WEBVTT
Kind: captions
Language: en-GB

1
00:00:00.149 --> 00:00:04.620
JEREMY KEITH: Ladies and gentlemen, please
give thunderous applause to Tom Scott!

2
00:00:04.620 --> 00:00:10.849
(APPLAUSE) TOM SCOTT: Thank you very much.
Thank you.

3
00:00:10.849 --> 00:00:18.209
So in the very first talk today, Warren Ellis
said that predicting the future was a mug's game.

4
00:00:18.209 --> 00:00:23.570
So there's been enough history today, let's
try and predict the future. (LAUGHTER)

5
00:00:23.589 --> 00:00:26.660
You don't really notice change happening.

6
00:00:26.660 --> 00:00:35.680
Google was incorporated 16 years ago yesterday.
1998. The easy access to information that

7
00:00:35.680 --> 00:00:41.570
it ushered in changed the world. And sure,
if it hadn't been Brin and Page, it would

8
00:00:41.570 --> 00:00:46.680
have been someone else with a similar algorithm,
but can you imagine taking someone from 1996,

9
00:00:46.680 --> 00:00:52.070
never mind Victorian times, which we had earlier,
just 1996 -- and showing them today? A small

10
00:00:52.070 --> 00:00:58.190
little magical device in your pocket that
can answer 99% of all your questions about

11
00:00:58.190 --> 00:01:05.190
the world, and 99% of all your requests for
media, but that reports your location to the

12
00:01:05.190 --> 00:01:11.880
police constantly so they can look you up
if you need to. And then, you've got any number

13
00:01:11.890 --> 00:01:17.990
of instant ways of communicating with people
around the world, nearly anyone in the world...

14
00:01:17.990 --> 00:01:23.490
all of which is monitored all the time by
shadowy government agencies.

15
00:01:23.490 --> 00:01:26.270
Would they think we're in a dystopia? Would
they think we're in some sort of utopia? Well,

16
00:01:26.270 --> 00:01:33.880
it'd be somewhere in the middle. So let's
plot a line. All the way over here, is 1998,

17
00:01:33.880 --> 00:01:40.490
the founding of Google. Right here in the
middle: is where we are now, 2014. Let's see

18
00:01:40.490 --> 00:01:45.450
what someone from 1998 would think about today,
because what I'm going to do, is go about

19
00:01:45.450 --> 00:01:51.840
sixteen years into the future, into 2030.
And let's say that we have roughly the same

20
00:01:51.840 --> 00:01:55.990
orders-of-magnitude increase in access to
information.

21
00:01:55.990 --> 00:02:00.840
Let's propose some slightly fanciful technology,
that is about as fanciful as Google would

22
00:02:00.840 --> 00:02:04.530
have been, all the way back over there.

23
00:02:04.530 --> 00:02:07.350
Let's do a talk from dConstruct 2030.

24
00:02:07.350 --> 00:02:11.999
And it would start a little bit something
like this:

25
00:02:11.999 --> 00:02:18.999
Teenagers are becoming telepathic, privacy
is dead, and we're in another tech bubble.

26
00:02:20.670 --> 00:02:24.159
Welcome to 2030. Or at least, one possible
2030.

27
00:02:24.159 --> 00:02:28.349
For the less-technical folks in the audience,
it might be a little bit difficult to work

28
00:02:28.349 --> 00:02:33.489
out what is in the heads of thirties teenagers
walking down the street, earbuds in,

29
00:02:33.489 --> 00:02:38.950
throat mics on, muttering to themselves constantly,
emotions flashing across their face faster

30
00:02:38.950 --> 00:02:45.469
than seems reasonable. And anyone who joins
an Engine network right now with no prior

31
00:02:45.469 --> 00:02:50.359
experience, will find it overwhelming. It
takes about four or five days of constant

32
00:02:50.359 --> 00:02:55.700
access before your brain starts to get used
to the enormous tidal wave of data that hits it

33
00:02:55.700 --> 00:03:00.309
every single second. Most people, particularly those
over the age of about 30, will give up within

34
00:03:00.309 --> 00:03:04.540
about four or five minutes. But there's a
reason that you hear testimonials from the

35
00:03:04.540 --> 00:03:09.379
people who've stuck around about slowly becoming
telepathic.

