1
00:00:06,320 --> 00:00:11,499
[Music]

2
00:00:17,119 --> 00:00:19,760
welcome back everyone

3
00:00:18,960 --> 00:00:23,039
um

4
00:00:19,760 --> 00:00:23,760
here we have lily ryan who is going to

5
00:00:23,039 --> 00:00:26,480
be

6
00:00:23,760 --> 00:00:29,679
presenting on

7
00:00:26,480 --> 00:00:32,640
an interesting segue down the

8
00:00:29,679 --> 00:00:34,160
misguided roads of

9
00:00:32,640 --> 00:00:37,120
how we're measuring

10
00:00:34,160 --> 00:00:39,920
our lives with smart devices lilly is a

11
00:00:37,120 --> 00:00:41,840
penetration tester a digitally digital

12
00:00:39,920 --> 00:00:43,120
security consultant and public speaker

13
00:00:41,840 --> 00:00:45,600
who serves on the board of digital

14
00:00:43,120 --> 00:00:47,840
rights watch lily specializes in web

15
00:00:45,600 --> 00:00:50,000
application security privacy education

16
00:00:47,840 --> 00:00:53,680
and the history of technology related

17
00:00:50,000 --> 00:00:53,680
issues no my lily

18
00:00:54,399 --> 00:00:59,039
hey everyone thank you for coming along

19
00:00:56,480 --> 00:01:00,960
to my talk today um

20
00:00:59,039 --> 00:01:02,320
took a sip of water at just the wrong

21
00:01:00,960 --> 00:01:04,479
time but anyway

22
00:01:02,320 --> 00:01:06,320
um thank you very much for being here

23
00:01:04,479 --> 00:01:08,479
before i wanted uh before i get into

24
00:01:06,320 --> 00:01:10,240
this i wanted to say that i'm speaking

25
00:01:08,479 --> 00:01:12,159
to you from the stolen lands of the

26
00:01:10,240 --> 00:01:15,439
warandra people and i acknowledged that

27
00:01:12,159 --> 00:01:15,439
sovereignty was never seeded

28
00:01:16,080 --> 00:01:18,000
so

29
00:01:17,040 --> 00:01:19,920
um

30
00:01:18,000 --> 00:01:22,240
my name's lily ryan this is how you can

31
00:01:19,920 --> 00:01:23,759
find me on twitter if you would like to

32
00:01:22,240 --> 00:01:25,360
i spend most of my time breaking

33
00:01:23,759 --> 00:01:28,000
computers and

34
00:01:25,360 --> 00:01:31,360
helping other people not break computers

35
00:01:28,000 --> 00:01:32,880
um and because i am or being a pen

36
00:01:31,360 --> 00:01:34,400
tester and i get paid to be

37
00:01:32,880 --> 00:01:35,759
professionally paranoid about things

38
00:01:34,400 --> 00:01:37,439
that means i ask these kinds of

39
00:01:35,759 --> 00:01:38,960
questions a lot

40
00:01:37,439 --> 00:01:41,119
about what it means to be human how we

41
00:01:38,960 --> 00:01:43,119
identify other humans authentication all

42
00:01:41,119 --> 00:01:45,360
of those kinds of things from that kind

43
00:01:43,119 --> 00:01:47,200
of security standpoint but because my

44
00:01:45,360 --> 00:01:49,439
background is in history i spend a lot

45
00:01:47,200 --> 00:01:51,040
of time asking the same questions from a

46
00:01:49,439 --> 00:01:53,439
philosophical and human rights based

47
00:01:51,040 --> 00:01:55,040
standpoint too so today i wanted to look

48
00:01:53,439 --> 00:01:57,119
at the history of the systems that we

49
00:01:55,040 --> 00:01:58,719
have tried to develop for identifying

50
00:01:57,119 --> 00:02:01,360
each other whether or not they're any

51
00:01:58,719 --> 00:02:02,719
good from a scientific point of view and

52
00:02:01,360 --> 00:02:04,240
then to look at some thoughtful ways

53
00:02:02,719 --> 00:02:05,759
that we can approach the topic of

54
00:02:04,240 --> 00:02:08,560
defining humans when it comes to

55
00:02:05,759 --> 00:02:10,319
building new apps and services i also

56
00:02:08,560 --> 00:02:11,760
want to say at the outset that the topic

57
00:02:10,319 --> 00:02:15,040
that we're getting into is heavy both

58
00:02:11,760 --> 00:02:18,319
emotionally and in terms of content

59
00:02:15,040 --> 00:02:20,480
i have uh not very long to cover about

60
00:02:18,319 --> 00:02:23,040
200 solid years of history philosophy

61
00:02:20,480 --> 00:02:25,440
and scientific development so i am going

62
00:02:23,040 --> 00:02:26,560
to have to skip some things and simplify

63
00:02:25,440 --> 00:02:28,959
others

64
00:02:26,560 --> 00:02:30,879
an entire ocean of ink has been spilt on

65
00:02:28,959 --> 00:02:32,239
this topic and this talk is only a drop

66
00:02:30,879 --> 00:02:34,879
so if you want to explore the rest of

67
00:02:32,239 --> 00:02:36,800
the ocean with me um please you know get

68
00:02:34,879 --> 00:02:38,319
in touch i'd be very happy to chat and

69
00:02:36,800 --> 00:02:39,760
with that out of the way let's get into

70
00:02:38,319 --> 00:02:41,760
it

71
00:02:39,760 --> 00:02:43,840
there's this really famous bit in sir

72
00:02:41,760 --> 00:02:46,480
arthur conan doyle's a study in scarlet

73
00:02:43,840 --> 00:02:48,080
where sherlock holmes meets dr watson

74
00:02:46,480 --> 00:02:49,599
for the very first time he looks him up

75
00:02:48,080 --> 00:02:52,800
and down and he says

76
00:02:49,599 --> 00:02:54,640
you have been in afghanistan i perceive

77
00:02:52,800 --> 00:02:56,080
and watson is really stunned that this

78
00:02:54,640 --> 00:02:58,159
stranger could know this about him

79
00:02:56,080 --> 00:02:59,840
without any advance warning

80
00:02:58,159 --> 00:03:01,440
later on we find out that sherlock

81
00:02:59,840 --> 00:03:04,159
holmes was able to deduce the fact that

82
00:03:01,440 --> 00:03:06,159
dr watson had been in afghanistan by

83
00:03:04,159 --> 00:03:09,040
observing the way he stands

84
00:03:06,159 --> 00:03:11,360
his skin color and his clothing all in a

85
00:03:09,040 --> 00:03:13,120
matter of seconds

86
00:03:11,360 --> 00:03:15,760
this kind of moment has been used over

87
00:03:13,120 --> 00:03:17,599
and over again to introduce detective

88
00:03:15,760 --> 00:03:20,080
characters in films and books and

89
00:03:17,599 --> 00:03:22,159
television shows in the years since then

90
00:03:20,080 --> 00:03:24,319
and it's known as the sherlock scan it's

91
00:03:22,159 --> 00:03:26,480
supposed to show it's supposed to show

92
00:03:24,319 --> 00:03:28,720
just how smart and observant the

93
00:03:26,480 --> 00:03:30,560
detective is how wise and knowledgeable

94
00:03:28,720 --> 00:03:32,879
about the world and how transparent the

95
00:03:30,560 --> 00:03:35,680
secrets of human nature are to such a

96
00:03:32,879 --> 00:03:37,200
superpowered deduction machine

97
00:03:35,680 --> 00:03:38,239
but if you've ever tried this out for

98
00:03:37,200 --> 00:03:39,519
yourself

99
00:03:38,239 --> 00:03:41,280
perhaps you've been trying to deduce

100
00:03:39,519 --> 00:03:43,440
something about a stranger at a coffee

101
00:03:41,280 --> 00:03:45,200
shop or a new acquaintance at a party

102
00:03:43,440 --> 00:03:48,640
you'll realize pretty quickly that it's

103
00:03:45,200 --> 00:03:50,080
also a complete load of junk

104
00:03:48,640 --> 00:03:52,239
in the real world

105
00:03:50,080 --> 00:03:53,760
things very rarely work this way

106
00:03:52,239 --> 00:03:55,760
while sherlock holmes said that he knew

107
00:03:53,760 --> 00:03:58,319
about watson's past partly because of

108
00:03:55,760 --> 00:03:59,680
how he was standing a person might stand

109
00:03:58,319 --> 00:04:01,439
in a particular way because they really

110
00:03:59,680 --> 00:04:03,040
need to pee or because they went too

111
00:04:01,439 --> 00:04:04,720
hard on leg day at the gym the day

112
00:04:03,040 --> 00:04:06,959
before or because they're wearing

113
00:04:04,720 --> 00:04:09,280
uncomfortable shoes or for an infinite

114
00:04:06,959 --> 00:04:11,280
number of other reasons

115
00:04:09,280 --> 00:04:13,680
the character of sherlock holmes grew

116
00:04:11,280 --> 00:04:15,920
out of a very particular very victorian

117
00:04:13,680 --> 00:04:18,160
english post-industrial revolution world

118
00:04:15,920 --> 00:04:19,840
view where scientific development

119
00:04:18,160 --> 00:04:21,840
scientific developments

120
00:04:19,840 --> 00:04:23,680
seem to be coming in thick and fast and

121
00:04:21,840 --> 00:04:26,560
new industries sprung up almost

122
00:04:23,680 --> 00:04:28,080
overnight and the world changed rapidly

123
00:04:26,560 --> 00:04:30,479
and those people who were in power who

124
00:04:28,080 --> 00:04:32,880
were leading this new scientific charge

125
00:04:30,479 --> 00:04:34,080
they were heady with success they began

126
00:04:32,880 --> 00:04:36,080
to believe

127
00:04:34,080 --> 00:04:37,600
that science really could tell us

128
00:04:36,080 --> 00:04:38,960
everything that we needed to know about

129
00:04:37,600 --> 00:04:40,880
the universe

130
00:04:38,960 --> 00:04:42,880
that there was nothing that a measuring

131
00:04:40,880 --> 00:04:44,720
tape and a calculating eye couldn't

132
00:04:42,880 --> 00:04:46,880
deduce about the world and the people in

133
00:04:44,720 --> 00:04:49,360
it and that it was only a matter of time

134
00:04:46,880 --> 00:04:51,840
before we hit upon this one unifying

135
00:04:49,360 --> 00:04:54,400
system that allowed us to definitively

136
00:04:51,840 --> 00:04:57,199
categorize everything

137
00:04:54,400 --> 00:05:00,240
sherlock holmes was their superhero the

138
00:04:57,199 --> 00:05:01,759
ultimate rational being a person who had

139
00:05:00,240 --> 00:05:03,360
it all figured out and who could read

140
00:05:01,759 --> 00:05:05,680
the world around him and translate its

141
00:05:03,360 --> 00:05:07,360
secrets back to the rest of us with

142
00:05:05,680 --> 00:05:09,360
unnerving ease

143
00:05:07,360 --> 00:05:11,919
the problem with this kind of thinking

144
00:05:09,360 --> 00:05:14,160
was that human beings remained and

145
00:05:11,919 --> 00:05:16,080
remain stubbornly difficult to

146
00:05:14,160 --> 00:05:17,199
categorize in any meaningful way

147
00:05:16,080 --> 00:05:19,919
although it hasn't stopped a lot of

148
00:05:17,199 --> 00:05:21,520
people from trying

149
00:05:19,919 --> 00:05:24,160
one of the main ways that we've tried to

150
00:05:21,520 --> 00:05:26,479
get the measure of our man so to speak

151
00:05:24,160 --> 00:05:28,160
is by using biometrics

152
00:05:26,479 --> 00:05:30,560
when i say biometrics in this talk i'm

153
00:05:28,160 --> 00:05:31,840
using the term to encompass a wide range

154
00:05:30,560 --> 00:05:34,240
of different methods that have been

155
00:05:31,840 --> 00:05:35,919
developed for measuring the human body

156
00:05:34,240 --> 00:05:37,360
historically we've used a lot of words

157
00:05:35,919 --> 00:05:38,880
for this kind of activity sometimes

158
00:05:37,360 --> 00:05:40,880
interchangeably

159
00:05:38,880 --> 00:05:42,720
anthropometry is another common word

160
00:05:40,880 --> 00:05:44,880
that has sometimes had the same meaning

161
00:05:42,720 --> 00:05:46,240
and sometimes been distinct

162
00:05:44,880 --> 00:05:48,400
there are also words that have been used

163
00:05:46,240 --> 00:05:50,639
to mean some of the things that we mean

164
00:05:48,400 --> 00:05:52,479
