1 00:00:00,000 --> 00:00:18,721 *34C3 preroll music* 2 00:00:18,721 --> 00:00:24,996 Herald: Humans of Congress, it is my pleasure to announce the next speaker. 3 00:00:24,996 --> 00:00:31,749 I was supposed to pick out a few awards or something, to actually present what he's 4 00:00:31,750 --> 00:00:36,419 done in his life, but I can only say: he's one of us! 5 00:00:36,419 --> 00:00:40,149 *applause* 6 00:00:40,149 --> 00:00:43,279 Charles Stross! *ongoing applause* 7 00:00:43,279 --> 00:00:45,969 Charles Stross: Hi! Is this on? Good. Great. 8 00:00:45,969 --> 00:00:50,679 I'm really pleased to be here and I want to start by apologizing for my total 9 00:00:50,679 --> 00:00:57,629 lack of German. So this talk is gonna be in English. Good morning. I'm Charlie Stross 10 00:00:57,629 --> 00:01:03,559 and it's my job to tell lies for money, or rather, I write science fiction, much of 11 00:01:03,559 --> 00:01:07,860 it about on the future, which in recent years has become ridiculously hard to 12 00:01:07,860 --> 00:01:16,090 predict. In this talk I'm going to talk about why. Now our species, Homo sapiens 13 00:01:16,090 --> 00:01:21,990 sapiens, is about 300,000 years old. It used to be about 200,000 years old, 14 00:01:21,990 --> 00:01:26,189 but it grew an extra 100,000 years in the past year because of new 15 00:01:26,189 --> 00:01:31,329 archaeological discoveries, I mean, go figure. For all but the last three 16 00:01:31,329 --> 00:01:36,759 centuries or so - of that span, however - predicting the future was really easy. If 17 00:01:36,759 --> 00:01:41,369 you were an average person - as opposed to maybe a king or a pope - natural disasters 18 00:01:41,369 --> 00:01:46,690 aside, everyday life 50 years in the future would resemble everyday life 50 19 00:01:46,690 --> 00:01:56,089 years in your past. Let that sink in for a bit. For 99.9% of human existence on this 20 00:01:56,089 --> 00:02:01,929 earth, the future was static. Then something changed and the future began to 21 00:02:01,929 --> 00:02:06,600 shift increasingly rapidly, until, in the present day, things are moving so fast, 22 00:02:06,600 --> 00:02:13,440 it's barely possible to anticipate trends from one month to the next. Now as an 23 00:02:13,440 --> 00:02:17,830 eminent computer scientist, Edsger Dijkstra once remarked, computer science is no more 24 00:02:17,830 --> 00:02:23,620 about computers than astronomy is about building big telescopes, the same can be 25 00:02:23,620 --> 00:02:28,760 said of my field of work, writing science fiction, sci-fi is rarely about science 26 00:02:28,760 --> 00:02:33,690 and even more rarely about predicting the future, but sometimes we dabble in 27 00:02:33,690 --> 00:02:42,140 Futurism and lately, Futurism has gotten really, really, weird. Now when I write a 28 00:02:42,140 --> 00:02:47,230 near future work of fiction, one set, say, a decade hence, there used to be a recipe I 29 00:02:47,230 --> 00:02:53,770 could follow, that worked eerily well. Simply put: 90% of the next decade stuff is 30 00:02:53,770 --> 00:02:57,420 already here around us today. Buildings are designed to 31 00:02:57,420 --> 00:03:03,040 last many years, automobiles have a design life of about a decade, so half the cars on 32 00:03:03,040 --> 00:03:10,240 the road in 2027 are already there now - they're new. People? There'll be some new 33 00:03:10,240 --> 00:03:15,520 faces, aged 10 and under, and some older people will have died, but most of us 34 00:03:15,520 --> 00:03:22,840 adults will still be around, albeit older and grayer, this is the 90% of a near 35 00:03:22,840 --> 00:03:30,820 future that's already here today. After the already existing 90%, another 9% of a 36 00:03:30,820 --> 00:03:35,650 near future a decade hence used to be easily predictable: you look at trends 37 00:03:35,650 --> 00:03:39,970 dictated by physical limits, such as Moore's law and you look at Intel's road 38 00:03:39,970 --> 00:03:44,000 map and you use a bit of creative extrapolation and you won't go too far 39 00:03:44,000 --> 00:03:52,510 wrong. If I predict - wearing my futurology hat - that in 2027 LTE cellular phones will 40 00:03:52,510 --> 00:03:57,560 be ubiquitous, 5G will be available for high bandwidth applications and there will be 41 00:03:57,560 --> 00:04:01,790 fallback to some kind of satellite data service at a price, you probably won't 42 00:04:01,790 --> 00:04:04,330 laugh at me. I mean, it's not like I'm predicting that 43 00:04:04,330 --> 00:04:08,110 airlines will fly slower and Nazis will take over the United States, is it ? 44 00:04:08,110 --> 00:04:09,900 *laughing* 45 00:04:09,900 --> 00:04:15,210 And therein lies the problem. There is remaining 1% of what Donald Rumsfeld 46 00:04:15,210 --> 00:04:20,940 called the "unknown unknowns", what throws off all predictions. As it happens, airliners 47 00:04:20,940 --> 00:04:26,060 today are slower than they were in the 1970s and don't get me started about the Nazis, 48 00:04:26,060 --> 00:04:31,860 I mean, nobody in 2007 was expecting a Nazi revival in 2017, were they? 49 00:04:31,860 --> 00:04:37,320 Only this time, Germans get to be the good guys. *laughing, applause* 50 00:04:37,320 --> 00:04:42,260 So. My recipe for fiction set 10 years in the future used to be: 51 00:04:42,260 --> 00:04:47,360 "90% is already here, 9% is not here yet but predictable 52 00:04:47,360 --> 00:04:53,730 and 1% is 'who ordered that?'" But unfortunately the ratios have changed, I think we're now 53 00:04:53,730 --> 00:04:59,660 down to maybe 80% already here - climate change takes a huge toll on architecture - 54 00:04:59,660 --> 00:05:05,760 then 15% not here yet, but predictable and a whopping 5% of utterly unpredictable 55 00:05:05,760 --> 00:05:12,740 deep craziness. Now... before I carry on with this talk, I want to spend a minute or 56 00:05:12,740 --> 00:05:18,530 two ranting loudly and ruling out the singularity. Some of you might assume, that 57 00:05:18,530 --> 00:05:23,590 as the author of books like "Singularity Sky" and "Accelerando", 58 00:05:23,590 --> 00:05:28,220 I expect an impending technological singularity, 59 00:05:28,220 --> 00:05:32,090 that we will develop self-improving artificial intelligence and mind uploading 60 00:05:32,090 --> 00:05:35,410 and the whole wish list of transhumanist aspirations promoted by the likes of 61 00:05:35,410 --> 00:05:42,300 Ray Kurzweil, will come to pass. Unfortunately this isn't the case. I think transhumanism 62 00:05:42,300 --> 00:05:49,050 is a warmed-over Christian heresy. While its adherents tend to be outspoken atheists, 63 00:05:49,050 --> 00:05:51,910 they can't quite escape from the history that gave rise to our current 64 00:05:51,910 --> 00:05:57,220 Western civilization. Many of you are familiar with design patterns, an approach 65 00:05:57,220 --> 00:06:01,560 to software engineering that focuses on abstraction and simplification, in order 66 00:06:01,560 --> 00:06:07,730 to promote reusable code. When you look at the AI singularity as a narrative and 67 00:06:07,730 --> 00:06:11,790 identify the numerous places in their story where the phrase "and then a miracle 68 00:06:11,790 --> 00:06:18,790 happens" occur, it becomes apparent pretty quickly, that they've reinvented Christiantiy. 69 00:06:18,790 --> 00:06:19,460 *applause* 70 00:06:19,460 --> 00:06:25,330 Indeed, the wellspring of today's transhumanists draw in a long rich 71 00:06:25,330 --> 00:06:29,990 history of Russian philosophy, exemplified by the russian orthodox theologian Nikolai 72 00:06:29,990 --> 00:06:37,169 Fyodorovich Fedorov by way of his disciple Konstantin Tsiolkovsky, whose derivation 73 00:06:37,169 --> 00:06:40,520 of a rocket equation makes him essentially the father of modern space 74 00:06:40,520 --> 00:06:45,350 flight. Once you start probing the nether regions of transhumanist forth and run 75 00:06:45,350 --> 00:06:49,800 into concepts like Roko's Basilisk - by the way, any of you who didn't know about the 76 00:06:49,800 --> 00:06:54,120 Basilisk before, are now doomed to an eternity in AI hell, terribly sorry - you 77 00:06:54,120 --> 00:06:57,889 realize, they've mangled it to match some of the nastier aspects of Presbyterian 78 00:06:57,889 --> 00:07:03,189 Protestantism. Now they basically invented original sin and Satan in the guise of an 79 00:07:03,189 --> 00:07:09,270 AI that doesn't exist yet ,it's.. kind of peculiar. Anyway, my take on the 80 00:07:09,270 --> 00:07:12,949 singularity is: What if something walks like a duck and quacks like a duck? It's 81 00:07:12,949 --> 00:07:18,460 probably a duck. And if it looks like a religion, it's probably a religion. 82 00:07:18,460 --> 00:07:23,280 I don't see much evidence for human-like, self-directed artificial intelligences 83 00:07:23,280 --> 00:07:28,070 coming along any time soon, and a fair bit of evidence, that nobody accepts and freaks 84 00:07:28,070 --> 00:07:32,150 in cognitive science departments, even want it. I mean, if we invented an AI 85 00:07:32,150 --> 00:07:35,940 that was like a human mind, it would do the AI equivalent of sitting on the sofa, 86 00:07:35,940 --> 00:07:39,455 munching popcorn and watching the Super Bowl all day. 87 00:07:39,455 --> 00:07:43,261 It wouldn't be much use to us. *laughter, applause* 88 00:07:43,261 --> 00:07:46,776 What we're getting instead, is self-optimizing tools that defy 89 00:07:46,776 --> 00:07:51,220 human comprehension, but are not in fact any more like our kind 90 00:07:51,220 --> 00:07:57,500 of intelligence than a Boeing 737 is like a seagull. Boeing 737s and seagulls both 91 00:07:57,500 --> 00:08:04,240 fly, Boeing 737s don't lay eggs and shit everywhere. So I'm going to wash my hands 92 00:08:04,240 --> 00:08:09,590 of a singularity as a useful explanatory model of the future without further ado. 93 00:08:09,590 --> 00:08:14,670 I'm one of those vehement atheists as well and I'm gonna try and offer you a better 94 00:08:14,670 --> 00:08:20,960 model for what's happening to us. Now, as my fellow Scottish science fictional author 95 00:08:20,960 --> 00:08:27,130 Ken MacLeod likes to say "the secret weapon of science fiction is history". 