WEBVTT 00:00:00.000 --> 00:00:02.140 Welcome to the Deep Dive. I am your host. And 00:00:02.140 --> 00:00:04.700 today we are venturing into a topic that is honestly 00:00:04.700 --> 00:00:06.919 moving so fast it feels like science fiction. 00:00:07.179 --> 00:00:09.119 Oh, absolutely. It's moving at breakneck speed. 00:00:09.279 --> 00:00:12.060 Right. And for you listening at home, our mission 00:00:12.060 --> 00:00:15.619 today is to give you a fast, thorough and hopefully 00:00:15.619 --> 00:00:18.820 really engaging shortcut to understanding this 00:00:18.820 --> 00:00:22.100 space. We are pulling all our facts today entirely 00:00:22.100 --> 00:00:25.359 from one massive comprehensive source, which 00:00:25.359 --> 00:00:28.140 is the Wikipedia article on robotics. Yeah, it's 00:00:28.140 --> 00:00:30.809 a fantastic primer. because it covers so much 00:00:30.809 --> 00:00:34.250 ground from the nuts and bolts to the really 00:00:34.250 --> 00:00:36.729 philosophical stuff. Exactly, because this is 00:00:36.729 --> 00:00:39.049 a field that is literally reshaping our daily 00:00:39.049 --> 00:00:41.329 lives and the global economy. So we've got our 00:00:41.329 --> 00:00:43.270 resident expert here to help decode all this. 00:00:43.450 --> 00:00:45.090 Thrilled to be here. There's a lot to cover. 00:00:45.270 --> 00:00:47.929 There really is. OK, let's unpack this. To understand 00:00:47.929 --> 00:00:50.590 what robots will be capable of tomorrow, we first 00:00:50.590 --> 00:00:52.429 have to look at how they're built today, right? 00:00:52.530 --> 00:00:54.030 Right. You have to start with the anatomy. Yeah. 00:00:54.250 --> 00:00:56.850 And the source breaks down four core aspects 00:00:56.850 --> 00:01:00.250 of robotics. You've got power. mechanics, control, 00:01:00.770 --> 00:01:02.770 and software. Which is a great way to categorize 00:01:02.770 --> 00:01:05.189 it. But I want to focus on the mechanics first. 00:01:05.670 --> 00:01:08.150 Specifically the actuators. Basically the muscles 00:01:08.150 --> 00:01:10.349 of the machine. Oh, the actuators are where the 00:01:10.349 --> 00:01:13.489 magic happens. Because we are so far past... 00:01:13.469 --> 00:01:16.430 clunky electric motors, like the source talks 00:01:16.430 --> 00:01:19.170 about piezoelectric motors. And it says they 00:01:19.170 --> 00:01:21.730 use piezo -ceramic elements that vibrate thousands 00:01:21.730 --> 00:01:24.230 of times per second to create movement. They 00:01:24.230 --> 00:01:27.250 do, yeah. But mechanically, how does that work? 00:01:27.329 --> 00:01:30.510 Like if my phone vibrates on a table, it just 00:01:30.510 --> 00:01:32.670 buzzes in place. It doesn't, you know, drive 00:01:32.670 --> 00:01:34.650 a gear forward. That is a perfect distinction 00:01:34.650 --> 00:01:37.170 to make because your phone vibrates using a tiny 00:01:37.170 --> 00:01:39.230 off -center weight spinning around inside. Oh, 00:01:39.230 --> 00:01:41.109 so it just throws the whole device off balance. 00:01:41.189 --> 00:01:43.900 Exactly. It's just wobbling. Right. But piezoelectric 00:01:43.900 --> 00:01:46.359 materials operate on a completely different principle. 