36
00:03:09.379 --> 00:03:15.980
What we have today is a confluence of two
things. First and foremost engines.

37
00:03:15.980 --> 00:03:22.480
Lin and Xiaoming's paper. Now legendary in
computer science. It's difficult to remember now

38
00:03:22.480 --> 00:03:28.299
just how difficult it was to get information
before Ganymede came along -- in the early

39
00:03:28.299 --> 00:03:32.189
days of Google, the same as in the early days
of Ganymede, it wasn't intuitive, it was a

40
00:03:32.189 --> 00:03:33.269
difficult process.

41
00:03:33.269 --> 00:03:39.219
You forget, now, that originally you had to
actually type. You had to spell things correctly.

42
00:03:39.219 --> 00:03:44.040
You had to load the right data, you had to
give it the right hints and ease it through

43
00:03:44.040 --> 00:03:47.489
the first stages of the process. So Ganymede,
and the other engines, could kind of work

44
00:03:47.489 --> 00:03:54.489
out what you were asking for. It wasn't intuitive,
just like typing that you just had to learn.

45
00:03:54.760 --> 00:03:59.519
Okay, you don't have to learn typing any more,
but once upon a time that was important.

46
00:03:59.519 --> 00:04:05.579
But it was still a light-speed leap forward.
Before Ganymede, if you wanted to go out with

47
00:04:05.579 --> 00:04:08.489
your friends to a restaurant, and you wanted
to find somewhere, you would have to open

48
00:04:08.489 --> 00:04:14.589
one app to find a restaurant, and then you
would open a second app to see if it was any good,

49
00:04:14.589 --> 00:04:18.620
and to see if it met your dietary requirements
and your budget, and then you would

50
00:04:18.620 --> 00:04:20.810
open a third app to see if there was a table.

51
00:04:20.810 --> 00:04:24.280
I mean, more than that, you might actually
have to use an old-school web page,

52
00:04:24.280 --> 00:04:26.650
if anyone remembers those.

53
00:04:26.650 --> 00:04:33.360
Well, you could try and wing it, of course,
plenty of people did -- but if you wing it,

54
00:04:33.360 --> 00:04:38.870
you were likely to be disappointed. And no-one's
disappointed any more. Ganymede, and later,

55
00:04:38.870 --> 00:04:46.400
the other engines' ability to take petabytes
of images, text, video, anything and to infer

56
00:04:46.400 --> 00:04:52.020
truths, to infer patterns, to infer faces
and dialogue and anything from them, in real time --

57
00:04:52.020 --> 00:04:59.020
that was a jump on the scale of the
mass printing press, on the scale of Google.

58
00:04:59.949 --> 00:05:04.979
No-one believed, in the mid-nineties, that
a search engine could realistically pull in

59
00:05:04.979 --> 00:05:08.759
results from the entire web and send them
out in real time. Not realistically.

60
00:05:08.760 --> 00:05:09.639
Couldn't be done.

61
00:05:09.639 --> 00:05:14.740
In the mid-tens, nobody thought that a search
engine could realistically answer queries

62
00:05:14.740 --> 00:05:19.419
like "where should we eat tonight"; not pull
in the meaning of "we" and "eat"

63
00:05:19.419 --> 00:05:23.599
and work out what you're actually asking.
And yet, here we are.

64
00:05:23.599 --> 00:05:29.009
Ganymede allowed that orders-of-magnitude
increase in access to information. Lin and Xiaoming

65
00:05:29.009 --> 00:05:34.180
had the perfect combination of philanthropy
and business acumen: they released

66
00:05:34.180 --> 00:05:38.310
the engine algorithm as open source -- 
that's what sparked the "engine explosion",

67
00:05:38.310 --> 00:05:45.729
that's why throat mics actually started working... -- but they kept their then-incredible exabyte

68
00:05:45.729 --> 00:05:50.650
of structured data, of inferences, of patterns, to themselves and their company.

69
00:05:50.650 --> 00:05:56.870
But that, on its own, doesn't explain the
muttering teenagers. The second part of what

70
00:05:56.870 --> 00:06:01.599
we have today is... is more a social change,
really.