by biometrics today like craniometry

165
00:05:50,639 --> 00:05:54,400
which is a system for measuring the

166
00:05:52,479 --> 00:05:56,240
shape of the human skull

167
00:05:54,400 --> 00:05:57,680
there's also physiognomy which is

168
00:05:56,240 --> 00:06:00,400
supposed to be the study of what you can

169
00:05:57,680 --> 00:06:02,880
tell about the kind of person someone is

170
00:06:00,400 --> 00:06:04,639
by looking at their physical features

171
00:06:02,880 --> 00:06:06,560
and as a subset of this you also have

172
00:06:04,639 --> 00:06:08,240
phrenology which has sometimes been used

173
00:06:06,560 --> 00:06:10,479
interchangeably with craniometry and

174
00:06:08,240 --> 00:06:12,319
sometimes not but has generally come to

175
00:06:10,479 --> 00:06:14,160
mean what you can tell about the kind of

176
00:06:12,319 --> 00:06:16,319
person someone is by measuring the shape

177
00:06:14,160 --> 00:06:17,840
of their skull

178
00:06:16,319 --> 00:06:21,919
a lot of these approaches have since

179
00:06:17,840 --> 00:06:22,960
been rightly discarded as pseudoscience

180
00:06:21,919 --> 00:06:25,280
um

181
00:06:22,960 --> 00:06:27,840
it's not generous to lump them in with

182
00:06:25,280 --> 00:06:30,160
other body measuring systems that we use

183
00:06:27,840 --> 00:06:32,080
for real benefit in modern medicine

184
00:06:30,160 --> 00:06:34,240
and in criminal science but my talk

185
00:06:32,080 --> 00:06:36,160
today is all about the problems that we

186
00:06:34,240 --> 00:06:37,199
have as we work out whether any

187
00:06:36,160 --> 00:06:40,080
particular

188
00:06:37,199 --> 00:06:43,840
system of measuring people is scientific

189
00:06:40,080 --> 00:06:43,840
and ethical or not

190
00:06:44,160 --> 00:06:48,319
the question of why we want to measure

191
00:06:46,080 --> 00:06:50,400
these things at all is that we're a

192
00:06:48,319 --> 00:06:51,919
species with large brains and way too

193
00:06:50,400 --> 00:06:54,240
much time on our hands

194
00:06:51,919 --> 00:06:56,080
and on a practical level the ability to

195
00:06:54,240 --> 00:06:58,000
recognize another human being is a

196
00:06:56,080 --> 00:06:59,280
safety measure you know is this person

197
00:06:58,000 --> 00:07:00,720
that you're seeing in your living room

198
00:06:59,280 --> 00:07:03,919
at night the person you share your house

199
00:07:00,720 --> 00:07:05,759
with or is it someone who's broken in

200
00:07:03,919 --> 00:07:07,840
we also have surveillance measures you

201
00:07:05,759 --> 00:07:10,000
know is this person the same person that

202
00:07:07,840 --> 00:07:12,000
you saw breaking into your house last

203
00:07:10,000 --> 00:07:13,599
month or just someone who kind of looks

204
00:07:12,000 --> 00:07:14,880
like that person

205
00:07:13,599 --> 00:07:16,720
and the more that a group of people is

206
00:07:14,880 --> 00:07:18,479
in a position of power over other groups

207
00:07:16,720 --> 00:07:20,800
of people the more it is important to

208
00:07:18,479 --> 00:07:22,479
them to know who someone is where

209
00:07:20,800 --> 00:07:24,720
they're going or who else they might be

210
00:07:22,479 --> 00:07:26,400
spending time with and for what purpose

211
00:07:24,720 --> 00:07:27,199
this kind of knowledge

212
00:07:26,400 --> 00:07:29,759
can

213
00:07:27,199 --> 00:07:31,919
arguably be used for public good to keep

214
00:07:29,759 --> 00:07:34,479
out intruders for example but

215
00:07:31,919 --> 00:07:36,639
in the wrong hands and even in ignorant

216
00:07:34,479 --> 00:07:37,680
hands it can be used as a form of social

217
00:07:36,639 --> 00:07:39,280
control

218
00:07:37,680 --> 00:07:41,199
this is why most authoritarian

219
00:07:39,280 --> 00:07:43,199
governments end up heavily invested in

220
00:07:41,199 --> 00:07:45,360
categorizing groups of humans and they

221
00:07:43,199 --> 00:07:46,720
make extensive use of whatever biometric

222
00:07:45,360 --> 00:07:48,000
tools and systems that they have

223
00:07:46,720 --> 00:07:49,840
available

224
00:07:48,000 --> 00:07:52,000
but also human beings

225
00:07:49,840 --> 00:07:54,639
just like to know more about themselves

226
00:07:52,000 --> 00:07:56,720
we just really like quizzes anything

227
00:07:54,639 --> 00:07:58,319
that claims to offer personalized

228
00:07:56,720 --> 00:08:00,319
insights into someone's life or

229
00:07:58,319 --> 00:08:01,919
character is gonna make money and it

230
00:08:00,319 --> 00:08:04,400
doesn't matter what kind of scientific

231
00:08:01,919 --> 00:08:06,080
evidence it's based on

232
00:08:04,400 --> 00:08:07,840
we have been trying to get this kind of

233
00:08:06,080 --> 00:08:10,639
insight from at least the fourth century

234
00:08:07,840 --> 00:08:13,120
bc when one of aristotle students wrote

235
00:08:10,639 --> 00:08:15,520
an early physionomic personality manual

236
00:08:13,120 --> 00:08:17,360
based on face types

237
00:08:15,520 --> 00:08:19,280
this was based on the idea that people

238
00:08:17,360 --> 00:08:21,280
who looked like certain kinds of animals

239
00:08:19,280 --> 00:08:23,120
would have traits associated with those

240
00:08:21,280 --> 00:08:25,120
animals so the person in this picture is

241
00:08:23,120 --> 00:08:27,199
supposed to be perceptive and determined

242
00:08:25,120 --> 00:08:28,800
like a sighthound

243
00:08:27,199 --> 00:08:31,120
students of medicine have taught have

244
00:08:28,800 --> 00:08:32,719
sought insight about humans by

245
00:08:31,120 --> 00:08:34,640
dissecting the bodies of the dead and

246
00:08:32,719 --> 00:08:36,800
mapping the bodies of the living testing

247
00:08:34,640 --> 00:08:39,519
out their theories and refining them

248
00:08:36,800 --> 00:08:41,440
and in the 1940s the myers-briggs type

249
00:08:39,519 --> 00:08:42,959
indicator test was released this is

250
00:08:41,440 --> 00:08:44,399
stuck around recruitment and hr

251
00:08:42,959 --> 00:08:46,160
departments to this day as a way of

252
00:08:44,399 --> 00:08:47,360
learning more about candidates and

253
00:08:46,160 --> 00:08:48,720
employees

254
00:08:47,360 --> 00:08:51,920
despite the fact that it does have

255
00:08:48,720 --> 00:08:54,560
well-documented methodological problems

256
00:08:51,920 --> 00:08:56,240
buzzfeed knows this about humans which

257
00:08:54,560 --> 00:08:57,440
is why the quiz section on the site is

258
00:08:56,240 --> 00:09:00,000
so huge

259
00:08:57,440 --> 00:09:01,680
they rake in money by placing a long

260
00:09:00,000 --> 00:09:03,680
line of ads on our screens right next to

261
00:09:01,680 --> 00:09:05,519
the place where we spend our down time

262
00:09:03,680 --> 00:09:07,920
working out whether we're more of a paul

263
00:09:05,519 --> 00:09:10,640
or a ringo or what our taste in 90

264
00:09:07,920 --> 00:09:12,399
sitcoms says about our fashion sense

265
00:09:10,640 --> 00:09:14,560
and i should add at this point that

266
00:09:12,399 --> 00:09:15,920
there is nothing wrong with enjoying

267
00:09:14,560 --> 00:09:17,360
taking a quiz or filling out a

268
00:09:15,920 --> 00:09:18,800
questionnaire it's just another part of

269
00:09:17,360 --> 00:09:20,560
being human

270
00:09:18,800 --> 00:09:23,760
the problems creep in

271
00:09:20,560 --> 00:09:26,080
when quizzes get weaponized

272
00:09:23,760 --> 00:09:28,240
the questionnaire was invented by a guy

273
00:09:26,080 --> 00:09:30,399
called francis galton this is him

274
00:09:28,240 --> 00:09:32,880
galton was obsessed with categorizing

275
00:09:30,399 --> 00:09:34,640
humans he made it his life's work the

276
00:09:32,880 --> 00:09:37,040
questionnaire was a tool that he used to

277
00:09:34,640 --> 00:09:39,360
help his research he was also charles

278
00:09:37,040 --> 00:09:40,560
darwin's cousin and also the guy who

279
00:09:39,360 --> 00:09:43,360
introduced the idea of using

280
00:09:40,560 --> 00:09:45,120
fingerprints as a uniquely uh a way of

281
00:09:43,360 --> 00:09:47,440
uniquely identifying criminals and

282
00:09:45,120 --> 00:09:50,000
matching them to crime scenes and also

283
00:09:47,440 --> 00:09:51,839
the guy who coined the term eugenics

284
00:09:50,000 --> 00:09:54,160
which is the idea that some humans are

285
00:09:51,839 --> 00:09:55,839
biologically superior to others and that

286
00:09:54,160 --> 00:09:57,680
we could create a master race of human

287
00:09:55,839 --> 00:09:59,519
beings by breeding superior humans with

288
00:09:57,680 --> 00:10:00,880
each other and killing off or enslaving

289
00:09:59,519 --> 00:10:01,920
the rest

290
00:10:00,880 --> 00:10:04,560
and if you're thinking that sounds

291
00:10:01,920 --> 00:10:06,320
familiar it's because yes galton's work

292
00:10:04,560 --> 00:10:09,440
was directly referenced by the nazi

293
00:10:06,320 --> 00:10:11,680
party in order to justify the holocaust

294
00:10:09,440 --> 00:10:14,480
it has also been used not just by the

295
00:10:11,680 --> 00:10:17,360
nazis to justify racial segregation the

296
00:10:14,480 --> 00:10:19,680
removal of children from their families

297
00:10:17,360 --> 00:10:22,000
uh the forced sterilization of disabled

298
00:10:19,680 --> 00:10:23,680
people and people with mental illnesses

299
00:10:22,000 --> 00:10:25,120
and prison inmates and queer people and

300
00:10:23,680 --> 00:10:26,800
people of color

301
00:10:25,120 --> 00:10:29,600
and other categories of people deemed to

302
00:10:26,800 --> 00:10:32,320
be quote unquote bad for humanity

303
00:10:29,600 --> 00:10:33,279
somehow

304
00:10:32,320 --> 00:10:36,079
yeah

305
00:10:33,279 --> 00:10:37,200
um so at around the same time as galton

306
00:10:36,079 --> 00:10:39,200
and his theories there were a lot of

307
00:10:37,200 --> 00:10:40,640
other people working on similar ideas in

308
00:10:39,200 --> 00:10:43,360
europe

309
00:10:40,640 --> 00:10:45,519
in france alphonse bertion this is him

310
00:10:43,360 --> 00:10:47,920
invented the mugshot and a system called

311
00:10:45,519 --> 00:10:49,519
anthropometrics this was mostly used to

312
00:10:47,920 --> 00:10:51,360
keep track of repeat criminals by

313
00:10:49,519 --> 00:10:52,640
measuring things like the circumference

314
00:10:51,360 --> 00:10:54,880
of their head the length of their

315
00:10:52,640 --> 00:10:56,000
forearms and taking those front and side

316
00:10:54,880 --> 00:10:58,240
on pictures

317
00:10:56,000 --> 00:11:00,240
for record keeping purposes this is a

318
00:10:58,240 --> 00:11:02,079
mug shot of bertion himself with some of

319
00:11:00,240 --> 00:11:03,920
his own measurements demonstrating his

320
00:11:02,079 --> 00:11:05,519
technique

321
00:11:03,920 --> 00:11:07,600
and in italy a guy called cesare

322
00:11:05,519 --> 00:11:09,600
lombroso was one of the first people to

323
00:11:07,600 --> 00:11:10,800
use the term criminology to define the

324
00:11:09,600 --> 00:11:12,480
study of criminals and their

325
00:11:10,800 --> 00:11:14,560
characteristics