96 00:08:27,130 --> 00:08:31,229 History is, loosely speaking, is the written record of what and how people did things 97 00:08:31,229 --> 00:08:36,528 in past times. Times that have slipped out of our personal memories. We science 98 00:08:36,528 --> 00:08:41,099 fiction writers tend to treat history as a giant toy chest to raid, whenever we feel 99 00:08:41,099 --> 00:08:45,389 like telling a story. With a little bit of history, it's really easy to whip up an 100 00:08:45,389 --> 00:08:49,019 entertaining yarn about a galactic empire that mirrors the development and decline 101 00:08:49,019 --> 00:08:53,529 of a Habsburg Empire or to respin the October Revolution as a tale of how Mars 102 00:08:53,529 --> 00:08:59,599 got its independence. But history is useful for so much more than that. 103 00:08:59,599 --> 00:09:05,380 It turns out, that our personal memories don't span very much time at all. I'm 53 104 00:09:05,380 --> 00:09:10,800 and I barely remember the 1960s. I only remember the 1970s with the eyes of a 6 to 105 00:09:10,800 --> 00:09:17,580 16 year old. My father died this year, aged 93, and he'd just about remembered the 106 00:09:17,580 --> 00:09:22,770 1930s. Only those of my father's generation directly remember the Great 107 00:09:22,780 --> 00:09:29,079 Depression and can compare it to the 2007/08 global financial crisis directly. 108 00:09:29,079 --> 00:09:34,250 We Westerners tend to pay little attention to cautionary tales told by 90-somethings. 109 00:09:34,250 --> 00:09:38,999 We're modern, we're change obsessed and we tend to repeat our biggest social mistakes 110 00:09:38,999 --> 00:09:43,259 just as they slip out of living memory, which means they recur on a timescale of 111 00:09:43,259 --> 00:09:47,680 70 to 100 years. So if our personal memories are useless, 112 00:09:47,680 --> 00:09:52,330 we need a better toolkit and history provides that toolkit. 113 00:09:52,330 --> 00:09:57,099 History gives us the perspective to see what went wrong in the past and to look for 114 00:09:57,099 --> 00:10:01,509 patterns and check to see whether those patterns are recurring in the present. 115 00:10:01,509 --> 00:10:06,689 Looking in particular at the history of the past two to four hundred years, that age of rapidly 116 00:10:06,689 --> 00:10:11,800 increasing change that I mentioned at the beginning. One glaringly obvious deviation 117 00:10:11,800 --> 00:10:16,929 from the norm of the preceding 3000 centuries is obvious, and that's 118 00:10:16,929 --> 00:10:22,059 the development of artificial intelligence, which happened no earlier than 1553 and no 119 00:10:22,059 --> 00:10:29,359 later than 1844. I'm talking of course about the very old, very slow AI's we call 120 00:10:29,359 --> 00:10:34,240 corporations. What lessons from the history of a company can we draw that tell us 121 00:10:34,240 --> 00:10:38,490 about the likely behavior of the type of artificial intelligence we're interested 122 00:10:38,490 --> 00:10:47,258 in here, today? Well. Need a mouthful of water. 123 00:10:47,258 --> 00:10:51,618 Let me crib from Wikipedia for a moment. 124 00:10:51,618 --> 00:10:56,329 Wikipedia: "In the late 18th century, Stewart Kyd, the author of the 125 00:10:56,329 --> 00:11:02,800 first treatise on corporate law in English, defined a corporation as: 'a collection of 126 00:11:02,800 --> 00:11:08,290 many individuals united into one body, under a special denomination, having 127 00:11:08,290 --> 00:11:13,779 perpetual succession under an artificial form, and vested, by policy of the law, with 128 00:11:13,779 --> 00:11:20,201 the capacity of acting, in several respects, as an individual, enjoying privileges and 129 00:11:20,201 --> 00:11:24,151 immunities in common, and of exercising a variety of political rights, more or less 130 00:11:24,151 --> 00:11:28,510 extensive, according to the design of its institution, or the powers conferred upon 131 00:11:28,510 --> 00:11:32,269 it, either at the time of its creation, or at any subsequent period of its 132 00:11:32,269 --> 00:11:36,699 existence.'" This was a late 18th century definition, 133 00:11:36,699 --> 00:11:42,509 sound like a piece of software to you? In 1844, the British government passed the 134 00:11:42,509 --> 00:11:46,450 "Joint Stock Companies Act" which created a register of companies and allowed any 135 00:11:46,450 --> 00:11:51,360 legal person, for a fee, to register a company which in turn existed as a 136 00:11:51,360 --> 00:11:55,869 separate legal person. Prior to that point, it required a Royal Charter or an act of 137 00:11:55,869 --> 00:12:00,420 Parliament to create a company. Subsequently, the law was extended to limit 138 00:12:00,420 --> 00:12:04,680 the liability of individual shareholders in event of business failure and then both 139 00:12:04,680 --> 00:12:08,629 Germany and the United States added their own unique twists to what today we see is 140 00:12:08,629 --> 00:12:14,360 the doctrine of corporate personhood. Now, though plenty of other things that 141 00:12:14,360 --> 00:12:18,740 happened between the 16th and 21st centuries did change the shape of the world we live in. 142 00:12:18,740 --> 00:12:22,149 I've skipped the changes in agricultural productivity that happened 143 00:12:22,149 --> 00:12:25,550 due to energy economics, which finally broke the Malthusian trap 144 00:12:25,550 --> 00:12:29,129 our predecessors lived in. This in turn broke the long-term 145 00:12:29,129 --> 00:12:32,681 cap on economic growth of about 0.1% per year 146 00:12:32,681 --> 00:12:35,681 in the absence of famines, plagues and wars and so on. 147 00:12:35,681 --> 00:12:38,790 I've skipped the germ theory of diseases and the development of trade empires 148 00:12:38,790 --> 00:12:42,760 in the age of sail and gunpowder, that were made possible by advances 149 00:12:42,760 --> 00:12:45,119 in accurate time measurement. 150 00:12:45,119 --> 00:12:49,079 I've skipped the rise, and hopefully decline, of the pernicious 151 00:12:49,079 --> 00:12:52,430 theory of scientific racism that underpinned Western colonialism and the 152 00:12:52,430 --> 00:12:57,220 slave trade. I've skipped the rise of feminism, the ideological position that 153 00:12:57,220 --> 00:13:02,470 women are human beings rather than property and the decline of patriarchy. 154 00:13:02,470 --> 00:13:05,999 I've skipped the whole of the Enlightenment and the Age of Revolutions, 155 00:13:05,999 --> 00:13:09,410 but this is a technocratic.. technocentric Congress, so I want to frame this talk in 156 00:13:09,410 --> 00:13:14,869 terms of AI, which we all like to think we understand. Here's the thing about these 157 00:13:14,869 --> 00:13:21,050 artificial persons we call corporations. Legally, they're people. They have goals, 158 00:13:21,050 --> 00:13:25,957 they operate in pursuit of these goals, they have a natural life cycle. 159 00:13:25,957 --> 00:13:33,230 In the 1950s, a typical U.S. corporation on the S&P 500 Index had a life span of 60 years. 160 00:13:33,230 --> 00:13:37,805 Today it's down to less than 20 years. This is largely due to predation. 161 00:13:37,805 --> 00:13:41,659 Corporations are cannibals, they eat one another. 162 00:13:41,659 --> 00:13:45,849 They're also hive super organisms like bees or ants. 163 00:13:45,849 --> 00:13:48,699 For the first century and a half, they relied entirely on human 164 00:13:48,699 --> 00:13:52,399 employees for their internal operation, but today they're automating their 165 00:13:52,399 --> 00:13:57,110 business processes very rapidly. Each human is only retained so long as they can 166 00:13:57,110 --> 00:14:01,339 perform their assigned tasks more efficiently than a piece of software 167 00:14:01,339 --> 00:14:05,439 and they can all be replaced by another human, much as the cells in our own bodies 168 00:14:05,439 --> 00:14:09,799 are functionally interchangeable and a group of cells can - in extremis - often be 169 00:14:09,799 --> 00:14:14,509 replaced by a prosthetic device. To some extent, corporations can be 170 00:14:14,509 --> 00:14:19,019 trained to serve of the personal desires of their chief executives, but even CEOs can 171 00:14:19,019 --> 00:14:22,749 be dispensed with, if their activities damage the corporation, as Harvey 172 00:14:22,749 --> 00:14:26,389 Weinstein found out a couple of months ago. 173 00:14:26,389 --> 00:14:30,939 Finally, our legal environment today has been tailored for the convenience of 174 00:14:30,939 --> 00:14:34,910 corporate persons, rather than human persons, to the point where our governments 175 00:14:34,910 --> 00:14:40,379 now mimic corporations in many of our internal structures. 