WEBVTT - Is Autonomous The Way To Go? 0:00:00.080 --> 0:00:02.720 Hey guys, this is Jonathan. I have a quick announcement 0:00:02.759 --> 0:00:05.400 before we get to today's episode, which is sort of 0:00:05.480 --> 0:00:08.760 a capper on the autonomous car episodes we listened to 0:00:09.119 --> 0:00:12.879 last week, and that is starting this week, Tech Stuff 0:00:12.920 --> 0:00:16.320 is going back to publishing two new episodes and one 0:00:16.360 --> 0:00:20.239 classic episode each week, so we're cutting back a little bit. 0:00:20.360 --> 0:00:22.920 This is so that I can spend more time putting 0:00:22.920 --> 0:00:25.720 these episodes together, making sure they are the absolute best 0:00:25.800 --> 0:00:28.400 I can make them, as well as making time to 0:00:28.520 --> 0:00:31.440 work on some other shows like The Brink. If you're 0:00:31.480 --> 0:00:33.159 not listening to The Brink, you should go check that 0:00:33.200 --> 0:00:36.280 out and some more shows that might be coming down 0:00:36.840 --> 0:00:39.360 in the near future once we get through pilot season. 0:00:39.560 --> 0:00:42.760 Here at How Stuff Works, and I just want to 0:00:42.840 --> 0:00:45.160 let you guys all be aware of that nothing bad 0:00:45.400 --> 0:00:47.839 is happening. These are good things. It means that I 0:00:47.880 --> 0:00:51.400 get to spend more time on this stuff and I'm 0:00:51.440 --> 0:00:55.320 not quite as a as rushed as I have been 0:00:55.360 --> 0:00:59.040 for the last half year. So enjoy and I'll see 0:00:59.080 --> 0:01:06.760 you guys again on Wednesday. Get in touch with technology 0:01:06.880 --> 0:01:13.520 with tech Stuff from how stuff works dot com. Hey there, 0:01:13.520 --> 0:01:16.800 and welcome to tech Stuff. I'm your host Jonathan Strickland. 0:01:16.840 --> 0:01:19.399 I'm an executive producer with How Stuff Works, and I 0:01:19.520 --> 0:01:23.759 heart radio and I love all things tech, and we 0:01:23.920 --> 0:01:29.760 are continuing our long journey with autonomous cars. Actually we 0:01:29.800 --> 0:01:33.319 are concluding it at this point, even though as I 0:01:33.360 --> 0:01:35.559 get to the end of the episode you will hear 0:01:36.120 --> 0:01:39.200 about how I could cover even more. But out of 0:01:39.240 --> 0:01:42.200 consideration for you, my dear listeners, I'll move on to 0:01:42.319 --> 0:01:44.319 other topics for a while. But last week I did 0:01:44.319 --> 0:01:47.960 a whole series of episodes about the history of autonomous cars, 0:01:48.400 --> 0:01:51.120 and we started with the science fiction visions of the 0:01:51.200 --> 0:01:55.440 nineteen twenties and radio controlled vehicles up to the upcoming 0:01:55.520 --> 0:01:59.360 launch of the Weymo one service in Phoenix, Arizona, which 0:02:00.120 --> 0:02:03.480 maybe in operation by the time you hear this. And 0:02:03.520 --> 0:02:07.160 I mentioned how over the last several years, in particular, 0:02:07.520 --> 0:02:11.320 a few accidents, a couple of them fatal accidents, have 0:02:11.440 --> 0:02:14.960 brought a lot more critical attention to driverless cars, even 0:02:15.000 --> 0:02:19.640 as numerous services prepared to unleash thousands of them on 0:02:19.680 --> 0:02:24.000 the streets of several cities across the world. In this episode, 0:02:24.280 --> 0:02:27.440 I'm going to talk about the arguments for and against 0:02:27.480 --> 0:02:32.799 autonomous cars, or at least against early autonomous car deployment. 0:02:33.240 --> 0:02:38.360 I'm gonna talk about how these things are complicating factors. 0:02:38.400 --> 0:02:41.680 It's not just a technological problem. In fact, I don't 0:02:41.680 --> 0:02:44.760 really get into it in this episode, but there are 0:02:45.040 --> 0:02:50.760 entire issues with regulation and legislation. Like many other technologies, 0:02:51.040 --> 0:02:56.160 autonomous cars has outpaced the law, and so there are 0:02:56.160 --> 0:02:59.440 a lot of places that are trying to take into 0:02:59.480 --> 0:03:03.280 account what driverless cars in the streets would actually mean, 0:03:03.360 --> 0:03:05.880 and how do you legislate that. Who do you find 0:03:05.880 --> 0:03:08.440 at fault in the event