WEBVTT - Ep68 "What if our brains worked a trillion times faster?" 0:00:05.200 --> 0:00:08.840 Why do we find movies so compelling given that they're 0:00:08.840 --> 0:00:13.319 just a series of photographs flash rapidly. What does this 0:00:13.480 --> 0:00:18.159 tell us about the slow speed of human brains? What 0:00:18.400 --> 0:00:22.120 percentage of stock market trades are taken care of by 0:00:22.239 --> 0:00:25.919 algorithms nowadays at timescales. 0:00:25.480 --> 0:00:28.000 That humans could not even conceive of. 0:00:28.600 --> 0:00:31.440 And what is it like to have the speed and 0:00:31.640 --> 0:00:37.080 power of a computer and be dealing with slow humans. 0:00:39.880 --> 0:00:43.280 Welcome to Inner Cosmos with me, David Eagleman. I'm a 0:00:43.360 --> 0:00:47.400 neuroscientist and in these episodes we sail deeply into our 0:00:47.520 --> 0:00:51.920 three pound universe to understand some of the most surprising 0:00:52.080 --> 0:01:06.600 aspects of our lives. Today's episode is about speed, the 0:01:06.640 --> 0:01:10.800 speed at which we humans operate and the speed at 0:01:10.800 --> 0:01:14.880 which our machines operate, and the future of this as 0:01:14.920 --> 0:01:17.360 the divergence grows even larger. 0:01:17.680 --> 0:01:18.720 So let's start with this. 0:01:19.000 --> 0:01:24.039 Imagine an extraterrestrial planet where they have an active society. 0:01:24.440 --> 0:01:28.520 These aliens have politics and division of labor, and philosophers 0:01:28.520 --> 0:01:31.360 and actors and athletes and artists, everything that makes up 0:01:31.680 --> 0:01:36.280 an active society. Okay, So, now imagine a second planet, 0:01:36.360 --> 0:01:40.000 Planet B. We land on this second planet in the 0:01:40.040 --> 0:01:44.600 near future, and we discover nothing like a society there. Instead, 0:01:44.640 --> 0:01:47.520 there are just things like trees, like these very old 0:01:47.600 --> 0:01:50.280 growths that don't really do much of anything. And we 0:01:50.480 --> 0:01:52.720 the astronauts, we stay there for a month, we take 0:01:52.720 --> 0:01:55.240 a lot of measurements, and finally we blast off when 0:01:55.280 --> 0:01:58.160 we go home. So two planets, one with an active 0:01:58.240 --> 0:02:02.600 society and the other with nothing but tree like things. 0:02:03.120 --> 0:02:08.040 Now imagine that these two descriptions are of the same planet. 0:02:08.120 --> 0:02:11.760 What we have stumbled on is a society of creatures 0:02:12.320 --> 0:02:17.040 that operate at a very different time scale from us. 0:02:17.360 --> 0:02:20.960 They move so slowly that we just can't see it. 0:02:21.360 --> 0:02:25.320 They have an active society, it just operates a trillion 0:02:25.480 --> 0:02:29.720 times slower than ours. And the question is, would we 0:02:29.840 --> 0:02:34.720 ever even notice that these tree people have a society 0:02:34.760 --> 0:02:38.400 happening in what we would think of as super slow motion. 0:02:38.800 --> 0:02:42.360 A conversation over a cup of coffee for them takes 0:02:42.600 --> 0:02:47.119 eons for us. So would we ever decode their language? 0:02:47.160 --> 0:02:50.480 Would we ever even have an opportunity to discover that 0:02:50.520 --> 0:02:53.560 there's a whole world going on there, but at a 0:02:53.720 --> 0:02:58.400 different temporal scale. So that's what today's episode is about. 0:02:58.800 --> 0:03:03.799 Hidden landscapes, parts of the time domain that are totally 0:03:03.840 --> 0:03:06.799 invisible to us. Now you've heard me talk in other 0:03:06.880 --> 0:03:10.400 episodes about all the information in the world that we 0:03:10.520 --> 0:03:14.080 just don't see. So is one example. We only see 0:03:14.520 --> 0:03:18.080 a small fraction of the light that is out there, 0:03:18.320 --> 0:03:21.960 and that little slice we call visible light. That's what 0:03:22.000 --> 0:03:25.799 we think of as red, orange, yellow, green, blue, indigo, violet. 0:03:26.080 --> 0:03:28.799 And it seems like that's all the light that's out there. 