WEBVTT

1
00:00:00.560 --> 00:00:03.720
<v Speaker 1>We've talked about true crime stories a lot. We talk

2
00:00:03.759 --> 00:00:07.519
<v Speaker 1>about DNA, and we talk about different ways that cases

3
00:00:07.559 --> 00:00:10.519
<v Speaker 1>can be built against people. One of those main ways

4
00:00:10.800 --> 00:00:13.279
<v Speaker 1>for the last one hundred years or so, the cases

5
00:00:13.320 --> 00:00:16.879
<v Speaker 1>have been built against criminals is fingerprints.

6
00:00:17.800 --> 00:00:20.239
<v Speaker 2>It's where we start True Crime Tuesday.

7
00:00:21.519 --> 00:00:25.199
<v Speaker 1>The story is true, that's true.

8
00:00:25.559 --> 00:00:30.280
<v Speaker 3>No, it sounds made up. I don't know. Garry and

9
00:00:30.320 --> 00:00:36.759
<v Speaker 3>Shannon present True Crime. The science of finger printing.

10
00:00:36.840 --> 00:00:41.560
<v Speaker 1>The pattern matching of fingerprinting started back in the eighteen seventies.

11
00:00:41.560 --> 00:00:44.960
<v Speaker 1>A guy named Henry Folds, a Scottish medical missionary who

12
00:00:45.039 --> 00:00:48.960
<v Speaker 1>was serving in Japan, started thinking about fingerprints and while

13
00:00:48.960 --> 00:00:51.600
<v Speaker 1>he was on an archaeological expedition with a friend, he

14
00:00:51.719 --> 00:00:55.600
<v Speaker 1>noticed that there was a print on a clay relic

15
00:00:55.679 --> 00:00:58.719
<v Speaker 1>and he reminded him of some stories he'd heard about merchants,

16
00:00:58.840 --> 00:01:02.600
<v Speaker 1>Chinese merchants who were used fingerprints to sign contracts, and

17
00:01:02.640 --> 00:01:07.200
<v Speaker 1>he said, our fingerprints really as unique as we think

18
00:01:07.239 --> 00:01:07.599
<v Speaker 1>they are.

19
00:01:08.200 --> 00:01:13.400
<v Speaker 4>So he started recording fingerprints using a method called nature printings.

20
00:01:13.519 --> 00:01:17.079
<v Speaker 4>Botanets use this to document plant leaves. You may have

21
00:01:17.159 --> 00:01:21.760
<v Speaker 4>seen this. So he smeared thin coats of printer's ink

22
00:01:21.959 --> 00:01:26.480
<v Speaker 4>over a tin plate and carefully pressed the finger print

23
00:01:26.680 --> 00:01:30.200
<v Speaker 4>of a fingertip of a volunteer onto the blotch, then

24
00:01:30.239 --> 00:01:33.519
<v Speaker 4>transfer their ink finger onto a piece of paper. Now,

25
00:01:33.599 --> 00:01:36.799
<v Speaker 4>not much has changed since good Old Henry's experience in

26
00:01:36.879 --> 00:01:40.840
<v Speaker 4>the nineteen seventies or eighteen seventies. Excuse me, that sounds

27
00:01:40.840 --> 00:01:43.680
<v Speaker 4>exactly like they do it, at least in the shows

28
00:01:43.680 --> 00:01:44.719
<v Speaker 4>that the police departments.

29
00:01:44.799 --> 00:01:47.640
<v Speaker 1>Sure, I mean, now, if you need to get fingerprinted,

30
00:01:47.680 --> 00:01:50.799
<v Speaker 1>they can do. You know, you throw your thumb print

31
00:01:50.840 --> 00:01:54.879
<v Speaker 1>down on a glass and they can scan the fingerprint

32
00:01:54.920 --> 00:01:58.480
<v Speaker 1>that way. That's probably the most advanced that they've gotten

33
00:01:58.480 --> 00:02:01.079
<v Speaker 1>when it comes to fingerprints. But over the next decade