326
00:11:12,480 --> 00:11:16,720
unlike criminology today however

327
00:11:14,560 --> 00:11:19,600
which covers a wide range of the study

328
00:11:16,720 --> 00:11:21,680
of criminal behavior lombroso was most

329
00:11:19,600 --> 00:11:23,839
interested in categorizing the humans

330
00:11:21,680 --> 00:11:26,640
who had committed crimes

331
00:11:23,839 --> 00:11:28,160
he was a proponent of the idea that

332
00:11:26,640 --> 00:11:29,760
uh well i mean it's been long since

333
00:11:28,160 --> 00:11:32,320
discredited he was a proponent of the

334
00:11:29,760 --> 00:11:34,240
idea that some people are born criminals

335
00:11:32,320 --> 00:11:36,320
who are biologically more likely to

336
00:11:34,240 --> 00:11:38,640
commit crimes no matter what because of

337
00:11:36,320 --> 00:11:40,800
this according to lombroso

338
00:11:38,640 --> 00:11:43,360
he said there would be physical ways to

339
00:11:40,800 --> 00:11:44,959
tell who those born criminals were so

340
00:11:43,360 --> 00:11:47,519
the system that he developed to tell

341
00:11:44,959 --> 00:11:48,640
criminals from normal people included a

342
00:11:47,519 --> 00:11:49,839
lot of the things that will sound

343
00:11:48,640 --> 00:11:51,200
familiar to you at this point like

344
00:11:49,839 --> 00:11:53,839
measuring the size of the head and

345
00:11:51,200 --> 00:11:55,680
analyzing the facial features

346
00:11:53,839 --> 00:11:57,600
lombroso's work incorporated a lot of

347
00:11:55,680 --> 00:12:00,160
classic physionomy of that sort of

348
00:11:57,600 --> 00:12:02,399
aristotle area era

349
00:12:00,160 --> 00:12:04,320
it drew from animal imagery

350
00:12:02,399 --> 00:12:06,800
so according to him people who look like

351
00:12:04,320 --> 00:12:08,160
lions were fierce and also gun leaders

352
00:12:06,800 --> 00:12:09,760
and people who looked like bulls were

353
00:12:08,160 --> 00:12:11,680
going to be stubborn and determined and

354
00:12:09,760 --> 00:12:13,279
so on

355
00:12:11,680 --> 00:12:15,519
you can likely imagine the comparisons

356
00:12:13,279 --> 00:12:17,360
that got used for people who looked less

357
00:12:15,519 --> 00:12:18,320
white skinned and european

358
00:12:17,360 --> 00:12:20,240
um

359
00:12:18,320 --> 00:12:22,079
they were far more distasteful and

360
00:12:20,240 --> 00:12:24,800
degrading and degrading and i'm not

361
00:12:22,079 --> 00:12:27,200
going to repeat them here

362
00:12:24,800 --> 00:12:29,360
sometimes physionomy also worked the

363
00:12:27,200 --> 00:12:31,680
other way as you can see here this dog

364
00:12:29,360 --> 00:12:33,360
is apparently ambitious and jealous

365
00:12:31,680 --> 00:12:35,200
don't ask me how you could know this i

366
00:12:33,360 --> 00:12:37,200
have no idea

367
00:12:35,200 --> 00:12:38,720
lombroso also believed left-handedness

368
00:12:37,200 --> 00:12:40,720
was a sign of the born criminal just

369
00:12:38,720 --> 00:12:41,920
like most spiral notebook manufacturers

370
00:12:40,720 --> 00:12:43,760
do

371
00:12:41,920 --> 00:12:45,200
and this would be more amusing if he

372
00:12:43,760 --> 00:12:48,000
wasn't trying to sell these ideas to

373
00:12:45,200 --> 00:12:50,959
actual law enforcement

374
00:12:48,000 --> 00:12:52,560
in germany franz josef gaul this dude

375
00:12:50,959 --> 00:12:54,639
came up with the pseudoscience of

376
00:12:52,560 --> 00:12:56,160
phrenology which was supposed to explain

377
00:12:54,639 --> 00:12:58,399
things about human character by

378
00:12:56,160 --> 00:13:00,880
measuring the shape of a person's skull

379
00:12:58,399 --> 00:13:02,720
his reasoning was that because the brain

380
00:13:00,880 --> 00:13:04,399
was the seed of the personality it would

381
00:13:02,720 --> 00:13:06,560
be possible to tell what sort of

382
00:13:04,399 --> 00:13:08,959
personality you had by which parts of

383
00:13:06,560 --> 00:13:10,560
your head were bigger than others

384
00:13:08,959 --> 00:13:12,720
larger forwards for example meant that

385
00:13:10,560 --> 00:13:14,560
you were supposed to be really smart

386
00:13:12,720 --> 00:13:16,160
and as i said this is a picture of goal

387
00:13:14,560 --> 00:13:19,440
so you could see why he might have a

388
00:13:16,160 --> 00:13:21,680
vested interest in saying that

389
00:13:19,440 --> 00:13:23,200
phrenology along with physionomy entered

390
00:13:21,680 --> 00:13:24,399
the realm of popular entertainment

391
00:13:23,200 --> 00:13:25,839
really quickly

392
00:13:24,399 --> 00:13:27,519
as the idea spread throughout europe and

393
00:13:25,839 --> 00:13:29,200
the united states people published

394
00:13:27,519 --> 00:13:30,000
handbooks that you could use as a guide

395
00:13:29,200 --> 00:13:32,639
to

396
00:13:30,000 --> 00:13:34,399
reading the personalities of others

397
00:13:32,639 --> 00:13:36,399
two american businessmen brothers by the

398
00:13:34,399 --> 00:13:38,160
name of fowler started manufacturing

399
00:13:36,399 --> 00:13:40,720
phrenology busts like the one you see

400
00:13:38,160 --> 00:13:42,399
here for use as a reference tool

401
00:13:40,720 --> 00:13:44,399
these kinds of busts are mostly seen

402
00:13:42,399 --> 00:13:47,040
these days as a prop in the office of a

403
00:13:44,399 --> 00:13:49,040
psychiatrist on tv even though most real

404
00:13:47,040 --> 00:13:51,519
psychiatrists wouldn't go near one of

405
00:13:49,040 --> 00:13:53,199
them with a 12-foot pole

406
00:13:51,519 --> 00:13:54,639
people held phrenology parties where

407
00:13:53,199 --> 00:13:56,160
they would literally have their heads

408
00:13:54,639 --> 00:13:58,480
examined and their personalities

409
00:13:56,160 --> 00:14:00,480
diagnosed you could even use phrenology

410
00:13:58,480 --> 00:14:02,639
to choose the ideal spouse as you can

411
00:14:00,480 --> 00:14:06,000
see in this page out of a phrenology

412
00:14:02,639 --> 00:14:06,000
handbook from 1902

413
00:14:06,240 --> 00:14:09,920
a lot of these images that you've been

414
00:14:07,680 --> 00:14:11,600
seeing as you will probably have noticed

415
00:14:09,920 --> 00:14:13,279
are extracts from different popular

416
00:14:11,600 --> 00:14:14,959
handbooks on phrenology and physionomy

417
00:14:13,279 --> 00:14:17,440
that are published that were published

418
00:14:14,959 --> 00:14:19,440
around this time i'm showing you these

419
00:14:17,440 --> 00:14:22,240
partly because some of the illustrations

420
00:14:19,440 --> 00:14:24,160
make no sense at all even in the context

421
00:14:22,240 --> 00:14:26,079
of the second weird suit of science but

422
00:14:24,160 --> 00:14:27,519
also because these illustrations

423
00:14:26,079 --> 00:14:30,720
demonstrate the kinds of things people

424
00:14:27,519 --> 00:14:32,560
sometimes have really genuinely believed

425
00:14:30,720 --> 00:14:33,920
about one another

426
00:14:32,560 --> 00:14:35,839
i should add that at the time this was

427
00:14:33,920 --> 00:14:37,839
happening most mainstream scientists

428
00:14:35,839 --> 00:14:40,720
were absolutely not on board with

429
00:14:37,839 --> 00:14:42,560
physionomy or phrenology or eugenics by

430
00:14:40,720 --> 00:14:44,480
the early 20th century as is the case

431
00:14:42,560 --> 00:14:46,480
today the general consensus among the

432
00:14:44,480 --> 00:14:48,959
scientific community was that none of

433
00:14:46,480 --> 00:14:51,120
these personality revealing methods were

434
00:14:48,959 --> 00:14:52,399
especially scientific

435
00:14:51,120 --> 00:14:54,399
there was no way that you could use

436
00:14:52,399 --> 00:14:56,000
methods like physionomy or phrenology to

437
00:14:54,399 --> 00:14:58,880
prove anything at all about a human

438
00:14:56,000 --> 00:15:00,320
being's personality a skull is a skull

439
00:14:58,880 --> 00:15:01,519
and while measuring a skull might tell

440
00:15:00,320 --> 00:15:03,839
you something about that skull's

441
00:15:01,519 --> 00:15:06,320
structure or density or shape

442
00:15:03,839 --> 00:15:08,959
that was all a skull has no moral value

443
00:15:06,320 --> 00:15:10,880
or social role it just is

444
00:15:08,959 --> 00:15:12,399
but it didn't really stop the popular

445
00:15:10,880 --> 00:15:14,720
appeal of the idea of some kind of

446
00:15:12,399 --> 00:15:16,639
universal system where you could look at

447
00:15:14,720 --> 00:15:17,760
a person's face and know all about their

448
00:15:16,639 --> 00:15:19,680
character

449
00:15:17,760 --> 00:15:21,279
and it didn't stop authoritarian regimes

450
00:15:19,680 --> 00:15:23,680
from seeing the potential in investing

451
00:15:21,279 --> 00:15:25,360
in these kinds of ideas

452
00:15:23,680 --> 00:15:27,360
one of the popular phrenology handbooks

453
00:15:25,360 --> 00:15:30,399
that i looked at during my research was

454
00:15:27,360 --> 00:15:32,880
this one it's literally called his power

455
00:15:30,399 --> 00:15:34,560
it was published in 1923 just as some

456
00:15:32,880 --> 00:15:36,240
world governments were using these bunk

457
00:15:34,560 --> 00:15:38,079
philosophies to establish their own

458
00:15:36,240 --> 00:15:39,759
political power

459
00:15:38,079 --> 00:15:41,199
and all of these ideas drifted in and

460
00:15:39,759 --> 00:15:43,120
out of each other's spheres in the late

461
00:15:41,199 --> 00:15:45,279
19th and early 20th centuries they

462
00:15:43,120 --> 00:15:46,800
bounced and branched off each other in

463
00:15:45,279 --> 00:15:48,000
countless ways

464
00:15:46,800 --> 00:15:49,600
you will have seen that some of the

465
00:15:48,000 --> 00:15:51,440
things that came out of the effort to

466
00:15:49,600 --> 00:15:53,920
measure and categorize human beings

467
00:15:51,440 --> 00:15:56,240
ended up being absolutely useless except

468
00:15:53,920 --> 00:15:57,839
for a quirky way for some people to pass

469
00:15:56,240 --> 00:15:59,680
the time

470
00:15:57,839 --> 00:16:01,440
other things that came out of this were

471
00:15:59,680 --> 00:16:03,440
among the greatest and most terrible

472
00:16:01,440 --> 00:16:05,440
atrocities perpetrated by human beings

473
00:16:03,440 --> 00:16:07,519
against each other in the history of our

474
00:16:05,440 --> 00:16:08,959
existence as a species

475
00:16:07,519 --> 00:16:10,959
and still other things that came out of

476
00:16:08,959 --> 00:16:13,600
this were things that have become an

477
00:16:10,959 --> 00:16:15,680
accepted or even arguably useful part of

478
00:16:13,600 --> 00:16:17,680
life today in many places

479
00:16:15,680 --> 00:16:19,199
the study of criminology the use of

480
00:16:17,680 --> 00:16:20,880
fingerprints and mug shots and police

481
00:16:19,199 --> 00:16:22,639
work questionnaires and surveys as

482
00:16:20,880 --> 00:16:24,079
research tools and the measurement of

483
00:16:22,639 --> 00:16:26,399
human bodies in order to learn more

484
00:16:24,079 --> 00:16:28,000
about them for medical reasons

485
00:16:26,399 --> 00:16:29,120