71
00:06:01.599 --> 00:06:07.569
One of the key points about engines is that
the more open you are, the more information

72
00:06:07.569 --> 00:06:15.260
you give them, then the better they are. That's
why those of us who remember privacy

73
00:06:15.260 --> 00:06:21.260
don't really get as much out of them: I mean, yes,
we keep throat mics on at work, we take them

74
00:06:21.270 --> 00:06:27.699
off when we get home, but the idea of allowing
any third party access to everything we say,

75
00:06:27.700 --> 00:06:31.760
everyone we meet, everything we do, everywhere
we are, is abhorrent.

76
00:06:31.760 --> 00:06:34.500
The idea that we should do that is just wrong.

77
00:06:34.509 --> 00:06:37.599
I remember clearing my cookies every day.

78
00:06:37.599 --> 00:06:41.539
Ask your parents. (LAUGHTER)

79
00:06:41.539 --> 00:06:47.819
Now, most of us, we suck it up, we deal with
it, because the results are worth it, the

80
00:06:47.819 --> 00:06:54.379
payoff that we get from being open is worth
what we perceive as a cost. But we still take

81
00:06:54.379 --> 00:06:58.449
our throat mics off when we get home. Hell,
I still use a keyboard every now and again

82
00:06:58.449 --> 00:07:05.449
to write code; some of us still have to, even
as typing becomes a more and more niche skill.

83
00:07:05.499 --> 00:07:09.770
The only way to stay connected to everyone
in your social circle, right now, no matter

84
00:07:09.770 --> 00:07:15.520
which engine you use, no matter which particular
company runs yours, the only way to

85
00:07:15.530 --> 00:07:23.039
stay connected to everyone is to be entirely open. That's the only way that the engines know

86
00:07:23.039 --> 00:07:29.639
which of your thoughts to send to other people, and which should come back.

87
00:07:29.639 --> 00:07:32.889
Do you remember the Facebook News Feed, and later
the Twitter News Feed?

88
00:07:32.889 --> 00:07:35.609
You remember the fuss that caused? (LAUGHTER)

89
00:07:35.609 --> 00:07:38.300
You remember how it prioritised information that was more important to you?

90
00:07:38.300 --> 00:07:44.320
I'm offline right now, at a conference, but as soon as I walk off and put my earbud back in,

91
00:07:44.320 --> 00:07:49.840
my engine will start
catching me up at 800 words a minute, faster

92
00:07:49.840 --> 00:07:56.099
than anyone could possibly speak but still
slow enough for people to understand, for

93
00:07:56.099 --> 00:08:00.259
the brain to understand, and it will start
catching me up just on the important things,

94
00:08:00.259 --> 00:08:05.219
just on the things that matter to me.

95
00:08:05.219 --> 00:08:08.419
That's what those muttering teenagers are
doing: as they walk down the street, they

96
00:08:08.419 --> 00:08:14.039
are talking to a thousand people at a time,
mediated not by an artificial intelligence,

97
00:08:14.039 --> 00:08:20.740
just by algorithms. They talk to the world,
and the world listens back, they are taking

98
00:08:20.740 --> 00:08:27.020
part in, I don't know, a giant 20th century
chat room? 18th century coffee shop? --

99
00:08:27.020 --> 00:08:29.900
It's something much much bigger than the sum of
its parts, certainly.

100
00:08:29.900 --> 00:08:36.660
I was telling worst-case-scenario scare stories about flashmobs back in the tens,

101
00:08:36.660 --> 00:08:40.040
and -- we hadn't seen the start of it.

102
00:08:40.040 --> 00:08:44.000
That's the crucial part of engines. You give,
they give back.

103
00:08:44.000 --> 00:08:47.120
All of those engines talk to each other --

104
00:08:47.120 --> 00:08:54.120
okay, no, the companies that control
the engines have data-sharing agreements between

105
00:08:54.150 --> 00:08:58.620
each other, because it's not a good idea to
anthropomorphise these things. If it needs to,

106
00:08:58.620 --> 00:09:04.170
the little engine on my phone, that I
carry in my pocket, it can access the downlinks

107
00:09:04.170 --> 00:09:08.580
from the military drones hovering over Brighton,
circling at 30,000 feet.