of an accident? These are 0:03:08.480 --> 0:03:12.280 still largely open questions. There are other ones that I 0:03:12.280 --> 0:03:15.440 haven't touched on in this episode either, For example, the 0:03:15.480 --> 0:03:19.440 trolley problem, the idea that if you have to build 0:03:19.480 --> 0:03:23.440 into your car system a decision making process to follow 0:03:23.560 --> 0:03:27.400 in the event that an injury is unavoidable, you've got 0:03:27.440 --> 0:03:31.800 and and a scenario where no matter what, someone is 0:03:31.800 --> 0:03:34.880 going to get hurt. How does the car make the 0:03:34.960 --> 0:03:40.600 choice of who gets hurt. That's a huge question in ethics, 0:03:40.640 --> 0:03:44.920 in artificial intelligence in general, and in autonomous cars in particular. 0:03:45.640 --> 0:03:47.480 I don't really go into that in this episode, so 0:03:47.520 --> 0:03:50.880 there are still many topics in this field that I 0:03:50.920 --> 0:03:53.640 will have to come back to at some future point. However, 0:03:53.800 --> 0:03:56.480 that being said, there's still a ton of stuff to 0:03:56.560 --> 0:03:59.240 talk about outside of that, and I'm going to start 0:03:59.320 --> 0:04:02.240 off by exam and some of the arguments for autonomous cars, 0:04:02.240 --> 0:04:06.880 as well as the criticisms leveled at those arguments now 0:04:06.880 --> 0:04:10.200 before we jump into this, because it's going to sound 0:04:10.240 --> 0:04:14.680 like I'm totally down on the concept, like I'm like, 0:04:14.720 --> 0:04:18.200 I'm really against the idea of driverless cars. That's what 0:04:18.279 --> 0:04:20.120 it might sound like to you as we go through 0:04:20.120 --> 0:04:22.760 this episode. But that is not the case. And I'm 0:04:22.800 --> 0:04:25.920 gonna be real with you, guys. I don't drive. I 0:04:25.920 --> 0:04:29.520 I do not have a driver's license. And I don't 0:04:29.560 --> 0:04:32.360 talk about this very much because it's hard to talk 0:04:32.360 --> 0:04:35.960 about because driving in the United States is such a 0:04:35.960 --> 0:04:41.440 a common experience among many many people. I mean, unless 0:04:41.440 --> 0:04:44.480 you're living in a city that has incredible public transportation 0:04:44.839 --> 0:04:50.000 and driving is incredibly expensive, then you probably drive a car, 0:04:50.160 --> 0:04:53.240 Especially in the Southeast where I live. Owning a car 0:04:53.320 --> 0:04:55.960 and driving a car is very much tied with the 0:04:55.960 --> 0:04:59.440 thought of independence. But I don't have a driver's license. 0:04:59.480 --> 0:05:03.000 I actually have a phobia about driving. I end up 0:05:03.160 --> 0:05:05.200 locking up when I get behind the wheel of a 0:05:05.240 --> 0:05:07.839 car and I become a danger on the road, and 0:05:07.880 --> 0:05:11.559 I'm aware of that, and then that also feeds into 0:05:11.600 --> 0:05:13.520 my fear and it gets worse and worse. So you 0:05:13.560 --> 0:05:18.480 see where I'm going. I I have realized that I 0:05:18.480 --> 0:05:20.760 am a danger to myself and others if I get 0:05:20.800 --> 0:05:22.400 behind the wheel of a car. I am not a 0:05:22.440 --> 0:05:27.200 good driver. So this is an incredibly frustrating condition for me, 0:05:27.279 --> 0:05:29.320 and it's difficult for me to talk about because there's 0:05:29.360 --> 0:05:33.039 a big stigma against this. But it means that I 0:05:33.120 --> 0:05:36.479 frequently have to rely upon other people to help me 0:05:36.520 --> 0:05:39.720 get around, and that means I often feel like a burden, 0:05:40.000 --> 0:05:42.480 even when people go out of their way to dismiss 0:05:42.560 --> 0:05:44.880 that idea, saying no, no, no, no, I want to 0:05:44.920 --> 0:05:46.799 give you a ride. I want you to be involved 0:05:46.839 --> 0:05:50.839 in this thing. I can't help but feel otherwise. So 0:05:51.200 --> 0:05:53.800 I very much want a future in which we have 0:05:53.960 --> 0:05:58.440 a reliable, safe, autonomous car service where I can call 0:05:58.560 --> 0:06:01.560 upon a car anytime need and it will show up. 0:06:01.600 --> 0:06:06.159 And I'm not putting anyone out right, I'm not making 0:06:06.200 --> 0:06:10.360 