[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:00.00,0:00:18.68,Default,,0000,0000,0000,,{\i1}35C3 preroll music{\i0}
Dialogue: 0,0:00:18.68,0:00:22.80,Default,,0000,0000,0000,,Herald Angel: Alright. Then it's my great\Npleasure to introduce Toni to you
Dialogue: 0,0:00:22.80,0:00:28.29,Default,,0000,0000,0000,,She's going to talk about "the Social Credit\NSystem," which is, kind of, feels to me
Dialogue: 0,0:00:28.29,0:00:36.01,Default,,0000,0000,0000,,like a Black Mirror episode coming to\Nlife. So, slightly nervous and really
Dialogue: 0,0:00:36.01,0:00:41.12,Default,,0000,0000,0000,,curious what we're going to learn today.\NSo please give a huge, warm round of
Dialogue: 0,0:00:41.12,0:00:45.61,Default,,0000,0000,0000,,applause and welcome Toni!\N{\i1}Applause{\i0}
Dialogue: 0,0:00:51.92,0:00:55.47,Default,,0000,0000,0000,,Toni: Good morning, everyone! Before I'm\Ngoing to be talking I'm going into my talk
Dialogue: 0,0:00:55.47,0:01:00.40,Default,,0000,0000,0000,,I'm just going to be presenting the\NChinese translation streams for everyone
Dialogue: 0,0:01:00.40,0:01:02.71,Default,,0000,0000,0000,,who doesn't speak English.
Dialogue: 0,0:01:02.71,0:01:17.33,Default,,0000,0000,0000,,{\i1}speaks chinese{\i0}
Dialogue: 0,0:01:17.33,0:01:23.25,Default,,0000,0000,0000,,{\i1}Applause{\i0}
Dialogue: 0,0:01:23.25,0:01:26.54,Default,,0000,0000,0000,,So because today's talk is about China we\Nfigured it would be good to have it
Dialogue: 0,0:01:26.54,0:01:31.78,Default,,0000,0000,0000,,in Chinese as well. And, I'm going to be\Ntalking today about
Dialogue: 0,0:01:31.78,0:01:35.44,Default,,0000,0000,0000,,the Social Credit system in China, where\N"the" Social Credit system that you
Dialogue: 0,0:01:35.44,0:01:38.97,Default,,0000,0000,0000,,always hear about in Western media\Nactually doesn't really exist
Dialogue: 0,0:01:38.97,0:01:43.47,Default,,0000,0000,0000,,and most of my talk will actually be\Ntalking about what all we don't know.
Dialogue: 0,0:01:43.47,0:01:47.03,Default,,0000,0000,0000,,Which could fill an entire hour or even
Dialogue: 0,0:01:47.03,0:01:51.47,Default,,0000,0000,0000,,more. But I'm just going to be focusing on\Nsome of the most interesting things for
Dialogue: 0,0:01:51.47,0:01:57.56,Default,,0000,0000,0000,,me. First of all, a little bit about me.\NI'm an economist, but I'm not I'm not only
Dialogue: 0,0:01:57.56,0:02:03.05,Default,,0000,0000,0000,,concerned with money. I'm kind of looking\Nat economy, at economics as the study of
Dialogue: 0,0:02:03.05,0:02:08.31,Default,,0000,0000,0000,,incentives, which means that what I'm\Nreally interested in is how humans respond
Dialogue: 0,0:02:08.31,0:02:12.48,Default,,0000,0000,0000,,to different kind of incentives. I don't\Nbelieve that humans are completely
Dialogue: 0,0:02:12.48,0:02:18.25,Default,,0000,0000,0000,,rational. But I do believe that humans do\Ntry to maximize what they think is their
Dialogue: 0,0:02:18.25,0:02:23.83,Default,,0000,0000,0000,,best interest. Now, some words about me: I\Nstudied math, economics and political
Dialogue: 0,0:02:23.83,0:02:29.19,Default,,0000,0000,0000,,science in a couple of different cities\Nall around the world. I spent overall 19
Dialogue: 0,0:02:29.19,0:02:35.14,Default,,0000,0000,0000,,months in China. Most recently I was there\Nin July on a government scholarship, which
Dialogue: 0,0:02:35.14,0:02:39.64,Default,,0000,0000,0000,,was really, really interesting, because\Nwhile there I read all of these Western
Dialogue: 0,0:02:39.64,0:02:43.71,Default,,0000,0000,0000,,newspaper articles about the Chinese\NSocial Credit system, and I went to a
Dialogue: 0,0:02:43.71,0:02:48.20,Default,,0000,0000,0000,,pretty good university and I asked them:\NSo what do you think about this system?
Dialogue: 0,0:02:48.20,0:02:52.36,Default,,0000,0000,0000,,And most of them basically looked at me\Nblankly, and were like: What system, I
Dialogue: 0,0:02:52.36,0:02:56.73,Default,,0000,0000,0000,,haven't even heard of this! So that was\Nkind of an interesting experience to me
Dialogue: 0,0:02:56.73,0:03:01.67,Default,,0000,0000,0000,,because in the West it's like this huge,\Nall-encompassing system. And in China,
Dialogue: 0,0:03:01.67,0:03:07.14,Default,,0000,0000,0000,,most people that aren't directly -- that\Naren't directly in touch with it actually
Dialogue: 0,0:03:07.14,0:03:11.32,Default,,0000,0000,0000,,don't know anything about this. I'm\Nbroadly interested in the impact of
Dialogue: 0,0:03:11.32,0:03:17.17,Default,,0000,0000,0000,,technology on society, life, and the\Neconomy, obviously, and in my free time I
Dialogue: 0,0:03:17.17,0:03:21.40,Default,,0000,0000,0000,,do a lot of data science and machine\Nlearning with Python and R. So, I thought
Dialogue: 0,0:03:21.40,0:03:27.18,Default,,0000,0000,0000,,it was quite interesting to look at the\NSocial Credit system, also from this point
Dialogue: 0,0:03:27.18,0:03:31.73,Default,,0000,0000,0000,,of view because you always heard that it's\Nlike this big data initiative, and then
Dialogue: 0,0:03:31.73,0:03:35.20,Default,,0000,0000,0000,,when coming down to it, what you actually\Nsee is that, they don't actually use
Dialogue: 0,0:03:35.20,0:03:40.02,Default,,0000,0000,0000,,machine learning all that much. They have,\Nbasically, a rule based catalog where, if
Dialogue: 0,0:03:40.02,0:03:44.46,Default,,0000,0000,0000,,you do this you get 50 points, if you do\Nthis you get 50 points, and then they
Dialogue: 0,0:03:44.46,0:03:48.96,Default,,0000,0000,0000,,actually have a lot of people that are\Nreporting on other people's behavior. I'm
Dialogue: 0,0:03:48.96,0:03:53.31,Default,,0000,0000,0000,,going to be talking about how exactly it\Nlooks, later on but I was very, very
Dialogue: 0,0:03:53.31,0:03:56.87,Default,,0000,0000,0000,,surprised after reading a lot of the\NWestern newspaper articles that were
Dialogue: 0,0:03:56.87,0:04:02.30,Default,,0000,0000,0000,,basically "Oh, this is this big dystopia,\NOrwellian, with big data working." And
Dialogue: 0,0:04:02.30,0:04:08.16,Default,,0000,0000,0000,,then, you read what's actually happening\Nand they have huge lists of "if you
Dialogue: 0,0:04:08.16,0:04:14.70,Default,,0000,0000,0000,,jaywalk, you get 10 points detracted from\Nyou," this kind of thing. If you want to
Dialogue: 0,0:04:14.70,0:04:19.00,Default,,0000,0000,0000,,get in touch with me you can use Twitter\Nbut you can also use different e-mails
Dialogue: 0,0:04:19.00,0:04:23.10,Default,,0000,0000,0000,,either my professional e-mail or my\Npersonal e-mail address, that you can both
Dialogue: 0,0:04:23.10,0:04:27.87,Default,,0000,0000,0000,,see there. If you have any thoughts on\Nthat or are interested in this a little
Dialogue: 0,0:04:27.87,0:04:31.76,Default,,0000,0000,0000,,more I can give you more resources as\Nwell, because obviously today's talk will
Dialogue: 0,0:04:31.76,0:04:38.42,Default,,0000,0000,0000,,only be scratching on the surface. So,\Nperceptions of the Social Credit System.
