42.239 developments that are made in that pursuit help us in 0:54:42.280 --> 0:54:45.600 other ways. And that wraps up this classic episode of 0:54:45.640 --> 0:54:47.680 tech Stuff. I hope you enjoyed it. If you have 0:54:47.800 --> 0:54:50.680 suggestions for topics I should cover in future episodes, please 0:54:50.719 --> 0:54:52.880 reach out to me on Twitter. The handle for the 0:54:52.880 --> 0:54:55.960 show is tech Stuff h SW and I'll tell to 0:54:56.000 --> 0:55:04.640 you again really soon. Text Stuff is an I Heart 0:55:04.760 --> 0:55:08.479 Radio production. For more podcasts from I Heart Radio, visit 0:55:08.520 --> 0:55:11.600 the I Heart Radio app, Apple Podcasts, or wherever you 0:55:11.680 --> 0:55:13.000 listen to your favorite shows.