0:03:28.840 --> 0:03:31.840 You open your eyes, you see reds and greens and 0:03:31.880 --> 0:03:34.640 blues and golds, and you think, okay, that's everything. And 0:03:34.680 --> 0:03:39.240 so it came as a surprise when scientists eventually realized 0:03:39.680 --> 0:03:43.200 that all these other signals that had been detected, things 0:03:43.240 --> 0:03:46.560 that are totally invisible to us, like infrared light or 0:03:46.680 --> 0:03:50.360 radio waves or X rays or ultraviolet or microwaves, all 0:03:50.400 --> 0:03:54.040 of these are light. They're exactly the same thing, is 0:03:54.040 --> 0:03:56.720 what we call visible light. It's just that we don't 0:03:56.720 --> 0:03:59.840 have the receptors in our eyes to pick up on them, 0:04:00.400 --> 0:04:03.360 and so they're totally invisible to us. And it turns 0:04:03.400 --> 0:04:07.480 out that what we call visible light is a very 0:04:07.680 --> 0:04:10.840 tiny part of the light spectrum. It's less than a 0:04:11.080 --> 0:04:14.840 ten trillionth of it. And once we understood this, we 0:04:14.920 --> 0:04:19.599 started to build technologies to burrow into those other parts 0:04:19.640 --> 0:04:24.000 of the spectrum. For example, we make little machines that 0:04:24.120 --> 0:04:27.920 sit in the dashboards of our cars, and these see 0:04:28.080 --> 0:04:31.680 in exactly the frequency that we call radio waves, and 0:04:31.720 --> 0:04:35.680 so we can transmit information in that frequency. Or we 0:04:35.720 --> 0:04:40.520 build night vision security cameras which see in the infrared range, 0:04:40.520 --> 0:04:43.560 so that we can detect in that frequency outside of 0:04:43.560 --> 0:04:46.000 what we can naturally see. And we build X ray 0:04:46.080 --> 0:04:49.839 machines and telescope dishes, and these all pick up on 0:04:50.080 --> 0:04:54.120 other frequencies that are invisible to us. We build all 0:04:54.240 --> 0:04:58.159 kinds of devices to see what we can't naturally see, 0:04:58.200 --> 0:05:01.840 and this has been the gen of scientific development over 0:05:01.880 --> 0:05:05.720 the last few hundred years. It's the understanding that the 0:05:05.760 --> 0:05:10.640 reality we inhabit is actually just a small fraction of 0:05:10.720 --> 0:05:13.960 what is out there, and there are many other things 0:05:14.160 --> 0:05:18.159 that can be taken advantage of. So we're very used 0:05:18.200 --> 0:05:23.320 to communicating in landscapes that are outside our biology. Our 0:05:23.320 --> 0:05:27.680 biology allows us to detect something like visible light, and 0:05:27.720 --> 0:05:29.720 then we figure out from there that there are other 0:05:30.240 --> 0:05:33.680 flavors of light that are beyond our biology, and we. 0:05:33.600 --> 0:05:34.880 Take advantage of those. 0:05:35.560 --> 0:05:39.240 So I want to use this as an analogy and 0:05:39.279 --> 0:05:42.680 an inroad for us thinking about somewhere else that we 0:05:42.760 --> 0:05:46.080 do this, and that is in the domain of time. 0:05:47.160 --> 0:05:50.280 The question is can we use different ranges of time 0:05:50.400 --> 0:05:55.359 to build technologies, perhaps a range of time that is 0:05:55.480 --> 0:05:59.920 essentially invisible in our normal perception. When I was a tea, 0:06:00.760 --> 0:06:04.360 I visited MIT and I met with Harold Edgerton who 0:06:04.360 --> 0:06:07.400 was known as Doc Edgerton. And he was a pioneer 0:06:07.600 --> 0:06:11.680 in high speed photography. And what he would do is 0:06:11.880 --> 0:06:16.480 capture moments that were otherwise totally invisible to us. So 0:06:16.680 --> 0:06:21.400 he froze ultrafast events using a quick strobe light and 0:06:21.640 --> 0:06:25.560 ultra high speed photography. One of his most famous photographs 0:06:25.720 --> 0:06:27.680 was taken in the nineteen sixties and you may have 0:06:27.720 --> 0:06:32.440 seen this. It shows a bullet piercing an apple. The 0:06:32.480 --> 0:06:36.159 image captures the exact moment of impact with the apple 0:06:36.240 --> 0:06:40.000 exploding into fragments, and the exposure time for images like 0:06:40.040 --> 0:06:43.680 