Dialogue: 0,0:04:38.42,0:04:43.31,Default,,0000,0000,0000,,One of the interesting things that I've\Ntalked about before was how, in the West
Dialogue: 0,0:04:43.31,0:04:48.41,Default,,0000,0000,0000,,and in China, the perception is completely\Ndifferent. So in the West, which is from
Dialogue: 0,0:04:48.41,0:04:54.55,Default,,0000,0000,0000,,financialtimes.com, you see this huge\Noverwhelming guy, and he basically puts
Dialogue: 0,0:04:54.55,0:04:59.28,Default,,0000,0000,0000,,every Chinese person under a microscope.\NThey're all kind of hunched over, and
Dialogue: 0,0:04:59.28,0:05:03.75,Default,,0000,0000,0000,,everyone has this score attached to them,\Nand they seem pretty sad and, like, very,
Dialogue: 0,0:05:03.75,0:05:08.70,Default,,0000,0000,0000,,very Orwellian concept. Whereas, in China,\Nthis is actually from a Chinese state
Dialogue: 0,0:05:08.70,0:05:13.81,Default,,0000,0000,0000,,media, and what it says is, well, we can\Nall live in harmony with this new system
Dialogue: 0,0:05:13.81,0:05:19.14,Default,,0000,0000,0000,,and all trust each other. And\Ninterestingly Chinese people actually
Dialogue: 0,0:05:19.14,0:05:24.31,Default,,0000,0000,0000,,believe that, to some degree. They believe\Nthat technology will fix all this current
Dialogue: 0,0:05:24.31,0:05:29.90,Default,,0000,0000,0000,,problems in society, especially because,\Nin China currently, trust is a rare
Dialogue: 0,0:05:29.90,0:05:39.40,Default,,0000,0000,0000,,commodity. And this new system will lead\Nto more efficiency and trust, and a better
Dialogue: 0,0:05:39.40,0:05:43.65,Default,,0000,0000,0000,,life. And I have a really, really\Ninteresting quote from a Western scholar,
Dialogue: 0,0:05:43.65,0:05:47.80,Default,,0000,0000,0000,,that really summarizes the Western\Nperspective: "What China is doing here is
Dialogue: 0,0:05:47.80,0:05:52.95,Default,,0000,0000,0000,,selectively breeding its population to\Nselect against the trait of critical,
Dialogue: 0,0:05:52.95,0:05:57.49,Default,,0000,0000,0000,,independent thinking. This may not be the\Npurpose, indeed I doubt it's the primary
Dialogue: 0,0:05:57.49,0:06:02.33,Default,,0000,0000,0000,,purpose, but it's nevertheless the effect\Nof giving only obedient people the social
Dialogue: 0,0:06:02.33,0:06:07.29,Default,,0000,0000,0000,,ability to have children, not to mention\Nsuccessful children." This, basically,
Dialogue: 0,0:06:07.29,0:06:11.96,Default,,0000,0000,0000,,plays with the idea that if you have a low\Nscore, currently, in the cities that are
Dialogue: 0,0:06:11.96,0:06:17.32,Default,,0000,0000,0000,,already testing this system, what happens\Nis, your children can't attend good
Dialogue: 0,0:06:17.32,0:06:23.19,Default,,0000,0000,0000,,schools. What happens is, you cannot take\Ntrains, you cannot take planes. You cannot
Dialogue: 0,0:06:23.19,0:06:31.14,Default,,0000,0000,0000,,book good hotels. Your life is just very,\Nvery inconvenient. And this is by design.
Dialogue: 0,0:06:31.14,0:06:38.62,Default,,0000,0000,0000,,This is kind of the plan. The Chinese\Ngovernment, they say it's a little
Dialogue: 0,0:06:38.62,0:06:42.34,Default,,0000,0000,0000,,different, the idea is about changing\Npeople's conduct by ensuring they are
Dialogue: 0,0:06:42.34,0:06:48.20,Default,,0000,0000,0000,,closely associated with it. One of the\Nmain things about this system is, there
Dialogue: 0,0:06:48.20,0:06:53.33,Default,,0000,0000,0000,,isn't very much new data being generated\Nfor the system. Instead, what's happening
Dialogue: 0,0:06:53.33,0:07:00.01,Default,,0000,0000,0000,,is, all the existing data that is already\Ncollected about you is, basically,
Dialogue: 0,0:07:00.01,0:07:06.37,Default,,0000,0000,0000,,combined into one big database for each\Nand every person by your ID number. So, in
Dialogue: 0,0:07:06.37,0:07:10.19,Default,,0000,0000,0000,,China, once you're born, you get an ID\Nnumber, which is similar to a Social
Dialogue: 0,0:07:10.19,0:07:13.96,Default,,0000,0000,0000,,Security number in the U.S. We don't\Nreally have a similar concept in Germany,
Dialogue: 0,0:07:13.96,0:07:19.31,Default,,0000,0000,0000,,and it used to be that your ID number was\Nonly necessary for public -- like for
Dialogue: 0,0:07:19.31,0:07:25.81,Default,,0000,0000,0000,,government stuff, but now you need your ID\Nnumber for getting a bank account, you
Dialogue: 0,0:07:25.81,0:07:29.58,Default,,0000,0000,0000,,need your ID number for buying a cell\Nphone, even if it's a prepaid cell phone,
Dialogue: 0,0:07:29.58,0:07:34.64,Default,,0000,0000,0000,,you still need your ID number. So all your\Nonline activity that happens with your
Dialogue: 0,0:07:34.64,0:07:38.43,Default,,0000,0000,0000,,cell phone is associated with your ID\Nnumber, which means you can't really do
Dialogue: 0,0:07:38.43,0:07:44.21,Default,,0000,0000,0000,,anything anonymously, because it's all\Ngoing back to your ID number. There's a
Dialogue: 0,0:07:44.21,0:07:49.45,Default,,0000,0000,0000,,couple of predecessors, some of them going\Nactually back to the 1990s, that are
Dialogue: 0,0:07:49.45,0:07:54.88,Default,,0000,0000,0000,,supposed to be integrated into the new\Nsystem. One of them, or like two of them
Dialogue: 0,0:07:54.88,0:08:00.80,Default,,0000,0000,0000,,are blacklists. One of them is a court\Nblacklist. So in China, courts work a
Dialogue: 0,0:08:00.80,0:08:06.56,Default,,0000,0000,0000,,little bit differently. They tend to like\Ngiving you fines, as they do in other
Dialogue: 0,0:08:06.56,0:08:11.52,Default,,0000,0000,0000,,countries, but they also like giving you\N"apologies to do." So one of the things,
Dialogue: 0,0:08:11.52,0:08:17.47,Default,,0000,0000,0000,,if you do something, for example you're a\Ncompany, your food safety wasn't up to par
Dialogue: 0,0:08:17.47,0:08:21.80,Default,,0000,0000,0000,,– you have to pay a fine. But in addition\Nto this fine you also have to write a
Dialogue: 0,0:08:21.80,0:08:27.15,Default,,0000,0000,0000,,public apology letter in the newspaper,\Nhow you are very sorry that this happened
Dialogue: 0,0:08:27.15,0:08:31.56,Default,,0000,0000,0000,,and it won't happen again, and it was a\Nmoral failing on your part, and it won't
Dialogue: 0,0:08:31.56,0:08:37.46,Default,,0000,0000,0000,,happen again. And if you don't do that,\Nyou go on this blacklist. Similarly, if
Dialogue: 0,0:08:37.46,0:08:42.70,Default,,0000,0000,0000,,you take out a line of credit and don't\Npay it back within three months, or like
Dialogue: 0,0:08:42.70,0:08:47.93,Default,,0000,0000,0000,,don't don't do any payments for three\Nmonths, you go on this debtors blacklist.
Dialogue: 0,0:08:47.93,0:08:52.01,Default,,0000,0000,0000,,If you're on this blacklist, which again\Nis associated with your {\i1}shēnfènzhèng{\i0}, so
Dialogue: 0,0:08:52.01,0:08:57.71,Default,,0000,0000,0000,,your ID number – what happens is you\Ncannot take trains you cannot take planes.
Dialogue: 0,0:08:57.71,0:09:02.31,Default,,0000,0000,0000,,Your life basically becomes very very\Ninconvenient, your children can't go to
Dialogue: 0,0:09:02.31,0:09:06.75,Default,,0000,0000,0000,,good public schools, your children can't\Ngo to private schools, your children can't
Dialogue: 0,0:09:06.75,0:09:14.59,Default,,0000,0000,0000,,go to universities, all of these issues\Nare suddenly coming up. There is also a
Dialogue: 0,0:09:14.59,0:09:19.83,Default,,0000,0000,0000,,company database that's called Credit\NChina which is basically similar to the
Dialogue: 0,0:09:19.83,0:09:24.96,Default,,0000,0000,0000,,public debtors blacklist but it's\Nbasically a credit system a credit score
Dialogue: 0,0:09:24.96,0:09:28.86,Default,,0000,0000,0000,,for companies. And then there's the credit\Nreference center of the People's Bank of
Dialogue: 0,0:09:28.86,0:09:34.57,Default,,0000,0000,0000,,China which is a credit score. It was\Nsupposed to be like Schufa or like the
Dialogue: 0,0:09:34.57,0:09:42.63,Default,,0000,0000,0000,,U.S. FICO for individuals. But one of the\Nbig problems in China is that there are a
Dialogue: 0,0:09:42.63,0:09:46.77,Default,,0000,0000,0000,,lot of people that aren't part of the\Nformal economy. A lot of people are
Dialogue: 0,0:09:46.77,0:09:52.53,Default,,0000,0000,0000,,migrant workers. They get their money in\Ncash. They do not have bank accounts. They
Dialogue: 0,0:09:52.53,0:10:00.41,Default,,0000,0000,0000,,do not have anything… they do not have rent\Nor utilities or anything like this because
Dialogue: 0,0:10:00.41,0:10:04.61,Default,,0000,0000,0000,,they live in the country. So they own\Ntheir own home which they built themselves
Dialogue: 0,0:10:04.61,0:10:10.92,Default,,0000,0000,0000,,so they didn't even finance it and their\Nhome isn't officially theirs because in
Dialogue: 0,0:10:10.92,0:10:16.07,Default,,0000,0000,0000,,China you can't actually own property.\NInstead the government leases it to you.