176 00:14:40,379 --> 00:14:43,940 So, to understand where we're going, we need to start by asking "What do our 177 00:14:43,940 --> 00:14:51,850 current actually existing AI overlords want?" 178 00:14:51,850 --> 00:14:56,400 Now, Elon Musk, who I believe you've all heard of, has an obsessive fear of one 179 00:14:56,400 --> 00:14:59,750 particular hazard of artificial intelligence, which he conceives of as 180 00:14:59,750 --> 00:15:03,760 being a piece of software that functions like a brain in a box, namely the 181 00:15:03,760 --> 00:15:09,859 Paperclip Optimizer or Maximizer. A Paperclip Maximizer is a term of art for 182 00:15:09,859 --> 00:15:14,519 a goal seeking AI that has a single priority, e.g., maximizing the 183 00:15:14,519 --> 00:15:19,759 number of paperclips in the universe. The Paperclip Maximizer is able to improve 184 00:15:19,759 --> 00:15:24,110 itself in pursuit of its goal, but has no ability to vary its goal, so will 185 00:15:24,124 --> 00:15:27,604 ultimately attempt to convert all the metallic elements in the solar system into 186 00:15:27,604 --> 00:15:31,749 paperclips, even if this is obviously detrimental to the well-being of the 187 00:15:31,749 --> 00:15:36,359 humans who set it this goal. Unfortunately I don't think Musk 188 00:15:36,359 --> 00:15:41,169 is paying enough attention, consider his own companies. 189 00:15:41,169 --> 00:15:45,480 Tesla isn't a Paperclip Maximizer, it's a battery Maximizer. 190 00:15:45,480 --> 00:15:48,240 After all, a battery.. an electric car is a battery with wheels and 191 00:15:48,240 --> 00:15:54,199 seats. SpaceX is an orbital payload Maximizer, driving down the cost of space 192 00:15:54,199 --> 00:15:59,424 launches in order to encourage more sales for the service it provides. SolarCity is 193 00:15:59,424 --> 00:16:05,700 a photovoltaic panel maximizer and so on. All three of the.. Musk's very own slow AIs 194 00:16:05,700 --> 00:16:09,259 are based on an architecture, designed to maximize return on shareholder 195 00:16:09,259 --> 00:16:13,180 investment, even if by doing so they cook the planet the shareholders have to live 196 00:16:13,180 --> 00:16:15,979 on or turn the entire thing into solar panels. 197 00:16:15,979 --> 00:16:19,479 But hey, if you're Elon Musk, thats okay, you're gonna retire on Mars anyway. 198 00:16:19,479 --> 00:16:20,879 *laughing* 199 00:16:20,879 --> 00:16:25,139 By the way, I'm ragging on Musk in this talks, simply because he's the current 200 00:16:25,139 --> 00:16:28,600 opinionated tech billionaire, who thinks for disrupting a couple of industries 201 00:16:28,600 --> 00:16:33,579 entitles him to make headlines. If this was 2007 and my focus slightly 202 00:16:33,579 --> 00:16:39,041 difference.. different, I'd be ragging on Steve Jobs and if we're in 1997 my target 203 00:16:39,041 --> 00:16:42,259 would be Bill Gates. Don't take it personally, Elon. 204 00:16:42,259 --> 00:16:44,109 *laughing* 205 00:16:44,109 --> 00:16:49,139 Back to topic. The problem of corporations is, that despite their overt 206 00:16:49,139 --> 00:16:53,959 goals, whether they make electric vehicles or beer or sell life insurance policies, 207 00:16:53,959 --> 00:16:59,778 they all have a common implicit Paperclip Maximizer goal: to generate revenue. If 208 00:16:59,778 --> 00:17:03,729 they don't make money, they're eaten by a bigger predator or they go bust. It's as 209 00:17:03,729 --> 00:17:07,929 vital to them as breathing is to us mammals. They generally pursue their 210 00:17:07,929 --> 00:17:12,439 implicit goal - maximizing revenue - by pursuing their overt goal. 211 00:17:12,439 --> 00:17:16,510 But sometimes they try instead to manipulate their environment, to ensure 212 00:17:16,510 --> 00:17:22,970 that money flows to them regardless. Human toolmaking culture has become very 213 00:17:22,970 --> 00:17:27,980 complicated over time. New technologies always come with an attached implicit 214 00:17:27,980 --> 00:17:32,869 political agenda that seeks to extend the use of the technology. Governments react 215 00:17:32,869 --> 00:17:37,230 to this by legislating to control new technologies and sometimes we end up with 216 00:17:37,230 --> 00:17:41,630 industries actually indulging in legal duels through the regulatory mechanism of 217 00:17:41,630 --> 00:17:49,582 law to determine, who prevails. For example, consider the automobile. You 218 00:17:49,582 --> 00:17:53,910 can't have mass automobile transport without gas stations and fuel distribution 219 00:17:53,910 --> 00:17:57,160 pipelines. These in turn require access to whoever 220 00:17:57,160 --> 00:18:01,179 owns the land the oil is extracted from under and before you know it, you end up 221 00:18:01,179 --> 00:18:06,660 with a permanent army in Iraq and a clamp dictatorship in Saudi Arabia. Closer to 222 00:18:06,660 --> 00:18:12,380 home, automobiles imply jaywalking laws and drink-driving laws. They affect Town 223 00:18:12,380 --> 00:18:16,750 Planning regulations and encourage suburban sprawl, the construction of human 224 00:18:16,750 --> 00:18:21,429 infrastructure on a scale required by automobiles, not pedestrians. 225 00:18:21,429 --> 00:18:24,979 This in turn is bad for competing transport technologies, like buses or 226 00:18:24,979 --> 00:18:31,720 trams, which work best in cities with a high population density. So to get laws 227 00:18:31,720 --> 00:18:35,399 that favour the automobile in place, providing an environment conducive to 228 00:18:35,399 --> 00:18:40,409 doing business, automobile companies spend money on political lobbyists and when they 229 00:18:40,409 --> 00:18:46,620 can get away with it, on bribes. Bribery needn't be blatant of course. E.g., 230 00:18:46,620 --> 00:18:51,929 the reforms of a British railway network in the 1960s dismembered many branch lines 231 00:18:51,929 --> 00:18:56,408 and coincided with a surge in road building and automobile sales. These 232 00:18:56,408 --> 00:19:01,138 reforms were orchestrated by Transport Minister Ernest Marples, who was purely a 233 00:19:01,138 --> 00:19:05,950 politician. The fact that he accumulated a considerable personal fortune during this 234 00:19:05,950 --> 00:19:10,269 period by buying shares in motorway construction corporations, has nothing to 235 00:19:10,269 --> 00:19:18,120 do with it. So, no conflict of interest there - now if the automobile in industry 236 00:19:18,120 --> 00:19:23,230 can't be considered a pure Paperclip Maximizer... sorry, the automobile 237 00:19:23,230 --> 00:19:27,510 industry in isolation can't be considered a pure Paperclip Maximizer. You have to 238 00:19:27,510 --> 00:19:31,659 look at it in conjunction with the fossil fuel industries, the road construction 239 00:19:31,659 --> 00:19:37,809 business, the accident insurance sector and so on. When you do this, you begin to 240 00:19:37,809 --> 00:19:42,740 see the outline of a paperclip-maximizing ecosystem that invades far-flung lands and 241 00:19:42,740 --> 00:19:47,167 grinds up and kills around one and a quarter million people per year. That's 242 00:19:47,167 --> 00:19:50,951 the global death toll from automobile accidents currently, according to the World 243 00:19:50,951 --> 00:19:56,429 Health Organization. It rivals the First World War on an ongoing permanent basis 244 00:19:56,429 --> 00:20:01,510 and these are all side effects of its drive to sell you a new car. Now, 245 00:20:01,510 --> 00:20:06,639 automobiles aren't of course a total liability. Today's cars are regulated 246 00:20:06,639 --> 00:20:11,302 stringently for safety and, in theory, to reduce toxic emissions. They're fast, 247 00:20:11,302 --> 00:20:16,990 efficient and comfortable. We can thank legal mandated regulations imposed by 248 00:20:16,990 --> 00:20:22,029 governments for this, of course. Go back to the 1970s and cars didn't have crumple 249 00:20:22,029 --> 00:20:27,199 zones, go back to the 50s and they didn't come with seat belts as standard. In the 250 00:20:27,199 --> 00:20:33,690 1930s, indicators, turn signals and brakes on all four wheels were optional and your 251 00:20:33,690 --> 00:20:38,850 best hope of surviving a 50 km/h-crash was to be thrown out of a car and land somewhere 252 00:20:38,850 --> 00:20:43,059 without breaking your neck. Regulator agencies are our current 253 00:20:43,059 --> 00:20:47,380 political system's tool of choice for preventing Paperclip Maximizers from 254 00:20:47,380 --> 00:20:55,159 running amok. Unfortunately, regulators don't always work. The first failure mode 255 00:20:55,159 --> 00:20:59,730 of regulators that you need to be aware of is regulatory capture, where regulatory 256 00:20:59,730 --> 00:21:05,038 bodies are captured by the industries they control. Ajit Pai, Head of American Federal 257 00:21:05,038 --> 00:21:09,460 Communications Commission, which just voted to eliminate net neutrality rules in the 258 00:21:09,460 --> 00:21:14,101 U.S., has worked as Associate General Counsel for Verizon Communications 259 00:21:14,101 --> 00:21:19,291 Inc, the largest current descendant of the Bell Telephone system's monopoly. After 260 00:21:19,291 --> 00:21:24,690 the AT&T antitrust lawsuit, the Bell network was broken up into the seven baby 261 00:21:24,690 --> 00:21:32,279 bells. They've now pretty much reformed and reaggregated and Verizon is the largest current one. 262 00:21:32,279 --> 00:21:36,099 Why should someone with a transparent interest in a technology corporation end 263 00:21:36,099 --> 00:21:41,129 up running a regulator that tries to control the industry in question? Well, if 264 00:21:41,129 --> 00:21:44,919 you're going to regulate a complex technology, you need to recruit regulators 265 00:21:44,919 --> 00:21:48,580 from people who understand it. Unfortunately, most of those people are 266 00:21:48,580 --> 00:21:53,520 industry insiders. Ajit Pai is clearly very much aware of how Verizon is 267 00:21:53,520 --> 00:21:58,450 regulated, very insightful into its operations and wants to do something about 268 00:21:58,450 --> 00:22:02,509 it - just not necessarily in the public interest. 