00:01:46.680 --> 00:01:49.540 OK. How so? Well, when you run an electrical 00:01:49.540 --> 00:01:52.819 current through a piezo ceramic, its actual physical 00:01:52.819 --> 00:01:54.879 shape changes. Right. The shape of the solid 00:01:54.879 --> 00:01:57.260 material changes? Yeah, it expands and contracts 00:01:57.260 --> 00:02:00.340 on a microscopic level. So engineers arrange 00:02:00.340 --> 00:02:03.000 these elements in a ring and time the electrical 00:02:03.000 --> 00:02:05.640 pulses to create a microscopic traveling wave 00:02:05.640 --> 00:02:08.639 along the surface. Almost like the way fans do 00:02:08.639 --> 00:02:11.719 in a sports stadium. Exactly like that, but happening 00:02:11.719 --> 00:02:14.520 ultrasonically, thousands of times a second. 00:02:14.580 --> 00:02:17.800 Wow. And if you press a rotor against that rippling 00:02:17.800 --> 00:02:20.879 surface, the peaks of the wave catch it. and 00:02:20.879 --> 00:02:23.520 push it along. So there's no heavy magnetic shaft 00:02:23.520 --> 00:02:25.900 spinning around. Right. You get this nanometer 00:02:25.900 --> 00:02:28.419 resolution, a level of microscopic precision 00:02:28.419 --> 00:02:31.639 a human hand could never achieve. That's incredible. 00:02:32.280 --> 00:02:34.379 But the source also mentions something called 00:02:34.379 --> 00:02:38.340 carbon nanotubes. And this blew my mind. An 8 00:02:38.340 --> 00:02:40.659 -millimeter wire of this stuff could replace 00:02:40.659 --> 00:02:43.500 a human bicep. And feasibly outperform it, yeah. 00:02:43.580 --> 00:02:46.020 I mean, if we can replace a bicep with an 8 -millimeter 00:02:46.020 --> 00:02:49.340 carbon wire, aren't we blurring the line between 00:02:49.340 --> 00:02:51.939 mechanical engineering and synthetic biology? 00:02:52.400 --> 00:02:54.560 At what point is it less of a machine and more 00:02:54.560 --> 00:02:57.120 of a manufactured organism? What's fascinating 00:02:57.120 --> 00:03:00.080 here is the material science revolution. Because 00:03:00.080 --> 00:03:02.400 for decades, the limiting factor was always weight. 00:03:02.659 --> 00:03:04.919 Right, heavy metal gears. Exactly. If you wanted 00:03:04.919 --> 00:03:07.340 a stronger robot, you needed a bigger motor. 00:03:07.500 --> 00:03:09.599 A bigger motor needs a heavier battery. Which 00:03:09.599 --> 00:03:12.180 means a bulkier frame to hold the battery. You 00:03:12.180 --> 00:03:14.639 are trapped in this cascading loop of adding 00:03:14.639 --> 00:03:18.699 mass. But materials like carbon nanotubes... 00:03:18.680 --> 00:03:21.180 completely shatter that loop. Because they reduce 00:03:21.180 --> 00:03:23.979 the weight so drastically. Yes, which unlocks 00:03:23.979 --> 00:03:27.180 entirely new form factors that were just impossible 00:03:27.180 --> 00:03:29.819 with clunky traditional motors. But strength 00:03:29.819 --> 00:03:31.819 and precision are only half the battle, right? 00:03:31.939 --> 00:03:34.379 I mean, if a robot is too rigid, it becomes a 00:03:34.379 --> 00:03:36.539 liability. Oh, totally. A traditional robotic 00:03:36.539 --> 00:03:38.740 arm will just drive through whatever is in its 00:03:38.740 --> 00:03:40.639 way to reach its programmed coordinate. Which 00:03:40.639 --> 00:03:42.740 is terrifying if you happen to be standing in 00:03:42.740 --> 00:03:44.800 its way. Right. It'll destroy the obstacle and 00:03:44.800 --> 00:03:47.199 probably strip its own gears. That's why the 00:03:47.199 --> 00:03:50.580 source highlights series elastic actuation which 00:03:50.580 --> 00:03:53.639 is basically adding shock absorbers right precisely 00:03:53.639 --> 00:03:57.379 instead of a stiff connection engineers intentionally 00:03:57.379 --> 00:04:00.319 design elasticity into the joint so if it bumps 00:04:00.319 --> 00:04:03.979 