34
00:02:01.200 --> 00:02:03.879
<v Speaker 1>or so, again we're talking about the eighteen seventies eighteen eighties,

35
00:02:04.239 --> 00:02:07.040
<v Speaker 1>he made thousands of these prints. He even enlisted the

36
00:02:07.040 --> 00:02:10.479
<v Speaker 1>staff of a hospital in Tokyo. Each fingerprint that he

37
00:02:10.560 --> 00:02:15.080
<v Speaker 1>captured appeared to be unique, and this is probably just

38
00:02:15.159 --> 00:02:20.000
<v Speaker 1>as important. It maintained its distinct pattern over time, so

39
00:02:20.159 --> 00:02:24.120
<v Speaker 1>as a kid got older, that pattern of loops and

40
00:02:24.199 --> 00:02:27.840
<v Speaker 1>whirls in that individual fingerprint never changed.

41
00:02:28.639 --> 00:02:33.240
<v Speaker 4>So his extensive and detailed collection of these fingerprints would

42
00:02:33.280 --> 00:02:38.439
<v Speaker 4>become a key piece of evidence. Tokyo police was investigating

43
00:02:38.520 --> 00:02:41.599
<v Speaker 4>a break in at the same hospital where he worked,

44
00:02:42.319 --> 00:02:46.520
<v Speaker 4>and authorities identified one of Henry's own employees as a

45
00:02:46.639 --> 00:02:52.159
<v Speaker 4>suspect and had even discovered sooty fingerprints that the intruder

46
00:02:52.199 --> 00:02:55.520
<v Speaker 4>had left on a wall. So that employee, it turned out,

47
00:02:55.599 --> 00:02:59.680
<v Speaker 4>had provided his fingerprints for Henry's study earlier, and when

48
00:02:59.719 --> 00:03:03.400
<v Speaker 4>the police compared the ones from the wall to the suspects,

49
00:03:04.120 --> 00:03:06.719
<v Speaker 4>I'd show they have the wrong person, the prince not

50
00:03:06.800 --> 00:03:07.240
<v Speaker 4>a match.

51
00:03:07.319 --> 00:03:10.240
<v Speaker 1>Well now, and this was exciting for the doctor. His

52
00:03:10.319 --> 00:03:13.280
<v Speaker 1>collection could be used to help solve crimes. Bloody finger

53
00:03:13.319 --> 00:03:17.439
<v Speaker 1>marks or impressions on clay or glass, whatever they may

54
00:03:17.560 --> 00:03:20.479
<v Speaker 1>lead to the scientific identification of criminals. This is a

55
00:03:20.560 --> 00:03:23.199
<v Speaker 1>letter that he wrote published in the journal Nature way

56
00:03:23.199 --> 00:03:27.199
<v Speaker 1>back when on the skin furrows of the hand. If

57
00:03:27.240 --> 00:03:29.000
<v Speaker 1>you're interested, you can go back and look it up.

58
00:03:29.599 --> 00:03:31.520
<v Speaker 1>But he said, there can be no doubt as to

59
00:03:31.560 --> 00:03:35.840
<v Speaker 1>the advantage of having, besides their photographs, a Nature copy

60
00:03:36.240 --> 00:03:41.800
<v Speaker 1>of the forever unchanging finger furrows of important criminals.

61
00:03:42.680 --> 00:03:44.360
<v Speaker 3>So he asked other scientists.

62
00:03:44.400 --> 00:03:45.759
<v Speaker 1>He thought that this was going to be a major

63
00:03:45.759 --> 00:03:48.680
<v Speaker 1>breakthrough and asked other scientists to help publicize his findings,

64
00:03:48.719 --> 00:03:53.280
<v Speaker 1>sent observations and classification system for fingerprints too, you know

65
00:03:53.439 --> 00:03:57.759
<v Speaker 1>a guy named Charles Darwin. Charles Darwin then shared it

66
00:03:57.800 --> 00:04:01.520
<v Speaker 1>with his cousin, the mathematics mathematician Francis Galton, and he

67
00:04:01.560 --> 00:04:04.479
<v Speaker 1>saw their potential of fingerprints to help solve crimes. He

68
00:04:04.560 --> 00:04:08.479
<v Speaker 1>decided to help develop a system to classify and to

69
00:04:08.639 --> 00:04:09.360
<v Speaker 1>analyze them.