this is the thing about biometrics

486
00:16:28,000 --> 00:16:30,959
generally

487
00:16:29,120 --> 00:16:33,199
while all measurements are themselves

488
00:16:30,959 --> 00:16:35,279
numbers on a page or in a database the

489
00:16:33,199 --> 00:16:37,040
questions that we try to answer and the

490
00:16:35,279 --> 00:16:38,720
problems that we try to solve using

491
00:16:37,040 --> 00:16:40,959
these measurements can have profound

492
00:16:38,720 --> 00:16:43,120
impacts and consequences whether they're

493
00:16:40,959 --> 00:16:45,680
good or evil

494
00:16:43,120 --> 00:16:47,279
so it's now the year 2022 and even

495
00:16:45,680 --> 00:16:48,959
though there are human beings living in

496
00:16:47,279 --> 00:16:51,360
space and we have the ability to hold

497
00:16:48,959 --> 00:16:54,079
events like this online in real time

498
00:16:51,360 --> 00:16:56,320
we haven't fundamentally changed we

499
00:16:54,079 --> 00:16:58,240
still want the same kinds of things like

500
00:16:56,320 --> 00:16:59,839
knowing about each other and ourselves

501
00:16:58,240 --> 00:17:01,680
to find a way of organizing the world in

502
00:16:59,839 --> 00:17:03,759
our heads that makes sense to us and for

503
00:17:01,680 --> 00:17:05,919
feeling like if we have enough data we

504
00:17:03,759 --> 00:17:07,839
might be able to establish some kind of

505
00:17:05,919 --> 00:17:10,480
control over our lives

506
00:17:07,839 --> 00:17:12,480
we even succeed at it some of the time

507
00:17:10,480 --> 00:17:14,240
wearable devices like the fitbit and the

508
00:17:12,480 --> 00:17:15,600
apple watch have become ways for us to

509
00:17:14,240 --> 00:17:17,839
measure the things that we want to know

510
00:17:15,600 --> 00:17:19,360
about ourselves like the number of steps

511
00:17:17,839 --> 00:17:21,679
that we take in a day or our heart rate

512
00:17:19,360 --> 00:17:23,280
or how well we slept last night

513
00:17:21,679 --> 00:17:24,400
some of us like to measure our weight

514
00:17:23,280 --> 00:17:26,640
and our bowel movements and our

515
00:17:24,400 --> 00:17:28,799
menstrual cycles and some of us lived

516
00:17:26,640 --> 00:17:30,640
vastly improved lives because we have

517
00:17:28,799 --> 00:17:33,679
the ability to do this i mean i'm i'm

518
00:17:30,640 --> 00:17:35,520
wearing fitbit right now

519
00:17:33,679 --> 00:17:37,520
the tools we have for

520
00:17:35,520 --> 00:17:39,679
uh for using this these days have become

521
00:17:37,520 --> 00:17:42,080
more complex than calipers and measuring

522
00:17:39,679 --> 00:17:42,080
tapes

523
00:17:43,120 --> 00:17:45,679
the current model of an iphone for

524
00:17:44,400 --> 00:17:47,760
example this is not a picture of a

525
00:17:45,679 --> 00:17:50,480
current iphone iphone users will know

526
00:17:47,760 --> 00:17:52,960
this but the current model of iphone has

527
00:17:50,480 --> 00:17:54,960
a true depth camera with infrared depth

528
00:17:52,960 --> 00:17:56,799
perception capabilities

529
00:17:54,960 --> 00:17:59,280
another 12 megapixel camera i think two

530
00:17:56,799 --> 00:18:00,960
now a microphone an ambient light sensor

531
00:17:59,280 --> 00:18:02,559
a moisture sensor an accelerometer a

532
00:18:00,960 --> 00:18:04,799
gyroscope a compass a barometer and a

533
00:18:02,559 --> 00:18:06,400
thumbprint sensor and we can carry it in

534
00:18:04,799 --> 00:18:08,000
our pocket

535
00:18:06,400 --> 00:18:09,679
if you add a specialized device like a

536
00:18:08,000 --> 00:18:12,080
smartwatch or an external camera our

537
00:18:09,679 --> 00:18:14,000
capability grows and as developers we

538
00:18:12,080 --> 00:18:16,559
can use these capabilities in our own

539
00:18:14,000 --> 00:18:17,840
software to help personalize them for

540
00:18:16,559 --> 00:18:20,640
our users and make the things that they

541
00:18:17,840 --> 00:18:22,080
do more relevant and meaningful

542
00:18:20,640 --> 00:18:23,440
some of you who are watching this might

543
00:18:22,080 --> 00:18:25,039
have built things using things like

544
00:18:23,440 --> 00:18:27,120
apple's research kit and care kit or

545
00:18:25,039 --> 00:18:29,600
health kit maybe all three these

546
00:18:27,120 --> 00:18:30,880
frameworks really super powerful they're

547
00:18:29,600 --> 00:18:32,799
useful in the potential that they

548
00:18:30,880 --> 00:18:35,039
represent for having people help for

549
00:18:32,799 --> 00:18:37,440
helping people sorry research and detect

550
00:18:35,039 --> 00:18:39,760
and manage human health using them

551
00:18:37,440 --> 00:18:41,600
people have done awesome things like

552
00:18:39,760 --> 00:18:43,679
learning to predict epileptic seizures

553
00:18:41,600 --> 00:18:45,520
and manage asthma and track medication

554
00:18:43,679 --> 00:18:48,240
use and care for patients after surgery

555
00:18:45,520 --> 00:18:50,160
and learn more about parkinson's disease

556
00:18:48,240 --> 00:18:51,600
but getting meaningful results usually

557
00:18:50,160 --> 00:18:53,840
comes down to whether or not we're

558
00:18:51,600 --> 00:18:55,919
collecting the right data and using it

559
00:18:53,840 --> 00:18:58,000
to answer questions that it can actually

560
00:18:55,919 --> 00:19:00,960
answer properly and then accepting

561
00:18:58,000 --> 00:19:02,400
feedback and reviewing our hypotheses

562
00:19:00,960 --> 00:19:03,280
if you've ever worked with user input

563
00:19:02,400 --> 00:19:04,799
data

564
00:19:03,280 --> 00:19:06,320
you probably know

565
00:19:04,799 --> 00:19:08,160
just how difficult it is to handle

566
00:19:06,320 --> 00:19:11,200
everything that your users throw at you

567
00:19:08,160 --> 00:19:14,000
things about humans are hard

568
00:19:11,200 --> 00:19:16,480
names genders addresses time zones

569
00:19:14,000 --> 00:19:19,120
geographic location language you name it

570
00:19:16,480 --> 00:19:21,039
you will find an exception to your rule

571
00:19:19,120 --> 00:19:22,880
this classic list of falsehoods that

572
00:19:21,039 --> 00:19:25,120
programmers believe is a fun way to dive

573
00:19:22,880 --> 00:19:27,280
in and see just how many ways that

574
00:19:25,120 --> 00:19:29,520
people can be

575
00:19:27,280 --> 00:19:30,799
human beings are impossibly varied and

576
00:19:29,520 --> 00:19:32,240
anyone we're not thinking about at the

577
00:19:30,799 --> 00:19:34,880
time when we build stuff will probably

578
00:19:32,240 --> 00:19:36,320
be left out

579
00:19:34,880 --> 00:19:38,960
some of you will know

580
00:19:36,320 --> 00:19:40,720
in 2014 apple became infamous for this

581
00:19:38,960 --> 00:19:42,559
because at its launch the health app

582
00:19:40,720 --> 00:19:44,559
contained no way for users to record

583
00:19:42,559 --> 00:19:46,720
menstrual data despite the fact that

584
00:19:44,559 --> 00:19:47,919
close to half of iphone users did

585
00:19:46,720 --> 00:19:50,160
menstruate

586
00:19:47,919 --> 00:19:51,679
this feature was added in 2015 and

587
00:19:50,160 --> 00:19:53,120
they've clearly taken this on board in

588
00:19:51,679 --> 00:19:55,120
the years since then some of you have

589
00:19:53,120 --> 00:19:57,280
probably seen that you know there's a

590
00:19:55,120 --> 00:19:58,799
cycle tracker by all accounts the

591
00:19:57,280 --> 00:20:00,559
features that that it now has the

592
00:19:58,799 --> 00:20:02,080
language it uses show that it's been

593
00:20:00,559 --> 00:20:03,360
given a lot of thought

594
00:20:02,080 --> 00:20:05,120
but even when you've accounted for the

595
00:20:03,360 --> 00:20:06,240
most standard use cases it's hard to

596
00:20:05,120 --> 00:20:08,720
come up with a system of human

597
00:20:06,240 --> 00:20:10,240
measurement that's true for everybody

598
00:20:08,720 --> 00:20:12,880
sentiment analysis from facial

599
00:20:10,240 --> 00:20:15,280
expressions is a particularly flawy area

600
00:20:12,880 --> 00:20:17,039
our faces don't all work the same way it

601
00:20:15,280 --> 00:20:18,240
might be all well and good as long as

602
00:20:17,039 --> 00:20:19,840
you're in total control of everything

603
00:20:18,240 --> 00:20:21,919
that your face is doing but not all of

604
00:20:19,840 --> 00:20:23,360
us are some folks will have ticks that

605
00:20:21,919 --> 00:20:25,440
they can't do anything about or they may

606
00:20:23,360 --> 00:20:27,600
have neurochemistry that makes it tricky

607
00:20:25,440 --> 00:20:30,159
to express emotions externally or

608
00:20:27,600 --> 00:20:31,840
perhaps not the expected ones

609
00:20:30,159 --> 00:20:33,039
this software also won't work well for

610
00:20:31,840 --> 00:20:34,559
you if you've had a stroke and you've

611
00:20:33,039 --> 00:20:36,240
lost control of some of your facial

612
00:20:34,559 --> 00:20:37,520
muscles or even if you've been to the

613
00:20:36,240 --> 00:20:39,760
dentist that day and you can't really

614
00:20:37,520 --> 00:20:41,440
feel your mouth

615
00:20:39,760 --> 00:20:43,200
we get similar problems with gait and

616
00:20:41,440 --> 00:20:45,039
posture analysis to try and predict

617
00:20:43,200 --> 00:20:46,080
things about someone's mood what an

618
00:20:45,039 --> 00:20:47,679
algorithm might have been trained to

619
00:20:46,080 --> 00:20:49,840
perceive as threatening or angry

620
00:20:47,679 --> 00:20:52,000
movement might also easily be the result

621
00:20:49,840 --> 00:20:53,760
of a panic attack or a lack of control

622
00:20:52,000 --> 00:20:55,440
over motor functions and the same is

623
00:20:53,760 --> 00:20:57,520
true of failing to account for the way

624
00:20:55,440 --> 00:21:00,000
that human beings change over time and

625
00:20:57,520 --> 00:21:02,080
sometimes rapidly there's been more than

626
00:21:00,000 --> 00:21:04,000
one example of pregnant people being

627
00:21:02,080 --> 00:21:05,200
hounded by their health trackers to lose

628
00:21:04,000 --> 00:21:07,360
weight when they were supposed to be

629
00:21:05,200 --> 00:21:09,440
gaining it and then having the health

630
00:21:07,360 --> 00:21:10,960
trackers yell at them again for not

631
00:21:09,440 --> 00:21:13,039
sleeping poorly

632
00:21:10,960 --> 00:21:14,400
for for sleeping poorly once the baby's

633
00:21:13,039 --> 00:21:16,480
been born

634
00:21:14,400 --> 00:21:18,400
there have been forums filled with

635
00:21:16,480 --> 00:21:20,640
first-hand accounts from people with

636
00:21:18,400 --> 00:21:22,720
mobility issues being berated for not

637
00:21:20,640 --> 00:21:24,799
completing their 10 000 steps when some

638
00:21:22,720 --> 00:21:26,000
people just want to celebrate doing 50

639
00:21:24,799 --> 00:21:27,679
steps

640
00:21:26,000 --> 00:21:29,440
and then we have issues with all of that

641
00:21:27,679 --> 00:21:31,679
data that we collect being used for

642
00:21:29,440 --> 00:21:33,200
purposes that have nothing to do with

643
00:21:31,679 --> 00:21:35,440
helping the people that the app is

644
