1
00:00:00,160 --> 00:00:02,639
Speaker 1: Driverless cars have been in the movies and on TV

2
00:00:02,800 --> 00:00:06,080
Speaker 1: since the seventies. Remember Night Rider in his Car the

3
00:00:06,280 --> 00:00:10,760
Speaker 1: Night Industries two thousand or kit if you prefer, Michael,

4
00:00:10,880 --> 00:00:12,239
Speaker 1: what do you think of a stunt where I could

5
00:00:12,240 --> 00:00:15,040
Speaker 1: some assault and over end it's never been done before?

6
00:00:15,280 --> 00:00:17,760
Speaker 1: Or better still, call about a flaming chariot. We can

7
00:00:17,840 --> 00:00:20,280
Speaker 1: set my wheels on fire into a jump at night.

8
00:00:20,320 --> 00:00:23,040
Speaker 1: It could be quite spectacular. Don't you think you can't

9
00:00:23,040 --> 00:00:28,040
Speaker 1: do anything about me? I think you'd better consider reprogramming him.

10
00:00:28,240 --> 00:00:30,640
Speaker 1: Driverless cars will be on the roads of our country

11
00:00:30,640 --> 00:00:33,360
Speaker 1: in the near future, and following the cars will be

12
00:00:33,440 --> 00:00:38,239
Speaker 1: lawsuits over auto accidents. That's certain. What's not certain is

13
00:00:38,280 --> 00:00:40,839
Speaker 1: who will pay for the damages. Joining us in the

14
00:00:40,880 --> 00:00:43,920
Speaker 1: studio is Eric Gordon, professor at the University of Michigan

15
00:00:44,000 --> 00:00:46,720
Speaker 1: Ross School of Business and Michigan Law School. Thanks for

16
00:00:46,760 --> 00:00:50,640
Speaker 1: coming in, Eric, Oh, it's nice. So driverless cars don't

17
00:00:50,680 --> 00:00:53,800
Speaker 1: get drunk or speed or text or a myriad of

18
00:00:53,800 --> 00:00:57,440
Speaker 1: other things, but they'll still get into accidents. So the

19
00:00:57,480 --> 00:01:01,960
Speaker 1: car's owner, the manufacturer, the software developer. Who's going to

20
00:01:02,000 --> 00:01:05,160
Speaker 1: be responsible for the damages? You know, the the most

21
00:01:05,280 --> 00:01:08,560
Speaker 1: likely answer and the most sensible answer is that initially

22
00:01:09,040 --> 00:01:11,600
Speaker 1: it'll be the manufacturer of the car. Now, the way

23
00:01:11,600 --> 00:01:14,160
Speaker 1: the law works, the manufacturer will turn around and sue

24
00:01:14,200 --> 00:01:16,959
Speaker 1: everybody else, try to blame it on the software person

25
00:01:17,160 --> 00:01:20,759
Speaker 1: or the person who did the chip. But initially it's

26
00:01:21,040 --> 00:01:25,640
Speaker 1: likely to be the manufacturer, because well, who else will

27
00:01:25,680 --> 00:01:27,960
Speaker 1: it be if if there's nobody driving the car, there's

28
00:01:28,000 --> 00:01:32,000
Speaker 1: nobody who's negligent, there's nobody failing to keep a lookout.

29
00:01:32,400 --> 00:01:35,480
Speaker 1: There's nobody who's you know, playing with the radio and

30
00:01:35,520 --> 00:01:37,920
Speaker 1: they should be looking at the road. So the only

31
00:01:38,000 --> 00:01:41,640
Speaker 1: person left standing is going to be the manufacturer. And

32
00:01:41,800 --> 00:01:44,560
Speaker 1: it's not a big leap legally under the sort of

33
00:01:44,600 --> 00:01:47,560
Speaker 1: the products liability law as it's developed for the last

34
00:01:47,640 --> 00:01:51,560
Speaker 1: thirty or forty years. Eric, does that mean that if

35
00:01:51,600 --> 00:01:54,160
Speaker 1: I'm driving a driverless car, I will no longer have

36
00:01:54,280 --> 00:01:57,720
Speaker 1: to have insurance? You know, there will be you'll still

37
00:01:57,760 --> 00:01:59,920
Speaker 1: want to have liability insurance because there could be some

38
00:02:00,040 --> 00:02:04,640
Speaker 1: stuff that you could do wrong. Um, they will probably

39
00:02:04,680 --> 00:02:08,919
Speaker 1: have some way of overriding as a safety feature. There

40
00:02:08,919 --> 00:02:11,240
Speaker 1: have to be a way of overriding the automatic stuff.