it inconvenient for somebody else. It's a service. I'm paying 0:06:10.400 --> 0:06:14.240 for it. Everything's cool. I'm eager to see that kind 0:06:14.240 --> 0:06:18.640 of thing across the world. That being said, I also 0:06:18.720 --> 0:06:23.320 recognize that I can't allow my desire for that technology 0:06:23.360 --> 0:06:26.640 to come to maturity to guide all my thinking on 0:06:26.680 --> 0:06:31.640 the matter to bias my point of view. The responsible 0:06:31.680 --> 0:06:35.080 thing to do is to hold up the facts to 0:06:35.279 --> 0:06:39.880 critical thought and consideration. After all, we're ultimately talking about 0:06:40.480 --> 0:06:43.960 enormous vehicles traveling at potentially very high speeds, and that's 0:06:43.960 --> 0:06:47.480 a lot of force. So it's only right that we 0:06:47.520 --> 0:06:51.239 look at all the facts before we leap to any conclusions. 0:06:52.120 --> 0:06:55.919 And knowing how badly I want autonomous cars to become 0:06:55.920 --> 0:06:59.120 a thing has to be a warning sign to me. 0:06:59.200 --> 0:07:01.599 I have to say, well, I need to be extra 0:07:01.760 --> 0:07:04.680 careful when I'm looking at the facts so that I 0:07:04.720 --> 0:07:10.040 don't just become latch on to the most optimistic outlooks 0:07:10.200 --> 0:07:12.840 and then just, you know, deny that there are any problems. 0:07:13.200 --> 0:07:17.440 So the first big argument for autonomous cars in favor 0:07:17.600 --> 0:07:20.760 of them tends to be that, assuming you have cars 0:07:20.800 --> 0:07:25.000 that perform safely and consistently, you would cut way back 0:07:25.360 --> 0:07:29.080 on the number of fatalities and injuries caused by car accidents. 0:07:29.400 --> 0:07:32.440 And the general line of reasoning is that machines can 0:07:32.480 --> 0:07:34.560 react in a fraction of the time that people can, 0:07:34.720 --> 0:07:39.360 which is absolutely true, and a well designed vehicle, one 0:07:39.400 --> 0:07:43.040 that has the proper sensors, could maintain a three D 0:07:43.240 --> 0:07:47.880 sixty degree awareness at all times, whereas a human can 0:07:47.960 --> 0:07:51.640 only focus on a small part of their overall field 0:07:51.640 --> 0:07:55.040 of view, which is locked to whichever way they happen 0:07:55.160 --> 0:07:58.400 to turn their face at any given time. That's also 0:07:58.520 --> 0:08:02.960 true assuming you've designed the sensors and the processing system 0:08:02.960 --> 0:08:07.880 properly so that it can accept that data and analyze 0:08:07.880 --> 0:08:09.960 it in real time and be able to react to 0:08:10.040 --> 0:08:15.480 it in real time. Now, this argument that autonomous cars 0:08:15.480 --> 0:08:20.040 will save lives also assumes that the software the algorithms 0:08:20.160 --> 0:08:24.280 guiding the car's behaviors have been properly designed and tested 0:08:24.360 --> 0:08:28.160 and they're reliable. Whether that means programming the car for 0:08:28.240 --> 0:08:32.760 specific scenarios or using a process like machine learning in 0:08:32.800 --> 0:08:36.120 which an autonomous system is quote unquote taught how to 0:08:36.280 --> 0:08:41.719 handle various driving scenarios. Now, that assumption is particularly important, 0:08:42.240 --> 0:08:45.599 and it also is one that we should be particularly 0:08:45.800 --> 0:08:49.080 critical of. Now, I don't mean to suggest that it 0:08:49.200 --> 0:08:54.200 is impossible to design a sufficiently adept system or a 0:08:54.240 --> 0:08:58.440 sufficiently a depth set of algorithms but rather we should 0:08:58.440 --> 0:09:01.040 be as certain as we can and be as we 0:09:01.120 --> 0:09:05.040 can reasonably expect us to be that such systems meet 0:09:05.080 --> 0:09:10.040 are required levels of reliability before we start getting into 0:09:10.200 --> 0:09:14.360 driverless vehicles for any sort of you know, given trip. Now, 0:09:14.400 --> 0:09:18.400 all that being said, statistics do show that humans are 0:09:18.679 --> 0:09:23.160 far from perfect drivers. The U s Department of Transportation 0:09:23.240 --> 0:09:28.520 produces reports from the Fatality Analysis Reporting System or FARS, 0:09:29.240 --> 0:09:31.960 and yes, that's