00:21:33,200 --> 00:21:36,720
ostensibly for

645
00:21:35,440 --> 00:21:38,400
once we got the hang of collecting

646
00:21:36,720 --> 00:21:39,840
menstrual data it was great for period

647
00:21:38,400 --> 00:21:41,919
tracking but we've also learned that

648
00:21:39,840 --> 00:21:43,440
companies like ovia health have

649
00:21:41,919 --> 00:21:45,520
developed family planning apps that

650
00:21:43,440 --> 00:21:47,280
people's employers can monitor so in the

651
00:21:45,520 --> 00:21:48,880
april of 2019

652
00:21:47,280 --> 00:21:51,360
there was a report that activision

653
00:21:48,880 --> 00:21:53,120
blizzard the video game company were

654
00:21:51,360 --> 00:21:54,880
encouraging menstruating employees to

655
00:21:53,120 --> 00:21:57,039
use this app which could provide their

656
00:21:54,880 --> 00:21:59,039
bosses with data about their menstrual

657
00:21:57,039 --> 00:22:00,080
cycles and pregnancy stages and fertile

658
00:21:59,039 --> 00:22:02,400
windows

659
00:22:00,080 --> 00:22:04,559
in exchange in exchange for getting a

660
00:22:02,400 --> 00:22:06,400
front row view to all this uterus

661
00:22:04,559 --> 00:22:08,480
activity for honestly god knows what

662
00:22:06,400 --> 00:22:11,200
reason the company paid these employees

663
00:22:08,480 --> 00:22:13,120
an extra one dollar a day

664
00:22:11,200 --> 00:22:14,400
given the fact that

665
00:22:13,120 --> 00:22:15,840
an entire

666
00:22:14,400 --> 00:22:17,760
californian government department has

667
00:22:15,840 --> 00:22:19,280
now filed a suit against activision

668
00:22:17,760 --> 00:22:21,919
blizzard for their despicable workplace

669
00:22:19,280 --> 00:22:23,280
culture i cannot imagine what extra hell

670
00:22:21,919 --> 00:22:25,120
it must have been like to work there

671
00:22:23,280 --> 00:22:27,360
while knowing your bosses were tracking

672
00:22:25,120 --> 00:22:29,280
your periods

673
00:22:27,360 --> 00:22:31,760
something some insurance companies like

674
00:22:29,280 --> 00:22:33,280
in in the united states in particular

675
00:22:31,760 --> 00:22:35,120
also have policies that mandate that

676
00:22:33,280 --> 00:22:36,640
their customers wear a fitbit or they'll

677
00:22:35,120 --> 00:22:37,520
pay extra premiums for the healthcare

678
00:22:36,640 --> 00:22:39,120
cover

679
00:22:37,520 --> 00:22:42,000
and this continues to ensure that the

680
00:22:39,120 --> 00:22:43,760
only people who can sidestep this this

681
00:22:42,000 --> 00:22:45,039
cycle of data gathering and surveillance

682
00:22:43,760 --> 00:22:47,360
and those who have enough money to

683
00:22:45,039 --> 00:22:49,840
afford it leaving the privacy risks for

684
00:22:47,360 --> 00:22:51,360
the most vulnerable people

685
00:22:49,840 --> 00:22:53,360
then we have the straight up old-school

686
00:22:51,360 --> 00:22:55,360
phrenological style uses the kinds of

687
00:22:53,360 --> 00:22:57,120
services where data is gathered about

688
00:22:55,360 --> 00:22:59,760
people and used to draw conclusions that

689
00:22:57,120 --> 00:23:01,440
there is no scientific basis for check

690
00:22:59,760 --> 00:23:03,760
out this screenshot here this is a

691
00:23:01,440 --> 00:23:05,679
company called face session face ception

692
00:23:03,760 --> 00:23:07,600
they've been around for a while now they

693
00:23:05,679 --> 00:23:09,600
have a whole business model centered

694
00:23:07,600 --> 00:23:11,120
around inferring things about people

695
00:23:09,600 --> 00:23:12,640
from an image of their face things like

696
00:23:11,120 --> 00:23:16,000
whether or not someone might be a poker

697
00:23:12,640 --> 00:23:18,400
player or a terrorist

698
00:23:16,000 --> 00:23:20,000
in 2017 two stanford university

699
00:23:18,400 --> 00:23:22,000
researchers conducted a study that

700
00:23:20,000 --> 00:23:23,120
claimed machines were better than humans

701
00:23:22,000 --> 00:23:26,159
at looking at someone's face and

702
00:23:23,120 --> 00:23:28,799
determining if they were gay or straight

703
00:23:26,159 --> 00:23:30,400
not only did this study lack rigor and

704
00:23:28,799 --> 00:23:33,679
completely ignore the existence of

705
00:23:30,400 --> 00:23:35,600
things like bisexuality or ethics it

706
00:23:33,679 --> 00:23:37,120
also re-popularized the idea that there

707
00:23:35,600 --> 00:23:38,640
was something inherent about looking at

708
00:23:37,120 --> 00:23:40,159
another person that could tell you

709
00:23:38,640 --> 00:23:43,360
whether or not they were attracted to

710
00:23:40,159 --> 00:23:44,799
people of the same gender as themselves

711
00:23:43,360 --> 00:23:46,400
a researcher at the university of

712
00:23:44,799 --> 00:23:48,559
pretoria in south africa tried to

713
00:23:46,400 --> 00:23:50,559
replicate the study in november 2018 and

714
00:23:48,559 --> 00:23:52,880
found that the results didn't bear up to

715
00:23:50,559 --> 00:23:54,480
much scientific scrutiny but that report

716
00:23:52,880 --> 00:23:56,080
didn't receive anywhere near as much

717
00:23:54,480 --> 00:23:58,559
publicity

718
00:23:56,080 --> 00:24:00,240
and then in the january of 2021 one of

719
00:23:58,559 --> 00:24:02,159
the authors of this weird gaydar study

720
00:24:00,240 --> 00:24:03,520
did it again claiming that facial

721
00:24:02,159 --> 00:24:06,720
recognition tech could tell you

722
00:24:03,520 --> 00:24:08,080
someone's political orientation

723
00:24:06,720 --> 00:24:09,919
if that's not enough during the late

724
00:24:08,080 --> 00:24:12,000
trump administration parts of the u.s

725
00:24:09,919 --> 00:24:13,440
government started considering whether

726
00:24:12,000 --> 00:24:15,440
they should create a national database

727
00:24:13,440 --> 00:24:16,720
of people with mental illnesses this

728
00:24:15,440 --> 00:24:18,400
would have been combined with a program

729
00:24:16,720 --> 00:24:20,000
to harvest data from health trackers

730
00:24:18,400 --> 00:24:22,080
like the fitbit and the apple watch to

731
00:24:20,000 --> 00:24:24,559
create a model that would watch for

732
00:24:22,080 --> 00:24:26,080
quote unquote neurobehavioral signs that

733
00:24:24,559 --> 00:24:27,440
might indicate someone with a mental

734
00:24:26,080 --> 00:24:28,720
illness was going to go on a shooting

735
00:24:27,440 --> 00:24:30,240
spree

736
00:24:28,720 --> 00:24:32,799
the group that pitched this called it

737
00:24:30,240 --> 00:24:35,360
the safe home project this is a backroom

738
00:24:32,799 --> 00:24:38,640
for stopping aberrant fatal events by

739
00:24:35,360 --> 00:24:40,720
helping overcome mental extremes

740
00:24:38,640 --> 00:24:42,320
and while the problem of gun violence in

741
00:24:40,720 --> 00:24:44,400
the united states is definitely

742
00:24:42,320 --> 00:24:46,080
something that needs addressing

743
00:24:44,400 --> 00:24:47,840
weaponizing biometric tech against

744
00:24:46,080 --> 00:24:50,400
people with mental illnesses is unlikely

745
00:24:47,840 --> 00:24:53,279
to do anything except create fear among

746
00:24:50,400 --> 00:24:54,720
a vulnerable population

747
00:24:53,279 --> 00:24:56,400
quite apart from the fact that there's

748
00:24:54,720 --> 00:24:58,080
no way we can tell whether someone's

749
00:24:56,400 --> 00:25:00,080
planning a violent rampage just from the

750
00:24:58,080 --> 00:25:01,760
readings off of like an apple watch

751
00:25:00,080 --> 00:25:03,760
there's also the problem that internet

752
00:25:01,760 --> 00:25:05,360
of things devices especially consumer

753
00:25:03,760 --> 00:25:07,520
grade personal health devices aren't

754
00:25:05,360 --> 00:25:08,799
that always that accurate that might be

755
00:25:07,520 --> 00:25:10,640
enough to tell you you got as much

756
00:25:08,799 --> 00:25:12,559
physical activity as you wanted compared

757
00:25:10,640 --> 00:25:14,880
to other days but there have also been

758
00:25:12,559 --> 00:25:17,279
cases like this one where the microsoft

759
00:25:14,880 --> 00:25:20,799
band made rest in peace was shown to be

760
00:25:17,279 --> 00:25:22,960
getting a heart rate of a chicken breast

761
00:25:20,799 --> 00:25:24,720
and there are also reports that using a

762
00:25:22,960 --> 00:25:26,960
green light as an optical sensor for

763
00:25:24,720 --> 00:25:28,559
tracking heart rates means that devices

764
00:25:26,960 --> 00:25:30,640
like the fitbit aren't accurately

765
00:25:28,559 --> 00:25:31,840
measuring this info for people with dark

766
00:25:30,640 --> 00:25:33,279
skin

767
00:25:31,840 --> 00:25:34,480
consumer grade health trackers will

768
00:25:33,279 --> 00:25:35,919
often go with green light sensors

769
00:25:34,480 --> 00:25:37,840
instead of infrared because they're

770
00:25:35,919 --> 00:25:40,400
cheaper and they lower the cost of the

771
00:25:37,840 --> 00:25:42,320
device but it's been shown that melanin

772
00:25:40,400 --> 00:25:44,320
will absorb green light so those with

773
00:25:42,320 --> 00:25:46,159
more melanin in their skin aren't going

774
00:25:44,320 --> 00:25:48,000
to get the same results as people with

775
00:25:46,159 --> 00:25:49,520
less of it

776
00:25:48,000 --> 00:25:51,440
so once again

777
00:25:49,520 --> 00:25:53,279
we're at a point where scientific and

778
00:25:51,440 --> 00:25:55,200
technological advances are coming at us

779
00:25:53,279 --> 00:25:56,720
thick and fast and the world is changing

780
00:25:55,200 --> 00:25:58,480
and we think we can solve all the

781
00:25:56,720 --> 00:26:00,720
world's problems by classifying and

782
00:25:58,480 --> 00:26:03,200
measuring the humans in it

783
00:26:00,720 --> 00:26:04,799
but by now you should be clear that i

784
00:26:03,200 --> 00:26:06,880
think some of the questions that we're

785
00:26:04,799 --> 00:26:09,120
trying to answer with biometrics aren't

786
00:26:06,880 --> 00:26:10,799
always the right ones and in some cases

787
00:26:09,120 --> 00:26:12,559
they're going to lead us or they are

788
00:26:10,799 --> 00:26:14,320
actually leading us down a really

789
00:26:12,559 --> 00:26:15,679
dangerous path

790
00:26:14,320 --> 00:26:18,080
we have a world

791
00:26:15,679 --> 00:26:19,440
where there are actual nazis again

792
00:26:18,080 --> 00:26:21,440
and people are writing apparently

793
00:26:19,440 --> 00:26:23,120
serious articles about craniomatry and

794
00:26:21,440 --> 00:26:25,279
race again

795
00:26:23,120 --> 00:26:27,840
and the last time around weaponized

796
00:26:25,279 --> 00:26:30,080
categorization was one of their most

797
00:26:27,840 --> 00:26:34,000
potent tools

798
00:26:30,080 --> 00:26:34,000
so knowing what we know what do we do

799
00:26:34,799 --> 00:26:38,960
i mean most of us mean well we want to

800
00:26:36,880 --> 00:26:40,720
do our best we want to build cool apps

801
00:26:38,960 --> 00:26:42,400
and software and all kinds of things

802
00:26:40,720 --> 00:26:43,840
that help people

803
00:26:42,400 --> 00:26:46,080