this it was less than a microsecond. Another one of 0:06:43.680 --> 0:06:48.640 his iconic images is the milk drop coronet that shows 0:06:48.680 --> 0:06:51.960 a drop of milk hitting a surface and forming this 0:06:52.480 --> 0:06:57.360 crown like splash, and the detail in this splash, frozen 0:06:57.440 --> 0:07:00.159 in time, gave all kinds of insights into flu with 0:07:00.279 --> 0:07:04.880 dynamics and surface tension. So the temporal resolution of these 0:07:04.920 --> 0:07:09.400 photos allowed everyone to study and to appreciate the beauty 0:07:09.960 --> 0:07:14.880 and the complexity of fast phenomenon. So he could freeze 0:07:15.000 --> 0:07:18.120 the motion of bullets in flight or splashing milk drops, 0:07:18.360 --> 0:07:22.760 and these images revealed details that were previously hidden to 0:07:23.000 --> 0:07:26.480 us due to the limitations of our very slow perception. 0:07:27.360 --> 0:07:29.600 And he also did this not just with still shots, 0:07:29.640 --> 0:07:31.960 but a series of still shots, so you get a 0:07:31.960 --> 0:07:35.040 little time lapse film. And in this way he could 0:07:35.040 --> 0:07:40.760 expand super brief events into viewable sequences. This was great 0:07:40.800 --> 0:07:44.360 for science and for art, and also for military applications 0:07:44.440 --> 0:07:48.720 like studying the behavior of bombs. So with these techniques 0:07:48.760 --> 0:07:52.760 he led us into new time domains which we suspected 0:07:52.840 --> 0:07:56.640 might exist, but which no human had ever actually seen. 0:07:57.120 --> 0:08:01.280 So he opened a window into a world of fleeting moments, 0:08:01.760 --> 0:08:05.200 capturing events that occur too quickly for our eyes to see, 0:08:05.840 --> 0:08:08.600 and reveling in a hidden beauty that we didn't even 0:08:08.680 --> 0:08:13.000 have the mental tools to imagine correctly. So the point 0:08:13.120 --> 0:08:17.080 is we can take advantage of these other timescales that 0:08:17.120 --> 0:08:19.800 we don't know much about, and we can do this 0:08:19.840 --> 0:08:22.480 by building the right kind of machinery. 0:08:22.800 --> 0:08:24.160 Now, this all works. 0:08:23.920 --> 0:08:29.280 Because although we humans are slow and pitifully limited in 0:08:29.440 --> 0:08:33.439 vision and in time, we have large prefrontal cortices. This 0:08:33.480 --> 0:08:36.840 is the part of the brain that allows us to extrapolate, 0:08:36.960 --> 0:08:39.439 to think about what's in the next room even though 0:08:39.480 --> 0:08:43.680 we're not there, to think about what we might do tomorrow, 0:08:44.040 --> 0:08:48.559 to think about possibilities that we haven't yet seen. And this, 0:08:48.840 --> 0:08:52.920 I suggest, is what has driven the development of new technologies, 0:08:53.000 --> 0:08:57.040 because we are constantly imagining things that are bigger or 0:08:57.040 --> 0:09:00.520 smaller than we can see. So we invent tell scopes 0:09:00.559 --> 0:09:05.320 and microscopes. We invent ways of seeing into other parts 0:09:05.360 --> 0:09:09.160 of the visible spectrum, like radios or X ray machines, 0:09:09.280 --> 0:09:12.640 or devices in the infrared or ultraviolet, and all over 0:09:12.679 --> 0:09:16.240 the place. We burrow into spaces that we can't see 0:09:16.280 --> 0:09:20.600 with our eyes, but we extrapolate that they must exist. 0:09:20.840 --> 0:09:24.160 And I suggest the prefrontal cortex is what allows us 0:09:24.200 --> 0:09:28.680 to extrapolate in the time domain. So here's an example. 