pretty grim, but it's very important work. 0:09:32.360 --> 0:09:35.320 The two thousand sixteen numbers, which were some of the 0:09:35.320 --> 0:09:38.240 more recent numbers that I could get hold of, say 0:09:38.240 --> 0:09:40.840 that in the United States in two thousand and sixteen, 0:09:41.160 --> 0:09:45.839 there were thirty four thousand, four hundred thirty nine recorded 0:09:45.920 --> 0:09:50.200 traffic accidents that had one or more fatalities. So if 0:09:50.200 --> 0:09:53.439 you look at all the actual deaths, you're looking at 0:09:53.480 --> 0:09:56.760 thirty seven thousand, four hundred sixty one people who lost 0:09:56.800 --> 0:09:59.880 their lives from car accidents. So when you take into 0:09:59.880 --> 0:10:03.400 a account the population of the United States, that averages 0:10:03.440 --> 0:10:07.400 out to eleven point six deaths due to car accidents 0:10:07.480 --> 0:10:10.679 per one hundred thousand people. When you break it out 0:10:10.720 --> 0:10:13.400 state by state and you look at their statistics, you 0:10:13.559 --> 0:10:17.679 see where the concentration is highest. Mississippi had the highest 0:10:17.760 --> 0:10:21.320 fatality rate by this metric, at twenty three point one 0:10:21.400 --> 0:10:25.520 fatalities per one hundred thousand people. The report also takes 0:10:25.520 --> 0:10:29.760 into account the number of vehicle miles traveled. It's very 0:10:29.800 --> 0:10:34.400 important as well, because that tells you how frequently people 0:10:34.400 --> 0:10:36.480 were on the roads and how far they went, and 0:10:36.480 --> 0:10:39.200 how much time they were spending on the roads. So 0:10:39.240 --> 0:10:44.440 in sixteen, the estimated total miles traveled in vehicles in 0:10:44.480 --> 0:10:49.840 the United States was three trillion, two hundred twenty billion, 0:10:50.120 --> 0:10:56.120 six hundred seventy seven million miles. It's a really important 0:10:56.200 --> 0:10:57.880 number that we're going to come back to in just 0:10:57.960 --> 0:11:01.760 a second. But another important number is that in the 0:11:01.880 --> 0:11:06.520 United States, experts estimate that human error causes somewhere between 0:11:08.000 --> 0:11:14.000 and of all car accidents. Uh, there's discrepancy because it 0:11:14.040 --> 0:11:18.160 all depends upon which study you're looking at, but the 0:11:18.160 --> 0:11:21.760 the message there is the overwhelming majority of car accidents 0:11:21.760 --> 0:11:25.080 in the United States are caused by human error, as 0:11:25.080 --> 0:11:29.120 opposed to say, mechanical failure or other extenuating circumstances like 0:11:29.520 --> 0:11:33.480 a tree falling on a car. So it's easy to 0:11:33.559 --> 0:11:37.319 see how the thought of removing the human element from 0:11:37.400 --> 0:11:41.960 driving could dramatically reduce the number of crashes. But again, 0:11:42.000 --> 0:11:45.360 this assumes that the driver list systems are in fact 0:11:45.720 --> 0:11:50.800 more reliable more safe than human drivers. Now they likely are, 0:11:51.520 --> 0:11:56.120 but it's hard for us to say we've definitively proven this. 0:11:56.400 --> 0:11:58.320 In fact, that would be misleading to say we have 0:11:58.440 --> 0:12:02.720 definitively proven these cars are more reliable and more safe 0:12:03.640 --> 0:12:08.640 under automated control than human control, And the reason for 0:12:08.720 --> 0:12:12.839 that is because of the amount of data. This is 0:12:12.880 --> 0:12:16.600 where we start to really get scientific with our thinking. 0:12:17.120 --> 0:12:20.560 So the safety argument says, imagine a future in which 0:12:20.600 --> 0:12:24.520 all cars are autonomous and can detect one another. They 0:12:24.520 --> 0:12:27.480 can react quickly to changes in the environment much much 0:12:27.559 --> 0:12:32.400 faster than humans can. So then imagine how drastic the 0:12:32.440 --> 0:12:36.240 reduction of fatalities would be. Those nearly forty thou people 0:12:36.640 --> 0:12:40.920 who died in would still be alive in a world 0:12:41.200 --> 0:12:44.760 where those cars were the way we got around. But 0:12:44.840 --> 0:12:48.880 this presupposes that the autonomous