Dialogue: 0,0:10:16.07,0:10:20.84,Default,,0000,0000,0000,,So there were a lot of people that were\Nnot covered in this system, and I think
Dialogue: 0,0:10:20.84,0:10:26.74,Default,,0000,0000,0000,,the last data that I had was that less\Nthan 10 percent of Chinese adult citizens
Dialogue: 0,0:10:26.74,0:10:32.70,Default,,0000,0000,0000,,were actually in the system and had any\Nsort of exposure to banks, which is very,
Dialogue: 0,0:10:32.70,0:10:38.09,Default,,0000,0000,0000,,very little. And that meant that people\Ncouldn't get credit because banks would
Dialogue: 0,0:10:38.09,0:10:42.22,Default,,0000,0000,0000,,only give credit to people that were in\Nthe system or people where they had
Dialogue: 0,0:10:42.22,0:10:46.65,Default,,0000,0000,0000,,some sort of handling on whether they\Nwould be paid back. Now, the
Dialogue: 0,0:10:46.65,0:10:51.84,Default,,0000,0000,0000,,implementation details of the new system\Nare very very scarce, but the basic idea
Dialogue: 0,0:10:51.84,0:10:56.08,Default,,0000,0000,0000,,is that Chinese citizens are divided into\Ntrustworthy individuals and what the
Dialogue: 0,0:10:56.08,0:11:02.31,Default,,0000,0000,0000,,Chinese call "trust breakers". Sometimes\Nyou have five different groups, sometimes
Dialogue: 0,0:11:02.31,0:11:05.46,Default,,0000,0000,0000,,you have two different groups, but in\Ngeneral there's sort of this cut-off:
Dialogue: 0,0:11:05.46,0:11:13.62,Default,,0000,0000,0000,,above this line it's good and beyond this\Nline it's bad. This is one graphic from
Dialogue: 0,0:11:13.62,0:11:20.41,Default,,0000,0000,0000,,the Wall Street Journal that just shows\Nsome of the inputs that go into the
Dialogue: 0,0:11:20.41,0:11:25.44,Default,,0000,0000,0000,,system. And one of the things that we see\Nis that the inputs are _crazy_ crazy
Dialogue: 0,0:11:25.44,0:11:35.01,Default,,0000,0000,0000,,varied. So it is: do you pay income taxes?\NDo you pay your utility bills on time?
Dialogue: 0,0:11:35.01,0:11:40.20,Default,,0000,0000,0000,,Do you respect your parents? However they\Nmeasure that. Do you have a criminal
Dialogue: 0,0:11:40.20,0:11:46.68,Default,,0000,0000,0000,,record? Do you pay for public\Ntransportation or have you been caught
Dialogue: 0,0:11:46.68,0:11:50.48,Default,,0000,0000,0000,,not paying? What about your friends?
Dialogue: 0,0:11:50.48,0:11:58.44,Default,,0000,0000,0000,,Do you retweet or use WeChat\Nto distribute sort of information against
Dialogue: 0,0:11:58.44,0:12:04.30,Default,,0000,0000,0000,,the party, which they call reliability. In\Nactuality it's not about whether it's
Dialogue: 0,0:12:04.30,0:12:08.46,Default,,0000,0000,0000,,factual, it's about whether it's against\Nthe party or not. Where do you buy and
Dialogue: 0,0:12:08.46,0:12:12.02,Default,,0000,0000,0000,,what do you buy, apparently if you buy\Ndiapers it's better than if you buy
Dialogue: 0,0:12:12.02,0:12:16.68,Default,,0000,0000,0000,,videogames. For your score. Because you\Nknow if you buy videogames obviously
Dialogue: 0,0:12:16.68,0:12:20.09,Default,,0000,0000,0000,,you're not very responsible. And if you\Nbuy diapers you have a kid, you are sort
Dialogue: 0,0:12:20.09,0:12:29.02,Default,,0000,0000,0000,,of conforming to the societal ideal. And\Nthen your score is supposed to go into all
Dialogue: 0,0:12:29.02,0:12:33.68,Default,,0000,0000,0000,,these different categories, you're\Nsupposed to have better access to social
Dialogue: 0,0:12:33.68,0:12:38.76,Default,,0000,0000,0000,,services if your score is good. You're\Nsupposed to have better access to internet
Dialogue: 0,0:12:38.76,0:12:43.31,Default,,0000,0000,0000,,services. So in theory the idea is that at\None point if your score is too bad, you're
Dialogue: 0,0:12:43.31,0:12:47.89,Default,,0000,0000,0000,,not allowed to use WeChat anymore. You're\Nnot allowed to use Alibaba anymore. You
Dialogue: 0,0:12:47.89,0:12:57.11,Default,,0000,0000,0000,,can't become a government worker. You can\Nnot take planes and high speed trains. You
Dialogue: 0,0:12:57.11,0:13:04.52,Default,,0000,0000,0000,,can not get a passport. And your insurance\Npremiums will go up. So it's supposed to
Dialogue: 0,0:13:04.52,0:13:12.54,Default,,0000,0000,0000,,be this really really big, overwhelming\Nsystem. But in actuality what they say
Dialogue: 0,0:13:12.54,0:13:19.46,Default,,0000,0000,0000,,their stated goals are, is "it's a\Nshorthand for a broad range of efforts to
Dialogue: 0,0:13:19.46,0:13:24.79,Default,,0000,0000,0000,,improve market security and public safety\Nby increasing integrity and mutual trust
Dialogue: 0,0:13:24.79,0:13:31.90,Default,,0000,0000,0000,,in society." So one idea is to allocate\Nresources more efficiently. Resource
Dialogue: 0,0:13:31.90,0:13:39.34,Default,,0000,0000,0000,,allocation in China is a pretty big\Nproblem, because people grow up with:
Dialogue: 0,0:13:39.34,0:13:43.94,Default,,0000,0000,0000,,There's 1.3 billion people. So there's -\Nit's always going to be scarce. And a lot
Dialogue: 0,0:13:43.94,0:13:50.16,Default,,0000,0000,0000,,of stuff is – people grow up with this\Nidea that it's just very very scarce, and
Dialogue: 0,0:13:50.16,0:13:54.64,Default,,0000,0000,0000,,current distribution strategies, which are\Nmostly financially based but also often
Dialogue: 0,0:13:54.64,0:13:59.36,Default,,0000,0000,0000,,{\i1}guanxi{\i0}-based, don't really seem fair. For\Nexample, public transport in China is
Dialogue: 0,0:13:59.36,0:14:04.77,Default,,0000,0000,0000,,highly subsidized, which means that the\Nprice does not reflect whether – does not
Dialogue: 0,0:14:04.77,0:14:10.25,Default,,0000,0000,0000,,reflect true scarcity. So currently the\Nway it works is in theory it's first come
Dialogue: 0,0:14:10.25,0:14:15.02,Default,,0000,0000,0000,,first serve, in practice there's people\Nthat are buying up all the tickets for,
Dialogue: 0,0:14:15.02,0:14:19.32,Default,,0000,0000,0000,,for example, the high speed train from\NShanghai to Beijing and then selling it at
Dialogue: 0,0:14:19.32,0:14:23.12,Default,,0000,0000,0000,,a profit, or selling it to certain\Ncompanies that have good ties to the
Dialogue: 0,0:14:23.12,0:14:28.04,Default,,0000,0000,0000,,government. That seems very unfair. So the\Nnew system is supposed to distribute them
Dialogue: 0,0:14:28.04,0:14:33.67,Default,,0000,0000,0000,,more fairly and more efficiently. The\Nother thing is restoring trust in people.
Dialogue: 0,0:14:33.67,0:14:38.21,Default,,0000,0000,0000,,Perceived inter-personal trust and trust\Nin institutions is extremely low in China.
Dialogue: 0,0:14:38.21,0:14:43.12,Default,,0000,0000,0000,,If you're from Germany, you might have\Nheard that there is Chinese gangs
Dialogue: 0,0:14:43.12,0:14:48.09,Default,,0000,0000,0000,,basically buying up German milk powder and\Nselling it in China. This is actually
Dialogue: 0,0:14:48.09,0:14:54.56,Default,,0000,0000,0000,,happening, because in 2008 there was a big\Nscandal with laced milk powder. And ever
Dialogue: 0,0:14:54.56,0:14:59.38,Default,,0000,0000,0000,,since then, anyone who can afford it does\Nnot use Chinese milk powder, because they
Dialogue: 0,0:14:59.38,0:15:03.73,Default,,0000,0000,0000,,don't trust the government, or the\Nregulations, the firms, enough to buy
Dialogue: 0,0:15:03.73,0:15:09.93,Default,,0000,0000,0000,,Chinese milk powder so they are actually\Nimporting this. And the big irony is:
Dialogue: 0,0:15:09.93,0:15:15.52,Default,,0000,0000,0000,,sometimes this milk powder is produced in\NChina, exported to Germany, and then
Dialogue: 0,0:15:15.52,0:15:22.25,Default,,0000,0000,0000,,exported back to China. The Social Credit\Nsystem is then supposed to identify those
Dialogue: 0,0:15:22.25,0:15:27.45,Default,,0000,0000,0000,,that deserve the trust. And the third\Npoint is sort of a reeducation of people.
Dialogue: 0,0:15:27.45,0:15:32.96,Default,,0000,0000,0000,,The idea is: they want to make people in\Nthe image that the Communist Party thinks
Dialogue: 0,0:15:32.96,0:15:40.53,Default,,0000,0000,0000,,people should be. And one additional way\Nto the punishments and rewards this could
Dialogue: 0,0:15:40.53,0:15:45.42,Default,,0000,0000,0000,,work, is the feeling of being surveyed.\NBecause you can't do anything anonymously,
Dialogue: 0,0:15:45.42,0:15:49.47,Default,,0000,0000,0000,,you will automatically adapt your behavior\Nbecause you know someone is watching you
Dialogue: 0,0:15:49.47,0:15:55.50,Default,,0000,0000,0000,,all the time, and this is how a lot of the\NChinese firewall actually works, because
Dialogue: 0,0:15:55.50,0:16:00.92,Default,,0000,0000,0000,,most people I know that are sort of more–\Nmore educated, they know ways to
Dialogue: 0,0:16:00.92,0:16:05.02,Default,,0000,0000,0000,,circumvent the Chinese firewall, but they\Nalso know that they're always being
Dialogue: 0,0:16:05.02,0:16:09.40,Default,,0000,0000,0000,,watched, so they don't do that because,\Nyou know, they're being watched, so they
Dialogue: 0,0:16:09.40,0:16:18.45,Default,,0000,0000,0000,,self– they censorship– they censor\Nthemselves. As I said before, allocation
Dialogue: 0,0:16:18.45,0:16:23.62,Default,,0000,0000,0000,,of scarce resources so far is mainly\Nthrough financial {\i1}guanxi{\i0} channels. {\i1}guanxi{\i0}
Dialogue: 0,0:16:23.62,0:16:29.29,Default,,0000,0000,0000,,is basically an all permeating network of\Nrelationships with a clear status
Dialogue: 0,0:16:29.29,0:16:34.40,Default,,0000,0000,0000,,hierarchy. So if I attend a school,\Neveryone who also attended this school
Dialogue: 0,0:16:34.40,0:16:40.32,Default,,0000,0000,0000,,will be sort of in my {\i1}guanxi{\i0} network. And\Nthere's this idea that we will have a
Dialogue: 0,0:16:40.32,0:16:45.68,Default,,0000,0000,0000,,system where we are all in-group, and in-\Ngroup we trust each other and we do favors
Dialogue: 0,0:16:45.68,0:16:51.12,Default,,0000,0000,0000,,for each other, and everyone who's outside\Nof my immediate group I don't trust and I
Dialogue: 0,0:16:51.12,0:16:56.35,Default,,0000,0000,0000,,don't do favors for. And in some ways the\N{\i1}guanxi{\i0} system right now is a substitute
Dialogue: 0,0:16:56.35,0:17:02.64,Default,,0000,0000,0000,,for formal institutions in China. For\Nexample if you want a passport right now.