but not everyone's thinking about the

804
00:26:43,840 --> 00:26:49,440
data or using that data to answer

805
00:26:46,080 --> 00:26:50,799
focused and answerable questions

806
00:26:49,440 --> 00:26:53,440
so how do we make sure that we get the

807
00:26:50,799 --> 00:26:54,720
best out of our scientific advancements

808
00:26:53,440 --> 00:26:56,559
and acknowledge our boundaries and our

809
00:26:54,720 --> 00:26:57,840
biases

810
00:26:56,559 --> 00:26:59,039
right now the main way of addressing

811
00:26:57,840 --> 00:27:00,720
this seems to be coming up with a

812
00:26:59,039 --> 00:27:02,640
proposal for an ethical framework that

813
00:27:00,720 --> 00:27:04,480
can be used to guide us

814
00:27:02,640 --> 00:27:06,480
so when i searched this a couple of

815
00:27:04,480 --> 00:27:08,480
hours ago google scholar returned me

816
00:27:06,480 --> 00:27:10,400
almost 700 hits for articles on

817
00:27:08,480 --> 00:27:12,880
biometric data ethics frameworks

818
00:27:10,400 --> 00:27:14,159
published just this year um i should

819
00:27:12,880 --> 00:27:16,720
remind you that i am delivering this

820
00:27:14,159 --> 00:27:18,559
talk two weeks into january

821
00:27:16,720 --> 00:27:20,080
in the last several years other ethical

822
00:27:18,559 --> 00:27:22,720
frameworks for artificial intelligence

823
00:27:20,080 --> 00:27:24,960
applications to massdev

824
00:27:22,720 --> 00:27:26,799
for artificial intelligence applications

825
00:27:24,960 --> 00:27:29,840
for mass data gathering have been

826
00:27:26,799 --> 00:27:31,279
proposed by microsoft and ibm

827
00:27:29,840 --> 00:27:32,720
and google

828
00:27:31,279 --> 00:27:34,480
and the australian government through

829
00:27:32,720 --> 00:27:36,399
the csiro

830
00:27:34,480 --> 00:27:38,000
and countless other research bodies and

831
00:27:36,399 --> 00:27:39,360
governments like the ethics institute

832
00:27:38,000 --> 00:27:40,880
which has published

833
00:27:39,360 --> 00:27:42,559
this responsible machine learning

834
00:27:40,880 --> 00:27:44,240
principle

835
00:27:42,559 --> 00:27:46,240
and there are technology consortiums

836
00:27:44,240 --> 00:27:47,919
like the partnership on ai and the

837
00:27:46,240 --> 00:27:49,440
information and technology industry

838
00:27:47,919 --> 00:27:50,960
council which have both published

839
00:27:49,440 --> 00:27:53,840
ethical framework principle statements

840
00:27:50,960 --> 00:27:55,840
that their members have to agree with

841
00:27:53,840 --> 00:27:58,640
a study that was done at the western

842
00:27:55,840 --> 00:28:00,080
sydney university in 2018 analyzed a

843
00:27:58,640 --> 00:28:01,760
whole lot of these ethics frameworks and

844
00:28:00,080 --> 00:28:04,399
distilled them into these five focus

845
00:28:01,760 --> 00:28:07,120
points privacy governance fairness

846
00:28:04,399 --> 00:28:08,720
shared benefit and transparency

847
00:28:07,120 --> 00:28:10,240
these are all great and important things

848
00:28:08,720 --> 00:28:12,080
to think about i'm glad that we have

849
00:28:10,240 --> 00:28:13,760
frameworks like this and more than one

850
00:28:12,080 --> 00:28:15,520
that i'm glad that peop there are people

851
00:28:13,760 --> 00:28:18,159
who want to talk about them

852
00:28:15,520 --> 00:28:20,240
but what most of them lack is any actual

853
00:28:18,159 --> 00:28:22,000
accountability or requirement or

854
00:28:20,240 --> 00:28:23,520
reporting or oversight

855
00:28:22,000 --> 00:28:25,039
they're mostly a list of nice things to

856
00:28:23,520 --> 00:28:26,880
have but they don't actually represent

857
00:28:25,039 --> 00:28:28,640
any obligation

858
00:28:26,880 --> 00:28:30,320
and in an industry where it is difficult

859
00:28:28,640 --> 00:28:32,080
for some organizations to realize that

860
00:28:30,320 --> 00:28:33,760
they could be doing the most good of all

861
00:28:32,080 --> 00:28:35,840
simply by paying tax in the countries

862
00:28:33,760 --> 00:28:37,440
where they operate the chances of these

863
00:28:35,840 --> 00:28:39,679
frameworks being adopted in practical

864
00:28:37,440 --> 00:28:40,880
terms may vary

865
00:28:39,679 --> 00:28:42,399
it also doesn't look great when

866
00:28:40,880 --> 00:28:43,600
companies fire the people they hire to

867
00:28:42,399 --> 00:28:44,399
help them with getting these things

868
00:28:43,600 --> 00:28:46,399
right

869
00:28:44,399 --> 00:28:47,840
like when google whose ceo i showed just

870
00:28:46,399 --> 00:28:50,159
before touting their ethical ai

871
00:28:47,840 --> 00:28:52,000
principles fired tim nitzgebru

872
00:28:50,159 --> 00:28:53,200
a an extremely prominent ai ethics

873
00:28:52,000 --> 00:28:54,799
researcher

874
00:28:53,200 --> 00:28:56,000
because she pointed out biases in their

875
00:28:54,799 --> 00:28:57,440
systems

876
00:28:56,000 --> 00:28:59,039
this slide isn't working but it is a

877
00:28:57,440 --> 00:29:00,720
screenshot of an article

878
00:28:59,039 --> 00:29:02,480
describing what i just said from the new

879
00:29:00,720 --> 00:29:03,840
york times please go and look that up

880
00:29:02,480 --> 00:29:06,640
it's a real thing that happened if you

881
00:29:03,840 --> 00:29:06,640
missed it at the time

882
00:29:07,039 --> 00:29:09,760
so

883
00:29:07,919 --> 00:29:11,440
legislating about against the worst of

884
00:29:09,760 --> 00:29:13,760
these abuses on the other hand this

885
00:29:11,440 --> 00:29:15,919
might have a stronger effect

886
00:29:13,760 --> 00:29:17,840
the gdpr is one of the more prominent

887
00:29:15,919 --> 00:29:19,760
examples of an attempt to address

888
00:29:17,840 --> 00:29:22,000
inappropriate data use

889
00:29:19,760 --> 00:29:24,480
and it includes biometric data

890
00:29:22,000 --> 00:29:26,240
it's improved things not just in europe

891
00:29:24,480 --> 00:29:27,760
where it applies even some of us in

892
00:29:26,240 --> 00:29:29,600
australia probably most of us at this

893
00:29:27,760 --> 00:29:31,760
point will have had to work on making

894
00:29:29,600 --> 00:29:33,520
sure that our systems comply with it

895
00:29:31,760 --> 00:29:35,520
if we do business in europe or we handle

896
00:29:33,520 --> 00:29:37,919
data about europeans and most of us do

897
00:29:35,520 --> 00:29:40,080
either or both of those things

898
00:29:37,919 --> 00:29:41,679
it's no coincidence that europe with its

899
00:29:40,080 --> 00:29:43,679
strong laws is where a lot of this

900
00:29:41,679 --> 00:29:46,000
phrenology and physiology stuff first

901
00:29:43,679 --> 00:29:48,240
started being developed and applied

902
00:29:46,000 --> 00:29:50,960
the attitude to data collection there is

903
00:29:48,240 --> 00:29:52,720
strongly informed by the fact that they

904
00:29:50,960 --> 00:29:54,880
lived through some of the worst of what

905
00:29:52,720 --> 00:29:56,880
can happen because of this

906
00:29:54,880 --> 00:29:59,360
but we still lack good legislation

907
00:29:56,880 --> 00:30:01,440
sometimes any relevant legislation in so

908
00:29:59,360 --> 00:30:03,279
many other countries

909
00:30:01,440 --> 00:30:04,799
sometimes older laws might be bent to

910
00:30:03,279 --> 00:30:06,480
apply to this situation but there isn't

911
00:30:04,799 --> 00:30:09,039
much that prevents biometric data from

912
00:30:06,480 --> 00:30:10,880
being abused this is partly because of

913
00:30:09,039 --> 00:30:12,399
the well-known problems with technology

914
00:30:10,880 --> 00:30:13,919
and law where tech capabilities have

915
00:30:12,399 --> 00:30:15,840
grown quicker than the law is able to

916
00:30:13,919 --> 00:30:18,480
keep up and partly because it's not

917
00:30:15,840 --> 00:30:20,640
always easy to understand how new tech

918
00:30:18,480 --> 00:30:22,240
might bring up old problems again

919
00:30:20,640 --> 00:30:24,080
there's also the fact that in australia

920
00:30:22,240 --> 00:30:26,080
we still don't have a bill of rights so

921
00:30:24,080 --> 00:30:27,440
we don't really have a legally enshrined

922
00:30:26,080 --> 00:30:29,440
right to privacy

923
00:30:27,440 --> 00:30:31,279
which is a concern given the absolute

924
00:30:29,440 --> 00:30:32,640
torrent of pro-surveillance legislative

925
00:30:31,279 --> 00:30:34,080
changes that have been made recently

926
00:30:32,640 --> 00:30:35,840
which affect the information that's

927
00:30:34,080 --> 00:30:38,480
gathered about us by both governments

928
00:30:35,840 --> 00:30:38,480
and companies

929
00:30:38,559 --> 00:30:42,240
lots of us are time poor and in the

930
00:30:40,399 --> 00:30:43,679
third year of a pandemic many of you may

931
00:30:42,240 --> 00:30:45,520
have bigger problems than campaigning

932
00:30:43,679 --> 00:30:46,720
about biometrics ethics which is fair

933
00:30:45,520 --> 00:30:48,480
enough

934
00:30:46,720 --> 00:30:49,600
there's another other crap going on in

935
00:30:48,480 --> 00:30:51,120
the world that we have to know how to

936
00:30:49,600 --> 00:30:52,640
pick our battles but if you want to

937
00:30:51,120 --> 00:30:54,080
contribute in some way to helping out

938
00:30:52,640 --> 00:30:55,600
you can always donate to the groups that

939
00:30:54,080 --> 00:30:57,679
are keeping their eye on these things

940
00:30:55,600 --> 00:30:59,360
holding organizations accountable and

941
00:30:57,679 --> 00:31:00,480
campaigning for proper legislative

942
00:30:59,360 --> 00:31:02,559
reform

943
00:31:00,480 --> 00:31:04,240
in australia specifically you can donate

944
00:31:02,559 --> 00:31:05,840
to both digital rights watch and

945
00:31:04,240 --> 00:31:08,080
electronic frontiers australia who are

946
00:31:05,840 --> 00:31:09,440
collectively doing that absolute best to

947
00:31:08,080 --> 00:31:10,720
look out for the interest of all of us

948
00:31:09,440 --> 00:31:13,279
in this space

949
00:31:10,720 --> 00:31:14,880
and because i am full disclosure on the

950
00:31:13,279 --> 00:31:16,880
board of digital rights watch i can

951
00:31:14,880 --> 00:31:18,960
attest wholeheartedly that we would not

952
00:31:16,880 --> 00:31:21,360
have been able to handle the amount of

953
00:31:18,960 --> 00:31:23,039
junk the last few years has thrown at us

954
00:31:21,360 --> 00:31:25,519
without the support of donors so we're

955
00:31:23,039 --> 00:31:27,519
actually able to hire staff now which is

956
00:31:25,519 --> 00:31:29,919
great because holding the government to

957
00:31:27,519 --> 00:31:32,480
account on these issues is literally

958
00:31:29,919 --> 00:31:34,320
several full-time jobs if you want to

959
00:31:32,480 --> 00:31:35,679
take action but you're not sure how or

960
00:31:34,320 --> 00:31:37,919
you don't have time but you do have a

961
00:31:35,679 --> 00:31:39,360
few spare dollars consider donating to

962
00:31:37,919 --> 00:31:42,720
either or both of these organizations

963
00:31:39,360 --> 00:31:44,720
because it actually helps seriously

964
00:31:42,720 --> 00:31:46,000