0:09:29.040 --> 0:09:31.120 I was in Germany a little while ago, and I 0:09:31.200 --> 0:09:35.760 was transfixed watching a river flowing through a very deep valley, 0:09:36.440 --> 0:09:38.920 and it was pretty effortless to look at the valley 0:09:39.040 --> 0:09:42.640 walls and think about this at the scale at which 0:09:42.800 --> 0:09:47.480 geologists look at a landscape. This exact moment of the 0:09:47.600 --> 0:09:51.600 river flowing had been happening for millions of years, and 0:09:51.640 --> 0:09:55.360 if I sped up the film in my head, that's 0:09:55.440 --> 0:09:59.360 how you get a valley of that depth. Now, to 0:09:59.400 --> 0:10:02.240 be clear, I wasn't imagining millions of years. I was 0:10:02.280 --> 0:10:06.000 imagining a film at my human speed. I just took 0:10:06.040 --> 0:10:09.960 the concept of huge time scales and squished them down 0:10:10.000 --> 0:10:12.680 to something that I could understand. This is the same 0:10:12.679 --> 0:10:17.200 way that someone using a microscope doesn't stare at the 0:10:17.280 --> 0:10:21.640 bacterium at the scale of a micron. Instead, she uses 0:10:21.800 --> 0:10:25.240 the machine, the microscope, to push the image up to 0:10:25.400 --> 0:10:29.360 a large size that her brain can handle. And this 0:10:29.440 --> 0:10:31.959 is all we ever do. We take things from different 0:10:32.000 --> 0:10:36.960 domains and translate them into something that we can understand. 0:10:37.240 --> 0:10:40.240 Now we're very good at that. Here's an example on 0:10:40.800 --> 0:10:43.720 the flip side of the river carving the valley. Open 0:10:43.760 --> 0:10:47.080 any book on let's say, molecular biology, and you'll see 0:10:47.400 --> 0:10:51.760 two strands of DNA unzipping and proteins buying to those 0:10:51.760 --> 0:10:56.000 strands and walk along it and this gets transcribed to RNA, 0:10:56.200 --> 0:11:00.120 and then that gets translated into little amino acids, and 0:11:00.160 --> 0:11:03.840 you can walk yourself through all the steps of transcription 0:11:03.920 --> 0:11:08.240 and translation. But all this takes place at a timescale 0:11:08.600 --> 0:11:12.720 that's faster than you can possibly conceive. It's about twenty 0:11:12.760 --> 0:11:15.960 amino acids get translated every second, and it's hard to 0:11:16.040 --> 0:11:19.840 picture that in real time, But no problem. 0:11:19.920 --> 0:11:21.800 We can apply the tools of our. 0:11:22.080 --> 0:11:25.800 Thinking to imagine it as though it were happening at 0:11:25.840 --> 0:11:29.319 our timescale, and of course we do this on crazier 0:11:29.320 --> 0:11:34.080 and crazier timescales. Let's say you're a physicist building a 0:11:34.200 --> 0:11:39.000 nuclear reactor and you're thinking about two nucleis smashing into 0:11:39.040 --> 0:11:42.720 one another, and as it busts up, the parts of 0:11:42.760 --> 0:11:46.080 it smash into other nuclei and bash those up, and. 0:11:46.000 --> 0:11:47.760 You get a chain reaction. 0:11:48.400 --> 0:11:50.680 And once you understand the simple model, you can picture 0:11:50.760 --> 0:11:53.840 each step of the chain reaction, even though what is 0:11:54.000 --> 0:11:59.320 happening takes place in one millionth of a second, a microsecond. 0:12:00.200 --> 0:12:03.120 If you made these calculations every day of your life 0:12:03.120 --> 0:12:05.560 on the whiteboard or on the computer, it would be 0:12:05.800 --> 0:12:11.400 absolutely impossible for you to ever perceive the event anywhere. 0:12:11.800 --> 0:12:16.160 But in your imagination, you couldn't perceive it directly. So 0:12:16.320 --> 0:12:21.559 in this way, You are like the biblical story of Moses, 0:12:21.600 --> 0:12:24.920 who leads his people through the desert but never gets 0:12:24.960 --> 0:12:30.000 to enter the Holy Land himself. You are the temporal Moses, 0:12:30.080 --> 0:12:33.800 knowing about these timescales, knowing about these other temporal worlds, 0:12:33.840 --> 0:12:38.679 but you are never able to enter the territory directly. 0:12:39.120 --> 0:12:42.640 Just think about the best theories of the beginning of 0:12:42.679 --> 0:12:45.360 the universe. The best models for this suggest that there 0:12:45.440 --> 0:12:48.880 was cosmic inflation where everything got bigger. There was this 0:12:49.200 --> 0:12:52.760 exponential expansion of space, and we say this all happened 0:12:52.800 --> 0:12:56.920 in the early universe. And when we look at the models, 0:12:56.960 --> 0:13:01.079 there was the Plank Epic, where physics were dominated by 0:13:01.120 --> 0:13:04.160 the quantum effects of gravity, and then that got followed 0:13:04.160 --> 