70
00:04:09.520 --> 00:04:12.520
<v Speaker 4>Birds of a feather, kids, This is why you should

71
00:04:12.560 --> 00:04:15.719
<v Speaker 4>have friends that are smart, because then you too will

72
00:04:15.759 --> 00:04:20.040
<v Speaker 4>be smart. He was the only person Henry studying fingerprints.

73
00:04:20.079 --> 00:04:22.040
<v Speaker 4>Around the same time, there was William Herschel. He was

74
00:04:22.079 --> 00:04:25.680
<v Speaker 4>a German born British astronomer. Astronomer excuse me, working in

75
00:04:25.680 --> 00:04:29.680
<v Speaker 4>India had long kept fingerprints, records of them, began using

76
00:04:29.720 --> 00:04:35.160
<v Speaker 4>them to seal contracts. And then Galton the mathematician, collected

77
00:04:35.160 --> 00:04:39.519
<v Speaker 4>the findings of fields in Herschel and in eighteen ninety

78
00:04:39.560 --> 00:04:43.959
<v Speaker 4>two published his own classification system. So anyway, that was

79
00:04:44.000 --> 00:04:46.519
<v Speaker 4>a hit, and then it became widely publicized that this

80
00:04:46.680 --> 00:04:47.079
<v Speaker 4>was a thing.

81
00:04:48.480 --> 00:04:53.120
<v Speaker 1>There are some problems with fingerprinting, however. The loops and

82
00:04:53.160 --> 00:04:57.360
<v Speaker 1>the whorls and the arches can shrink the pool of

83
00:04:57.439 --> 00:04:59.879
<v Speaker 1>possible matches to somebody, but they're not nearly enough for

84
00:05:00.120 --> 00:05:05.079
<v Speaker 1>identification for some people. The second level of analysis includes

85
00:05:05.360 --> 00:05:06.959
<v Speaker 1>the ridges of a print.

86
00:05:07.439 --> 00:05:11.199
<v Speaker 4>For every partial fingerprint found at a crime scene, you

87
00:05:11.240 --> 00:05:13.839
<v Speaker 4>can find an expert witness that will take the sand

88
00:05:13.920 --> 00:05:16.600
<v Speaker 4>and tell you that that fingerprint is not reliable.

89
00:05:17.079 --> 00:05:18.600
<v Speaker 2>Well, now, not only.

90
00:05:18.360 --> 00:05:22.040
<v Speaker 4>Will you have the advice of whatever expert witness was

91
00:05:22.439 --> 00:05:25.160
<v Speaker 4>handsomely paid by the defense to take the stand and

92
00:05:25.199 --> 00:05:29.319
<v Speaker 4>tell you that fingerprint evidence was not reliable, AI has

93
00:05:29.360 --> 00:05:34.360
<v Speaker 4>stepped into the conversation and says that fingerprints are not unique.

94
00:05:35.160 --> 00:05:40.439
<v Speaker 4>How will this affect cases in the future and retroactively perhaps.

95
00:05:40.240 --> 00:05:41.680
<v Speaker 2>Hey, Gary and Shannon's Karen.

96
00:05:41.959 --> 00:05:46.040
<v Speaker 4>Yes, the clicking of the pen super annoying, and that's

97
00:05:46.120 --> 00:05:46.720
<v Speaker 4>all I heard.

98
00:05:46.759 --> 00:05:48.720
<v Speaker 2>I had annoyed. It was Gary was saying.

99
00:05:49.319 --> 00:05:52.839
<v Speaker 3>And then also, Gary, can you please talk like you're.

100
00:05:52.759 --> 00:05:54.959
<v Speaker 2>Drunk all the time. It's great.

101
00:05:55.519 --> 00:06:00.240
<v Speaker 4>It kills me, truly, and it just makes my date better.