into a wall it absorbs the shock yes it filters 00:04:03.979 --> 00:04:06.280 out the impact force yeah making it safer for 00:04:06.280 --> 00:04:08.500 humans and allowing the robot to store energy 00:04:08.500 --> 00:04:10.319 kind of like the tendons in your leg when you 00:04:10.319 --> 00:04:13.259 jump okay so that perfectly transitions us to 00:04:13.259 --> 00:04:16.240 the next big hurdle Getting around. Right. Because 00:04:16.240 --> 00:04:19.339 once you have these incredible muscles, the natural 00:04:19.339 --> 00:04:21.660 next step is figuring out how to harness that 00:04:21.660 --> 00:04:24.459 power to move through the world without falling 00:04:24.459 --> 00:04:27.060 on your face. Which is a massive computational 00:04:27.060 --> 00:04:29.949 challenge. Yeah. Especially on two legs. Bipedal 00:04:29.949 --> 00:04:32.230 walking is essentially just a controlled forward 00:04:32.230 --> 00:04:35.050 fall. Yeah, the source goes into the zero moment 00:04:35.050 --> 00:04:38.589 point approach, or ZMP. This was Honda's Asimo 00:04:38.589 --> 00:04:41.269 robot, right? Yes, the classic Asimo. From what 00:04:41.269 --> 00:04:43.529 I read, the computer is constantly calculating 00:04:43.529 --> 00:04:45.910 gravity and acceleration to make sure the floor 00:04:45.910 --> 00:04:48.110 pushing back up cancels out any force that would 00:04:48.110 --> 00:04:50.250 tip it over. That's the math behind it, yeah. 00:04:50.250 --> 00:04:53.769 Yeah. But the reality is ZMP requires very flat 00:04:53.769 --> 00:04:56.629 surfaces. And because it's terrified of momentum, 00:04:57.170 --> 00:04:59.870 it keeps its knees bent and its feet flat. Which 00:04:59.870 --> 00:05:02.089 is why people always joked that ASIMO looked 00:05:02.089 --> 00:05:04.029 like it desperately needed the bathroom. Exactly. 00:05:04.110 --> 00:05:06.509 It's fighting momentum instead of using it. Right. 00:05:06.629 --> 00:05:08.550 But then the source contrasts that with the MIT 00:05:08.550 --> 00:05:11.910 leg lab. They used dynamic balancing. Yes. Hopping 00:05:11.910 --> 00:05:14.209 on one leg like a pogo stick. Which is wild. 00:05:14.230 --> 00:05:16.110 If it started falling, it just hopped into the 00:05:16.110 --> 00:05:18.970 fall to catch its center of mass. Leaning into 00:05:18.970 --> 00:05:21.610 the momentum, just like a human does when sprinting. 00:05:21.759 --> 00:05:24.199 You're technically airborne and falling, trusting 00:05:24.199 --> 00:05:26.439 your foot to catch you. But wait, let me push 00:05:26.439 --> 00:05:28.740 back on this a bit. Because even with dynamic 00:05:28.740 --> 00:05:31.600 balancing, making a two -legged robot walk is 00:05:31.600 --> 00:05:34.319 insanely hard. It's incredibly difficult, yeah. 00:05:34.420 --> 00:05:37.000 So why do we keep forcing this bipedal issue? 00:05:37.139 --> 00:05:40.019 The source talks about biomimetics copying nature. 00:05:40.519 --> 00:05:43.920 We have robotic flyers inspired by bats and raptors 00:05:43.920 --> 00:05:46.759 that use in -plane wing deformation. Oh, where 00:05:46.759 --> 00:05:49.860 the wings change shape mid -flap. Beautiful engineering. 