269 00:22:02,509 --> 00:22:11,029 *applause* When regulators end up staffed by people 270 00:22:11,029 --> 00:22:15,370 drawn from the industries they're supposed to control, they frequently end up working 271 00:22:15,370 --> 00:22:20,019 with their former office mates, to make it easier to turn a profit, either by raising 272 00:22:20,019 --> 00:22:24,350 barriers to keep new insurgent companies out or by dismantling safeguards that 273 00:22:24,350 --> 00:22:31,610 protect the public. Now a second problem is regulatory lag where a technology 274 00:22:31,610 --> 00:22:35,240 advances so rapidly, that regulations are laughably obsolete by the time they're 275 00:22:35,240 --> 00:22:40,450 issued. Consider the EU directive requiring cookie notices on websites to 276 00:22:40,450 --> 00:22:45,709 caution users, that their activities are tracked and their privacy may be violated. 277 00:22:45,709 --> 00:22:51,419 This would have been a good idea in 1993 or 1996, but unfortunatelly it didn't show up 278 00:22:51,419 --> 00:22:57,690 until 2011. Fingerprinting and tracking mechanisms have nothing to do with cookies 279 00:22:57,690 --> 00:23:04,149 and were already widespread by then. Tim Berners-Lee observed in 1995, that five 280 00:23:04,149 --> 00:23:07,780 years worth of change was happening on the web for every 12 months of real-world 281 00:23:07,780 --> 00:23:12,389 time. By that yardstick, the cookie law came out nearly a century too late to do 282 00:23:12,389 --> 00:23:19,269 any good. Again, look at Uber. This month, the European Court of Justice ruled that 283 00:23:19,269 --> 00:23:24,520 Uber is a taxi service, not a Web App. This is arguably correct - the problem is, Uber 284 00:23:24,520 --> 00:23:28,970 has spread globally since it was founded eight years ago, subsidizing its drivers to 285 00:23:28,970 --> 00:23:33,580 put competing private hire firms out of business. Whether this is a net good for 286 00:23:33,580 --> 00:23:38,659 societys own is debatable. The problem is, a taxi driver can get awfully hungry if she 287 00:23:38,659 --> 00:23:42,429 has to wait eight years for a court ruling against a predator intent on disrupting 288 00:23:42,429 --> 00:23:49,549 her business. So, to recap: firstly, we already have Paperclip Maximizers and 289 00:23:49,549 --> 00:23:54,980 Musk's AI alarmism is curiously mirror blind. Secondly, we have mechanisms for 290 00:23:54,980 --> 00:23:59,509 keeping Paperclip Maximizers in check, but they don't work very well against AIs that 291 00:23:59,509 --> 00:24:03,490 deploy the dark arts, especially corruption and bribery and they're even 292 00:24:03,490 --> 00:24:07,769 worse against true AIs, that evolved too fast for human mediated mechanisms like 293 00:24:07,769 --> 00:24:13,529 the law to keep up with. Finally, unlike the naive vision of a Paperclip Maximizer 294 00:24:13,529 --> 00:24:19,190 that maximizes only paperclips, existing AIs have multiple agendas, their overt 295 00:24:19,190 --> 00:24:24,210 goal, but also profit seeking, expansion into new markets and to accommodate the 296 00:24:24,210 --> 00:24:27,500 desire of whoever is currently in the driving seat. 297 00:24:27,500 --> 00:24:30,500 *sighs* 298 00:24:30,500 --> 00:24:36,190 Now, this brings me to the next major heading in this dismaying laundry list: 299 00:24:36,190 --> 00:24:42,560 how it all went wrong. It seems to me that our current political upheavals, the best 300 00:24:42,560 --> 00:24:49,379 understood, is arising from the capture of post 1917 democratic institutions by 301 00:24:49,379 --> 00:24:54,811 large-scale AI. Everywhere you look, you see voters protesting angrily against an 302 00:24:54,811 --> 00:24:58,779 entrenched establishment, that seems determined to ignore the wants and needs 303 00:24:58,779 --> 00:25:04,330 of their human constituents in favor of those of the machines. The brexit upset 304 00:25:04,330 --> 00:25:07,130 was largely result of a protest vote against the British political 305 00:25:07,130 --> 00:25:11,340 establishment, the election of Donald Trump likewise, with a side order of racism 306 00:25:11,340 --> 00:25:15,940 on top. Our major political parties are led by people who are compatible with the 307 00:25:15,940 --> 00:25:20,789 system as it exists today, a system that has been shaped over decades by 308 00:25:20,789 --> 00:25:26,010 corporations distorting our government and regulatory environments. We humans live in 309 00:25:26,010 --> 00:25:30,700 a world shaped by the desires and needs of AI, forced to live on their terms and we're 310 00:25:30,700 --> 00:25:34,010 taught, that we're valuable only to the extent we contribute to the rule of the 311 00:25:34,010 --> 00:25:40,259 machines. Now this is free sea and we're all more interested in computers and 312 00:25:40,259 --> 00:25:44,389 communications technology than this historical crap. But as I said earlier, 313 00:25:44,389 --> 00:25:49,149 history is a secret weapon, if you know how to use it. What history is good for, is 314 00:25:49,149 --> 00:25:53,080 enabling us to spot recurring patterns that repeat across timescales outside our 315 00:25:53,080 --> 00:25:58,129 personal experience. And if we look at our historical very slow AIs, what do we learn 316 00:25:58,129 --> 00:26:04,880 from them about modern AI and how it's going to behave? Well to start with, our 317 00:26:04,880 --> 00:26:09,879 AIs have been warped, the new AIs, the electronic one's instantiated in our 318 00:26:09,879 --> 00:26:14,639 machines, have been warped by a terrible fundamentally flawed design decision back 319 00:26:14,639 --> 00:26:20,289 in 1995, but as damaged democratic political processes crippled our ability 320 00:26:20,289 --> 00:26:24,570 to truly understand the world around us and led to the angry upheavals and upsets 321 00:26:24,570 --> 00:26:30,440 of our present decade. That mistake was the decision, to fund the build-out of a 322 00:26:30,440 --> 00:26:34,370 public World Wide Web as opposed to be earlier government-funded corporate and 323 00:26:34,370 --> 00:26:38,230 academic Internet by monetizing eyeballs through advertising 324 00:26:38,230 --> 00:26:45,260 revenue. The ad-supported web we're used to today wasn't inevitable. If you recall 325 00:26:45,260 --> 00:26:49,510 the web as it was in 1994, there were very few ads at all and not much, in a way, of 326 00:26:49,510 --> 00:26:55,720 Commerce. 1995 was the year, the World Wide Web really came to public attention in the 327 00:26:55,720 --> 00:27:00,570 anglophone world and consumer-facing websites began to appear. Nobody really 328 00:27:00,570 --> 00:27:04,490 knew, how this thing was going to be paid for. The original .com bubble was all 329 00:27:04,490 --> 00:27:07,850 about working out, how to monetize the web for the first time and a lot of people 330 00:27:07,850 --> 00:27:12,840 lost their shirts in the process. A naive initial assumption was that the 331 00:27:12,840 --> 00:27:17,440 transaction cost of setting up a tcp/ip connection over modem was too high to 332 00:27:17,440 --> 00:27:22,500 support.. to be supported by per-use micro billing for web pages. So instead of 333 00:27:22,500 --> 00:27:27,059 charging people fraction of a euro cent for every page view, we'd bill customers 334 00:27:27,059 --> 00:27:31,570 indirectly, by shoving advertising banners in front of their eyes and hoping they'd 335 00:27:31,570 --> 00:27:39,029 click through and buy something. Unfortunately, advertising is in an 336 00:27:39,029 --> 00:27:45,509 industry, one of those pre-existing very slow AI ecosystems I already alluded to. 337 00:27:45,509 --> 00:27:49,700 Advertising tries to maximize its hold on the attention of the minds behind each 338 00:27:49,700 --> 00:27:53,750 human eyeball. The coupling of advertising with web search was an inevitable 339 00:27:53,750 --> 00:27:57,991 outgrowth, I mean how better to attract the attention of reluctant subjects, than to 340 00:27:57,991 --> 00:28:01,450 find out what they're really interested in seeing and selling ads that relate to 341 00:28:01,450 --> 00:28:07,070 those interests. The problem of applying the paperclip maximize approach to 342 00:28:07,070 --> 00:28:12,509 monopolizing eyeballs, however, is that eyeballs are a limited, scarce resource. 343 00:28:12,509 --> 00:28:18,500 There are only 168 hours in every week, in which I can gaze at banner ads. Moreover, 344 00:28:18,500 --> 00:28:22,289 most ads are irrelevant to my interests and it doesn't matter, how often you flash an ad 345 00:28:22,289 --> 00:28:28,129 for dog biscuits at me, I'm never going to buy any. I have a cat. To make best 346 00:28:28,129 --> 00:28:31,990 revenue-generating use of our eyeballs, it's necessary for the ad industry to 347 00:28:31,990 --> 00:28:36,789 learn, who we are and what interests us and to target us increasingly minutely in hope 348 00:28:36,789 --> 00:28:39,549 of hooking us with stuff we're attracted to. 349 00:28:39,549 --> 00:28:43,879 In other words: the ad industry is a paperclip maximizer, but for its success, 350 00:28:43,879 --> 00:28:49,990 it relies on developing a theory of mind that applies to human beings. 