cars would perform as we 0:12:48.920 --> 0:12:51.959 would like them too, and that still remains to be seen. 0:12:53.040 --> 0:12:57.800 Google's way MOW, which is the company that I think 0:12:57.920 --> 0:13:02.600 is the the foremost authority on driverless cars as far 0:13:02.640 --> 0:13:07.280 as public awareness is concerned. At least, they like to 0:13:07.320 --> 0:13:10.560 point out that their self driving vehicles have passed the 0:13:10.720 --> 0:13:14.720 ten million mile mark. In other words, autonomous vehicles have 0:13:14.880 --> 0:13:18.760 driven more than ten million miles and there have been 0:13:18.800 --> 0:13:23.040 no accidents in Weymo vehicles that resulted in a fatality 0:13:23.280 --> 0:13:27.200 from their operation in autonomous mode. And ten million is 0:13:27.200 --> 0:13:29.959 a big number. Ten million miles. That's a lot of distance. 0:13:30.040 --> 0:13:35.520 But remember that in humans in the United States drove 0:13:35.880 --> 0:13:40.960 three trillion miles in one year. You know, Weymo is 0:13:41.000 --> 0:13:45.280 looking at the entire history of their autonomous car program 0:13:45.320 --> 0:13:49.400 in one year. Because of the number of drivers and 0:13:49.559 --> 0:13:53.400 how much we drive, humans racked up more than three 0:13:53.520 --> 0:13:58.520 trillion miles. So we have more human drivers on the streets, 0:13:58.800 --> 0:14:02.480 which also makes it difficult to draw any sort of 0:14:02.520 --> 0:14:05.880 comparison between human drivers and autonomous drivers. There are more 0:14:05.960 --> 0:14:09.520 human drivers out there and they're driving more miles than 0:14:09.559 --> 0:14:13.040 autonomous cars are, so it's not an apples to apples 0:14:13.160 --> 0:14:17.680 kind of comparison. We cannot say definitively that autonomous cars 0:14:17.679 --> 0:14:21.320 are safer. They just haven't been driving enough and there 0:14:21.360 --> 0:14:23.760 aren't enough of them out there for us to make 0:14:23.800 --> 0:14:27.520 that kind of general conclusion. Does that mean it's wrong? 0:14:27.840 --> 0:14:31.520 Not necessarily. It just means we lack the information to 0:14:31.600 --> 0:14:35.480 be able to say it definitively, and without that information, 0:14:36.000 --> 0:14:39.720 there's doubt, and when there's doubt, you have problems. In 0:14:39.760 --> 0:14:42.240 the last couple of years, there have been some pretty 0:14:42.320 --> 0:14:47.160 high profile incidents involving autonomous cars and accidents, and these 0:14:47.200 --> 0:14:51.720 accidents don't necessarily mean that the safety argument for autonomous 0:14:51.720 --> 0:14:55.840 cars is invalid. We haven't seen a rash of driver 0:14:55.920 --> 0:14:59.000 list car accidents. We've only seen a couple like a 0:14:59.040 --> 0:15:04.040 few dozen, and many of those aren't even the fault 0:15:04.160 --> 0:15:06.640 of the autonomous system, but rather the fault of human 0:15:06.720 --> 0:15:08.320 drivers who happened to be on the road at the 0:15:08.400 --> 0:15:12.360 same time. However, even if we were to sift through 0:15:12.520 --> 0:15:17.640 all of the data and prove definitively, statistically speaking, that 0:15:17.760 --> 0:15:22.000 driverless cars are safer than one driven by an average person, 0:15:22.800 --> 0:15:26.320 it might not matter at the end of the day. Now, 0:15:26.360 --> 0:15:30.360 this is because we humans are not entirely rational, and 0:15:30.440 --> 0:15:33.800 I am guessing this does not come as a shocking 0:15:34.320 --> 0:15:38.960 revelation to you guys. Psychologically, it can be easier for 0:15:39.040 --> 0:15:44.760 us to accept that driving a car on roads dominated 0:15:44.760 --> 0:15:49.680 by other human drivers is risky. We understand this because 0:15:49.720 --> 0:15:53.760 we know that driving, while often a mundane task that 0:15:53.800 --> 0:15:57.000 we can kind of do without even devoting our full 0:15:57.040 --> 0:16:00.760 attention to it, can sometimes be punctuated by moments of 0:16:00.920 --> 0:16:05.560 unique circumstances, and that in those circumstances, the wrong reaction 0:16:05.880 --> 0:16:09.200 could cause an accident. This is something we can wrap 0:16:09.240 --> 0:16:13.120 our minds around. We can accept this. We know accidents happen, 0:16:13.280 --> 0:16:17.240 sometimes we are in them, sometimes we cause them, and 0:16:17.280 --> 0:16:20.880 that is something that we can accept. But there's a 0:16:20.920 --> 0:16:23.600 bit of a leap for an awful lot of people 0:16:23.920 --> 0:16:27.520 when you shift that two vehicles that are controlled not 0:16:27.640 --> 0:16:32.440 by human drivers but by machines and algorithms. Suddenly the 0:16:32.480 --> 0:16:36.400 same people who accept that driving can be inherently dangerous 0:16:36.440 --> 0:16:40.040 in certain situations find it much more difficult to accept 0:16:40.320 --> 0:16:45.240 when humans are no longer directly responsible. The journal Risk 0:16:45.360 --> 0:16:49.720 Analysis published the results of a survey in which people 0:16:49.760 --> 0:16:54.040 were asked to determine how safe an autonomous vehicle would 0:16:54.080 --> 0:16:57.800 have to be before they would feel comfortable with the 0:16:57.840 --> 0:17:01.680 idea of them being on the road. The results were 0:17:01.680 --> 0:17:05.720 really interesting. They showed that, on average, people felt that 0:17:05.800 --> 0:17:08.639 autonomous cars would have to be proven to be four 0:17:08.840 --> 0:17:13.560 to five times safer than human operated vehicles for them 0:17:13.560 --> 0:17:18.600 to feel comfortable four to five times safer. So let 0:17:18.640 --> 0:17:23.199 that sink in because by this argument, even if the 0:17:23.320 --> 0:17:28.280 driverless car was twice as safe as the average human driver, 0:17:28.880 --> 0:17:31.760 which by definition would mean that if you had enough 0:17:31.760 --> 0:17:34.560 of these replacing human drivers out on the road, would 0:17:34.640 --> 0:17:38.800 mean you would have fewer injuries, fewer cases of vehicle 0:17:38.960 --> 0:17:43.440 or property damage, fewer fatalities. But that wouldn't be enough 0:17:43.760 --> 0:17:47.600 for these people to feel comfortable even if they knew that, 0:17:47.760 --> 0:17:51.760 as a result, fewer people were being hurt or killed, 0:17:52.080 --> 0:17:55.200 there was less damage. And when you turn this argument 0:17:55.200 --> 0:17:58.520 on its side, it's kind of like saying it's okay 0:17:58.560 --> 0:18:01.520 if a certain number of people are hurt or killed 0:18:01.760 --> 0:18:06.840 because we didn't shift to autonomous cars. It's only when 0:18:06.840 --> 0:18:12.159 autonomous cars are far more safe and capable four to 0:18:12.280 --> 0:18:16.200 five times than human drivers that people start to feel 0:18:16.200 --> 0:18:19.119 better about them. And I think you could argue that 0:18:19.160 --> 0:18:23.639 it's a pretty extreme double standard. We're not requiring humans 0:18:24.320 --> 0:18:29.320 to undergo driving tests more frequently to prove that they 0:18:29.480 --> 0:18:33.040 are just as safe or not if not safer, than 0:18:33.080 --> 0:18:34.960 they were the last time they took a test. We're 0:18:34.960 --> 0:18:38.280 not requiring people to do that. But it is the 0:18:38.320 --> 0:18:42.040 sort of stuff we're demanding from the machines, and it's 0:18:42.160 --> 0:18:45.680 also something I can kind of understand because it means 0:18:45.760 --> 0:18:48.639 taking the humans out of the equation, or seeming to, 0:18:49.080 --> 0:18:52.160 since these cars and algorithms are ultimately designed by humans 0:18:53.119 --> 0:18:58.280 and that's scary. The idea of entrusting lives to machines 0:18:59.320 --> 0:19:02.480 so is a bit of a paradox. So to summarize 0:19:02.520 --> 0:19:05.879 the safety part of the argument, we don't yet have 0:19:06.040 --> 0:19:09.680 enough information to really know how safe autonomous cars are 0:19:09.720 --> 0:19:13.920 compared to human drivers, although I suspect they are as 0:19:14.000 --> 0:19:17.880 a whole safer, but again hard to tell because you're 0:19:17.920 --> 0:19:22.800 extrapolating a lot from a very small data sample size. 0:19:23.400 --> 0:19:26.879 We don't know because we lack that information. There just 0:19:26.960 --> 0:19:29.960 aren't enough driverless cars in the road that haven't driven 0:19:30.040 --> 0:19:34.040 