41
00:02:11,280 --> 00:02:15,680
Speaker 1: So I think you'll have liability insurance. Um, but the cost,

42
00:02:15,840 --> 00:02:18,320
Speaker 1: the direct cost to you will go down. You're still

43
00:02:18,360 --> 00:02:20,160
Speaker 1: gonna pay for it because it's going to be rolled

44
00:02:20,160 --> 00:02:23,119
Speaker 1: into the price of the car. There still could be

45
00:02:23,600 --> 00:02:27,000
Speaker 1: an element of human error, such as a person crossing

46
00:02:27,040 --> 00:02:31,360
Speaker 1: against a light. So how will liability be decided? What

47
00:02:31,480 --> 00:02:34,560
Speaker 1: will have to be proved in a court case? You know,

48
00:02:34,639 --> 00:02:38,240
Speaker 1: there has to be some defect in the actual card

49
00:02:38,400 --> 00:02:42,160
Speaker 1: mechanism and how it's manufactured or how it was designed.

50
00:02:42,200 --> 00:02:44,440
Speaker 1: You know, maybe it was manufactured just the way it

51
00:02:44,480 --> 00:02:47,560
Speaker 1: was supposed to, but it was designed poorly. But uh,

52
00:02:47,760 --> 00:02:50,760
Speaker 1: there has to be some defect in the car, because yeah,

53
00:02:50,760 --> 00:02:53,800
Speaker 1: they're there always can be crazy things. There could be

54
00:02:53,919 --> 00:02:57,000
Speaker 1: a drunk person or a college student. I live in

55
00:02:57,040 --> 00:03:00,760
Speaker 1: college town and college students tend to sort of randomly

56
00:03:01,200 --> 00:03:05,959
Speaker 1: wander out into the streets. Um, there's probably no mechanism

57
00:03:06,360 --> 00:03:11,120
Speaker 1: human or autonomous that can be held responsible for something

58
00:03:11,160 --> 00:03:13,840
Speaker 1: that you really can't prevent somebody jumps in front of you,

59
00:03:13,960 --> 00:03:17,000
Speaker 1: So that there will have to be something that was wrong.

60
00:03:17,160 --> 00:03:20,560
Speaker 1: It won't be there's damage. The car company pays for it.

61
00:03:20,760 --> 00:03:23,040
Speaker 1: You're gonna look for a defect in how the car

62
00:03:23,160 --> 00:03:26,160
Speaker 1: is manufactured. You're gonna look for a defect in how

63
00:03:26,200 --> 00:03:32,079
Speaker 1: the software was designed. Um. Or there's something that I

64
00:03:32,160 --> 00:03:35,600
Speaker 1: think is going on now maybe with Tesla. Uh, there's

65
00:03:35,600 --> 00:03:39,320
Speaker 1: going to have to be some sort of misleading statements

66
00:03:39,320 --> 00:03:42,200
Speaker 1: from the car company about how risky it is. So

67
00:03:42,440 --> 00:03:46,440
Speaker 1: you know, this came up with Tesla is autopilot misleading

68
00:03:46,480 --> 00:03:50,280
Speaker 1: people into thinking it's really an autopilot where you don't

69
00:03:50,280 --> 00:03:53,520
Speaker 1: have to pay attention. So you know, there has to

70
00:03:53,520 --> 00:03:56,880
Speaker 1: be something that was wrong with the car, just as

71
00:03:56,920 --> 00:03:59,160
Speaker 1: now there has to be something with the way you're driving.

72
00:03:59,560 --> 00:04:02,280
Speaker 1: Your drive being down the street and you're in an accident,

73
00:04:03,080 --> 00:04:06,560
Speaker 1: you're not liable just because you're driving. Well, Eric, let

74
00:04:06,560 --> 00:04:08,880
Speaker 1: me as ask the question in a different way. If

75
00:04:09,320 --> 00:04:14,240
Speaker 1: two driverless cars getting a crash, by definition, is one

76
00:04:14,280 --> 00:04:17,560
Speaker 1: of those two cars going to be at fault? Well,

77
00:04:17,600 --> 00:04:21,040
Speaker 1: you know, maybe maybe we'll have no fault amongst autonomous

78
00:04:21,080 --> 00:04:24,960
Speaker 1: cars as we have with drivers now. Um. Uh, so

79
00:04:25,000 --> 00:04:27,440
Speaker 1: that it'll be up to the company's to battle back

80
00:04:27,480 --> 00:04:31,400
Speaker 1: and forth. Um. But you know, if it's two cars,

81
00:04:31,839 --> 00:04:36,160
Speaker 1: two autonomous cars, if they're properly designed, one or the

82
00:04:36,200 --> 00:04:40,279
Speaker 1: other of them should have stopped. Um. But I think

83
00:04:40,520 --> 00:04:43,480
Speaker 1: the liability will work an awful lot like it works

84
00:04:43,480 --> 00:04:46,680
Speaker 1: with the drivers. So think about this. There's a car

85
00:04:46,920 --> 00:04:49,960
Speaker 1: that has a defect and it's going through an intersection

86
00:04:50,000 --> 00:04:53,279
Speaker 1: where it should have stopped. It collides with another car.

87
00:04:54,440 --> 00:04:57,800
Speaker 1: If the other car, even though it wasn't initially at fault,

88
00:04:57,920 --> 00:05:02,640
Speaker 1: could have avoided the accident and famous last clear chance

89
00:05:02,720 --> 00:05:07,920
Speaker 1: doctrine in law, you know, maybe the second car is liable. So,

90
00:05:08,040 --> 00:05:09,960
Speaker 1: you know, I don't think we're gonna have to see

91
00:05:10,000 --> 00:05:14,640
Speaker 1: a lot of big changes in legal theory, um, but

92
00:05:14,720 --> 00:05:19,080
Speaker 1: it's gonna be kind of weird because we're talking about machines.