0:13:08.120 by the Grand Unification Epic, and that was followed by 0:13:08.160 --> 0:13:12.760 the inflationary epic, in which the universe expanded by trillions 0:13:12.760 --> 0:13:16.360 and trillions of times, and then the electroweek Epic, and 0:13:16.400 --> 0:13:18.760 the quark Epic and the hadron epic and so on. 0:13:18.800 --> 0:13:19.640 Now here's the key. 0:13:20.120 --> 0:13:25.120 The Plank Epic lasted less then ten to the negative 0:13:25.200 --> 0:13:29.000 forty three seconds. That's less than a trillionth of a 0:13:29.040 --> 0:13:31.360 trillionth of a trillion of a trillionth of a second, 0:13:32.040 --> 0:13:35.880 and the grand unification epic was over by ten to 0:13:35.960 --> 0:13:39.800 the negative thirty six seconds, and the inflationary epic was 0:13:40.000 --> 0:13:43.400 over in the first peko second of cosmic time, that's 0:13:43.440 --> 0:13:47.280 ten to the negative twelve seconds. And eventually protons formed. 0:13:47.440 --> 0:13:49.680 And oh, by the way, that happened by the first 0:13:50.000 --> 0:13:53.240 microsecond one millionth of a second after the Big Bang, 0:13:53.640 --> 0:13:56.440 and then nuclear fusion began, and that was a ten 0:13:56.520 --> 0:13:59.040 milliseconds after the start of the universe, or one one 0:13:59.120 --> 0:14:01.240 hundredth of a second. No, how do people even put 0:14:01.280 --> 0:14:04.920 together theories like this, Well, you can't even run experiments 0:14:04.920 --> 0:14:07.720 at those temperatures or densities. But you can do is 0:14:07.800 --> 0:14:12.440 extrapolate the known physical laws to extremely high temperatures, and 0:14:12.480 --> 0:14:14.600 you get a picture of what was likely to have 0:14:14.720 --> 0:14:18.040 happened there. And again, this is something that we can 0:14:18.120 --> 0:14:24.240 never experience directly, but we can extrapolate our local understanding 0:14:24.280 --> 0:14:28.640 of time to imagine it. Presumably, the squirrels running around 0:14:28.640 --> 0:14:32.880 in your backyard simply don't have the neural capacity to 0:14:33.120 --> 0:14:36.840 imagine time at any scale other than their own. So 0:14:37.160 --> 0:14:39.560 a squirrel is never going to come up with the 0:14:39.640 --> 0:14:58.080 Plank epic. Now, what I want to talk about is 0:14:58.120 --> 0:15:01.360 the way that we do these extra appellations, and then 0:15:01.840 --> 0:15:05.320 we build technologies to get us there, like Doc Edgerton 0:15:05.400 --> 0:15:10.040 at MIT did and translating things into something useful for 0:15:10.200 --> 0:15:14.920 our pitifully slow time scales. So I'll give you an 0:15:14.960 --> 0:15:18.320 example of this. I was in Shanghai a few years ago, 0:15:18.600 --> 0:15:20.880 and on the edge of the river where a lot 0:15:20.920 --> 0:15:23.680 of tourists walk around, they had this terrific display where 0:15:23.680 --> 0:15:26.800 it was like a big plant wall, and from a 0:15:26.840 --> 0:15:31.800 distance you could see these beautiful, slowly flapping butterflies on 0:15:31.840 --> 0:15:34.080 the wall. But when you go up close and really 0:15:34.120 --> 0:15:37.400 examine this, you see that each butterfly is actually an 0:15:37.600 --> 0:15:43.120 illusion produced from a rapidly spinning bar of led lights. 0:15:43.560 --> 0:15:45.640 So the bar is attached at one end to a 0:15:45.760 --> 0:15:48.120 motor and it spins around and around in a circle 0:15:48.240 --> 0:15:48.800 very quickly. 0:15:49.080 --> 0:15:52.000 You may have seen these before. You can display. 0:15:51.760 --> 0:15:56.400 Cartoon characters or words or animals. The rod is a 0:15:56.560 --> 0:15:59.240 row of led lights, and these are all turning on 0:15:59.320 --> 0:16:03.120 and off with different colors of different intensities at exactly 0:16:03.160 --> 0:16:06.480 the right time in their rotations, so that they draw 0:16:06.680 --> 0:16:09.840 out what appears to be a picture of an object. 