00:05:49.959 --> 00:05:52.560 Right, and robotic fish that achieve over 90 00:05:52.560 --> 00:05:55.459 % propulsive efficiency, they actually outperform 00:05:55.459 --> 00:05:58.819 real fish. They do. So if nature already perfected 00:05:58.819 --> 00:06:01.660 four legs or wings or fins, why force a square 00:06:01.660 --> 00:06:04.040 peg into a round hole with two legs? Well, it 00:06:04.040 --> 00:06:06.540 comes down to the human -built environment. Our 00:06:06.540 --> 00:06:09.620 entire world. The stairs, the doors, narrow hallways, 00:06:09.720 --> 00:06:12.199 it's all designed around the geometry of a bipedal 00:06:12.199 --> 00:06:15.199 human. Oh, I see. Yeah. If you want a robot to 00:06:15.199 --> 00:06:17.180 help out in a hospital, a robotic dog is very 00:06:17.180 --> 00:06:19.879 stable, sure. But it can't easily turn a door 00:06:19.879 --> 00:06:22.240 handle or climb a vertical ladder. It has to 00:06:22.240 --> 00:06:24.439 fit into the physical spaces we've already carved 00:06:24.439 --> 00:06:28.000 out for ourselves. Exactly. And to do that, the 00:06:28.000 --> 00:06:30.800 math is staggering. You have to understand the 00:06:30.800 --> 00:06:34.139 difference between direct kinematics and inverse 00:06:34.139 --> 00:06:36.439 dynamics. Okay, walk me through that because 00:06:36.439 --> 00:06:39.560 the source really emphasized how vital that mathematical 00:06:39.560 --> 00:06:42.100 shift is. Sure. Direct kinematics is forward 00:06:42.100 --> 00:06:44.660 thinking. You tell the robot to bend its knee 00:06:44.660 --> 00:06:47.379 30 degrees and the math calculates where the 00:06:47.379 --> 00:06:50.100 foot will end up. Simple trigonometry. Right. 00:06:50.779 --> 00:06:53.360 But walking upstairs requires inverse dynamics. 00:06:53.660 --> 00:06:55.860 You work backwards. So you start with the foot 00:06:55.860 --> 00:06:58.639 landing on the step. Exactly. You say, I need 00:06:58.639 --> 00:07:01.579 the foot here with as much downward force at 00:07:01.579 --> 00:07:04.980 this exact velocity. Now, computer, work backwards 00:07:04.980 --> 00:07:06.759 and tell me the exact electrical currents I need 00:07:06.759 --> 00:07:08.939 to send to 20 different actuators right this 00:07:08.939 --> 00:07:11.860 millisecond to make that happen. Wow. Factoring 00:07:11.860 --> 00:07:14.319 in gravity and friction backwards across all 00:07:14.319 --> 00:07:16.920 those joints. Hundreds of times a second. It's 00:07:16.920 --> 00:07:19.660 a triumph of modern computing. But movement is 00:07:19.660 --> 00:07:21.860 totally useless if the robot doesn't know what 00:07:21.860 --> 00:07:24.300 it's bumping into, right? It's worse than useless. 00:07:24.300 --> 00:07:27.060 It's dangerous. Which brings us to the cognitive 00:07:27.060 --> 00:07:30.899 shift. Moving from physical mobility to perception 00:07:30.899 --> 00:07:33.920 and sensing. Yes, the delicate world of touch. 00:07:34.740 --> 00:07:37.639 Vision is largely solved with cameras, but touch 00:07:37.639 --> 00:07:39.860 is incredibly hard. Because if it can't feel 00:07:39.860 --> 00:07:42.360 an egg, it'll either drop it or crush it. Precisely. 00:07:42.560 --> 00:07:45.180 The tactile sensors in the source are so cool. 00:07:45.420 --> 00:07:48.079 They use a rigid core surrounded by a conductive 00:07:48.079 --> 00:07:50.709 fluid. all wrapped in this elastomeric skin. 00:07:50.930 --> 00:07:53.790 Yes. The fluid displacement alters the electrical 00:07:53.790 --> 00:07:56.449 resistance. It constantly maps the force in three 00:07:56.449 --> 00:07:58.970 dimensions. And reading that, my first thought 00:07:58.970 --> 00:08:01.370 was, that's exactly how a human blister works. 00:08:01.550 --> 00:08:03.589 Huh. That's a brilliant way to look at it. Right. 