Dialogue: 0,0:17:02.64,0:17:07.36,Default,,0000,0000,0000,,You can of course apply for passports\Nthrough regular channels, which might take
Dialogue: 0,0:17:07.36,0:17:12.39,Default,,0000,0000,0000,,months and months. Or you can apply for a\Npassport through knowing someone and
Dialogue: 0,0:17:12.39,0:17:16.75,Default,,0000,0000,0000,,knowing someone, which might take only two\Ndays. Whereas in Germany you have these
Dialogue: 0,0:17:16.75,0:17:22.51,Default,,0000,0000,0000,,very regular, formal institutions, in\NChina they still use {\i1}guanxi{\i0}. But,
Dialogue: 0,0:17:22.51,0:17:26.85,Default,,0000,0000,0000,,increasingly especially young people find\Nthat {\i1}guanxi{\i0} are very unfair, because a lot
Dialogue: 0,0:17:26.85,0:17:31.56,Default,,0000,0000,0000,,of these are: where you went to school,\Nwhich is determined by where you're born,
Dialogue: 0,0:17:31.56,0:17:37.20,Default,,0000,0000,0000,,who your parents are, and all these\Nthings. Another thing that's important to
Dialogue: 0,0:17:37.20,0:17:42.17,Default,,0000,0000,0000,,understand because: the system works\Nthrough public shaming. And in a lot of
Dialogue: 0,0:17:42.17,0:17:48.43,Default,,0000,0000,0000,,western society we can't really imagine\Nthat, like, I wouldn't really care if my
Dialogue: 0,0:17:48.43,0:17:54.70,Default,,0000,0000,0000,,name was in a newspaper of someone who\Njaywalked for example. It would be: oh
Dialogue: 0,0:17:54.70,0:17:59.63,Default,,0000,0000,0000,,well, that's okay. But in China this is\Nactually a very very serious thing. So
Dialogue: 0,0:17:59.63,0:18:05.03,Default,,0000,0000,0000,,saving face is very very important in\NChina. And when I went to school there I
Dialogue: 0,0:18:05.03,0:18:13.08,Default,,0000,0000,0000,,actually – we had this dormitory, and it\Nwas an all foreigners dormitory, where the
Dialogue: 0,0:18:13.08,0:18:17.68,Default,,0000,0000,0000,,staff that were responsible for the\Ndormitory felt that foreigners were not
Dialogue: 0,0:18:17.68,0:18:24.66,Default,,0000,0000,0000,,behaving in the way they should. So their\Nidea was to put the names, the pictures,
Dialogue: 0,0:18:24.66,0:18:29.25,Default,,0000,0000,0000,,and the offenses of the foreigners in the\Nelevator to shame them publicly. So for
Dialogue: 0,0:18:29.25,0:18:33.69,Default,,0000,0000,0000,,example if you brought a person of the\Nopposite sex to your room, they would put
Dialogue: 0,0:18:33.69,0:18:39.91,Default,,0000,0000,0000,,your name, your offense and your room\Nnumber in the elevator. And of course this
Dialogue: 0,0:18:39.91,0:18:44.63,Default,,0000,0000,0000,,didn't work because for a lot of western\Npeople it was basically like: "oh well I'm
Dialogue: 0,0:18:44.63,0:18:49.14,Default,,0000,0000,0000,,going to try to be there as often as\Npossible because this is like a badge of
Dialogue: 0,0:18:49.14,0:18:53.67,Default,,0000,0000,0000,,honor for me" and the Chinese people they\Nfigured "well this is really really shame
Dialogue: 0,0:18:53.67,0:18:58.08,Default,,0000,0000,0000,,and I'm losing my face". She brought\Nalcohol. So this didn't really work at
Dialogue: 0,0:18:58.08,0:19:06.39,Default,,0000,0000,0000,,all. But this is kind of the mindset that\Nis behind a lot of these initiatives. As I
Dialogue: 0,0:19:06.39,0:19:12.23,Default,,0000,0000,0000,,said there's a lot of problems with – we\Ndon't really know what's going to happen.
Dialogue: 0,0:19:12.23,0:19:17.00,Default,,0000,0000,0000,,And one of the ways that we can see what\Nmight happen is actually to look at pilot
Dialogue: 0,0:19:17.00,0:19:24.00,Default,,0000,0000,0000,,systems. China has – or like ever since\Nthe Communist Party took hold – the
Dialogue: 0,0:19:24.00,0:19:28.63,Default,,0000,0000,0000,,Chinese government has tried a lot of\Npolicy experimentation. So whenever they
Dialogue: 0,0:19:28.63,0:19:33.85,Default,,0000,0000,0000,,try a new policy, they don't roll it out\Nall over, but they choose different pilot
Dialogue: 0,0:19:33.85,0:19:38.89,Default,,0000,0000,0000,,cities or pilot districts, and then they\Nchoose "oh well this is the district where
Dialogue: 0,0:19:38.89,0:19:42.90,Default,,0000,0000,0000,,I'm going to be trying this system and I'm\Ngoing to be trying another system in
Dialogue: 0,0:19:42.90,0:19:48.36,Default,,0000,0000,0000,,another district or city". And this is\Nalso what they did for the, or what
Dialogue: 0,0:19:48.36,0:19:52.52,Default,,0000,0000,0000,,they're doing for the Social Credit\Nsystem. Now I have three systems that I
Dialogue: 0,0:19:52.52,0:19:58.30,Default,,0000,0000,0000,,looked at intensively for this\Npresentation, overall there's about 70
Dialogue: 0,0:19:58.30,0:20:07.03,Default,,0000,0000,0000,,that I know of - the Suining system,\NSuining is a city in China, the Rongcheng
Dialogue: 0,0:20:07.03,0:20:11.41,Default,,0000,0000,0000,,system, another city in China and Sesame\NCredit. Sesame Credit is a commercial
Dialogue: 0,0:20:11.41,0:20:16.43,Default,,0000,0000,0000,,system from Alibaba - I assume everyone\Nknows Alibaba, the're basically the
Dialogue: 0,0:20:16.43,0:20:22.77,Default,,0000,0000,0000,,Chinese Amazon, except they're bigger and\Nhave more users and make more money,
Dialogue: 0,0:20:22.77,0:20:27.40,Default,,0000,0000,0000,,actually. And they have their own little\Nsystem. One of the problems with this kind
Dialogue: 0,0:20:27.40,0:20:31.62,Default,,0000,0000,0000,,of system that I found when I tried\Nmodeling it, was that it's a very very
Dialogue: 0,0:20:31.62,0:20:38.93,Default,,0000,0000,0000,,complex system and small changes in input\Nactually changed the output significantly.