108
00:09:08.580 --> 00:09:15.260
Okay, the outsourced, privatised military
drones hovering over Brighton at 30,000 feet.

109
00:09:15.260 --> 00:09:19.680
Maybe it needs to guide me past an obstruction,
maybe, more likely, it wants to magically

110
00:09:19.680 --> 00:09:24.120
route me towards a friend, or maybe it wants
to magically route me away from a cluster

111
00:09:24.120 --> 00:09:28.750
of people with ASBOs, or whatever ASBOs are
called now. Or maybe it just wants to answer

112
00:09:28.750 --> 00:09:33.570
my questions about finding a pub where I can
see the sunset, and no-one's asked that question yet.

113
00:09:33.570 --> 00:09:39.950
In exchange for me knowing that, right
now, then other engines, from the government

114
00:09:39.950 --> 00:09:44.640
and police, can always see what I am, who
I am, what I'm doing, and what I'm going to be up to.

115
00:09:44.640 --> 00:09:51.430
They know everything about me.
But I also know everything about them.

116
00:09:51.430 --> 00:09:57.830
Those of us who weren't able to make that sacrifice...
don't get to use engines.

117
00:09:57.840 --> 00:10:01.180
That's not a choice, the same way not having a smartphone wasn’t really an option,

118
00:10:01.180 --> 00:10:04.160
neither is not having an engine connection.

119
00:10:04.160 --> 00:10:09.710
I mentioned people with ASBOs, by the way:
we’ve seen a lot of press about those excluded

120
00:10:09.710 --> 00:10:15.060
by the algorithms. Cory Doctorow, who’s
up next, talked about Whuffie in his book

121
00:10:15.060 --> 00:10:21.230
Down and Out in the Magic Kingdom. A magical
currency based on reputation score.

122
00:10:21.230 --> 00:10:27.470
Now he wrote about neural implants and brain backups,
and essentially an automated system

123
00:10:27.470 --> 00:10:31.410
that worked out what other people thought of you and used that as an actual currency.

124
00:10:31.410 --> 00:10:34.850
Bruce Sterling, we heard this
referenced earlier, Maneki Neko.

125
00:10:34.850 --> 00:10:38.790
It was used as an example of plausible deniability earlier,

126
00:10:38.790 --> 00:10:44.180
but it was artificial intelligence
that was mediating that, a network that actively

127
00:10:44.180 --> 00:10:49.440
attacked people who were trying to bring it
down, and made their lives very difficult.

128
00:10:49.440 --> 00:10:56.440
What we have is so much more mundane than
either of those options, so much less explicit.

129
00:10:56.770 --> 00:11:02.830
Your engine notices the time that it takes
you to respond to things, notices the reactions

130
00:11:02.830 --> 00:11:09.030
you have to people, notices who you avoid,
and your biases start to get taken on by the engine.

131
00:11:09.030 --> 00:11:15.200
And it was designed by engineers in
Silicon Valley and Zhongguancun, so,

132
00:11:15.200 --> 00:11:18.180
as Fox News so delicately put it, and I apologise
for the accent,

133
00:11:18.180 --> 00:11:20.680
“there’s some folks it just don’t work so well on”.

134
00:11:20.680 --> 00:11:27.440
Do you remember the racist webcam, did anyone
see that, from 2009? It was an HP laptop with

135
00:11:27.440 --> 00:11:32.470
an automatic tracking camera, and it wouldn’t
follow you unless – and I’m quoting here

136
00:11:32.470 --> 00:11:38.440
– you had enough “difference in intensity
of contrast between the eyes and the upper cheek".

137
00:11:38.440 --> 00:11:44.300
Yeah. It’s that, but worse. Because
all the implicit associations, all the hidden biases,

138
00:11:44.300 --> 00:11:50.020
that study after study after study
after study have shown are in everyone's heads?

139
00:11:50.020 --> 00:11:52.230
They get picked up by engines.

140
00:11:52.230 --> 00:11:56.570
The fact that college professors are more
likely to respond to inquiries from men? Picked

141
00:11:56.570 --> 00:11:58.920
Picked up by the engines.