102
00:06:00.879 --> 00:06:02.000
<v Speaker 1>Okay, Okay.

103
00:06:02.680 --> 00:06:03.560
<v Speaker 3>The irony of that.

104
00:06:03.680 --> 00:06:06.439
<v Speaker 1>Is my wife says that she can never tell, she

105
00:06:06.519 --> 00:06:08.480
<v Speaker 1>can never tell when I've had too much a drink ever,

106
00:06:08.600 --> 00:06:09.560
<v Speaker 1>vocally like that.

107
00:06:09.639 --> 00:06:11.600
<v Speaker 3>I don't I don't slur my words.

108
00:06:11.399 --> 00:06:14.240
<v Speaker 2>Or part of it is that you don't speak.

109
00:06:14.399 --> 00:06:15.120
<v Speaker 3>That's part of it.

110
00:06:15.199 --> 00:06:18.040
<v Speaker 1>Because I'm super self conscious of that if I ever

111
00:06:18.120 --> 00:06:20.079
<v Speaker 1>start to feel like, you know what, that was probably

112
00:06:20.240 --> 00:06:22.120
<v Speaker 1>one too many and I'm going to start slurring my

113
00:06:22.120 --> 00:06:26.879
<v Speaker 1>words here, I'll use words with hard consonants, and that's it.

114
00:06:27.639 --> 00:06:30.920
<v Speaker 3>You don't start with a year or a. You got

115
00:06:30.920 --> 00:06:32.560
<v Speaker 3>to start with a or a or a.

116
00:06:34.040 --> 00:06:37.160
<v Speaker 4>Just knowing that alone is you're already a head of

117
00:06:37.199 --> 00:06:43.079
<v Speaker 4>the game of the game about fingerprint I was just

118
00:06:43.120 --> 00:06:43.759
<v Speaker 4>going to give you that.

119
00:06:44.000 --> 00:06:46.360
<v Speaker 3>Oh, thank you God. I knew a lot about fingerprints,

120
00:06:46.639 --> 00:06:49.480
<v Speaker 3>but you just educated me. Thank you so much, Thank

121
00:06:49.519 --> 00:06:50.079
<v Speaker 3>you so much.

122
00:06:50.120 --> 00:06:52.319
<v Speaker 1>Buckle up, buttercup, because it's going to get a little

123
00:06:52.319 --> 00:06:56.439
<v Speaker 1>bit more interesting. So we're talking about fingerprint analysis for

124
00:06:56.439 --> 00:07:00.360
<v Speaker 1>true crime Tuesday, right. Long been one of the the

125
00:07:00.399 --> 00:07:05.920
<v Speaker 1>benchmarks of crime scene investigation is looking for fingerprints, and

126
00:07:06.079 --> 00:07:10.480
<v Speaker 1>there has been a belief since fingerprinting began basically one

127
00:07:10.519 --> 00:07:15.040
<v Speaker 1>hundred and fifty years ago, that individuals you and I,

128
00:07:15.519 --> 00:07:19.920
<v Speaker 1>on our individual fingers have different fingerprints. For example, if

129
00:07:19.959 --> 00:07:25.680
<v Speaker 1>I leave my thumb print on a glass and they

130
00:07:25.720 --> 00:07:29.480
<v Speaker 1>are searching for my index finger print, that those things

131
00:07:29.519 --> 00:07:30.439
<v Speaker 1>will not match.

132
00:07:31.000 --> 00:07:31.800
<v Speaker 3>So that would be.