0:16:09.920 --> 0:16:13.600 Now why does this work, Well, it's because of something 0:16:13.680 --> 0:16:17.400 called persistence of vision. And this is what happens when 0:16:17.440 --> 0:16:20.200 you flash a bunch of images really rapidly, one after 0:16:20.240 --> 0:16:23.120 the other. The bar is in this position, and then 0:16:23.160 --> 0:16:26.320 the next moment it's displaying it slightly differently one degree away, 0:16:26.440 --> 0:16:28.640 and then differently again when it ticks. 0:16:28.360 --> 0:16:30.080 Forward one more degree, and so on. 0:16:30.240 --> 0:16:32.640 But because each of these little flashes of light are 0:16:32.760 --> 0:16:37.080 so fast, they last in your vision. So the light 0:16:37.160 --> 0:16:39.680 pops on and off in a fraction of a millisecond, 0:16:39.720 --> 0:16:42.520 but your brain just can't see something that brief, So 0:16:42.560 --> 0:16:44.880 as far as you're concerned, it comes on and it 0:16:44.960 --> 0:16:47.480 stays on for a little while. So when the bar 0:16:47.640 --> 0:16:51.480 spins around and around, it essentially paints in the air. 0:16:52.360 --> 0:16:54.680 Now here's the key. If you could see faster, if 0:16:54.680 --> 0:16:57.480 you could see what is actually happening out there, you'd 0:16:57.480 --> 0:16:59.760 see a little bar spinning in a circle and changing 0:16:59.760 --> 0:17:03.400 color as it does so. But that's not what you see. 0:17:03.440 --> 0:17:06.679 You see a butterfly in the air, you see a 0:17:06.720 --> 0:17:10.040 cartoon character, you see a flag waving. But here's the 0:17:10.119 --> 0:17:13.639 thing to appreciate. The speed at which the rod needs 0:17:13.680 --> 0:17:17.680 to spin is very fast, and therefore the speed at 0:17:17.720 --> 0:17:21.000 which the lights need to know precisely when to turn 0:17:21.040 --> 0:17:24.120 on and off is even faster, and all of this 0:17:24.200 --> 0:17:27.880 is computed at rocket ship speed, and you just look 0:17:27.920 --> 0:17:31.640 at it and appreciate that there's a nice butterfly out there. 0:17:32.000 --> 0:17:34.680 So on the one hand, you have lightning fast calculations 0:17:34.680 --> 0:17:37.480 on the level of sub millisecond, and on the other 0:17:37.640 --> 0:17:40.160 level we get a nice picture. 0:17:40.400 --> 0:17:41.440 And persistence of. 0:17:41.440 --> 0:17:45.760 Vision is the reason why movies appear to be continuous motion, 0:17:45.920 --> 0:17:49.720 even though they are actually just a series of still images. 0:17:49.840 --> 0:17:54.199 In other words, after photography was invented, people realize that 0:17:54.240 --> 0:17:57.159 if you flash one photograph and then there's a blank, 0:17:57.200 --> 0:18:00.200 and then you flash another photo right after, like twenty 0:18:00.200 --> 0:18:05.520 milliseconds after, your brain will interpret that as continuous motion. 0:18:06.080 --> 0:18:11.760 And this was the birth of moving photography or movies. Now, 0:18:11.960 --> 0:18:14.320 the challenge was tougher with televisions. 0:18:14.680 --> 0:18:17.640 Entrepreneurs wanted to get movies into the homes of. 0:18:17.640 --> 0:18:21.159 People, but you couldn't flicker photographs forty eight times per 0:18:21.200 --> 0:18:23.560 second because you didn't have a film strip that you 0:18:23.600 --> 0:18:24.680 were just shining light through. 0:18:24.720 --> 0:18:26.200 So how do you do it? Well? 0:18:26.240 --> 0:18:29.600 The genius there was to use a cathode ray tube, 0:18:29.600 --> 0:18:33.920 which points a beam of light at a phosphorescent screen, 0:18:34.400 --> 0:18:38.040 and that causes a little spot to glow from dark 0:18:38.080 --> 0:18:40.879 to bright and then you move that little light gun 0:18:40.920 --> 0:18:44.399 over to the next spot over and set the next brightness, 0:18:44.640 --> 0:18:46.199 and then you go to the next spot in the 0:18:46.200 --> 0:18:46.960 next spot. 0:18:46.680 --> 0:18:47.080 And so on. 0:18:47.320 --> 0:18:50.359 And you had to cover the entire screen one dot 0:18:50.400 --> 0:18:52.600 at a time, and the whole trick was to do 0:18:52.640 --> 0:18:55.720 it really fast, so that you paint the screen in 0:18:56.200 --> 0:18:59.240 one sixtieth of a second, and then you