142
00:11:58.930 --> 00:12:02.230
It’s not clear enough to quantify, it’s
not clear enough to defend against.

143
00:12:02.230 --> 00:12:06.880
It's not one thing that you can pull out of
the enormous web of connections and patterns

144
00:12:06.880 --> 00:12:12.770
that the engines pick up, and you should have
seen the fury when one CEO said well,

145
00:12:12.770 --> 00:12:16.890
that's the way it should be, engines should reflect
the biases of the people who own them.

146
00:12:16.890 --> 00:12:22.970
And then what about the people on the other
end of the scale? The people who want to be

147
00:12:22.970 --> 00:12:29.440
popular, or don't want to be popular. We were
starting to see sponsored lives, back in 2014.

148
00:12:29.440 --> 00:12:35.260
Just the very start of it. The people who
were -- yeah, I hated the word back then --

149
00:12:35.260 --> 00:12:41.570
the people who were vlogging. The people who were
talking into their webcam once a day and uploading

150
00:12:41.570 --> 00:12:48.880
their lives and their thoughts and their experiences.
And kids, teenagers, kids younger than that,

151
00:12:48.880 --> 00:12:52.950
were flocking to these people. And old media,
television in particular couldn't understand

152
00:12:52.950 --> 00:12:58.670
it, why would you want to just watch someone
talk into a webcam. And they were trying to

153
00:12:58.670 --> 00:13:03.070
ask that even as they were updating their
antiquated reality-show formats to become

154
00:13:03.070 --> 00:13:08.570
ever more and more artificial. Saturday night
entertainment, 2014, BBC1:

155
00:13:08.570 --> 00:13:12.340
celebrities doing acrobatics.

156
00:13:12.340 --> 00:13:15.580
No wonder the kids flocked off to YouTube and
its successors.

157
00:13:15.580 --> 00:13:17.710
And no wonder the advertisers went with them.

158
00:13:19.760 --> 00:13:25.060
Streaming lifelogs, of course, had been tried
for years, back in 1996 was arguably the first one,

159
00:13:25.070 --> 00:13:31.000
the first real-time feed, that was pre-Google.
But the trouble was, with the exception of

160
00:13:31.000 --> 00:13:35.310
some obsessives, and folks who were happy
to watch people play video games 24 hours a day,

161
00:13:35.310 --> 00:13:41.050
you still need an editor. No-one has
time to watch lots of people's entire lives,

162
00:13:41.050 --> 00:13:46.130
so you still needed an editor to package up
the highlights, tell a story, work out the

163
00:13:46.130 --> 00:13:51.750
details, and put it together for consumption.
Well, engines can do that. Powered by measuring

164
00:13:51.750 --> 00:13:56.550
attention responses, constant, automated A-B
testing, and all its successors.

165
00:13:56.550 --> 00:13:59.680
They can do that. They did it.

166
00:13:59.690 --> 00:14:05.760
Almost everyone born after 2010 is now providing
a lifelog for their friends, and if they're

167
00:14:05.760 --> 00:14:09.940
good at it, if they're popular, they're providing
it for an audience beyond that as well.

168
00:14:09.940 --> 00:14:15.470
No wonder they're the ones getting sponsorship and product placement -- and for advertisers, there are

169
00:14:15.470 --> 00:14:21.330
no limits on how sneakily or subtly you can start
putting product placement into there.

170
00:14:21.330 --> 00:14:28.330
The regulation hasn't caught up with it yet at all. And it’s notable that there are people, folks in the news,

171
00:14:28.360 --> 00:14:31.380
folks who really don't want to be there, they
have that audience

172
00:14:31.380 --> 00:14:32.800
even though they really don't want it.

173
00:14:32.800 --> 00:14:39.040
There used to be that concept of the A-list
celebrity and the Z-list celebrity, of course,

174
00:14:39.040 --> 00:14:44.910
the idea that Hollywood and movies and television
and gossip magazines, they were the arbiters

175
00:14:44.910 --> 00:14:49.970
of who was a celebrity and who the public
cared about. That was the idea, right?

176
00:14:49.970 --> 00:14:55.590
You appear in a movie, you appear on television,
you are in. You are part of that celebrity.