133
00:07:33.720 --> 00:07:37.240
<v Speaker 1>You could not pin me at the crime scene because

134
00:07:37.240 --> 00:07:43.199
<v Speaker 1>my thumb print wouldn't match my index fingerprint. So an

135
00:07:43.279 --> 00:07:48.199
<v Speaker 1>undergraduate senior at Columbia Engineering named Gabe Gwo put together

136
00:07:48.240 --> 00:07:52.480
<v Speaker 1>a study where he fed into this thing about sixty

137
00:07:52.920 --> 00:07:56.959
<v Speaker 1>into a database about sixty thousand fingerprints, and there were

138
00:07:57.079 --> 00:08:01.079
<v Speaker 1>certain pairs of fingerprints that did belong to the same

139
00:08:01.120 --> 00:08:04.439
<v Speaker 1>person from different fingers, but that they belonged to the

140
00:08:04.480 --> 00:08:09.360
<v Speaker 1>same person. The AI system was able to tell at

141
00:08:09.360 --> 00:08:14.439
<v Speaker 1>a surprisingly high rate when prints that looked different actually

142
00:08:14.480 --> 00:08:17.560
<v Speaker 1>came from the same person, so that you would be

143
00:08:17.639 --> 00:08:22.480
<v Speaker 1>able to tell from a thumb print what the index

144
00:08:22.639 --> 00:08:26.199
<v Speaker 1>print looked like, or which of those was the match

145
00:08:26.759 --> 00:08:27.920
<v Speaker 1>from the same person.

146
00:08:28.720 --> 00:08:33.039
<v Speaker 3>And that's just with sixty thousand fingerprints. They're saying.

147
00:08:33.080 --> 00:08:39.320
<v Speaker 1>Now, imagine how specific and how accurate it would be

148
00:08:39.840 --> 00:08:43.120
<v Speaker 1>if it had access to millions of fingerprints and not

149
00:08:43.240 --> 00:08:44.279
<v Speaker 1>just tens of thousands.

150
00:08:44.960 --> 00:08:48.480
<v Speaker 4>This is a problem when it comes to the peer

151
00:08:48.799 --> 00:08:55.080
<v Speaker 4>review process. This was rejected this project by a well

152
00:08:55.240 --> 00:08:59.759
<v Speaker 4>established forensics journal did not accept the suggestion that different

153
00:09:00.080 --> 00:09:05.759
<v Speaker 4>fingers might produce prints with shared characteristics. The paper was

154
00:09:05.799 --> 00:09:11.519
<v Speaker 4>turned away again when it went to get a second opinion.

155
00:09:11.600 --> 00:09:15.000
<v Speaker 4>So to speak, this is more than one hundred years

156
00:09:15.039 --> 00:09:18.279
<v Speaker 4>of accepted practice that you're trying to disrupt and overturn,

157
00:09:19.120 --> 00:09:22.039
<v Speaker 4>but it was finally recognized, the study was and published

158
00:09:22.080 --> 00:09:28.000
<v Speaker 4>and the peer reviewed journal Science Advances. So reticent to

159
00:09:28.120 --> 00:09:33.879
<v Speaker 4>change the scientific community is, but nevertheless open to the possibility.

160
00:09:34.639 --> 00:09:37.440
<v Speaker 4>It's funny that the scientific community would push back on

161
00:09:37.480 --> 00:09:38.320
<v Speaker 4>something like this.

162
00:09:38.639 --> 00:09:40.000
<v Speaker 3>They don't like to be proven wrong.

163
00:09:40.039 --> 00:09:42.039
<v Speaker 1>But that's the whole point of science is to then

164
00:09:42.320 --> 00:09:46.480
<v Speaker 1>ask the questions of is this right? Does this work?

165
00:09:46.759 --> 00:09:49.120
<v Speaker 1>Can we make it better? All of those things right?

166
00:09:49.799 --> 00:09:52.000
<v Speaker 1>And if I die is going to help us do that,

167
00:09:52.080 --> 00:09:54.159
<v Speaker 1>then that's probably a good use of it.

168
00:09:54.440 --> 00:09:56.720
<v Speaker 2>I wonder what you would say about this if you

169
00:09:56.799 --> 00:09:57.679
<v Speaker 2>were drunk.

170
00:09:58.799 --> 00:10:03.399
<v Speaker 1>I would say I would say that the if you

171
00:10:03.440 --> 00:10:06.360
<v Speaker 1>if you get your stop looking at my fingers.

172
00:10:08.039 --> 00:10:09.679
<v Speaker 3>I don't know what I would say a drunk