paint the 0:18:59.280 --> 0:19:02.239 next screen again one dot at a time, and it 0:19:02.280 --> 0:19:06.280 has to get done in the next sixtieth of a second. Now, 0:19:06.320 --> 0:19:08.639 the genius of coming up with the technology like that 0:19:09.160 --> 0:19:12.680 is realizing that you can do something very very fast 0:19:12.720 --> 0:19:16.000 with a machine, and that the person at the receiving 0:19:16.160 --> 0:19:20.360 end is incredibly slow and has a sluggish brain and 0:19:20.480 --> 0:19:24.400 can't understand anything at these millisecond time scales and instead 0:19:24.520 --> 0:19:28.840 just blends everything together. And that's how you tell stories 0:19:28.920 --> 0:19:32.480 about James Cagney or Fat Albert or the family in Bonanza, 0:19:33.040 --> 0:19:36.360 and the sluggish humans on the receiving end are simply 0:19:36.440 --> 0:19:40.959 registering the emotion of the story instead of seeing the 0:19:41.359 --> 0:19:45.840 invisibly fast mechanics that are zipping along in their own 0:19:46.000 --> 0:19:50.520 timescale range, painting the screen over and over sixty times 0:19:50.560 --> 0:19:55.640 every second in a way that is completely invisible to us. 0:19:55.960 --> 0:19:58.919 So we build machines to leverage the fact that the 0:19:59.000 --> 0:20:03.119 visual brain integrates what it sees over a little window 0:20:03.200 --> 0:20:06.879 of time. So if two or more images arrive within 0:20:06.920 --> 0:20:09.800 a window of about a tenth of a second, they're 0:20:09.840 --> 0:20:13.560 smeared together. And by the way, building devices to take 0:20:13.600 --> 0:20:16.920 advantage of that isn't even so new. We've been building 0:20:17.000 --> 0:20:21.600 primitive things for centuries. For example, think of that little 0:20:21.680 --> 0:20:26.040 toy known as a thaumatrope. It's a flat disc with 0:20:26.160 --> 0:20:28.280 a picture on each side. So you might have a 0:20:28.320 --> 0:20:30.440 picture of a bird on one side of the disk 0:20:30.720 --> 0:20:33.320 and a picture of a tree branch on the other. 0:20:33.680 --> 0:20:35.840 And what you do is you wind up the disk 0:20:36.000 --> 0:20:37.960 with a piece of string, and when you pull the 0:20:38.000 --> 0:20:42.280 string tight, the disc spins rapidly so that both sides 0:20:42.480 --> 0:20:45.119 are seen in rapid alternation. You see the bird, you 0:20:45.160 --> 0:20:46.960 see the branch. You see the bird, you see the branch, 0:20:47.359 --> 0:20:50.600 and the bird appears to be sitting on the branch. 0:20:50.720 --> 0:20:54.520 The images fuse. This is an example of an old 0:20:54.560 --> 0:20:58.960 technology taking advantage of the slowness of our brains. Now 0:20:59.000 --> 0:21:02.360 I find all this credible, from the butterflies to the 0:21:02.400 --> 0:21:06.040 fact that movies move, to the fact that television works. 0:21:06.600 --> 0:21:10.080 But with modern electronics, combined with the slowness of our 0:21:10.160 --> 0:21:13.600 visual systems, we can build these kinds of devices to 0:21:13.840 --> 0:21:17.679 paint pictures for us. Our machines can be fast, a 0:21:17.680 --> 0:21:21.040 lot faster than our poor little brains. Now, if you 0:21:21.119 --> 0:21:24.840 were the machine, you would think, what am I doing? 0:21:24.920 --> 0:21:28.440 I'm just changing the colors of the light as I spin. 0:21:28.960 --> 0:21:33.040 Surely no one out there is falling for this. And 0:21:33.080 --> 0:21:35.199 this is the thing I really want to dig into today, 0:21:35.240 --> 0:21:39.960 the difference in the time worlds between us and our machines. 0:21:40.720 --> 0:21:43.600 So the best place to start is with our computers, 0:21:44.200 --> 0:21:49.199 which perform a certain number of floating point operations per second. 0:21:49.320 --> 0:21:54.199 This measure is called flops floating point operations per second. Now, 0:21:54.200 --> 0:21:56.840 don't worry about the details, except to say a floating 0:21:56.840 --> 0:22:00.440 point operation is a calculation like adding two. 0:22:00.840 --> 0:22:02.000 Very big numbers. 