Dialogue: 0,0:20:38.93,0:20:43.55,Default,,0000,0000,0000,,So when they try– usually when they try\Nthis pilot system they basically have a
Dialogue: 0,0:20:43.55,0:20:47.85,Default,,0000,0000,0000,,couple of pilots, then they choose the\Npilot that is best and they roll it out
Dialogue: 0,0:20:47.85,0:20:52.56,Default,,0000,0000,0000,,all over. But for this kind of thing,\Nwhere you have a lot of complex issues, it
Dialogue: 0,0:20:52.56,0:20:59.41,Default,,0000,0000,0000,,might not be the best way to do that. The\NSuining system is actually considered the
Dialogue: 0,0:20:59.41,0:21:05.98,Default,,0000,0000,0000,,predecessor of all current systems. It had\Na focus on punishment, and it was quite
Dialogue: 0,0:21:05.98,0:21:12.00,Default,,0000,0000,0000,,interesting. At the beginning of the trial\Nperiod they published a catalogue of
Dialogue: 0,0:21:12.00,0:21:17.26,Default,,0000,0000,0000,,scores and consequences. Here is an\Nexample. This is basically taken from this
Dialogue: 0,0:21:17.26,0:21:22.64,Default,,0000,0000,0000,,catalog. So if you took out bank loans and\Ndidn't repay them, you got deducted 50
Dialogue: 0,0:21:22.64,0:21:26.91,Default,,0000,0000,0000,,points. Everyone started with 1000\Npoints for this system. If you didn't pay
Dialogue: 0,0:21:26.91,0:21:32.69,Default,,0000,0000,0000,,back your credit cards you also got\Ndeducted 50 points. If you evaded taxes,
Dialogue: 0,0:21:32.69,0:21:43.89,Default,,0000,0000,0000,,also 50 points. If you sold fake goods, 35\Npoints were deducted. And actually the
Dialogue: 0,0:21:43.89,0:21:50.45,Default,,0000,0000,0000,,system was abolished I think in 2015,\N2016, because all the Chinese state media
Dialogue: 0,0:21:50.45,0:21:56.49,Default,,0000,0000,0000,,and also a lot of Internet citizens talked\Nabout how it's an Orwellian system and how
Dialogue: 0,0:21:56.49,0:22:03.02,Default,,0000,0000,0000,,it's not a good system, because it's all\Nvery centralized and everything that you
Dialogue: 0,0:22:03.02,0:22:08.88,Default,,0000,0000,0000,,do is basically recorded centrally. But\NCreemers writes: "Nonetheless, the Suining
Dialogue: 0,0:22:08.88,0:22:13.09,Default,,0000,0000,0000,,system already contained the embryonic\Nforms of several elements of subsequent
Dialogue: 0,0:22:13.09,0:22:17.29,Default,,0000,0000,0000,,social credit initiatives: The notion of\Ndisproportional disincentives against rule
Dialogue: 0,0:22:17.29,0:22:22.50,Default,,0000,0000,0000,,breaking, public naming and shaming of\Nwrongdoers, and most importantly, the
Dialogue: 0,0:22:22.50,0:22:26.90,Default,,0000,0000,0000,,expansion of the credit mechanism outside\Nof the market economic context, also
Dialogue: 0,0:22:26.90,0:22:29.76,Default,,0000,0000,0000,,encompassing compliance with\Nadministrative regulations and urban
Dialogue: 0,0:22:29.76,0:22:34.30,Default,,0000,0000,0000,,management rules." So one of the things\Nthat is difficult for especially German
Dialogue: 0,0:22:34.30,0:22:40.19,Default,,0000,0000,0000,,speakers is that credit in Chinese,\N{\i1}xìnyòng{\i0}, means credit as in "loan", but
Dialogue: 0,0:22:40.19,0:22:46.35,Default,,0000,0000,0000,,also means credit as in "trust". So the\NSocial Credit System is one way of trying
Dialogue: 0,0:22:46.35,0:22:52.24,Default,,0000,0000,0000,,to conflate those two – the economic\Ncredit and the trust credit – into one big
Dialogue: 0,0:22:52.24,0:23:01.67,Default,,0000,0000,0000,,system. But the Suining system basically\Nfailed. So, they adapted the system and
Dialogue: 0,0:23:01.67,0:23:07.22,Default,,0000,0000,0000,,are now practicing a new kind of system,\Nthe Rongcheng system. Whenever you read a
Dialogue: 0,0:23:07.22,0:23:11.31,Default,,0000,0000,0000,,newspaper article on the social credit\Nsystem in the west, most people went to
Dialogue: 0,0:23:11.31,0:23:15.06,Default,,0000,0000,0000,,Rongcheng because they just received a\Ncouple of awards from the Chinese
Dialogue: 0,0:23:15.06,0:23:22.23,Default,,0000,0000,0000,,government for being so advanced at this\Nsocial credit thing. But it's very
Dialogue: 0,0:23:22.23,0:23:25.99,Default,,0000,0000,0000,,difficult to call this "one system"\Nbecause there's actually many many
Dialogue: 0,0:23:25.99,0:23:31.02,Default,,0000,0000,0000,,intertwined systems. There is one city\Nlevel system, where city level offenses
Dialogue: 0,0:23:31.02,0:23:38.52,Default,,0000,0000,0000,,are recorded. For example tax evasion, and\Nthere's a couple of rules. If you evade
Dialogue: 0,0:23:38.52,0:23:43.73,Default,,0000,0000,0000,,taxes your score goes down 50. But then if\Nyou live in one neighborhood your score
Dialogue: 0,0:23:43.73,0:23:47.62,Default,,0000,0000,0000,,might go up for volunteering with the\Nelderly. If you live in another
Dialogue: 0,0:23:47.62,0:23:54.20,Default,,0000,0000,0000,,neighborhood your score might go up for,\Nfor example, planting some trees in
Dialogue: 0,0:23:54.20,0:23:59.18,Default,,0000,0000,0000,,your garden or backyard. So depending on\Nyour neighborhood, your score might be
Dialogue: 0,0:23:59.18,0:24:05.59,Default,,0000,0000,0000,,different. If you work for a– if you work\Nfor a taxi cab company, for example, they
Dialogue: 0,0:24:05.59,0:24:10.78,Default,,0000,0000,0000,,also have their own little score system\Nand your score might go up if you get good
Dialogue: 0,0:24:10.78,0:24:16.57,Default,,0000,0000,0000,,reviews from your drive…, from your\Npassengers. Your score might go down if
Dialogue: 0,0:24:16.57,0:24:25.78,Default,,0000,0000,0000,,you don't follow traffic rules, these\Nkinds of things. There are designated
Dialogue: 0,0:24:25.78,0:24:31.55,Default,,0000,0000,0000,,scorekeepers at each level. So, each\Ndistrict chooses a couple of people who
Dialogue: 0,0:24:31.55,0:24:38.33,Default,,0000,0000,0000,,are responsible for passing on the\Ninformation to the next higher level,
Dialogue: 0,0:24:38.33,0:24:42.87,Default,,0000,0000,0000,,about who did what. There is supposed to\Nbe an official appeals procedure, so
Dialogue: 0,0:24:42.87,0:24:46.94,Default,,0000,0000,0000,,whenever you score changes you're supposed\Nto be notified, but apparently that's not
Dialogue: 0,0:24:46.94,0:24:54.49,Default,,0000,0000,0000,,happening at this point for most people.\NAgain, it's a system of data sharing, and
Dialogue: 0,0:24:54.49,0:24:58.52,Default,,0000,0000,0000,,one thing that they haven't really\Ndisclosed yet is what kind of data is
Dialogue: 0,0:24:58.52,0:25:04.76,Default,,0000,0000,0000,,shared. Are they only sharing the points,\Nso if I'm in a district and I plant some
Dialogue: 0,0:25:04.76,0:25:11.48,Default,,0000,0000,0000,,trees, does the central system get the\Ninformation "person A planted some trees,"
Dialogue: 0,0:25:11.48,0:25:16.30,Default,,0000,0000,0000,,or does the central system get the\Ninformation "person A got 5 points?" We
Dialogue: 0,0:25:16.30,0:25:21.57,Default,,0000,0000,0000,,don't know at this point. And it would\Nmean something very different for how the
Dialogue: 0,0:25:21.57,0:25:27.03,Default,,0000,0000,0000,,system could be used. But still the end\Nresult, at this point, is that there's one
Dialogue: 0,0:25:27.03,0:25:31.36,Default,,0000,0000,0000,,score. So you have one central score and\Nit's kind of– there's all these different
Dialogue: 0,0:25:31.36,0:25:35.78,Default,,0000,0000,0000,,smaller systems that go into this score.\NBut at the end, everyone has one central
Dialogue: 0,0:25:35.78,0:25:44.53,Default,,0000,0000,0000,,score, and currently about 85 percent of\Npeople are between 950 and 1050. So you
Dialogue: 0,0:25:44.53,0:25:49.99,Default,,0000,0000,0000,,start off with a thousand – and those are\Nbasically the normal people – and then
Dialogue: 0,0:25:49.99,0:25:58.94,Default,,0000,0000,0000,,anyone above a 1050 is considered a\Ntrustworthy person, and anyone below 1050
Dialogue: 0,0:25:58.94,0:26:05.34,Default,,0000,0000,0000,,is considered a trust-breaker. And, as\NI've said before, with the naming and
Dialogue: 0,0:26:05.34,0:26:12.31,Default,,0000,0000,0000,,shaming and all these things, what you can\Nactually see here is a billboard with the
Dialogue: 0,0:26:12.31,0:26:17.52,Default,,0000,0000,0000,,best trustworthy families in Rongcheng. So\Nthese are the families that have the
Dialogue: 0,0:26:17.52,0:26:23.59,Default,,0000,0000,0000,,highest scores, for example. Sesame Credit\Nis a little different. It's the only
Dialogue: 0,0:26:23.59,0:26:27.99,Default,,0000,0000,0000,,system that actually uses machine learning\Nand artificial intelligence to determine
Dialogue: 0,0:26:27.99,0:26:32.98,Default,,0000,0000,0000,,the outputs. In Rongcheng, for example,\Nthey have artificial intelligence, they
Dialogue: 0,0:26:32.98,0:26:37.12,Default,,0000,0000,0000,,have computer vision, for the most part,\Nand the computer vision cameras they
Dialogue: 0,0:26:37.12,0:26:42.22,Default,,0000,0000,0000,,decide– they try to recognize you when you\Njaywalk. And then when they recognize you
Dialogue: 0,0:26:42.22,0:26:48.25,Default,,0000,0000,0000,,when jaywalking, you get a small SMS;\N"well, we just saw you jaywalking, your
Dialogue: 0,0:26:48.25,0:26:56.81,Default,,0000,0000,0000,,score is now dropping." But how the score\Ndevelops, depending on your jaywalking,
Dialogue: 0,0:26:56.81,0:27:01.12,Default,,0000,0000,0000,,isn't really determined by machine\Nlearning or artificial intelligence.
Dialogue: 0,0:27:01.12,0:27:06.84,Default,,0000,0000,0000,,Instead, it's determined by rules. You\Nknow: one time jaywalking deducts five
Dialogue: 0,0:27:06.84,0:27:12.46,Default,,0000,0000,0000,,points, and this is stated somewhere.\NSesame Credit doesn't work like that.