nearly enough miles and enough different environments in order for 0:19:34.119 --> 0:19:37.080 us to draw any firm conclusions. We also have to 0:19:37.080 --> 0:19:41.000 remember that different companies are taking very different approaches to 0:19:41.160 --> 0:19:44.520 driverless cars, so we can't make any sweeping statements because 0:19:44.520 --> 0:19:47.480 one approach might be better in some situations and worse 0:19:47.560 --> 0:19:50.800 than others, with no single approach being the best for 0:19:50.880 --> 0:19:55.240 all scenarios or regions. We do know human drivers are 0:19:55.280 --> 0:19:58.840 fallible and that we are at fault in the overwhelming 0:19:58.880 --> 0:20:03.560 majority of car accidents, and we suspect that eventually will 0:20:03.600 --> 0:20:06.520 have autonomous cars that can operate more safely than people. 0:20:07.520 --> 0:20:11.679 We just don't know if we're there already or if 0:20:11.720 --> 0:20:14.159 we still have a ways to go. Now. I'll have 0:20:14.240 --> 0:20:17.400 more to say about the arguments for and the arguments 0:20:17.400 --> 0:20:19.840 against autonomous cars in just a second, but first let's 0:20:19.880 --> 0:20:29.879 take a quick break to thank our sponsor. So, the 0:20:29.960 --> 0:20:34.400 safety argument also presents a couple of other possibilities. Besides 0:20:34.640 --> 0:20:38.719 an improvement in safety, assuming that everything is working properly. Uh, 0:20:39.600 --> 0:20:43.640 it also presents the possibility that autonomous cars might require 0:20:44.200 --> 0:20:48.040 less material and less energy to operate. So here here's 0:20:48.040 --> 0:20:50.719 the thought. If your vehicles are proven to be safe 0:20:51.440 --> 0:20:54.760 and you don't have to worry about them getting into accidents, 0:20:54.840 --> 0:20:56.840 let's say that you have reached a point where you're 0:20:56.840 --> 0:21:01.159 ready to really deploy these, then you don't have to 0:21:01.200 --> 0:21:05.359 build your vehicles quite as sturdily. Right. One of the 0:21:05.400 --> 0:21:09.960 reasons why cars and trucks and other you know, consumer 0:21:10.040 --> 0:21:13.960 vehicles way as much as they do is because the 0:21:14.000 --> 0:21:16.639 materials they're made out of are in part there to 0:21:16.720 --> 0:21:20.479 protect the people inside in case of accident. But if 0:21:20.520 --> 0:21:23.560 you've removed the possibility of accident, then you don't have 0:21:23.640 --> 0:21:29.600 to build them quite as Hardy will say, so it 0:21:29.600 --> 0:21:33.120 means you could go to lighter materials to make your vehicles. Well, 0:21:33.200 --> 0:21:36.600 lighter vehicles need less energy to move around than a 0:21:36.600 --> 0:21:39.640 heavier vehicle does. You don't have to put as much 0:21:39.680 --> 0:21:42.240 force to get the vehicle moving or to stop it. 0:21:42.760 --> 0:21:45.200 So if you reduce the weight of vehicles on the road, 0:21:45.960 --> 0:21:48.600 you also reduce the amount of energy you need to 0:21:48.680 --> 0:21:53.280 move them around. That also means less fuel consumption, which 0:21:53.320 --> 0:21:56.000 could be in the vehicle itself if it's a gasoline 0:21:56.000 --> 0:21:59.879 powered vehicle or hydrogen powered vehicle, or in a power 0:22:00.040 --> 0:22:03.640 plant that produces the electricity that is going to charge 0:22:03.720 --> 0:22:07.800 those electric vehicles batteries. And that sounds pretty good, except 0:22:07.840 --> 0:22:12.199 I would argue we can't really afford to get fully 0:22:12.320 --> 0:22:17.199 behind switching over to light vehicles until we reach a 0:22:17.280 --> 0:22:22.040 real tipping point of the percentage of driverless cars versus 0:22:22.200 --> 0:22:25.920 human operated ones, because even if we assume robot cars 0:22:25.960 --> 0:22:29.720 are perfect, which again this is just like hypothetical, just 0:22:29.840 --> 0:22:32.359 assume that they work exactly the way they're supposed to. 0:22:32.960 --> 0:22:34.720 You still have humans on the road and they could 0:22:34.760 --> 0:22:38.320 still cause accidents. And for that reason, we can't really 0:22:38.359 --> 0:22:42.680 afford to essue safety features that add weight to vehicles 0:22:42.720 --> 0:22:46.000 just yet. We need to still have those considerations in place. 