177
00:14:55.590 --> 00:15:02.050
That's not how it works now. Celebrity 
works like the music industry used to.

178
00:15:02.050 --> 00:15:07.220
Sure, everyone knows that a folk-rock-sea-shanty
band is not going to make the top of the

179
00:15:07.240 --> 00:15:13.160
Radio 1 playlist -- they’re friends of mine, they
played Glastonbury, they're really good --

180
00:15:13.160 --> 00:15:17.300
but that's fine, because they know that a folk-rock-sea-shanty band's not going to be there,

181
00:15:17.300 --> 00:15:18.860
they're not aiming for that.

182
00:15:18.870 --> 00:15:24.030
They're aiming for the portion of their audience
that likes folk-rock-sea-shanty bands.

183
00:15:24.030 --> 00:15:26.970
They're aspiring to be the greatest folk-rock-sea-shanty
band they can be.

184
00:15:26.970 --> 00:15:30.430
That's what they're aiming for. That's what their audience wants.

185
00:15:30.430 --> 00:15:34.100
And as the music industry always was, they can talk directly to those people,

186
00:15:34.100 --> 00:15:37.360
they can get the fans, they can target a niche.

187
00:15:37.370 --> 00:15:42.760
That's how celebrity works now. You can target a
niche. It doesn't matter, you may never be

188
00:15:42.760 --> 00:15:47.870
a Hollywood A-list celebrity, but it doesn't
matter. You may be more interested in...

189
00:15:47.870 --> 00:15:53.110
it's always video games -- some form of video game,
some specific form. You can aspire to be the

190
00:15:53.110 --> 00:15:58.770
celebrity player of that particular genre
and you will find an audience, and they will

191
00:15:58.770 --> 00:16:02.210
know where you are, and the advertisers will
follow.

192
00:16:03.460 --> 00:16:05.760
Of course, we've seen the dangers of this,
haven't we?

193
00:16:07.260 --> 00:16:09.760
Cardinal Richelieu's quote, supposed quote,

194
00:16:09.760 --> 00:16:13.120
"Give me six lines written by the most
honest man in the world,

195
00:16:13.120 --> 00:16:16.160
"and I will find something in them to hang him."

196
00:16:16.160 --> 00:16:19.680
Which has never been
more true. We've seen modern engine technology

197
00:16:19.680 --> 00:16:24.840
used by oppressive governments just as often
as we've seen it used to take them down: yes,

198
00:16:24.840 --> 00:16:29.540
knowing the identity of every protester in
a mob is devastating knowledge, but it will

199
00:16:29.540 --> 00:16:32.720
not stop that mob storming the gates of your
palace

200
00:16:32.720 --> 00:16:36.520
If they're already there, it's a little too late.

201
00:16:36.520 --> 00:16:42.120
We've also seen what a mob of angry,
bitter keyboard vigilantes will do to someone

202
00:16:42.120 --> 00:16:47.640
that's offended them when they really do have
access to every aspect of your life.

203
00:16:47.640 --> 00:16:54.250
Hiding from abusers, from stalkers, is significantly
more difficult than it ever was.

204
00:16:54.280 --> 00:16:59.060
There is no off button, no block button for face recognition technology,

205
00:16:59.060 --> 00:17:02.440
despite the EU's ham-fisted attempt
to force one into law.

206
00:17:04.620 --> 00:17:10.340
Politicians. Politicans! Politicans have found
that not only are their current lives open

207
00:17:10.350 --> 00:17:16.380
to detailed analysis, but their previous ones
are too: Facebook survived long enough to

208
00:17:16.380 --> 00:17:21.780
get its data hooked into engines, just long
enough. And of course, your privacy settings

209
00:17:21.780 --> 00:17:27.050
don't matter that much when every tourist
photo you've ever appeared in is now available

210
00:17:27.050 --> 00:17:31.920
for download as well. It odesn't matter whetehr
you tagged that as "remove me from this photo"

211
00:17:31.920 --> 00:17:34.900
because that doesn't really apply any more.

212
00:17:34.900 --> 00:17:39.220
When all of your users have blanket access
and just give it over?