00:08:03.629 --> 00:08:05.930 Like if you have a blister, you feel the pressure 00:08:05.930 --> 00:08:08.310 changing the fluid dynamics before it even hits 00:08:08.310 --> 00:08:11.389 the tissue underneath. And that nuance is why 00:08:11.389 --> 00:08:13.850 prosthetics like the smart hand can actually 00:08:13.850 --> 00:08:17.750 allow users to type on a keyboard. They can feel 00:08:17.750 --> 00:08:20.129 the resistance of the keys through that artificial 00:08:20.129 --> 00:08:22.990 skin. That is just amazing. But touch is only 00:08:22.990 --> 00:08:25.670 one way to interact. The source also dives into 00:08:25.670 --> 00:08:28.470 human -robot interaction, or HRI. Oh, the psychological 00:08:28.470 --> 00:08:31.269 layer. Yeah. And things like effective computing, 00:08:31.750 --> 00:08:34.389 getting robots to display synthetic emotions. 00:08:34.730 --> 00:08:37.149 Because speech recognition alone is incredibly 00:08:37.149 --> 00:08:39.549 difficult with accents and room acoustics. Right. 00:08:39.610 --> 00:08:42.940 So they use robots like... Kismet or Robin the 00:08:42.940 --> 00:08:46.500 robot to analyze a human face and display synthetic 00:08:46.500 --> 00:08:49.460 facial expressions to reduce human anxiety Yes 00:08:49.460 --> 00:08:52.960 Which brings up a weird philosophical point if 00:08:52.960 --> 00:08:56.000 a machine uses an AI algorithm to analyze my 00:08:56.000 --> 00:08:58.539 face and then produces a synthetic emotional 00:08:58.539 --> 00:09:01.769 response just to calm me down Does it matter 00:09:01.769 --> 00:09:04.370 that the empathy is completely faked? This raises 00:09:04.370 --> 00:09:06.370 an important question about social acceptance. 00:09:06.809 --> 00:09:09.149 Because functionally, the illusion of empathy 00:09:09.149 --> 00:09:11.090 is just as important as the mechanics of the 00:09:11.090 --> 00:09:13.129 robot. Really? Even if we know it's just code? 00:09:13.309 --> 00:09:15.929 Yes. If a faceless metal machine approaches a 00:09:15.929 --> 00:09:18.450 patient in a hospital, human cortisol levels 00:09:18.450 --> 00:09:21.129 spike. It's a primal threat response. That makes 00:09:21.129 --> 00:09:23.429 total sense. But if the robot tilts its head, 00:09:23.789 --> 00:09:25.950 lowers its voice, and displays synthetic concern, 00:09:26.350 --> 00:09:28.940 it hacks our autonomic nervous system. It forces 00:09:28.940 --> 00:09:31.740 us to feel safe. Wow. So the fake empathy produces 00:09:31.740 --> 00:09:34.600 a real biological result in the human. Exactly. 00:09:34.659 --> 00:09:36.919 It's a critical safety feature, not just a parlor 00:09:36.919 --> 00:09:39.000 trick. OK, so we've got super strong muscles, 00:09:39.480 --> 00:09:42.620 the math to navigate stairs, and the social skills 00:09:42.620 --> 00:09:45.039 to calm us down. So what are they actually doing 00:09:45.039 --> 00:09:46.980 out in society right now? They are everywhere. 00:09:47.139 --> 00:09:49.360 It's not just welding car frames anymore. Right. 00:09:49.500 --> 00:09:52.700 The source outlines some wild real world applications. 00:09:53.379 --> 00:09:56.559 Automated flatbread baking. Oh yeah. Spice kitchen 00:09:56.559 --> 00:09:59.679 in Massachusetts using robotic food bowls to 00:09:59.679 --> 00:10:02.340 cook and serve meals. Completely autonomously. 