0:22:46.800 --> 0:22:50.520 Another big argument for autonomous cars is that they could 0:22:50.560 --> 0:22:55.000 have a positive influence on traffic, meaning a reduction in 0:22:55.080 --> 0:22:59.159 traffic and traffic jams and congestion. But the studies on 0:22:59.200 --> 0:23:04.520 this subject of conflicting answers. The National Science Foundation funded 0:23:04.560 --> 0:23:07.240 a study to look into this matter, and the study 0:23:07.320 --> 0:23:11.359 consulted with auto industry experts and government officials. And in 0:23:11.359 --> 0:23:15.160 that study, the researchers created a simulation in which one 0:23:15.240 --> 0:23:20.439 autonomous vehicle, a computer controlled simulated vehicle, circled a track 0:23:20.720 --> 0:23:23.560 with at least twenty vehicles driven by humans. It was 0:23:23.600 --> 0:23:27.159 a simulation that used VR, so these were VR cars 0:23:27.240 --> 0:23:31.400 driven by human beings and not just you know, twenty 0:23:31.720 --> 0:23:34.480 simulated human beings, because then you're talking about two different 0:23:34.480 --> 0:23:37.960 types of autonomous cars, and that wouldn't tell you anything anyway. 0:23:38.119 --> 0:23:41.440 This experiment saw that humans create a stop and go 0:23:41.680 --> 0:23:47.480 oscillation effect in traffic, but that autonomous cars don't cause 0:23:47.680 --> 0:23:51.000 this kind of oscillation effect, and this can happen in 0:23:51.040 --> 0:23:54.080 things like when cars change lanes or they need to 0:23:54.840 --> 0:23:57.760 exit the highway or into the highway, and this can 0:23:57.840 --> 0:23:59.840 lead to traffic jams. So if you've ever been on 0:24:00.080 --> 0:24:03.200 highway and you hit a super slow section of traffic 0:24:03.720 --> 0:24:06.520 and then the traffic starts to clear up, and you 0:24:06.600 --> 0:24:10.640 never see any reason for the traffic jam, right, there's 0:24:10.640 --> 0:24:14.240 no accident, there's no reason, nothing to rubber neck at. 0:24:15.119 --> 0:24:17.919 You might wonder what the heck happened, Well, that's this phenomenon, 0:24:18.000 --> 0:24:23.320 this oscillating effect. It probably means that someone somewhere did 0:24:23.359 --> 0:24:27.879 this stop and go approach and it ended up perpetuating 0:24:28.040 --> 0:24:33.320 throughout the traffic from that point behind the car. So 0:24:33.359 --> 0:24:37.679 the researchers concluded that if even only five percent of 0:24:37.720 --> 0:24:40.399 the cars on the road were to be autonomous, it 0:24:40.440 --> 0:24:43.359 would cut down on traffic problems and it would reduce 0:24:43.560 --> 0:24:46.280 fuel consumption because you wouldn't be doing the stop and go, 0:24:46.440 --> 0:24:49.400 you'd be using fuel more efficiently, and that in turn 0:24:49.440 --> 0:24:52.720 would mean less pollution admitted as a result. So it 0:24:52.800 --> 0:24:55.280 has this ripple effect, which sounds great. It's kind of 0:24:55.320 --> 0:24:58.080 similar to the safety argument. And then the fact that 0:24:58.080 --> 0:25:01.359 that would mean less fuel consumption, less volution. You start 0:25:01.400 --> 0:25:05.720 to see real promise here. However, there's a flip side 0:25:05.720 --> 0:25:09.320 to this argument that we cannot ignore. While various studies 0:25:09.480 --> 0:25:13.800 have argued that concentrations of between five to of autonomous 0:25:13.880 --> 0:25:17.160 cars on the road would really help with traffic problems, 0:25:17.480 --> 0:25:20.880 and a full conversion to autonomous cars would be presumably 0:25:21.000 --> 0:25:25.119 even better, this often presupposes that the number of cars 0:25:25.119 --> 0:25:27.880 on the road would remain pretty much the same as 0:25:27.920 --> 0:25:32.359 they are today, but some studies suggest that just isn't 0:25:32.480 --> 0:25:37.320 necessarily the case. For example, the World Economic Forum and 0:25:37.400 --> 0:25:41.080 the Boston Consulting Group partnered to conduct a study about 0:25:41.119 --> 0:25:46.040 how driverless cars might impact traffic in and around Boston. 0:25:46.680 --> 0:25:50.120