213
00:17:39.220 --> 00:17:45.360
Well, your life, from the advent of smartphones onwards, is perfectly searchable.

214
00:17:49.560 --> 00:17:50.910
(LAUGHTER)

215
00:17:50.920 --> 00:17:54.560
I can’t even claim I was young, I was, like,
mid-twenties at that point.

216
00:17:55.040 --> 00:18:00.530
And even if, somehow, you manage to come through
all that squeaky clean, without a stain on

217
00:18:00.530 --> 00:18:05.750
your character, without a bad picture anywhere,
can you claim that for everyone

218
00:18:05.750 --> 00:18:09.730
you were ever associated with? Everyone you've ever appeared next to?

219
00:18:09.730 --> 00:18:13.130
Can you say for every word you ever

220
00:18:13.130 --> 00:18:20.120
anonymously wrote on every web forum anywhere,
that all of those were fine?

221
00:18:20.120 --> 00:18:23.200
When IP address logs, which are in the engines,

222
00:18:23.200 --> 00:18:26.290
and prose analysis, which is really easy to do,

223
00:18:26.290 --> 00:18:30.010
can trace it back to you with 95% accuracy?

224
00:18:30.010 --> 00:18:37.030
We're at panopticon now. We're there. At least,
for everyone who lives in a city, or uses

225
00:18:37.030 --> 00:18:42.860
anything other than the postal service, and
that's 85% of the UK population and rising.

226
00:18:42.860 --> 00:18:49.530
But... have you seen the crime statistics?

227
00:18:49.530 --> 00:18:53.930
Pretty much every type of crime is down in
the last ten years.

228
00:18:53.930 --> 00:18:56.900
That's not a statistical blip, it's not under-reporting,

229
00:18:56.900 --> 00:19:03.040
the percentage of crimes going unreported has gone down too.

230
00:19:03.040 --> 00:19:06.200
And yes, there's a methodology for that, you
can look it up.

231
00:19:06.200 --> 00:19:08.960
I know correlation is not causation,

232
00:19:08.960 --> 00:19:14.380
I know that the two don't go together,
but if you go out into the street and just

233
00:19:14.380 --> 00:19:20.150
ask people, people who grew up post-Google,
if they would trade everything that the last 16,

234
00:19:20.150 --> 00:19:26.000
the last 32 years has given them for some
nebulous concept of privacy? You are going

235
00:19:26.000 --> 00:19:30.630
to wait a long time before you find someone
who thinks that's a good idea.

236
00:19:30.630 --> 00:19:36.700
And if you talk to the kids who grew up post-Engines,
then privacy's this strange thing that their

237
00:19:36.710 --> 00:19:41.610
grandparents believe in. Why wouldn't you
share exactly what you're doing?

238
00:19:41.610 --> 00:19:44.430
That's just the right thing to do.

239
00:19:44.430 --> 00:19:51.680
Imagine, for a minute, that you're sitting
back here in 2014, the day before Lin and Xiaoming

240
00:19:51.680 --> 00:19:56.380
released that paper. Nothing seems
to have changed for a little while.

241
00:19:56.380 --> 00:20:01.880
The last big revolution, back in 2014? The last big
revolution was Twitter,

242
00:20:01.880 --> 00:20:06.720
and that was eight years old. It's ancient history.

243
00:20:06.720 --> 00:20:10.320
There was just enough complacency, just enough thought of stability

244
00:20:10.320 --> 00:20:15.030
as they came out of a recession, that they'd gone and got themselves into another bubble.

245
00:20:15.030 --> 00:20:20.800
Apps this time, smartphone apps. Previously,
back then, it was the dot-com bubble, but

246
00:20:20.800 --> 00:20:30.140
now, apps. There were companies investing
billions into little toys on phones, they'd

247
00:20:30.140 --> 00:20:33.620
They'd never get their money back. They knew, somewhere they must have known

248
00:20:33.620 --> 00:20:37.440
they were never going to get their money back, but it meant their competitors

249
00:20:37.440 --> 00:20:40.160
weren't getting that little toy on the phone.

250
00:20:40.160 --> 00:20:44.980
And maybe, maybe they'd make enough money
back with advertising or whatever idea they had

251
00:20:44.980 --> 00:20:50.620
that they could start to make something
out of it. That's exactly what's going on now.