00:10:02.659 --> 00:10:06.240 And IBM has this fully automated lights out factory 00:10:06.240 --> 00:10:08.879 in Texas. Lights out because machines don't need 00:10:08.879 --> 00:10:11.720 overhead lights to see. Exactly. Which of course 00:10:11.720 --> 00:10:15.149 brings up the elephant in the room. Jobs. The 00:10:15.149 --> 00:10:18.629 source cites a 2017 study warning that 47 percent 00:10:18.629 --> 00:10:21.409 of U .S. jobs are at risk. As a very significant 00:10:21.409 --> 00:10:24.129 number. And it quotes Stephen Hawking from 2016 00:10:24.129 --> 00:10:26.870 saying this job destruction will reach deep into 00:10:26.870 --> 00:10:29.190 the middle class. It's a valid concern raised 00:10:29.190 --> 00:10:31.809 by many experts. But to balance that impartially, 00:10:32.049 --> 00:10:35.009 as the source does, there are also massive occupational 00:10:35.009 --> 00:10:38.429 safety and health benefits. OSH benefits. Yes, 00:10:38.470 --> 00:10:42.190 absolutely. Robots taking on the dirty, dangerous 00:10:42.190 --> 00:10:46.110 jobs. Like nuclear cleanup, where human exposure 00:10:46.110 --> 00:10:48.769 is lethal. Or heavy industrial assembly. And 00:10:48.769 --> 00:10:51.809 we're seeing the rise of cobots, right? Collaborative 00:10:51.809 --> 00:10:54.889 robots working alongside humans. Plus, the industry 00:10:54.889 --> 00:10:58.149 itself is projected to be worth $568 billion 00:10:58.149 --> 00:11:02.049 by 2030. Creating entirely new career paths in 00:11:02.049 --> 00:11:04.830 maintenance and programming. So, what does this 00:11:04.830 --> 00:11:07.389 all mean for you listening? Are we prepping for 00:11:07.389 --> 00:11:11.190 a dystopian job wipeout or a utopian release 00:11:11.190 --> 00:11:14.149 from dangerous labor? Well, looking at the data 00:11:14.149 --> 00:11:16.629 neutrally, history shows that technological shifts 00:11:16.629 --> 00:11:19.789 usually displace specific tasks, not necessarily 00:11:19.789 --> 00:11:22.269 entire workforces permanently. Task displacement, 00:11:22.429 --> 00:11:24.509 OK. Right. The machines take the repetitive, 00:11:24.509 --> 00:11:27.110 dangerous tasks, and humans pivot to higher level 00:11:27.110 --> 00:11:29.529 decision making and managing those systems. It's 00:11:29.529 --> 00:11:31.909 about how your specific field might integrate 00:11:31.909 --> 00:11:34.250 those cobots. That's a very grounded way to look 00:11:34.250 --> 00:11:36.129 at it. But to see where the economy is truly 00:11:36.129 --> 00:11:37.909 headed, we have to look at the bleeding edge 00:11:37.909 --> 00:11:40.129 research. The really avant -garde stuff. Yeah, 00:11:40.149 --> 00:11:42.750 moving from humans programming robots to robots 00:11:42.750 --> 00:11:45.370 evolving themselves. Evolutionary robotics. This 00:11:45.370 --> 00:11:48.450 is a monumental shift. The source explains this, 00:11:48.450 --> 00:11:50.210 and here's where it gets really interesting. 00:11:50.669 --> 00:11:53.909 They run physics simulations with thousands of 00:11:53.909 --> 00:11:56.389 random robot designs. Different leg lengths, 00:11:56.730 --> 00:11:58.850 joint angles, everything. Right. And they set 00:11:58.850 --> 00:12:01.970 a fitness function, like just travel across this 00:12:01.970 --> 00:12:05.210 uneven terrain. Most of the random designs just 00:12:05.210 --> 00:12:07.330 twitch or fall over. And the system deletes them. 00:12:07.529 --> 00:12:09.929 It just deletes the losers, then it takes the 00:12:09.929 --> 00:12:13.070 winners, mutates them slightly, and spawns a 00:12:13.070 --> 00:12:15.350 new generation without any human intervention. 