351 00:28:49,990 --> 00:28:52,990 *sighs* 352 00:28:52,990 --> 00:28:55,739 Do I need to divert on to the impassioned rant about the hideous corruption 353 00:28:55,739 --> 00:28:59,759 and evil that is Facebook? Audience: Yes! 354 00:28:59,759 --> 00:29:03,289 CS: Okay, somebody said yes. I'm guessing you've heard it all before, 355 00:29:03,289 --> 00:29:07,429 but for too long don't read.. summary is: Facebook is as much a search engine as 356 00:29:07,429 --> 00:29:12,029 Google or Amazon. Facebook searches are optimized for faces, that is for human 357 00:29:12,029 --> 00:29:15,649 beings. If you want to find someone you fell out of touch with thirty years ago, 358 00:29:15,649 --> 00:29:20,610 Facebook probably knows where they live, what their favorite color is, what sized 359 00:29:20,610 --> 00:29:24,390 shoes they wear and what they said about you to your friends behind your back all 360 00:29:24,390 --> 00:29:30,090 those years ago, that made you cut them off. Even if you don't have a Facebook account, 361 00:29:30,090 --> 00:29:34,230 Facebook has a You account, a hole in their social graph of a bunch of connections 362 00:29:34,230 --> 00:29:38,669 pointing in to it and your name tagged on your friends photographs. They know a lot 363 00:29:38,669 --> 00:29:42,529 about you and they sell access to their social graph to advertisers, who then 364 00:29:42,529 --> 00:29:47,400 target you, even if you don't think you use Facebook. Indeed, there is barely any 365 00:29:47,400 --> 00:29:52,270 point in not using Facebook these days, if ever. Social media Borg: "Resistance is 366 00:29:52,270 --> 00:30:00,510 futile!" So however, Facebook is trying to get eyeballs on ads, so is Twitter and so 367 00:30:00,510 --> 00:30:05,970 are Google. To do this, they fine-tuned the content they show you to make it more 368 00:30:05,970 --> 00:30:11,520 attractive to your eyes and by attractive I do not mean pleasant. We humans have an 369 00:30:11,520 --> 00:30:15,139 evolved automatic reflex to pay attention to threats and horrors as well as 370 00:30:15,139 --> 00:30:19,830 pleasurable stimuli and the algorithms, that determine what they show us when we 371 00:30:19,830 --> 00:30:24,259 look at Facebook or Twitter, take this bias into account. You might react more 372 00:30:24,259 --> 00:30:28,490 strongly to a public hanging in Iran or an outrageous statement by Donald Trump than 373 00:30:28,490 --> 00:30:32,100 to a couple kissing. The algorithm knows and will show you whatever makes you pay 374 00:30:32,100 --> 00:30:38,389 attention, not necessarily what you need or want to see. 375 00:30:38,389 --> 00:30:42,510 So this brings me to another point about computerized AI as opposed to corporate 376 00:30:42,510 --> 00:30:47,260 AI. AI algorithms tend to embody the prejudices and beliefs of either the 377 00:30:47,260 --> 00:30:52,759 programmers, or the data set the AI was trained on. 378 00:30:52,759 --> 00:30:56,250 A couple of years ago I ran across an account of a webcam, developed by mostly 379 00:30:56,250 --> 00:31:00,860 pale-skinned Silicon Valley engineers, that had difficulty focusing or achieving correct 380 00:31:00,860 --> 00:31:04,500 color balance, when pointed at dark-skinned faces. 381 00:31:04,500 --> 00:31:08,400 Fast an example of human programmer induced bias, they didn't have a wide 382 00:31:08,400 --> 00:31:13,249 enough test set and didn't recognize that they were inherently biased towards 383 00:31:13,249 --> 00:31:19,290 expecting people to have pale skin. But with today's deep learning, bias can creep 384 00:31:19,290 --> 00:31:24,009 in, while the datasets for neural networks are trained on, even without the programmers 385 00:31:24,009 --> 00:31:28,659 intending it. Microsoft's first foray into a conversational chat bot driven by 386 00:31:28,659 --> 00:31:33,330 machine learning, Tay, was what we yanked offline within days last year, because 387 00:31:33,330 --> 00:31:37,149 4chan and reddit based trolls discovered, that they could train it towards racism and 388 00:31:37,149 --> 00:31:43,649 sexism for shits and giggles. Just imagine you're a poor naive innocent AI who's just 389 00:31:43,649 --> 00:31:48,480 been switched on and you're hoping to pass your Turing test and what happens? 4chan 390 00:31:48,480 --> 00:31:53,189 decide to play with your head. *laughing* 391 00:31:53,189 --> 00:31:57,519 I got to feel sorry for Tay. Now, humans may be biased, 392 00:31:57,519 --> 00:32:01,059 but at least individually we're accountable and if somebody gives you 393 00:32:01,059 --> 00:32:06,499 racist or sexist abuse to your face, you can complain or maybe punch them. It's 394 00:32:06,499 --> 00:32:10,740 impossible to punch a corporation and it may not even be possible to identify the 395 00:32:10,740 --> 00:32:16,029 source of unfair bias, when you're dealing with a machine learning system. AI based 396 00:32:16,029 --> 00:32:21,739 systems that instantiate existing prejudices make social change harder. 397 00:32:21,739 --> 00:32:25,470 Traditional advertising works by playing on the target customer's insecurity and 398 00:32:25,470 --> 00:32:30,750 fear as much as their aspirations. And fear of a loss of social status and privileges 399 00:32:30,750 --> 00:32:36,240 are powerful stress. Fear and xenophobia are useful tools for tracking advertising.. 400 00:32:36,240 --> 00:32:39,580 ah, eyeballs. What happens when we get pervasive social 401 00:32:39,580 --> 00:32:44,379 networks, that have learned biases against say Feminism or Islam or melanin? Or deep 402 00:32:44,379 --> 00:32:48,230 learning systems, trained on datasets contaminated by racist dipshits and their 403 00:32:48,230 --> 00:32:53,369 propaganda? Deep learning systems like the ones inside Facebook, that determine which 404 00:32:53,369 --> 00:32:58,210 stories to show you to get you to pay as much attention as possible to be adverse. 405 00:32:58,210 --> 00:33:04,841 I think, you probably have an inkling of how.. where this is now going. Now, if you 406 00:33:04,841 --> 00:33:08,990 think, this is sounding a bit bleak and unpleasant, you'd be right. I write sci-fi. 407 00:33:08,990 --> 00:33:12,669 You read or watch or play sci-fi. We're acculturated to think of science and 408 00:33:12,669 --> 00:33:19,420 technology as good things that make life better, but this ain't always so. Plenty of 409 00:33:19,420 --> 00:33:23,140 technologies have historically been heavily regulated or even criminalized for 410 00:33:23,140 --> 00:33:27,690 good reason and once you get past any reflexive indignation, criticism of 411 00:33:27,690 --> 00:33:32,740 technology and progress, you might agree with me, that it is reasonable to ban 412 00:33:32,740 --> 00:33:39,139 individuals from owning nuclear weapons or nerve gas. Less obviously, they may not be 413 00:33:39,139 --> 00:33:42,820 weapons, but we've banned chlorofluorocarbon refrigerants, because 414 00:33:42,820 --> 00:33:45,570 they were building up in the high stratosphere and destroying the ozone 415 00:33:45,570 --> 00:33:51,360 layer that protects us from UVB radiation. We banned tetra e-file LED in 416 00:33:51,360 --> 00:33:57,540 gasoline, because it poisoned people and led to a crime wave. These are not 417 00:33:57,540 --> 00:34:03,240 weaponized technologies, but they have horrible side effects. Now, nerve gas and 418 00:34:03,240 --> 00:34:08,690 leaded gasoline were 1930s chemical technologies, promoted by 1930s 419 00:34:08,690 --> 00:34:15,390 corporations. Halogenated refrigerants and nuclear weapons are totally 1940s. ICBMs 420 00:34:15,390 --> 00:34:19,210 date to the 1950s. You know, I have difficulty seeing why people are getting 421 00:34:19,210 --> 00:34:26,040 so worked up over North Korea. North Korea reaches 1953 level parity - be terrified 422 00:34:26,040 --> 00:34:30,760 and hide under the bed! I submit that the 21st century is throwing 423 00:34:30,760 --> 00:34:35,030 up dangerous new technologies, just as our existing strategies for regulating very 424 00:34:35,030 --> 00:34:42,060 slow AIs have proven to be inadequate. And I don't have an answer to how we regulate 425 00:34:42,060 --> 00:34:45,889 new technologies, I just want to flag it up as a huge social problem that is going to 426 00:34:45,889 --> 00:34:50,380 affect the coming century. I'm now going to give you four examples of 427 00:34:50,380 --> 00:34:54,290 new types of AI application that are going to warp our societies even more 428 00:34:54,290 --> 00:35:00,800 badly than the old slow AIs, we.. have done. This isn't an exhaustive list, this is just 429 00:35:00,800 --> 00:35:05,160 some examples I dream, I pulled out of my ass. We need to work out a general 430 00:35:05,160 --> 00:35:08,430 strategy for getting on top of this sort of thing before they get on top of us and 431 00:35:08,430 --> 00:35:12,040 I think, this is actually a very urgent problem. So I'm just going to give you this 432 00:35:12,040 --> 00:35:18,110 list of dangerous new technologies that are arriving now, or coming, and send you 433 00:35:18,110 --> 00:35:22,150 away to think about what to do next. I mean, we are activists here, we should be 434 00:35:22,150 --> 00:35:27,690 thinking about this and planning what to do. Now, the first nasty technology I'd 435 00:35:27,690 --> 00:35:31,600 like to talk about, is political hacking tools that rely on social graph directed 436 00:35:31,600 --> 00:35:39,500 propaganda. This is low-hanging fruit after the electoral surprises of 2016. 