252
00:20:50.620 --> 00:20:54.290
Engines, the engine I use, the engine
those muttering teenagers use -- has just

253
00:20:54.290 --> 00:21:01.170
been bought for twenty billion dollars. It'll
last, what, a couple of years? The investors

254
00:21:01.170 --> 00:21:04.610
are banking they can make that money back
in that time, they won't make their money back,

255
00:21:04.610 --> 00:21:07.460
just like last time, and just like the
time before that.

256
00:21:07.460 --> 00:21:13.559
But my point is not to complain about people being people and making some poor investment choices.

257
00:21:13.559 --> 00:21:20.370
And my point is not to mourn my old-fashioned
ideas of privacy: because even back in 2014,

258
00:21:20.370 --> 00:21:25.450
that ship was sailing and it would have taken
an incredible effort to stop it.

259
00:21:25.450 --> 00:21:30.141
Privacy was already dead, it's just that not everyone agreed with it yet.

260
00:21:30.141 --> 00:21:32.960
And my point is not to be someone

261
00:21:32.960 --> 00:21:36.320
from the previous generation complaining
about muttering kids and their damned distractions.

262
00:21:36.320 --> 00:21:40.590
Of course they're distracted, they're part
of what may be the greatest thing ever to

263
00:21:40.590 --> 00:21:42.340
happen to this planet.

264
00:21:42.340 --> 00:21:47.940
My point is this: the dot-com crash, all the
way back then, didn't kill dot-coms.

265
00:21:47.940 --> 00:21:51.080
The app crash did not kill apps. Not the concept.

266
00:21:51.080 --> 00:21:54.140
It killed a lot of the individual ones, but we're still using them.

267
00:21:54.140 --> 00:21:57.360
And the engine crash
won't kill engines. But -- you're sitting here,

268
00:21:57.360 --> 00:22:02.510
and you think the world is always going
to be like this. Everyone does. I do.

269
00:22:02.510 --> 00:22:06.870
We've always had engines, the same time we've always
had smartphones, back in 2014.

270
00:22:06.870 --> 00:22:13.260
If I think back to my time in university, I had a smartphone then. I just remember -- no, I didn't.

271
00:22:13.270 --> 00:22:16.070
I had an old brick of a Nokia, because that's when I went to university.

272
00:22:16.070 --> 00:22:19.940
But my brain's casually edited in that, of course, we always had smartphones.

273
00:22:19.950 --> 00:22:25.730
But here's the thing: it might not have been
engines. It might have been anything.

274
00:22:25.730 --> 00:22:31.360
It might have been nanotech, or biotech, or some weird
thing I can't think of even now.

275
00:22:31.360 --> 00:22:35.460
Predictions of the future never ever turn out to be true.

276
00:22:35.460 --> 00:22:40.420
But that next thing, whatever it is, it is
right around the corner.

277
00:22:40.420 --> 00:22:43.240
It always is, no matter what year it is.

278
00:22:43.240 --> 00:22:47.700
Whatever that next thing is will change the world, and it will arrive so fast,

279
00:22:47.700 --> 00:22:52.100
faster than anyone can think.
But it won't be shocking. It won't be startling.

280
00:22:52.100 --> 00:22:57.309
It won't plunge us into a horrible dystopia
or elevate us to a glorious, technicolour

281
00:22:57.309 --> 00:23:04.309
future world, the same way that our world,
right now, isn't really either of those things.

282
00:23:04.660 --> 00:23:09.490
Whatever comes next will change our lives,
and we have the choice to shape how it'll

283
00:23:09.490 --> 00:23:15.640
change our lives, but it will change them.
And those lives will, by and large, get a

284
00:23:15.640 --> 00:23:21.050
little bit better, just like they usually
do. So whatever that next thing is, and whatever

285
00:23:21.050 --> 00:23:25.580
year it actually turns out to be, I hope I'm
with you to see it.

286
00:23:25.580 --> 00:23:28.280
My name's Tom Scott. Enjoy the rest of the
show.

287
00:23:29.760 --> 00:23:31.760
[Translating these subtitles? Add your name here!]