93
00:05:19,920 --> 00:05:24,599
Speaker 1: Will it be the courts that decide the liability standards

94
00:05:24,800 --> 00:05:29,120
Speaker 1: or state laws being passed? So so far it looks

95
00:05:29,120 --> 00:05:33,080
Speaker 1: as if it's going to be courts relying on state

96
00:05:33,160 --> 00:05:36,640
Speaker 1: law um and and the laws hopefully won't be that

97
00:05:36,800 --> 00:05:38,720
Speaker 1: different from state to state, but they're they're a little

98
00:05:38,720 --> 00:05:41,400
Speaker 1: different from state to state. Now when it's when it's drivers,

99
00:05:41,920 --> 00:05:45,440
Speaker 1: the federal government has not stepped in, and it's the

100
00:05:45,560 --> 00:05:51,760
Speaker 1: National Highway Transportation Safety Administration, the government regulatory agency so

101
00:05:51,800 --> 00:05:55,599
Speaker 1: far has said that no, UM, it's not going to

102
00:05:55,760 --> 00:05:58,680
Speaker 1: step in and try to preempt and have a national rule.

103
00:05:58,839 --> 00:06:00,839
Speaker 1: It is so far are going to leave it with

104
00:06:00,880 --> 00:06:03,599
Speaker 1: the states, So it'll be the same kind of patchworky

105
00:06:03,680 --> 00:06:07,080
Speaker 1: thing that we now have with human beings. That that

106
00:06:07,440 --> 00:06:09,360
Speaker 1: I mean, should the machines be treated any better than

107
00:06:09,440 --> 00:06:14,200
Speaker 1: human beings? And is all this complicated enough that we

108
00:06:14,200 --> 00:06:17,600
Speaker 1: should worry that it might actually delay the introduction of

109
00:06:17,720 --> 00:06:22,520
Speaker 1: or the widespread introduction of driverless cars? You know? Uh?

110
00:06:22,560 --> 00:06:26,600
Speaker 1: I think what's going to happen is that the insurance

111
00:06:26,640 --> 00:06:29,440
Speaker 1: companies are going to catch up, because at the end

112
00:06:29,480 --> 00:06:31,800
Speaker 1: of the day, this is an insurance question. Who's going

113
00:06:31,839 --> 00:06:35,840
Speaker 1: to carry the big insurance uh? And if it's going

114
00:06:35,880 --> 00:06:38,920
Speaker 1: to be the manufacturers for most likely to be liable,

115
00:06:39,240 --> 00:06:42,120
Speaker 1: they're going to carry the big policies, and the insurance

116
00:06:42,120 --> 00:06:44,760
Speaker 1: companies are gonna bid for that business, so they have

117
00:06:44,880 --> 00:06:48,080
Speaker 1: a lot of reason to sort it out. The system

118
00:06:48,160 --> 00:06:50,720
Speaker 1: might actually end up being sort of cheaper per person

119
00:06:50,920 --> 00:06:54,560
Speaker 1: because the the car companies are buying the insurance, you know,

120
00:06:54,600 --> 00:07:00,800
Speaker 1: they're buying it in bulk. About thirty seconds here, what

121
00:07:00,880 --> 00:07:04,760
Speaker 1: about laws like the safe distance between vehicles differing from

122
00:07:04,800 --> 00:07:08,280
Speaker 1: state to state. Supposed the states can't get it together. Yeah,

123
00:07:08,520 --> 00:07:10,520
Speaker 1: that's kind of interesting. I think the cars are going

124
00:07:10,560 --> 00:07:13,120
Speaker 1: to have to know where they are. Of course, no geolocating.

125
00:07:13,160 --> 00:07:15,960
Speaker 1: They have GPS systems and they'll have to be programmed

126
00:07:16,000 --> 00:07:19,280
Speaker 1: so that when you go from state to state, your

127
00:07:19,000 --> 00:07:21,600
Speaker 1: your your programs sort of changes the way you go

128
00:07:21,720 --> 00:07:25,520
Speaker 1: from you know, fifty to the cars will have to

129
00:07:25,520 --> 00:07:28,560
Speaker 1: adjust and they'll have to do it automatically. It sounds

130
00:07:28,760 --> 00:07:32,480
Speaker 1: very complex to me, Eric, and uh, sounds like we

131
00:07:32,560 --> 00:07:34,679
Speaker 1: have a while to go to figure this out. Thanks

132
00:07:34,680 --> 00:07:37,360
Speaker 1: so much. That's Eric Gordon. He's a professor at both

133
00:07:37,400 --> 00:07:41,400
Speaker 1: the University of Michigan Ross School of Business and Michigan

134
00:07:41,520 --> 00:07:43,160
Speaker 1: Law School. Double threat