and in terms of what else you can do if

965
00:31:44,720 --> 00:31:47,760
you work in tech and you're asked to

966
00:31:46,000 --> 00:31:49,440
implement systems that you think are

967
00:31:47,760 --> 00:31:51,600
feeding into the same problem

968
00:31:49,440 --> 00:31:53,679
push back say no

969
00:31:51,600 --> 00:31:55,200
we are the people with the skills needed

970
00:31:53,679 --> 00:31:57,679
to make these things real and we get to

971
00:31:55,200 --> 00:31:59,360
choose how we use those skills

972
00:31:57,679 --> 00:32:01,760
that said i realize it's not that easy

973
00:31:59,360 --> 00:32:04,000
for a lot of us to just do this you know

974
00:32:01,760 --> 00:32:05,519
it requires enormous privilege and

975
00:32:04,000 --> 00:32:07,840
confidence to be able to stand up and

976
00:32:05,519 --> 00:32:10,399
say no in a workplace not everybody's in

977
00:32:07,840 --> 00:32:12,480
a position to do that that is okay

978
00:32:10,399 --> 00:32:14,399
but if you want support in how to speak

979
00:32:12,480 --> 00:32:16,640
up in your workplace this is where

980
00:32:14,399 --> 00:32:18,960
unions can help you out there are unions

981
00:32:16,640 --> 00:32:20,880
supporting tech workers in australia

982
00:32:18,960 --> 00:32:22,640
like professionals australia and we're

983
00:32:20,880 --> 00:32:24,880
seeing growing numbers of tech workers

984
00:32:22,640 --> 00:32:26,240
all around the world unionize because

985
00:32:24,880 --> 00:32:27,919
collective action is the only way that

986
00:32:26,240 --> 00:32:29,360
we can really fight against mega

987
00:32:27,919 --> 00:32:30,799
corporations owned by men who have

988
00:32:29,360 --> 00:32:33,679
enough money to fire themselves into

989
00:32:30,799 --> 00:32:35,919
this into space for the lulls

990
00:32:33,679 --> 00:32:37,679
they didn't stay there anyway

991
00:32:35,919 --> 00:32:39,600
to fight back against them as well as

992
00:32:37,679 --> 00:32:42,559
against all the systems and standards

993
00:32:39,600 --> 00:32:44,080
that they prop up as normal

994
00:32:42,559 --> 00:32:45,919
the other thing that we can do to stop

995
00:32:44,080 --> 00:32:47,360
biometric abuse becoming part of

996
00:32:45,919 --> 00:32:49,200
everyday life

997
00:32:47,360 --> 00:32:51,039
is to educate each other about where

998
00:32:49,200 --> 00:32:53,200
humanity's already been and the plethora

999
00:32:51,039 --> 00:32:55,360
of futures that are possible that don't

1000
00:32:53,200 --> 00:32:57,120
involve perpetuating these systems

1001
00:32:55,360 --> 00:32:58,559
this is why i'm here right now i talk

1002
00:32:57,120 --> 00:33:00,159
about history so that i can help give

1003
00:32:58,559 --> 00:33:02,159
historical context to other software

1004
00:33:00,159 --> 00:33:04,559
developers because knowing more about

1005
00:33:02,159 --> 00:33:05,919
what came before helps us to build

1006
00:33:04,559 --> 00:33:07,440
better futures

1007
00:33:05,919 --> 00:33:09,039
so this talk is one resource i'm adding

1008
00:33:07,440 --> 00:33:11,760
to the pile that you can use to help

1009
00:33:09,039 --> 00:33:13,120
other people learn about these issues

1010
00:33:11,760 --> 00:33:14,480
and here are some other resources by

1011
00:33:13,120 --> 00:33:16,320
people who have studied this topic in

1012
00:33:14,480 --> 00:33:18,559
much further depth

1013
00:33:16,320 --> 00:33:20,640
they cover off everything from the

1014
00:33:18,559 --> 00:33:22,399
history of scientific racism to the bias

1015
00:33:20,640 --> 00:33:23,919
in google search results

1016
00:33:22,399 --> 00:33:25,760
some of them most of them will make you

1017
00:33:23,919 --> 00:33:27,360
angry but also most of them will give

1018
00:33:25,760 --> 00:33:28,640
you hope

1019
00:33:27,360 --> 00:33:30,240
if you want to check out some of the old

1020
00:33:28,640 --> 00:33:31,600
phrenology and physionomy handbooks like

1021
00:33:30,240 --> 00:33:32,640
the ones i've been showing you pieces of

1022
00:33:31,600 --> 00:33:34,559
today

1023
00:33:32,640 --> 00:33:36,640
honestly maybe don't

1024
00:33:34,559 --> 00:33:38,720
a lot of the content in these books is

1025
00:33:36,640 --> 00:33:40,240
disgustingly racist and sexist and

1026
00:33:38,720 --> 00:33:42,320
classist and just about every other kind

1027
00:33:40,240 --> 00:33:44,399
of ist you can imagine there are better

1028
00:33:42,320 --> 00:33:46,320
things to read

1029
00:33:44,399 --> 00:33:47,519
if this talk is about anything it's

1030
00:33:46,320 --> 00:33:48,799
really about the fact that biometric

1031
00:33:47,519 --> 00:33:50,240
measurements and data don't always give

1032
00:33:48,799 --> 00:33:52,240
us good answers for the problems that

1033
00:33:50,240 --> 00:33:54,000
we're trying to solve no matter how much

1034
00:33:52,240 --> 00:33:56,080
that we've been told that data will save

1035
00:33:54,000 --> 00:33:57,840
everything solve everything sorry and

1036
00:33:56,080 --> 00:33:59,120
save everything i guess

1037
00:33:57,840 --> 00:34:00,960
we have been trying to get somewhere

1038
00:33:59,120 --> 00:34:02,720
with the whole categorizing humans thing

1039
00:34:00,960 --> 00:34:04,559
since literally before the time of

1040
00:34:02,720 --> 00:34:06,320
aristotle and we've been repeating the

1041
00:34:04,559 --> 00:34:08,000
same questions to ourselves over and

1042
00:34:06,320 --> 00:34:10,079
over again

1043
00:34:08,000 --> 00:34:11,839
there is an interesting irony in us

1044
00:34:10,079 --> 00:34:13,919
wanting to harness biometric data

1045
00:34:11,839 --> 00:34:15,679
because we want neat answers and systems

1046
00:34:13,919 --> 00:34:17,839
for our lives but that the answer that

1047
00:34:15,679 --> 00:34:19,599
we keep coming up with is that there is

1048
00:34:17,839 --> 00:34:21,599
no one good system

1049
00:34:19,599 --> 00:34:23,760
everything works on a case-by-case basis

1050
00:34:21,599 --> 00:34:25,599
human beings are all kinds of strange

1051
00:34:23,760 --> 00:34:28,960
and wonderful things and they can't be

1052
00:34:25,599 --> 00:34:30,480
categorized we can't be categorized

1053
00:34:28,960 --> 00:34:32,480
sherlock holmes for all that he was the

1054
00:34:30,480 --> 00:34:35,119
pinnacle of the victorian rationalist he

1055
00:34:32,480 --> 00:34:37,919
knew this too in the short story a case

1056
00:34:35,119 --> 00:34:40,000
of identity he observed that life is

1057
00:34:37,919 --> 00:34:42,240
infinitely stranger than anything which

1058
00:34:40,000 --> 00:34:44,240
the mind of man could invent

1059
00:34:42,240 --> 00:34:45,839
people are weird and messy and if we

1060
00:34:44,240 --> 00:34:47,359
write software that respects and works

1061
00:34:45,839 --> 00:34:49,760
with that

1062
00:34:47,359 --> 00:34:51,119
maybe we'll do a bit better

1063
00:34:49,760 --> 00:34:52,879
hopefully we can make decisions about

1064
00:34:51,119 --> 00:34:54,560
what's right by being informed about the

1065
00:34:52,879 --> 00:34:56,639
past and by learning from the worst of

1066
00:34:54,560 --> 00:34:57,760
our mistakes

1067
00:34:56,639 --> 00:34:59,760
if you want to keep talking about the

1068
00:34:57,760 --> 00:35:01,440
topics that i've covered today

1069
00:34:59,760 --> 00:35:03,520
feel free to get in touch

1070
00:35:01,440 --> 00:35:06,240
um here's me on twitter again

1071
00:35:03,520 --> 00:35:06,240
thank you very much

1072
00:35:07,680 --> 00:35:11,359
to knock away lily that was uh amazing

1073
00:35:10,560 --> 00:35:14,160
um

1074
00:35:11,359 --> 00:35:17,599
and i think you know

1075
00:35:14,160 --> 00:35:19,760
so much of what you've put there is as

1076
00:35:17,599 --> 00:35:21,599
conversations i've had for the last five

1077
00:35:19,760 --> 00:35:24,560
years with friends and family like they

1078
00:35:21,599 --> 00:35:27,119
don't understand why i don't have a

1079
00:35:24,560 --> 00:35:29,839
fitness tracker or a smart watch or why

1080
00:35:27,119 --> 00:35:31,520
i don't opt into these things and it's a

1081
00:35:29,839 --> 00:35:33,200
really difficult conversation to have

1082
00:35:31,520 --> 00:35:35,200
because you end up

1083
00:35:33,200 --> 00:35:36,960
going into this rabbit hole of this is

1084
00:35:35,200 --> 00:35:39,280
what's going on this is how it's been

1085
00:35:36,960 --> 00:35:41,119
collected and harvested and yes there's

1086
00:35:39,280 --> 00:35:42,800
some benefit to for having that cloud

1087
00:35:41,119 --> 00:35:44,800
service

1088
00:35:42,800 --> 00:35:47,359
because you get some insight but where

1089
00:35:44,800 --> 00:35:49,760
is where is the rest of it going yeah

1090
00:35:47,359 --> 00:35:52,079
it's funny

1091
00:35:49,760 --> 00:35:53,280
yeah my fitbit went off during this talk

1092
00:35:52,079 --> 00:35:54,640
because it's like your heart rate's

1093
00:35:53,280 --> 00:35:58,720
really high and i was like i know i'm

1094
00:35:54,640 --> 00:36:00,640
giving a talk it's nerve-wracking

1095
00:35:58,720 --> 00:36:02,480
um we do have a couple of questions

1096
00:36:00,640 --> 00:36:04,000
coming in um and

1097
00:36:02,480 --> 00:36:05,839
i think there's more coming in as i'm

1098
00:36:04,000 --> 00:36:07,040
speaking so that's good sign um the

1099
00:36:05,839 --> 00:36:08,960
feedback in the channel has been amazing

1100
00:36:07,040 --> 00:36:10,720
i know you can't see it but people no i

1101
00:36:08,960 --> 00:36:13,119
got that close engaged

1102
00:36:10,720 --> 00:36:15,359
and really engaged so i'll come back and

1103
00:36:13,119 --> 00:36:18,960
check it later and i'll be um i'll be in

1104
00:36:15,359 --> 00:36:20,160
the post uh talk conversation room

1105
00:36:18,960 --> 00:36:22,160
um

1106
00:36:20,160 --> 00:36:24,240
in venulis after this

1107
00:36:22,160 --> 00:36:26,240
um so question

1108
00:36:24,240 --> 00:36:28,640
what can we do in a digitally connected

1109
00:36:26,240 --> 00:36:32,240
world to reduce our risk exposure as

1110
00:36:28,640 --> 00:36:33,680
individuals and as societies

1111
00:36:32,240 --> 00:36:36,560
okay just starting off with the small

1112
00:36:33,680 --> 00:36:36,560
small questions yeah

1113
00:36:37,280 --> 00:36:39,839
i think you know individuals are

1114
00:36:38,640 --> 00:36:42,720
probably the thing that we have the most

1115
00:36:39,839 --> 00:36:44,320
control over societies are harder being

1116
00:36:42,720 --> 00:36:46,320
that they're larger scale

1117
00:36:44,320 --> 00:36:47,520
um part of that can be about thinking

1118
00:36:46,320 --> 00:36:49,520
about the tech that you want to bring

1119
00:36:47,520 --> 00:36:51,040
into your lives and whether that's worth

1120
00:36:49,520 --> 00:36:52,960
it clearly i made a decision that i

1121
00:36:51,040 --> 00:36:54,720
wanted to use a fitbit

1122
00:36:52,960 --> 00:36:55,800
i use that within my own specific

1123
00:36:54,720 --> 00:36:57,119
parameters and

1124
00:36:55,800 --> 00:36:58,640
[Music]