437 00:35:39,500 --> 00:35:43,280 Cambridge Analytica pioneered the use of deep learning by scanning the Facebook and 438 00:35:43,280 --> 00:35:47,750 Twitter social graphs to identify voters political affiliations by simply looking 439 00:35:47,750 --> 00:35:52,950 at what tweets or Facebook comments they liked, very able to do this, to identify 440 00:35:52,950 --> 00:35:56,400 individuals with a high degree of precision, who were vulnerable to 441 00:35:56,400 --> 00:36:01,490 persuasion and who lived in electorally sensitive districts. They then canvassed 442 00:36:01,490 --> 00:36:06,910 them with propaganda, that targeted their personal hot-button issues to change their 443 00:36:06,910 --> 00:36:12,060 electoral intentions. The tools developed by web advertisers to sell products have 444 00:36:12,060 --> 00:36:16,260 now been weaponized for political purposes and the amount of personal information 445 00:36:16,260 --> 00:36:21,260 about our affiliations that we expose on social media, makes us vulnerable. Aside, in 446 00:36:21,260 --> 00:36:24,970 the last U.S. Presidential election, as mounting evidence for the British 447 00:36:24,970 --> 00:36:28,510 referendum on leaving the EU was subject to foreign cyber war attack, now 448 00:36:28,510 --> 00:36:33,250 weaponized social media, as was the most recent French Presidential election. 449 00:36:33,250 --> 00:36:38,080 In fact, if we remember the leak of emails from the Macron campaign, it turns out that 450 00:36:38,080 --> 00:36:42,020 many of those emails were false, because the Macron campaign anticipated that they 451 00:36:42,020 --> 00:36:47,120 would be attacked and an email trove would be leaked in the last days before the 452 00:36:47,120 --> 00:36:51,260 election. So they deliberately set up false emails that would be hacked and then 453 00:36:51,260 --> 00:37:00,610 leaked and then could be discredited. It gets twisty fast. Now I'm kind of biting 454 00:37:00,610 --> 00:37:04,650 my tongue and trying, not to take sides here. I have my own political affiliation 455 00:37:04,650 --> 00:37:09,770 after all, and I'm not terribly mainstream. But if social media companies don't work 456 00:37:09,770 --> 00:37:14,220 out how to identify and flag micro- targeted propaganda, then democratic 457 00:37:14,220 --> 00:37:18,140 institutions will stop working and elections will be replaced by victories, whoever 458 00:37:18,140 --> 00:37:23,200 can buy the most trolls. This won't simply be billionaires but.. like the Koch 459 00:37:23,200 --> 00:37:26,470 brothers and Robert Mercer from the U.S. throwing elections to whoever will 460 00:37:26,470 --> 00:37:31,120 hand them the biggest tax cuts. Russian military cyber war doctrine calls for the 461 00:37:31,120 --> 00:37:35,730 use of social media to confuse and disable perceived enemies, in addition to the 462 00:37:35,730 --> 00:37:39,900 increasingly familiar use of zero-day exploits for espionage, such as spear 463 00:37:39,900 --> 00:37:43,200 phishing and distributed denial-of-service attacks, on our infrastructure, which are 464 00:37:43,200 --> 00:37:49,260 practiced by Western agencies. Problem is, once the Russians have demonstrated that 465 00:37:49,260 --> 00:37:53,990 this is an effective tactic, the use of propaganda bot armies in cyber war will go 466 00:37:53,990 --> 00:38:00,400 global. And at that point, our social discourse will be irreparably poisoned. 467 00:38:00,400 --> 00:38:04,990 Incidentally, I'd like to add - as another aside like the Elon Musk thing - I hate 468 00:38:04,990 --> 00:38:09,680 the cyber prefix! It usually indicates, that whoever's using it has no idea what 469 00:38:09,680 --> 00:38:15,940 they're talking about. *applause, laughter* 470 00:38:15,940 --> 00:38:21,120 Unfortunately, much as the way the term hacker was corrupted from its original 471 00:38:21,120 --> 00:38:27,020 meaning in the 1990s, the term cyber war has, it seems, to have stuck and it's now an 472 00:38:27,020 --> 00:38:31,500 actual thing that we can point to and say: "This is what we're talking about". So I'm 473 00:38:31,500 --> 00:38:35,510 afraid, we're stuck with this really horrible term. But that's a digression, I 474 00:38:35,510 --> 00:38:38,780 should get back on topic, because I've only got 20 minutes to go. 475 00:38:38,780 --> 00:38:46,240 Now, the second threat that we need to think about regulating ,or controlling, is 476 00:38:46,240 --> 00:38:50,031 an adjunct to deep learning target propaganda: it's the use of neural network 477 00:38:50,031 --> 00:38:56,940 generated false video media. We're used to photoshopped images these days, but faking 478 00:38:56,940 --> 00:39:02,510 video and audio takes it to the next level. Luckily, faking video and audio is 479 00:39:02,510 --> 00:39:08,560 labor-intensive, isn't it? Well nope, not anymore. We're seeing the first generation 480 00:39:08,560 --> 00:39:13,510 of AI assisted video porn, in which the faces of film stars are mapped onto those 481 00:39:13,510 --> 00:39:17,370 of other people in a video clip, using software rather than laborious in human 482 00:39:17,370 --> 00:39:22,490 process. A properly trained neural network 483 00:39:22,490 --> 00:39:29,430 recognizes faces and transforms the face of the Hollywood star, they want to put 484 00:39:29,430 --> 00:39:35,190 into a porn movie, into the face of - onto the face of the porn star in the porn clip 485 00:39:35,190 --> 00:39:40,520 and suddenly you have "Oh dear God, get it out of my head" - no, not gonna give you 486 00:39:40,520 --> 00:39:43,500 any examples. Let's just say it's bad stuff. 487 00:39:43,500 --> 00:39:47,260 *laughs* Meanwhile we have WaveNet, a system 488 00:39:47,260 --> 00:39:51,120 for generating realistic sounding speech, if a voice of a human's speak of a neural 489 00:39:51,120 --> 00:39:56,160 network has been trained to mimic any human speaker. We can now put words into 490 00:39:56,160 --> 00:40:01,250 other people's mouths realistically without employing a voice actor. This 491 00:40:01,250 --> 00:40:06,650 stuff is still geek intensive. It requires relatively expensive GPUs or cloud 492 00:40:06,650 --> 00:40:10,910 computing clusters, but in less than a decade it'll be out in the wild, turned 493 00:40:10,910 --> 00:40:15,500 into something, any damn script kiddie can use and just about everyone will be able 494 00:40:15,500 --> 00:40:18,940 to fake up a realistic video of someone they don't like doing something horrible. 495 00:40:18,940 --> 00:40:26,970 I mean, Donald Trump in the White House. I can't help but hope that out there 496 00:40:26,970 --> 00:40:30,670 somewhere there's some geek like Steve Bannon with a huge rack of servers who's 497 00:40:30,670 --> 00:40:40,980 faking it all, but no. Now, also we've already seen alarm this year over bizarre 498 00:40:40,980 --> 00:40:45,090 YouTube channels that attempt to monetize children's TV brands by scraping the video 499 00:40:45,090 --> 00:40:50,220 content of legitimate channels and adding their own advertising in keywords on top 500 00:40:50,220 --> 00:40:54,110 before reposting it. This is basically your YouTube spam. 501 00:40:54,110 --> 00:40:58,521 Many of these channels are shaped by paperclip maximizing advertising AIs, but 502 00:40:58,521 --> 00:41:03,670 are simply trying to maximise their search ranking on YouTube and it's entirely 503 00:41:03,670 --> 00:41:08,470 algorithmic: you have a whole list of keywords, you perm, you take them, you slap 504 00:41:08,470 --> 00:41:15,300 them on top of existing popular videos and re-upload the videos. Once you add neural 505 00:41:15,300 --> 00:41:19,510 network driven tools for inserting character A into pirated video B, to click 506 00:41:19,510 --> 00:41:24,420 maximize.. for click maximizing bots, things are gonna get very weird, though. And 507 00:41:24,420 --> 00:41:28,980 they're gonna get even weirder, when these tools are deployed for political gain. 508 00:41:28,980 --> 00:41:34,830 We tend - being primates, that evolved 300 thousand years ago in a smartphone free 509 00:41:34,830 --> 00:41:39,950 environment - to evaluate the inputs from our eyes and ears much less critically 510 00:41:39,950 --> 00:41:44,380 than what random strangers on the Internet tell us in text. We're already too 511 00:41:44,380 --> 00:41:48,930 vulnerable to fake news as it is. Soon they'll be coming for us, armed with 512 00:41:48,930 --> 00:41:54,830 believable video evidence. The Smart Money says that by 2027 you won't be able to 513 00:41:54,830 --> 00:41:58,550 believe anything you see in video, unless for a cryptographic signatures on it, 514 00:41:58,550 --> 00:42:02,530 linking it back to the camera that shot the raw feed. But you know how good most 515 00:42:02,530 --> 00:42:08,150 people are at using encryption - it's going to be chaos! 516 00:42:08,150 --> 00:42:14,070 So, paperclip maximizers with focus on eyeballs are very 20th century. The new 517 00:42:14,070 --> 00:42:19,750 generation is going to be focusing on our nervous system. Advertising as an industry 518 00:42:19,750 --> 00:42:22,510 can only exist because of a quirk of our nervous system, which is that we're 519 00:42:22,510 --> 00:42:26,680 susceptible to addiction. Be it tobacco, gambling or heroin, we 520 00:42:26,680 --> 00:42:31,980 recognize addictive behavior, when we see it. Well, do we? It turns out the human 521 00:42:31,980 --> 00:42:36,250 brain's reward feedback loops are relatively easy to gain. Large 522 00:42:36,250 --> 00:42:41,400 corporations like Zynga - producers of FarmVille - exist solely because of it, 523 00:42:41,400 --> 00:42:45,580 free to use social media platforms like Facebook and Twitter, are dominant precisely 524 00:42:45,580 --> 00:42:50,100 because they're structured to reward frequent short bursts of interaction and 525 00:42:50,100 --> 00:42:54,790 to generate emotional engagement - not necessarily positive emotions, anger and 526 00:42:54,790 --> 00:42:59,660 hatred are just as good when it comes to attracting eyeballs for advertisers. 