Dialogue: 0,0:27:12.46,0:27:19.81,Default,,0000,0000,0000,,Instead it uses a secret algorithm, and\Nthe way– I talked to some people that work
Dialogue: 0,0:27:19.81,0:27:25.33,Default,,0000,0000,0000,,for Sesame Credit or for Alibaba, and the\Nway they described it was; they basically
Dialogue: 0,0:27:25.33,0:27:33.26,Default,,0000,0000,0000,,clustered people based on behavior, then\Ngave scores to these clusters, and
Dialogue: 0,0:27:33.26,0:27:40.70,Default,,0000,0000,0000,,then afterwards, did basically reverse\Nengineered their own score, using machine
Dialogue: 0,0:27:40.70,0:27:45.95,Default,,0000,0000,0000,,learning, so that whenever something new\Nhappens, you can move to a different
Dialogue: 0,0:27:45.95,0:27:54.33,Default,,0000,0000,0000,,cluster. This Sesame Credit was actually\Nrefused accreditation as a credit score in
Dialogue: 0,0:27:54.33,0:28:03.40,Default,,0000,0000,0000,,2017, so banks are not allowed to use the\NSesame Credit score for your– to use the
Dialogue: 0,0:28:03.40,0:28:09.19,Default,,0000,0000,0000,,Sesame Credit score to determine whether\Nthey give you loans or not. Because Sesame
Dialogue: 0,0:28:09.19,0:28:13.77,Default,,0000,0000,0000,,Credit is quite ingenious – obviously\NAlibaba wants to keep you within their
Dialogue: 0,0:28:13.77,0:28:19.92,Default,,0000,0000,0000,,platform – so if you buy using Alibaba and\Nusing Alipay, your score goes up. If you
Dialogue: 0,0:28:19.92,0:28:29.09,Default,,0000,0000,0000,,buy using Weechatpay, which is a competing\Nplatform, your score goes down. This uses
Dialogue: 0,0:28:29.09,0:28:34.00,Default,,0000,0000,0000,,many of the same rewards mechanisms of the\Nofficial government systems, and this is
Dialogue: 0,0:28:34.00,0:28:38.39,Default,,0000,0000,0000,,just an illustration of what kind of\Nscores you can have, apparently your
Dialogue: 0,0:28:38.39,0:28:46.07,Default,,0000,0000,0000,,scores can go between 350 and 850, and in\NChinese there's basically five different
Dialogue: 0,0:28:46.07,0:28:55.81,Default,,0000,0000,0000,,levels. So 385 is a "trust-breaker" or\N"missing trust". And then 731 is "trust is
Dialogue: 0,0:28:55.81,0:29:06.17,Default,,0000,0000,0000,,exceedingly high". So one way I tried to\Napproach this issue was through agent-
Dialogue: 0,0:29:06.17,0:29:10.10,Default,,0000,0000,0000,,based modeling. Social Credit System is\Nindividual level, but what we're really
Dialogue: 0,0:29:10.10,0:29:13.40,Default,,0000,0000,0000,,interested in, or what I'm really\Ninterested in, is actually societal-level
Dialogue: 0,0:29:13.40,0:29:19.04,Default,,0000,0000,0000,,consequences. So if everyone gets this\Nscore, what does that mean for society?
Dialogue: 0,0:29:19.04,0:29:24.05,Default,,0000,0000,0000,,And agent-based modeling works quite well\Nfor that, because it allows us to imbue
Dialogue: 0,0:29:24.05,0:29:28.96,Default,,0000,0000,0000,,agents with some sort of rationality, but\Nwith a bounded rationality. What does
Dialogue: 0,0:29:28.96,0:29:32.80,Default,,0000,0000,0000,,bounded rationality mean? Usually in\Neconomics people assume agents are
Dialogue: 0,0:29:32.80,0:29:38.07,Default,,0000,0000,0000,,completely rational, so they are profit\Nmaximizers, they have all the information.
Dialogue: 0,0:29:38.07,0:29:45.38,Default,,0000,0000,0000,,But in reality, agents don't have all the\Ninformation, they have a lot of issues
Dialogue: 0,0:29:45.38,0:29:51.10,Default,,0000,0000,0000,,with keeping stuff in their mind. So a lot\Nof the time, they won't choose the best
Dialogue: 0,0:29:51.10,0:29:57.76,Default,,0000,0000,0000,,thing in the world, but they choose the\Nbest thing that they see. And bounded
Dialogue: 0,0:29:57.76,0:30:02.04,Default,,0000,0000,0000,,rationality allows us to account for this\Nthing. It allows us to account for
Dialogue: 0,0:30:02.04,0:30:08.38,Default,,0000,0000,0000,,heuristics and these things. And what I\Ndid is I took the propensity for specific
Dialogue: 0,0:30:08.38,0:30:12.32,Default,,0000,0000,0000,,behavior from current state of the art\Nresearch, mostly from behavioral
Dialogue: 0,0:30:12.32,0:30:17.90,Default,,0000,0000,0000,,economics. For example, I looked at tax\Nevasion, and I looked at who is likely to
Dialogue: 0,0:30:17.90,0:30:24.63,Default,,0000,0000,0000,,evade taxes in a system, and then\Nobviously there was some stochastic –
Dialogue: 0,0:30:24.63,0:30:30.81,Default,,0000,0000,0000,,some chance element. But the\Ndistribution that I chose is related to
Dialogue: 0,0:30:30.81,0:30:37.68,Default,,0000,0000,0000,,the current research. And I also checked\Nthat my model has similar results to the
Dialogue: 0,0:30:37.68,0:30:45.40,Default,,0000,0000,0000,,Rongcheng model, which I modeled at at the\Nbeginning. So on average 87% of my users
Dialogue: 0,0:30:45.40,0:30:49.24,Default,,0000,0000,0000,,have a score of within 10 percent of the\Noriginal score, which is also the data
Dialogue: 0,0:30:49.24,0:31:00.74,Default,,0000,0000,0000,,that Rongcheng city actually publishes.\NNow, for the most part, I compared design
Dialogue: 0,0:31:00.74,0:31:05.22,Default,,0000,0000,0000,,choices in two axes. One of them was a\Ncentralized system versus a multi-level
Dialogue: 0,0:31:05.22,0:31:10.71,Default,,0000,0000,0000,,system, and a rule-based system versus a\Nmachine learning system. The centralized
Dialogue: 0,0:31:10.71,0:31:20.65,Default,,0000,0000,0000,,system is basically: you have a central –\Nall the information is kept centrally, and
Dialogue: 0,0:31:20.65,0:31:27.19,Default,,0000,0000,0000,,everyone in China, or wherever, in\NRongcheng has the exact same scoring
Dialogue: 0,0:31:27.19,0:31:35.52,Default,,0000,0000,0000,,opportunities. Now, if you have a\Ncentralized system the clear expectations
Dialogue: 0,0:31:35.52,0:31:39.60,Default,,0000,0000,0000,,were pretty good. But, at the same time,\Nthe acceptance from the population was
Dialogue: 0,0:31:39.60,0:31:46.59,Default,,0000,0000,0000,,really, really low, which they found\Nduring the Suining experiment. And
Dialogue: 0,0:31:46.59,0:31:50.54,Default,,0000,0000,0000,,there's also the problem of a single point\Nof failure. Who decides the central
Dialogue: 0,0:31:50.54,0:31:59.08,Default,,0000,0000,0000,,catalog, and, depending on who, sort of,\Nhas the power, it kind of, just,
Dialogue: 0,0:31:59.08,0:32:05.10,Default,,0000,0000,0000,,reproduces power structures. So because\Nyou have this central catalog, the same
Dialogue: 0,0:32:05.10,0:32:10.76,Default,,0000,0000,0000,,people that are in power centrally, they\Nare basically deciding some sort of score
Dialogue: 0,0:32:10.76,0:32:15.34,Default,,0000,0000,0000,,mechanism that works for them very well,\Nso that they and their family will have
Dialogue: 0,0:32:15.34,0:32:23.28,Default,,0000,0000,0000,,high scores. And multi-level system has\Nthe advantage that local adaptation kind
Dialogue: 0,0:32:23.28,0:32:29.08,Default,,0000,0000,0000,,of works, and there's sort of many points\Nof failure. But in my model, when I
Dialogue: 0,0:32:29.08,0:32:36.86,Default,,0000,0000,0000,,allowed locals to basically set their own\Nrules, what happened was that they
Dialogue: 0,0:32:36.86,0:32:42.52,Default,,0000,0000,0000,,competed. So, it started out being this\Ndistrict of Rongcheng, for example, and
Dialogue: 0,0:32:42.52,0:32:46.12,Default,,0000,0000,0000,,this district of Rongcheng, they compete\Nfor the best people that they want to
Dialogue: 0,0:32:46.12,0:32:52.00,Default,,0000,0000,0000,,attract, and suddenly you have this kind\Nof race to the bottom, where people want
Dialogue: 0,0:32:52.00,0:32:58.09,Default,,0000,0000,0000,,to move where they wouldn't be prosecuted,\Nso they move to places where there's less
Dialogue: 0,0:32:58.09,0:33:04.00,Default,,0000,0000,0000,,cameras, for example. At the same time,\Nthere's many points of failure, especially
Dialogue: 0,0:33:04.00,0:33:13.50,Default,,0000,0000,0000,,the way it's currently set up, with people\Nreporting data to the next high level.
Dialogue: 0,0:33:13.50,0:33:19.45,Default,,0000,0000,0000,,And, a lot of the time, what we have\Nactually seen in Rongcheng, was that they
Dialogue: 0,0:33:19.45,0:33:24.36,Default,,0000,0000,0000,,reported data on people they didn't like\Nmore than data on people they did like.