1125
00:36:57,119 --> 00:37:00,480
control as much as i can where that

1126
00:36:58,640 --> 00:37:02,000
information goes but

1127
00:37:00,480 --> 00:37:03,040
certainly asking yourself the questions

1128
00:37:02,000 --> 00:37:05,280
about the kinds of things that you want

1129
00:37:03,040 --> 00:37:07,040
to be using in your life is important

1130
00:37:05,280 --> 00:37:08,880
and if you have no choice

1131
00:37:07,040 --> 00:37:10,240
um besides using some of this stuff in

1132
00:37:08,880 --> 00:37:12,000
your life because i know that there are

1133
00:37:10,240 --> 00:37:13,440
people who for various medical reasons

1134
00:37:12,000 --> 00:37:15,760
are going to have to

1135
00:37:13,440 --> 00:37:16,960
um and probably for other reasons too

1136
00:37:15,760 --> 00:37:18,720
um

1137
00:37:16,960 --> 00:37:20,320
it's worth thinking about the ways that

1138
00:37:18,720 --> 00:37:22,480
you you might be able to limit that and

1139
00:37:20,320 --> 00:37:24,480
if you can't also getting in touch with

1140
00:37:22,480 --> 00:37:26,320
the companies and rattling a few cages

1141
00:37:24,480 --> 00:37:28,960
to ask more about why there aren't as

1142
00:37:26,320 --> 00:37:31,359
many options

1143
00:37:28,960 --> 00:37:32,880
yeah i mean one thing i found is that

1144
00:37:31,359 --> 00:37:35,119
it's almost as if this is the new

1145
00:37:32,880 --> 00:37:37,680
digital divide those that are privileged

1146
00:37:35,119 --> 00:37:39,760
positions enough to be able to push back

1147
00:37:37,680 --> 00:37:41,760
against these sort of technologies and

1148
00:37:39,760 --> 00:37:43,680
this data metric that's almost the new

1149
00:37:41,760 --> 00:37:46,320
digital divide it's not access anymore

1150
00:37:43,680 --> 00:37:47,680
it's about the ability to opt out

1151
00:37:46,320 --> 00:37:49,200
yeah and whether or not definitely

1152
00:37:47,680 --> 00:37:50,800
position to do that

1153
00:37:49,200 --> 00:37:52,800
yeah definitely because it's always

1154
00:37:50,800 --> 00:37:54,320
about you know paying

1155
00:37:52,800 --> 00:37:55,359
for that extra bit and if you don't have

1156
00:37:54,320 --> 00:37:57,280
the money

1157
00:37:55,359 --> 00:38:00,240
you can't

1158
00:37:57,280 --> 00:38:02,000
and and that's really you know

1159
00:38:00,240 --> 00:38:03,440
the the main thing that causes those

1160
00:38:02,000 --> 00:38:05,920
issues that the people who don't who

1161
00:38:03,440 --> 00:38:08,720
can't afford it are going to have

1162
00:38:05,920 --> 00:38:10,000
this sort of inescapable path towards

1163
00:38:08,720 --> 00:38:11,599
having all of their data gathered and

1164
00:38:10,000 --> 00:38:14,160
harvested and being surveilled in all of

1165
00:38:11,599 --> 00:38:14,160
these ways

1166
00:38:14,800 --> 00:38:19,839
um another question

1167
00:38:17,599 --> 00:38:22,079
is there a standard for new software

1168
00:38:19,839 --> 00:38:23,359
engineers who might not be aware that

1169
00:38:22,079 --> 00:38:25,520
they code

1170
00:38:23,359 --> 00:38:27,599
the code that they write may have

1171
00:38:25,520 --> 00:38:29,920
ethical issues

1172
00:38:27,599 --> 00:38:31,040
good questions that's a good question

1173
00:38:29,920 --> 00:38:32,320
i mean

1174
00:38:31,040 --> 00:38:33,920
i know that there are quite a few

1175
00:38:32,320 --> 00:38:38,160
frameworks i run through a lot of them

1176
00:38:33,920 --> 00:38:40,079
um and there are far far more than i

1177
00:38:38,160 --> 00:38:42,560
than i was able to put into this talk or

1178
00:38:40,079 --> 00:38:45,200
we would have been here forever um

1179
00:38:42,560 --> 00:38:46,960
i think just you know even

1180
00:38:45,200 --> 00:38:49,839
sending them a link to this talk

1181
00:38:46,960 --> 00:38:52,160
um after after it's published or

1182
00:38:49,839 --> 00:38:53,760
giving people one of those books that i

1183
00:38:52,160 --> 00:38:56,000
listed earlier might be a way of doing

1184
00:38:53,760 --> 00:38:58,720
it um i know that um particularly for

1185
00:38:56,000 --> 00:39:00,800
new graduates um there are there is a

1186
00:38:58,720 --> 00:39:02,240
lot more focus now on bringing ethics

1187
00:39:00,800 --> 00:39:05,119
into computer science and software

1188
00:39:02,240 --> 00:39:06,880
development courses which is great um

1189
00:39:05,119 --> 00:39:08,240
but it's it doesn't mean that it's going

1190
00:39:06,880 --> 00:39:10,800
to reach everybody in quite the same

1191
00:39:08,240 --> 00:39:13,200
ways and unfortunately a lot of the time

1192
00:39:10,800 --> 00:39:15,040
most people only get that realization

1193
00:39:13,200 --> 00:39:16,800
when something happens to them what

1194
00:39:15,040 --> 00:39:18,320
happens to someone they know that they

1195
00:39:16,800 --> 00:39:20,720
realize is unfair or they haven't been

1196
00:39:18,320 --> 00:39:22,320
able to opt out of it or it's creepy um

1197
00:39:20,720 --> 00:39:24,320
i know a few people who had that aha

1198
00:39:22,320 --> 00:39:27,040
moment when the taliban took back over

1199
00:39:24,320 --> 00:39:28,480
in afghanistan um recently and

1200
00:39:27,040 --> 00:39:30,400
they started using all the biometric

1201
00:39:28,480 --> 00:39:32,640
data that the us had gathered

1202
00:39:30,400 --> 00:39:34,800
about afghani citizens and particularly

1203
00:39:32,640 --> 00:39:37,200
afghani us collaborators to start

1204
00:39:34,800 --> 00:39:39,680
persecuting those people so

1205
00:39:37,200 --> 00:39:41,359
yeah that i mean that is one one example

1206
00:39:39,680 --> 00:39:43,440
of how it's not just this academic

1207
00:39:41,359 --> 00:39:44,960
slippery slope argument you see the same

1208
00:39:43,440 --> 00:39:46,720
thing with the way that facial

1209
00:39:44,960 --> 00:39:48,240
recognition is used on the uyghurs in

1210
00:39:46,720 --> 00:39:50,079
china as well

1211
00:39:48,240 --> 00:39:52,320
that these things do get used in these

1212
00:39:50,079 --> 00:39:53,839
ways and there are ways of stopping them

1213
00:39:52,320 --> 00:39:56,000
and thinking about them that start with

1214
00:39:53,839 --> 00:39:56,960
the very small scale stuff and by

1215
00:39:56,000 --> 00:39:59,839
thinking about what we're going to

1216
00:39:56,960 --> 00:39:59,839
accept in our lives

1217
00:40:00,000 --> 00:40:05,200
um at the risk of possibly opening a

1218
00:40:03,040 --> 00:40:07,839
hornet's nest that i think you've very

1219
00:40:05,200 --> 00:40:10,240
very well avoided but i mean this

1220
00:40:07,839 --> 00:40:13,040
pandemic obviously has has raised a lot

1221
00:40:10,240 --> 00:40:15,359
of the similar questions around the data

1222
00:40:13,040 --> 00:40:17,520
privacy and and you know there's the

1223
00:40:15,359 --> 00:40:20,480
same kind of

1224
00:40:17,520 --> 00:40:22,240
mold as we've been discussing um and i

1225
00:40:20,480 --> 00:40:23,839
guess yeah i think that that certainly

1226
00:40:22,240 --> 00:40:25,440
has created a bit more of a zeitgeist

1227
00:40:23,839 --> 00:40:27,680
than the general public

1228
00:40:25,440 --> 00:40:29,440
around these issues

1229
00:40:27,680 --> 00:40:30,960
what's really gotten me has been

1230
00:40:29,440 --> 00:40:33,200
particularly the way that it's affected

1231
00:40:30,960 --> 00:40:35,040
kids during the pandemic with academic

1232
00:40:33,200 --> 00:40:37,280
monitoring software that has the camera

1233
00:40:35,040 --> 00:40:38,400
on all the time and you know they're

1234
00:40:37,280 --> 00:40:39,839
trying to use it to make sure that

1235
00:40:38,400 --> 00:40:41,760
attendance is assured and that people

1236
00:40:39,839 --> 00:40:43,599
aren't cheating and copying answers from

1237
00:40:41,760 --> 00:40:46,240
each other in a remote way all this

1238
00:40:43,599 --> 00:40:49,119
proctoring software that um gets loaded

1239
00:40:46,240 --> 00:40:50,720
on into like children's laptops and

1240
00:40:49,119 --> 00:40:52,480
turned on and

1241
00:40:50,720 --> 00:40:54,240
you know you can't opt out of it

1242
00:40:52,480 --> 00:40:56,160
um and i know that that happens too in a

1243
00:40:54,240 --> 00:40:57,599
lot of workplaces where

1244
00:40:56,160 --> 00:40:59,280
particularly for those and again it's

1245
00:40:57,599 --> 00:41:01,200
usually people who aren't in

1246
00:40:59,280 --> 00:41:03,680
particularly high-paying jobs that

1247
00:41:01,200 --> 00:41:05,599
aren't trusted as much for no particular

1248
00:41:03,680 --> 00:41:07,280
reason but their work their employers

1249
00:41:05,599 --> 00:41:08,960
don't trust them and then they get

1250
00:41:07,280 --> 00:41:10,800
monitored through their cameras all the

1251
00:41:08,960 --> 00:41:12,079
time every day while they're working to

1252
00:41:10,800 --> 00:41:13,440
see whether they're working to see what

1253
00:41:12,079 --> 00:41:14,880
they're clicking on to see how fast

1254
00:41:13,440 --> 00:41:15,760
they're typing to make sure it's really

1255
00:41:14,880 --> 00:41:18,880
them

1256
00:41:15,760 --> 00:41:20,560
um it's creepy and it's insidious

1257
00:41:18,880 --> 00:41:22,319
and that's that's just one of the ways

1258
00:41:20,560 --> 00:41:24,720
that the pandemic has made things worse

1259
00:41:22,319 --> 00:41:26,160
for a lot of people in these ways

1260
00:41:24,720 --> 00:41:28,240
yeah i mean there's some comments in the

1261
00:41:26,160 --> 00:41:30,160
chat remote proctoring is weird for

1262
00:41:28,240 --> 00:41:32,720
professional certs i i mean we do it at

1263
00:41:30,160 --> 00:41:34,960
red hat and i when i learned that i was

1264
00:41:32,720 --> 00:41:36,720
just like well i guess i'm not getting a

1265
00:41:34,960 --> 00:41:38,079
red hat cert now because it's just

1266
00:41:36,720 --> 00:41:40,720
ridiculous i mean there's actually a

1267
00:41:38,079 --> 00:41:42,480
bunch of studies and academia looking at

1268
00:41:40,720 --> 00:41:45,119
how it doesn't really achieve what it

1269
00:41:42,480 --> 00:41:47,440
sets out to do right and i think that's

1270
00:41:45,119 --> 00:41:48,960
it's just silly silly costs but

1271
00:41:47,440 --> 00:41:50,560
someone's selling something and

1272
00:41:48,960 --> 00:41:52,240
someone's buying it right

1273
00:41:50,560 --> 00:41:53,920
yeah thank you very much for your talk

1274
00:41:52,240 --> 00:41:56,400
lily that was really good and really

1275
00:41:53,920 --> 00:41:58,480
engaging um as i said there's been some

1276
00:41:56,400 --> 00:41:59,359
really good conversation going on on the

1277
00:41:58,480 --> 00:42:00,400
channel

1278
00:41:59,359 --> 00:42:02,560
um

1279
00:42:00,400 --> 00:42:03,839
again thank you

1280
00:42:02,560 --> 00:42:05,760
no worries

1281
00:42:03,839 --> 00:42:08,400
um i'll be in the chat later

1282
00:42:05,760 --> 00:42:10,480
um just a couple of notes before uh

1283
00:42:08,400 --> 00:42:13,440
before i hit off um so professional

1284
00:42:10,480 --> 00:42:15,280
delegates uh networking session um will

1285
00:42:13,440 --> 00:42:16,480
be happening in about an hour so that's

1286
00:42:15,280 --> 00:42:17,359
at eight

1287
00:42:16,480 --> 00:42:20,160
sorry

1288
00:42:17,359 --> 00:42:23,040
6 30 australian eastern standard time

1289
00:42:20,160 --> 00:42:25,359
um so if you come back into venulis then

1290
00:42:23,040 --> 00:42:27,839
there will be a special room open if

1291
00:42:25,359 --> 00:42:29,200
you're a professional delegate ticket

1292
00:42:27,839 --> 00:42:30,640
holder

1293
00:42:29,200 --> 00:42:32,960
and for speakers and i think the

1294
00:42:30,640 --> 00:42:34,880
volunteers as well so please join us

1295
00:42:32,960 --> 00:42:37,920
back to there we have a an interesting

1296
00:42:34,880 --> 00:42:40,079
talk by anthony green who i had no i had

1297
00:42:37,920 --> 00:42:42,640
no idea who this person i'm kiwi so

1298
00:42:40,079 --> 00:42:44,640
apparently he's quite a a prolific

1299
00:42:42,640 --> 00:42:46,480
political commentator

1300
00:42:44,640 --> 00:42:48,640
um that most australians know so that

1301
00:42:46,480 --> 00:42:50,240
should be really interesting um to hear

1302
00:42:48,640 --> 00:42:52,560
his thoughts on

1303
00:42:50,240 --> 00:42:53,520
politics analysis stuff which sounds

1304
00:42:52,560 --> 00:42:55,359
cool

1305
00:42:53,520 --> 00:42:57,599
all right thank you very much for today

1306
00:42:55,359 --> 00:42:59,200
everyone it's um been great being your

1307
00:42:57,599 --> 00:43:02,920
host in this room

1308
00:42:59,200 --> 00:43:02,920
enjoy the rest of your evenings