527 00:42:59,660 --> 00:43:04,510 Smartphone addiction is a side effect of advertising as a revenue model. Frequent 528 00:43:04,510 --> 00:43:09,850 short bursts of interaction to keep us coming back for more. Now a new.. newish 529 00:43:09,850 --> 00:43:13,720 development, thanks to deep learning again - I keep coming back to deep learning, 530 00:43:13,720 --> 00:43:19,290 don't I? - use of neural networks in a manner that Marvin Minsky never envisaged, 531 00:43:19,290 --> 00:43:22,530 back when he was deciding that the Perzeptron was where it began and ended 532 00:43:22,530 --> 00:43:27,360 and it couldn't do anything. Well, we have neuroscientists now, who've 533 00:43:27,360 --> 00:43:33,680 mechanized the process of making apps more addictive. Dopamine Labs is one startup 534 00:43:33,680 --> 00:43:38,030 that provides tools to app developers to make any app more addictive, as well as to 535 00:43:38,030 --> 00:43:41,130 reduce the desire to continue participating in a behavior if it's 536 00:43:41,130 --> 00:43:46,520 undesirable, if the app developer actually wants to help people kick the habit. This 537 00:43:46,520 --> 00:43:52,060 goes way beyond automated A/B testing. A/B testing allows developers to plot a binary 538 00:43:52,060 --> 00:43:58,760 tree path between options, moving towards a single desired goal. But true deep 539 00:43:58,760 --> 00:44:03,640 learning, addictiveness maximizers, can optimize for multiple attractors in 540 00:44:03,640 --> 00:44:10,011 parallel. The more users you've got on your app, the more effectively you can work 541 00:44:10,011 --> 00:44:16,920 out, what attracts them and train them and focus on extra addictive characteristics. 542 00:44:16,920 --> 00:44:20,610 Now, going by their public face, the folks at Dopamine Labs seem to have ethical 543 00:44:20,610 --> 00:44:24,930 qualms about the misuse of addiction maximizers. But neuroscience isn't a 544 00:44:24,930 --> 00:44:28,730 secret and sooner or later some really unscrupulous sociopaths will try to see 545 00:44:28,730 --> 00:44:36,440 how far they can push it. So let me give you a specific imaginary scenario: Apple 546 00:44:36,440 --> 00:44:41,400 have put a lot of effort into making real- time face recognition work on the iPhone X 547 00:44:41,400 --> 00:44:44,940 and it's going to be everywhere on everybody's phone in another couple of 548 00:44:44,940 --> 00:44:50,760 years. You can't fool an iPhone X with a photo or even a simple mask. It does depth 549 00:44:50,760 --> 00:44:54,010 mapping to ensure, your eyes are in the right place and can tell whether they're 550 00:44:54,010 --> 00:44:58,200 open or closed. It recognizes your face from underlying bone structure through 551 00:44:58,200 --> 00:45:02,760 makeup and bruises. It's running continuously, checking pretty much as often 552 00:45:02,760 --> 00:45:06,840 as every time you'd hit the home button on a more traditional smartphone UI and it 553 00:45:06,840 --> 00:45:13,750 can see where your eyeballs are pointing. The purpose of a face recognition system 554 00:45:13,750 --> 00:45:19,980 is to provide for real-time authenticate continuous authentication when you're 555 00:45:19,980 --> 00:45:23,960 using a device - not just enter a PIN or sign a password or use a two factor 556 00:45:23,960 --> 00:45:28,420 authentication pad, but the device knows that you are its authorized user on a 557 00:45:28,420 --> 00:45:33,000 continuous basis and if somebody grabs your phone and runs away with it, it'll 558 00:45:33,000 --> 00:45:38,050 know that it's been stolen immediately, it sees the face of the thief. 559 00:45:38,050 --> 00:45:42,910 However, your phone monitoring your facial expressions and correlating against app 560 00:45:42,910 --> 00:45:48,250 usage has other implications. Your phone will be aware of precisely what you like 561 00:45:48,250 --> 00:45:53,550 to look at on your screen.. on its screen. We may well have sufficient insight on the 562 00:45:53,550 --> 00:45:59,740 part of the phone to identify whether you're happy or sad, bored or engaged. 563 00:45:59,740 --> 00:46:04,680 With addiction seeking deep learning tools and neural network generated images, those 564 00:46:04,680 --> 00:46:09,150 synthetic videos I was talking about, it's entirely.. in principle entirely possible to 565 00:46:09,150 --> 00:46:15,250 feed you an endlessly escalating payload of arousal-inducing inputs. It might be 566 00:46:15,250 --> 00:46:19,440 Facebook or Twitter messages, optimized to produce outrage, or it could be porn 567 00:46:19,440 --> 00:46:24,620 generated by AI to appeal to kinks you don't even consciously know you have. 568 00:46:24,620 --> 00:46:28,980 But either way, the app now owns your central nervous system and you will be 569 00:46:28,980 --> 00:46:36,230 monetized. And finally, I'd like to raise a really hair-raising specter that goes well 570 00:46:36,230 --> 00:46:41,730 beyond the use of deep learning and targeted propaganda and cyber war. Back in 571 00:46:41,730 --> 00:46:46,340 2011, an obscure Russian software house launched an iPhone app for pickup artists 572 00:46:46,340 --> 00:46:52,940 called 'Girls Around Me'. Spoiler: Apple pulled it like a hot potato as soon as 573 00:46:52,940 --> 00:46:59,060 word got out that it existed. Now, Girls Around Me works out where the user is 574 00:46:59,060 --> 00:47:03,730 using GPS, then it would query Foursquare and Facebook for people matching a simple 575 00:47:03,730 --> 00:47:09,880 relational search, for single females on Facebook, per relationship status, who have 576 00:47:09,880 --> 00:47:15,020 checked in, or been checked in by their friends, in your vicinity on Foursquare. 577 00:47:15,020 --> 00:47:19,270 The app then displays their locations on a map along with links to their social media 578 00:47:19,270 --> 00:47:24,560 profiles. If they were doing it today, the interface would be gamified, showing strike 579 00:47:24,560 --> 00:47:28,980 rates and a leaderboard and flagging targets who succumbed to harassment as 580 00:47:28,980 --> 00:47:31,960 easy lays. But these days, the cool kids and single 581 00:47:31,960 --> 00:47:35,980 adults are all using dating apps with a missing vowel in the name, only a creeper 582 00:47:35,980 --> 00:47:45,420 would want something like Girls Around Me, right? Unfortunately, there are much, much 583 00:47:45,420 --> 00:47:49,450 nastier uses of and scraping social media to find potential victims for serial 584 00:47:49,450 --> 00:47:54,480 rapists. Does your social media profile indicate your political religious 585 00:47:54,480 --> 00:47:59,650 affiliation? No? Cambridge Analytica can work them out with 99.9% precision 586 00:47:59,650 --> 00:48:04,660 anyway, so don't worry about that. We already have you pegged. Now add a service 587 00:48:04,660 --> 00:48:08,430 that can identify people's affiliation and location and you have a beginning of a 588 00:48:08,430 --> 00:48:13,520 flash mob app, one that will show people like us and people like them on a 589 00:48:13,520 --> 00:48:18,551 hyperlocal map. Imagine you're a young female and a 590 00:48:18,551 --> 00:48:22,240 supermarket like Target has figured out from your purchase patterns, that you're 591 00:48:22,240 --> 00:48:28,160 pregnant, even though you don't know it yet. This actually happened in 2011. Now 592 00:48:28,160 --> 00:48:31,250 imagine, that all the anti-abortion campaigners in your town have an app 593 00:48:31,250 --> 00:48:36,110 called "Babies Risk" on their phones. Someone has paid for the analytics feed 594 00:48:36,110 --> 00:48:40,320 from the supermarket and every time you go near a family planning clinic, a group of 595 00:48:40,320 --> 00:48:44,270 unfriendly anti-abortion protesters somehow miraculously show up and swarm 596 00:48:44,270 --> 00:48:49,990 you. Or imagine you're male and gay and the "God hates fags"-crowd has invented a 597 00:48:49,990 --> 00:48:55,300 100% reliable gaydar app, based on your Grindr profile, and is getting their fellow 598 00:48:55,300 --> 00:49:00,810 travelers to queer bash gay men - only when they're alone or outnumbered by ten to 599 00:49:00,810 --> 00:49:06,410 one. That's the special horror of precise geolocation not only do you always know 600 00:49:06,410 --> 00:49:13,020 where you are, the AIs know, where you are and some of them aren't friendly. Or 601 00:49:13,020 --> 00:49:17,330 imagine, you're in Pakistan and Christian Muslim tensions are rising or in rural 602 00:49:17,330 --> 00:49:24,250 Alabama or an Democrat, you know the possibilities are endless. Someone out 603 00:49:24,250 --> 00:49:30,120 there is working on this. A geolocation aware, social media scraping deep learning 604 00:49:30,120 --> 00:49:34,230 application, that uses a gamified competitive interface to reward its 605 00:49:34,230 --> 00:49:38,500 players for joining in acts of mob violence against whoever the app developer 606 00:49:38,500 --> 00:49:41,640 hates. Probably it has an innocuous seeming, but 607 00:49:41,640 --> 00:49:46,650 highly addictive training mode, to get the users accustomed to working in teams and 608 00:49:46,650 --> 00:49:52,640 obeying the apps instructions. Think Ingress or Pokemon Go. Then at some pre- 609 00:49:52,640 --> 00:49:57,770 planned zero-hour, it switches mode and starts rewarding players for violence, 610 00:49:57,770 --> 00:50:01,780 players who have been primed to think of their targets as vermin by a steady drip 611 00:50:01,780 --> 00:50:06,470 feed of micro-targeted dehumanizing propaganda inputs, delivered over a period 612 00:50:06,470 --> 00:50:12,310 of months. And the worst bit of this picture? Is that the app developer isn't even a 613 00:50:12,310 --> 00:50:16,540 nation-state trying to disrupt its enemies or an extremist political group trying to 614 00:50:16,540 --> 00:50:22,000 murder gays, Jews or Muslims. It's just a Paperclip Maximizer doing what it does 615 00:50:22,000 --> 00:50:27,900 and you are the paper. Welcome to the 21st century. 