Dialogue: 0,0:33:24.36,0:33:29.39,Default,,0000,0000,0000,,Or, their families got better scores than\Npeople they didn't know. So it also kind
Dialogue: 0,0:33:29.39,0:33:41.32,Default,,0000,0000,0000,,of reproduced these biases. The rule based\Nsystem has the advantage that people were
Dialogue: 0,0:33:41.32,0:33:46.53,Default,,0000,0000,0000,,more prone to adapt their behavior,\Nbecause they actually knew what they
Dialogue: 0,0:33:46.53,0:33:50.63,Default,,0000,0000,0000,,needed to do in order to adapt their\Nbehavior. But the score didn't really
Dialogue: 0,0:33:50.63,0:33:54.05,Default,,0000,0000,0000,,correlate with the important\Ncharacteristics that they actually cared
Dialogue: 0,0:33:54.05,0:34:01.61,Default,,0000,0000,0000,,about. And, as opposed to in this machine\Nlearning system, you know how in Germany
Dialogue: 0,0:34:01.61,0:34:06.91,Default,,0000,0000,0000,,we don't really know the Schufa algorithm.\NAnd I, for example, don't exactly know
Dialogue: 0,0:34:06.91,0:34:12.14,Default,,0000,0000,0000,,what I could do in order to improve my\NSchufa score. And this is a similar system
Dialogue: 0,0:34:12.14,0:34:16.97,Default,,0000,0000,0000,,in China with the Sesame Credit score. A\Nlot of people don't really – they say,
Dialogue: 0,0:34:16.97,0:34:21.80,Default,,0000,0000,0000,,"well I really want to adapt my behavior\Nto the score, to improve my score, but
Dialogue: 0,0:34:21.80,0:34:29.12,Default,,0000,0000,0000,,when I tried doing that my score actually\Ngot worse." And you can have different
Dialogue: 0,0:34:29.12,0:34:36.44,Default,,0000,0000,0000,,biases, that I'm going to be talking about\Nin a little bit. There's also this big
Dialogue: 0,0:34:36.44,0:34:42.58,Default,,0000,0000,0000,,problem of incentive mismatch. So, the\Ndecentralized, rules-based systems like
Dialogue: 0,0:34:42.58,0:34:47.30,Default,,0000,0000,0000,,Rongcheng, which is the system that I\Nanalyzed the most. Why, because I believe
Dialogue: 0,0:34:47.30,0:34:51.53,Default,,0000,0000,0000,,this is the system that we're moving\Ntowards right now. Because Rongcheng won a
Dialogue: 0,0:34:51.53,0:34:57.02,Default,,0000,0000,0000,,lot of awards. So the Chinese government,\Nthe way they usually work is, they try
Dialogue: 0,0:34:57.02,0:35:01.71,Default,,0000,0000,0000,,pilots, then they choose the best couple\Nof systems, they give them awards, and
Dialogue: 0,0:35:01.71,0:35:06.89,Default,,0000,0000,0000,,then they roll out the system nationwide.\NSo I assume that the system that's going
Dialogue: 0,0:35:06.89,0:35:12.99,Default,,0000,0000,0000,,to be – the system in the end will be\Nsimilar to the Rongcheng system. Now, one
Dialogue: 0,0:35:12.99,0:35:19.67,Default,,0000,0000,0000,,problem that I actually saw in my\Nsimulation was that you could have this
Dialogue: 0,0:35:19.67,0:35:24.96,Default,,0000,0000,0000,,possible race to the bottom. There's also\Nthis conflict of interest in those that
Dialogue: 0,0:35:24.96,0:35:29.71,Default,,0000,0000,0000,,set the rules, because a lot of the time,\Nthe way it works is, you have your
Dialogue: 0,0:35:29.71,0:35:35.94,Default,,0000,0000,0000,,company, and your company, you, in\Ncombination with your party leaders,
Dialogue: 0,0:35:35.94,0:35:43.76,Default,,0000,0000,0000,,actually decide on the rules for the score\Nsystem. But the scores of all your
Dialogue: 0,0:35:43.76,0:35:48.32,Default,,0000,0000,0000,,employees actually determines your\Ncompany's score. If you employ a lot of
Dialogue: 0,0:35:48.32,0:35:52.59,Default,,0000,0000,0000,,people with high scores you get a better\Nscore. So you will have this incentive to
Dialogue: 0,0:35:52.59,0:35:57.40,Default,,0000,0000,0000,,give out high scores and to make sure that\Neveryone gets high scores. But at the same
Dialogue: 0,0:35:57.40,0:36:04.72,Default,,0000,0000,0000,,time the government has an incentive for\Nscores to be comparable. So there's a lot
Dialogue: 0,0:36:04.72,0:36:10.02,Default,,0000,0000,0000,,of incentives mismatch. The government\Nalso has the incentive to keep false
Dialogue: 0,0:36:10.02,0:36:15.54,Default,,0000,0000,0000,,negatives down, but they actually, the way\Nthe Chinese system currently works is,
Dialogue: 0,0:36:15.54,0:36:22.59,Default,,0000,0000,0000,,they emphasize catching trust-breakers\Nmore than rewarding trust-follow... or
Dialogue: 0,0:36:22.59,0:36:28.75,Default,,0000,0000,0000,,trustworthy people. So, false positives,\Nfor them, are less important, but false
Dialogue: 0,0:36:28.75,0:36:34.36,Default,,0000,0000,0000,,positives erode the trust in the system,\Nand they lead to a lot less behavioral
Dialogue: 0,0:36:34.36,0:36:40.74,Default,,0000,0000,0000,,adaptation. I was actually able to show\Nthis using some nudging research that
Dialogue: 0,0:36:40.74,0:36:47.79,Default,,0000,0000,0000,,showed that as soon as you introduce an\Nerror probability and you can be caught
Dialogue: 0,0:36:47.79,0:36:54.53,Default,,0000,0000,0000,,for something that you didn't do, your\Nprobability of changing your behavior
Dialogue: 0,0:36:54.53,0:37:02.12,Default,,0000,0000,0000,,based on this score is actually lower. And\Nin Rongcheng, one of the perverse things
Dialogue: 0,0:37:02.12,0:37:09.88,Default,,0000,0000,0000,,that they're doing is, you can donate\Nmoney to the party or to, like, party
Dialogue: 0,0:37:09.88,0:37:17.34,Default,,0000,0000,0000,,affiliated social services, and this will\Ngive you points, which is kind of an
Dialogue: 0,0:37:17.34,0:37:24.27,Default,,0000,0000,0000,,indulgence system. Which is quite\Ninteresting, especially because a lot of
Dialogue: 0,0:37:24.27,0:37:32.69,Default,,0000,0000,0000,,these donation systems work in a way that\Nyou can donate 50000 renminbi and you get
Dialogue: 0,0:37:32.69,0:37:36.74,Default,,0000,0000,0000,,50 points, and then you donate another\N50000 renminbi and you get another 50
Dialogue: 0,0:37:36.74,0:37:44.40,Default,,0000,0000,0000,,points. So you can basically donate a lot\Nof money and then behave however you want,
Dialogue: 0,0:37:44.40,0:37:56.22,Default,,0000,0000,0000,,and still get a good score. And the trust\Nin other people can actually go down even
Dialogue: 0,0:37:56.22,0:38:00.56,Default,,0000,0000,0000,,more in this system, because suddenly you\Nonly trust them because of their scores,
Dialogue: 0,0:38:00.56,0:38:05.46,Default,,0000,0000,0000,,and the current system is set up so that\Nyou can actually look up scores of
Dialogue: 0,0:38:05.46,0:38:09.73,Default,,0000,0000,0000,,everyone that you want to work with, and\Nif they don't have a score high enough
Dialogue: 0,0:38:09.73,0:38:15.14,Default,,0000,0000,0000,,then suddenly you don't want to work with\Nthem. The trust in the legal system can
Dialogue: 0,0:38:15.14,0:38:22.26,Default,,0000,0000,0000,,also decrease, actually. Why? Because trust\Nin the legal system in China is already
Dialogue: 0,0:38:22.26,0:38:26.04,Default,,0000,0000,0000,,low, and a lot of the things, like\Njaywalking, they're already illegal in
Dialogue: 0,0:38:26.04,0:38:30.38,Default,,0000,0000,0000,,China, as they are here, but no one cares.\NAnd suddenly, you have this parallel
Dialogue: 0,0:38:30.38,0:38:37.27,Default,,0000,0000,0000,,system that punishes you for whatever.\NBut, why don't you just try to fix the
Dialogue: 0,0:38:37.27,0:38:45.19,Default,,0000,0000,0000,,legal system, which would be my approach.\NSuddenly, illegal activity could happen
Dialogue: 0,0:38:45.19,0:38:51.72,Default,,0000,0000,0000,,more offline, and this is one of those\Nthings that is quite interesting. In
Dialogue: 0,0:38:51.72,0:38:58.06,Default,,0000,0000,0000,,countries that we've seen that have moved\Ntowards mobile payments, and away
Dialogue: 0,0:38:58.06,0:39:05.07,Default,,0000,0000,0000,,from cash, you see less robberies but you\Ndon't actually see less crime. Instead you
Dialogue: 0,0:39:05.07,0:39:12.43,Default,,0000,0000,0000,,see more new types of crime. So, you see\Nmore credit card fraud, you see more phone
Dialogue: 0,0:39:12.43,0:39:19.39,Default,,0000,0000,0000,,robberies, these kinds of things. And this\Nis also where things could move in the
Dialogue: 0,0:39:19.39,0:39:29.71,Default,,0000,0000,0000,,Chinese case. One major problem is also\Nthat this new system – I've talked a
Dialogue: 0,0:39:29.71,0:39:34.50,Default,,0000,0000,0000,,little bit about this one, but – it can\Nintroduce a lot of new bias, and reproduce
Dialogue: 0,0:39:34.50,0:39:44.71,Default,,0000,0000,0000,,the bias even more. So, for example, China\Nis a country of 55 minorities. The Han are
Dialogue: 0,0:39:44.71,0:39:50.25,Default,,0000,0000,0000,,a big majority, they have about 94 percent\Nof the population. So any computer vision
Dialogue: 0,0:39:50.25,0:39:58.02,Default,,0000,0000,0000,,task, we've shown, that they are really,\Nreally bad at discriminating between
Dialogue: 0,0:39:58.02,0:40:04.93,Default,,0000,0000,0000,,individuals in smaller ethnic groups. In\Nthe U.S., most computer vision tasks
Dialogue: 0,0:40:04.93,0:40:09.92,Default,,0000,0000,0000,,perform worse for African-Americans, they\Nperform worse for women, because all of
Dialogue: 0,0:40:09.92,0:40:16.93,Default,,0000,0000,0000,,the training sets are male and white, and\Nmaybe Asian. In China, all of these tasks
Dialogue: 0,0:40:16.93,0:40:26.61,Default,,0000,0000,0000,,are actually performing worse for ethnic\Nminorities, for the Uyghurs, for example.