0:22:02.160 --> 0:22:04.240 So this is a standard way to measure how fast 0:22:04.280 --> 0:22:08.480 the computer can chug along. Now, think about how long 0:22:08.520 --> 0:22:12.320 it would take you to add two very large numbers. 0:22:12.640 --> 0:22:16.760 It's unlikely that you could do even one floating point 0:22:16.800 --> 0:22:20.800 operation in a second. But a computer from the last 0:22:20.800 --> 0:22:27.480 century could do a gigaflop that's one billion floating point 0:22:27.520 --> 0:22:32.359 operations in a second. Just think about sitting there with 0:22:32.400 --> 0:22:35.600 a paper and pencil and adding two big numbers and 0:22:35.640 --> 0:22:40.600 doing that a billion times every second. Then in this century, 0:22:40.640 --> 0:22:47.399 computers hit terraflops that's one trillion floating point operations in 0:22:47.480 --> 0:22:51.840 a second. And then it wasn't long before computers hit petaflops, 0:22:51.880 --> 0:22:56.240 which is a thousand terraflops. In other words, a thousand 0:22:56.680 --> 0:23:02.199 trillion or a quadrillion operations person second. Nvidia just released 0:23:02.200 --> 0:23:06.480 a new chip that clocks in at twenty petaflops, So 0:23:06.640 --> 0:23:09.240 just imagine how long it would take you to do 0:23:10.160 --> 0:23:15.879 twenty thousand trillion calculations. It does that a second. The 0:23:16.359 --> 0:23:20.560 current record is the US supercomputer called Frontier, which does 0:23:21.160 --> 0:23:26.760 eleven hundred petaflops or one point one exaFLOPS. These numbers 0:23:26.760 --> 0:23:29.720 are so mind boggling to think about. Just imagine a 0:23:29.800 --> 0:23:35.520 machine performing one quintillion floating point operations per second. That's 0:23:35.560 --> 0:23:39.320 a one followed by eighteen zeros. That's a lot faster 0:23:39.720 --> 0:23:42.120 than all the humans in the world could do if 0:23:42.119 --> 0:23:44.800 they all sat with paper and pencil and worked on 0:23:44.840 --> 0:23:49.399 calculations for a million years. That's what gets accomplished in 0:23:49.600 --> 0:23:53.560 one second by a computer like that. We are living 0:23:53.840 --> 0:23:58.240 on completely different time domains, and this is what allows 0:23:58.280 --> 0:24:02.040 you to enjoy a song or YouTube video or an 0:24:02.080 --> 0:24:06.440 Inner Cosmos podcast. Now, something I think is quite stunning 0:24:06.600 --> 0:24:08.760 is to consider what it is like. 0:24:08.840 --> 0:24:11.240 To be your computer keyboard. 0:24:11.640 --> 0:24:14.240 Even if you think you are a fast typist, your 0:24:14.359 --> 0:24:20.560 keyboard actually goes to sleep in between your slow keystrokes. 0:24:20.920 --> 0:24:24.040 So I'm gonna do something here. Let's imagine that you 0:24:24.119 --> 0:24:27.679 and I are down inside your computer, living at the 0:24:27.720 --> 0:24:32.200 speed of your computer chip. Now, the poor slow human 0:24:32.359 --> 0:24:35.640 on the outside is typing, and let's say they're typing 0:24:35.680 --> 0:24:40.000 relatively fast, so when they strike a key on the keyboard, 0:24:41.080 --> 0:24:47.040 it's going to sound like this to us inside the computer. Now, 0:24:47.119 --> 0:24:51.879 let's keep talking and we'll see when the next keystroke arrives. 0:24:52.119 --> 0:24:55.920 Imagine that this human is typing relatively fast, like sixty 0:24:55.960 --> 0:24:59.000 words per minute, which is about three hundred keystrokes per minute, 0:24:59.200 --> 0:25:02.960 or a keystroke every two hundred milliseconds from our point 0:25:03.040 --> 0:25:07.240 of view from inside the computer. The next key will 0:25:07.280 --> 0:25:11.280 not even get hit until about the end of this podcast, 0:25:11.680 --> 0:25:15.760 And that's why the software routine that monitors the keyboard 0:25:16.560 --> 0:25:20.760 goes to sleep while it waits for the next key 0:25:21.320 --> 0:25:24.240 to be struck. And I'll tell you something else fascinating 0:25:24.280 --> 0:25:28.960 about our computers and our interaction with them. You remember capture, 0:25:29.119 --> 0:25:30.960