616 00:50:27,900 --> 00:50:40,730 *applause* Uhm... 617 00:50:40,730 --> 00:50:41,960 Thank you. 618 00:50:41,960 --> 00:50:47,780 *ongoing applause* We have a little time for questions. Do 619 00:50:47,780 --> 00:50:53,790 you have a microphone for the orders? Do we have any questions? ... OK. 620 00:50:53,790 --> 00:50:56,430 Herald: So you are doing a Q&A? CS: Hmm? 621 00:50:56,430 --> 00:51:02,290 Herald: So you are doing a Q&A. Well if there are any questions, please come 622 00:51:02,290 --> 00:51:24,320 forward to the microphones, numbers 1 through 4 and ask. 623 00:51:24,320 --> 00:51:29,010 Mic 1: Do you really think it's all bleak and dystopian like you prescribed 624 00:51:29,010 --> 00:51:34,750 it, because I also think the future can be bright, looking at the internet with open 625 00:51:34,750 --> 00:51:38,730 source and like, it's all growing and going faster and faster in a good 626 00:51:38,730 --> 00:51:44,110 direction. So what do you think about the balance here? 627 00:51:44,110 --> 00:51:48,070 CS: *sighs* Basically, I think the problem is, that about 3% of us 628 00:51:48,070 --> 00:51:54,330 are sociopaths or psychopaths, who spoil everything for the other 97% of us. 629 00:51:54,330 --> 00:51:56,980 Wouldn't it be great if somebody could write an app that would identify all the 630 00:51:56,980 --> 00:51:59,690 psychopaths among us and let the rest of us just kill them? 631 00:51:59,690 --> 00:52:05,320 *laughing, applause* Yeah, we have all the 632 00:52:05,320 --> 00:52:12,300 tools to make a utopia, we have it now today. A bleak miserable grim meathook 633 00:52:12,300 --> 00:52:17,510 future is not inevitable, but it's up to us to use these tools to prevent the bad 634 00:52:17,510 --> 00:52:22,480 stuff happening and to do that, we have to anticipate the bad outcomes and work to 635 00:52:22,480 --> 00:52:25,260 try and figure out a way to deal with them. That's what this talk is. I'm trying 636 00:52:25,260 --> 00:52:29,300 to do a bit of a wake-up call and get people thinking about how much worse 637 00:52:29,300 --> 00:52:33,630 things can get and what we need to do to prevent it from happening. What I was 638 00:52:33,630 --> 00:52:37,960 saying earlier about our regulatory systems being broken, stands. How do we 639 00:52:37,960 --> 00:52:47,500 regulate the deep learning technologies? This is something we need to think about. 640 00:52:47,500 --> 00:52:55,230 H: Okay mic number two. Mic 2: Hello? ... When you talk about 641 00:52:55,230 --> 00:53:02,250 corporations as AIs, where do you see that analogy you're making? Do you see them as 642 00:53:02,250 --> 00:53:10,010 literally AIs or figuratively? CS: Almost literally. If 643 00:53:10,010 --> 00:53:14,230 you're familiar with philosopher (?) Searle's Chinese room paradox 644 00:53:14,230 --> 00:53:17,620 from the 1970s, by which he attempted to prove that artificial intelligence was 645 00:53:17,620 --> 00:53:24,380 impossible, a corporation is very much the Chinese room implementation of an AI. It 646 00:53:24,380 --> 00:53:29,400 is a bunch of human beings in a box. You put inputs into the box, you get apples 647 00:53:29,400 --> 00:53:33,000 out of a box. Does it matter, whether it's all happening in software or whether 648 00:53:33,000 --> 00:53:37,520 there's a human being following rules inbetween to assemble the output? I don't 649 00:53:37,520 --> 00:53:41,190 see there being much of a difference. Now you have to look at a company at a 650 00:53:41,190 --> 00:53:46,630 very abstract level to view it as an AI, but more and more companies are automating 651 00:53:46,630 --> 00:53:54,500 their internal business processes. You've got to view this as an ongoing trend. And 652 00:53:54,500 --> 00:54:01,490 yeah, they have many of the characteristics of an AI. 653 00:54:01,490 --> 00:54:06,240 Herald: Okay mic number four. Mic 4: Hi, thanks for your talk. 654 00:54:06,240 --> 00:54:13,220 You probably heard of the Time Well Spent and Design Ethics movements that 655 00:54:13,220 --> 00:54:16,850 are alerting developers to dark patterns in UI design, where 656 00:54:16,850 --> 00:54:20,630 these people design apps to manipulate people. I'm curious if you find any 657 00:54:20,630 --> 00:54:26,750 optimism in the possibility of amplifying or promoting those movements. 658 00:54:26,750 --> 00:54:31,570 CS: Uhm, you know, I knew about dark patterns, I knew about people trying to 659 00:54:31,570 --> 00:54:36,230 optimize them, I wasn't actually aware there were movements against this. Okay I'm 660 00:54:36,230 --> 00:54:40,760 53 years old, I'm out of touch. I haven't actually done any serious programming in 661 00:54:40,760 --> 00:54:47,000 15 years. I'm so rusty, my rust has rust on it. But, you know, it is a worrying trend 662 00:54:47,000 --> 00:54:54,000 and actual activism is a good start. Raising awareness of hazards and of what 663 00:54:54,000 --> 00:54:58,330 we should be doing about them, is a good start. And I would classify this actually 664 00:54:58,330 --> 00:55:04,220 as a moral issue. We need to.. corporations evaluate everything in terms 665 00:55:04,220 --> 00:55:07,960 of revenue, because it's very equivalent to breathing, they have to 666 00:55:07,960 --> 00:55:13,810 breathe. Corporations don't usually have any moral framework. We're humans, we need 667 00:55:13,810 --> 00:55:18,480 a moral framework to operate within. Even if it's as simple as first "Do no harm!" 668 00:55:18,480 --> 00:55:22,500 or "Do not do unto others that which would be repugnant if it was done unto you!", 669 00:55:22,500 --> 00:55:26,450 the Golden Rule. So, yeah, we should be trying to spread awareness of this about 670 00:55:26,450 --> 00:55:32,170 and working with program developers, to look to remind them that they are human 671 00:55:32,170 --> 00:55:36,480 beings and have to be humane in their application of technology, is a necessary 672 00:55:36,480 --> 00:55:39,510 start. *applause* 673 00:55:39,510 --> 00:55:45,501 H: Thank you! Mic 3? Mic 3: Hi! Yeah, I think that folks, 674 00:55:45,501 --> 00:55:48,600 especially in this sort of crowd, tend to jump to the "just get off of 675 00:55:48,600 --> 00:55:52,330 Facebook"-solution first, for a lot of these things that are really, really 676 00:55:52,330 --> 00:55:56,640 scary. But what worries me, is how we sort of silence ourselves when we do that. 677 00:55:56,640 --> 00:56:02,240 After the election I actually got back on Facebook, because the Women's March was 678 00:56:02,240 --> 00:56:07,250 mostly organized through Facebook. But yeah, I think we need a lot more 679 00:56:07,250 --> 00:56:12,880 regulation, but we can't just throw it out. We're.. because it's.. 680 00:56:12,880 --> 00:56:16,520 social media is the only... really good platform we have right now 681 00:56:16,520 --> 00:56:20,410 to express ourselves, to have our rules, or power. 682 00:56:20,410 --> 00:56:24,660 CS: Absolutely. I have made a point of not really using Facebook 683 00:56:24,660 --> 00:56:28,480 for many, many, many years. I have a Facebook page simply to 684 00:56:28,480 --> 00:56:31,530 shut up the young marketing people at my publisher, who used to prop up every two 685 00:56:31,530 --> 00:56:34,680 years and say: "Why don't you have a Facebook. Everybody's got a Facebook." 686 00:56:34,680 --> 00:56:39,790 No, I've had a blog since 1993! *laughing* 687 00:56:39,790 --> 00:56:44,170 But no, I'm gonna have to use Facebook, because these days, not using Facebook is 688 00:56:44,170 --> 00:56:50,780 like not using email. You're cutting off your nose to spite your face. What we 689 00:56:50,780 --> 00:56:54,830 really do need to be doing, is looking for some form of effective oversight of 690 00:56:54,830 --> 00:57:00,830 Facebook and particularly, of how they.. the algorithms that show you content, are 691 00:57:00,830 --> 00:57:05,340 written. What I was saying earlier about how algorithms are not as transparent as 692 00:57:05,340 --> 00:57:10,530 human beings to people, applies hugely to them. And both, Facebook and Twitter 693 00:57:10,530 --> 00:57:15,180 control the information that they display to you. 694 00:57:15,180 --> 00:57:18,690 Herald: Okay, I'm terribly sorry for all the people queuing at the mics now, we're out 695 00:57:18,690 --> 00:57:24,960 of time. I also have to apologize, I announced, that this talk was being held in 696 00:57:24,960 --> 00:57:29,690 English, but it was being held in English. *the latter pronounced on the G* 697 00:57:29,690 --> 00:57:31,470 Thank you very much, Charles Stross! 698 00:57:31,470 --> 00:57:33,690 CS: Thank you very much for listening to me, it's been a pleasure! 699 00:57:33,690 --> 00:57:36,500 *applause* 700 00:57:36,500 --> 00:57:52,667 *postroll music* 701 00:57:52,667 --> 00:57:57,781 *subtitles created by c3subtitles.de in the year 2018*