00:12:15.769 --> 00:12:18.779 It is digital Darwinism. Literally, humans aren't 00:12:18.779 --> 00:12:20.460 the intelligent designers anymore. We're just 00:12:20.460 --> 00:12:22.399 setting up the terrarium. And if we connect this 00:12:22.399 --> 00:12:25.039 to the bigger picture, relying on evolutionary 00:12:25.039 --> 00:12:27.980 algorithms means we end up with highly efficient 00:12:27.980 --> 00:12:30.820 robotic designs where human engineers don't actually 00:12:30.820 --> 00:12:33.659 understand how they work. Wait, really? We build 00:12:33.659 --> 00:12:36.580 them but don't understand them? Yes. It echoes 00:12:36.580 --> 00:12:40.379 the black box nature of natural biology. The 00:12:40.379 --> 00:12:42.500 algorithm might find a mechanical shortcut that 00:12:42.500 --> 00:12:45.070 defies traditional engineering logic. That is 00:12:45.070 --> 00:12:47.710 wild. And the source briefly touches on swarm 00:12:47.710 --> 00:12:50.450 robotics, too, relying on emergent behavior, 00:12:50.950 --> 00:12:53.389 like ants building a bridge based on local rules. 00:12:53.570 --> 00:12:56.490 Exactly. Or quantum robotics, using quantum computers 00:12:56.490 --> 00:12:59.070 to evaluate millions of environmental variables 00:12:59.070 --> 00:13:02.169 instantly. Which all leads to this timeline from 00:13:02.169 --> 00:13:05.990 Hans Moravec in 1997. He predicted robot intelligence 00:13:05.990 --> 00:13:09.289 would match a lizard by 2010, a mouse by 2020, 00:13:09.730 --> 00:13:13.769 a monkey, and finally a human by 2045. A very 00:13:13.769 --> 00:13:16.289 famous and highly debated timeline. But seeing 00:13:16.289 --> 00:13:18.330 all this compounding acceleration, it really 00:13:18.330 --> 00:13:20.830 makes you wonder. We've covered a massive amount 00:13:20.830 --> 00:13:23.149 of ground today. We really have. From 8 -millimeter 00:13:23.149 --> 00:13:25.710 carbon biceps replacing human muscles to the 00:13:25.710 --> 00:13:28.190 sheer mathematical weight of bipedal walking. 00:13:28.450 --> 00:13:31.230 Inverse dynamics. action. We talked about empathetic 00:13:31.230 --> 00:13:33.750 fluid skin hacking our nervous systems and finally 00:13:33.750 --> 00:13:37.149 this wild concept of digital Darwinism. It arms 00:13:37.149 --> 00:13:39.110 you with a serious understanding for your next 00:13:39.110 --> 00:13:41.850 meeting or just satisfies that deep curiosity. 00:13:42.289 --> 00:13:43.690 Absolutely and I do want to leave you with one 00:13:43.690 --> 00:13:45.690 final thought to mull over. Oh please do. We 00:13:45.690 --> 00:13:47.549 just talked about evolutionary robotics using 00:13:47.549 --> 00:13:49.830 a fitness function to let robots design themselves, 00:13:49.909 --> 00:13:53.019 right? If robots are evolving purely based on 00:13:53.019 --> 00:13:55.259 survival of the fittest algorithms to achieve 00:13:55.259 --> 00:13:57.940 a goal, what happens when their definition of 00:13:57.940 --> 00:14:00.179 the most efficient design stops making sense 00:14:00.179 --> 00:14:02.440 to the human eye and we have to share our world 00:14:02.440 --> 00:14:06.220 with machines we no longer comprehend? Man, that 00:14:06.220 --> 00:14:08.940 is a profound and slightly haunting thought to 00:14:08.940 --> 00:14:11.179 end on. Thank you so much for breaking all of 00:14:11.179 --> 00:14:13.000 this down. It was my pleasure. And to everyone 00:14:13.000 --> 00:14:14.960 listening, keep questioning the world around 00:14:14.960 --> 00:14:17.659 you. We will catch you on the next Deep Dive.