Dialogue: 0,0:40:26.61,0:40:32.46,Default,,0000,0000,0000,,And one way that they could try to abuse\Nthe system is to basically just – what
Dialogue: 0,0:40:32.46,0:40:38.21,Default,,0000,0000,0000,,they're also doing already in Xinjiang is\N– to basically just identify, "oh this is
Dialogue: 0,0:40:38.21,0:40:45.98,Default,,0000,0000,0000,,a person of the minority, well I'm just\Ngoing to go and check him or her more
Dialogue: 0,0:40:45.98,0:40:50.21,Default,,0000,0000,0000,,thoroughly." This is actually what happens\Nin Xinjiang. If you're in Xinjiang and you
Dialogue: 0,0:40:50.21,0:40:59.25,Default,,0000,0000,0000,,look like a Turkish person, or like from\NTurkmenistan, from a Turkish people, you
Dialogue: 0,0:40:59.25,0:41:04.21,Default,,0000,0000,0000,,are a lot more likely to be questioned.\NYou're a lot more likely to be stopped and
Dialogue: 0,0:41:04.21,0:41:12.58,Default,,0000,0000,0000,,they ask you or require you to download\Nspyware on your phone. And this is
Dialogue: 0,0:41:12.58,0:41:17.66,Default,,0000,0000,0000,,currently what happens and this new kind\Nof system can actually help you with that.
Dialogue: 0,0:41:17.66,0:41:24.71,Default,,0000,0000,0000,,I've said that it can reproduce these kind\Nof power structures, and now obviously we
Dialogue: 0,0:41:24.71,0:41:29.89,Default,,0000,0000,0000,,all know neutral technology doesn't really\Nexist, but in the Chinese case, in the
Dialogue: 0,0:41:29.89,0:41:33.53,Default,,0000,0000,0000,,social credit case, they don't even\Npretend – they always say "well, this
Dialogue: 0,0:41:33.53,0:41:37.35,Default,,0000,0000,0000,,is neutral technology and it's all a lot\Nbetter," but actually it's the people
Dialogue: 0,0:41:37.35,0:41:43.97,Default,,0000,0000,0000,,currently in power, they decide on what\Ngives you point and what deducts points
Dialogue: 0,0:41:43.97,0:41:50.03,Default,,0000,0000,0000,,for you. Another problem, currently the\Nentire system is set up in a way that it
Dialogue: 0,0:41:50.03,0:41:54.62,Default,,0000,0000,0000,,all goes together with your {\i1}shēnfènzhèng{\i0},\Nwith your I.D. card. What if you don't
Dialogue: 0,0:41:54.62,0:41:59.41,Default,,0000,0000,0000,,have an I.D. card? That's foreigners for\None. But it's also people in China that
Dialogue: 0,0:41:59.41,0:42:05.48,Default,,0000,0000,0000,,were born during the one child policy and\Nwere not registered. There's quite a lot
Dialogue: 0,0:42:05.48,0:42:09.20,Default,,0000,0000,0000,,of them, actually. They're not registered\Nanywhere and suddenly they can't do
Dialogue: 0,0:42:09.20,0:42:13.70,Default,,0000,0000,0000,,anything, because they don't have a score,\Nthey can't get a phone, they can't do
Dialogue: 0,0:42:13.70,0:42:20.88,Default,,0000,0000,0000,,anything, really. And part of the push\Nwith this social credit system is to go
Dialogue: 0,0:42:20.88,0:42:26.78,Default,,0000,0000,0000,,away from cash, actually. So if you need\Nto use your phone to pay, but for your
Dialogue: 0,0:42:26.78,0:42:29.52,Default,,0000,0000,0000,,phone you need your {\i1}shēnfènzhèng{\i0}.\NIf you don't have a {\i1}shēnfènzhèng{\i0},
Dialogue: 0,0:42:29.52,0:42:32.57,Default,,0000,0000,0000,,well, tough luck for you.
Dialogue: 0,0:42:32.57,0:42:38.68,Default,,0000,0000,0000,,And currently the system in Rongcheng\Nis set up in a way that you can check
Dialogue: 0,0:42:38.68,0:42:44.90,Default,,0000,0000,0000,,other people's scores and you can also see\Nwhat they lose points for. So you can
Dialogue: 0,0:42:44.90,0:42:49.78,Default,,0000,0000,0000,,actually, sort of, choose to discriminate\Nagainst people that are gay, for example,
Dialogue: 0,0:42:49.78,0:42:53.05,Default,,0000,0000,0000,,because they might have lost points for\Ngoing to a gay bar, which you can lose
Dialogue: 0,0:42:53.05,0:43:02.13,Default,,0000,0000,0000,,points for. Another big issue, currently,\Nis data privacy and security. Personal
Dialogue: 0,0:43:02.13,0:43:07.25,Default,,0000,0000,0000,,data is grossly undervalued in China. If\Nyou ask a Chinese person, "what do you
Dialogue: 0,0:43:07.25,0:43:14.69,Default,,0000,0000,0000,,think, how much is your data worth?," they\Nsay "what data? I don't have data." And,
Dialogue: 0,0:43:14.69,0:43:19.13,Default,,0000,0000,0000,,currently, the way it works is, if you\Nhave someone's ID number, which is quite
Dialogue: 0,0:43:19.13,0:43:24.89,Default,,0000,0000,0000,,easy to find out, you can actually buy\Naccess to a lot of personal information
Dialogue: 0,0:43:24.89,0:43:31.36,Default,,0000,0000,0000,,for a small fee. So you pay about 100\Neuros and you get all hotel bookings of
Dialogue: 0,0:43:31.36,0:43:35.96,Default,,0000,0000,0000,,the last year, you get information of who\Nbooked these hotels with them, you get
Dialogue: 0,0:43:35.96,0:43:40.83,Default,,0000,0000,0000,,information of where they stay, you get\Ntrain bookings, you get access to all of
Dialogue: 0,0:43:40.83,0:43:47.20,Default,,0000,0000,0000,,the official databases for this one\Nperson. And for another 700 renminbi you
Dialogue: 0,0:43:47.20,0:43:52.83,Default,,0000,0000,0000,,can actually get live location data, so\Nyou can get the data of where this person
Dialogue: 0,0:43:52.83,0:43:56.40,Default,,0000,0000,0000,,is right now, or where his or her phone is\Nright now, but if you've ever been to
Dialogue: 0,0:43:56.40,0:44:03.44,Default,,0000,0000,0000,,China you know that where the phone is,\Nusually, the people aren't far. Supchina
Dialogue: 0,0:44:03.44,0:44:08.50,Default,,0000,0000,0000,,actually did an experiment where a couple\Nof journalists tried buying that, because
Dialogue: 0,0:44:08.50,0:44:14.19,Default,,0000,0000,0000,,it's actually these kind of services are\Noffered on weechat, pretty publicly. And
Dialogue: 0,0:44:14.19,0:44:26.36,Default,,0000,0000,0000,,you can just buy them, quite easily. So\None additional thing that I looked at is,
Dialogue: 0,0:44:26.36,0:44:30.58,Default,,0000,0000,0000,,because one of the things that is quite\Ninteresting is, you have this idea of
Dialogue: 0,0:44:30.58,0:44:39.31,Default,,0000,0000,0000,,credit as twofold. Credit is trust credit\Nbut credit is also loan credit, and what
Dialogue: 0,0:44:39.31,0:44:43.73,Default,,0000,0000,0000,,if credit institutions actually use this\Nunified credit score to determine credit
Dialogue: 0,0:44:43.73,0:44:49.06,Default,,0000,0000,0000,,distribution? The idea is that it's\Nsupposed to lead to reduced information
Dialogue: 0,0:44:49.06,0:44:55.47,Default,,0000,0000,0000,,asymmetry, obviously, so fewer defaults\Nand overall more credit creation. New
Dialogue: 0,0:44:55.47,0:44:59.55,Default,,0000,0000,0000,,people are supposed to get access to\Ncredit, and there's supposed to be less
Dialogue: 0,0:44:59.55,0:45:04.59,Default,,0000,0000,0000,,shadow banking. But what actually happens?\NI'm not going to be talking about how I
Dialogue: 0,0:45:04.59,0:45:08.62,Default,,0000,0000,0000,,set up the model but just about my\Nresults. If you have this kind of score
Dialogue: 0,0:45:08.62,0:45:14.37,Default,,0000,0000,0000,,that includes credit information but also\Nincludes morally good – or measures of
Dialogue: 0,0:45:14.37,0:45:18.78,Default,,0000,0000,0000,,being morally good – what you have is, in\Nthe beginning, about 30 percent more
Dialogue: 0,0:45:18.78,0:45:23.96,Default,,0000,0000,0000,,agents get access to credit, and\Nespecially people that previously have not
Dialogue: 0,0:45:23.96,0:45:29.71,Default,,0000,0000,0000,,gotten credit access suddenly have credit\Naccess. But the problem is that this
Dialogue: 0,0:45:29.71,0:45:36.17,Default,,0000,0000,0000,,social credit score that correlates all of\Nthese different issues, it correlates only
Dialogue: 0,0:45:36.17,0:45:41.81,Default,,0000,0000,0000,,very, very weakly with repayment ability\Nor repayment wishes, and thus suddenly you
Dialogue: 0,0:45:41.81,0:45:47.88,Default,,0000,0000,0000,,have all of these non-performing loans.\NYou have – and what we see is sort of
Dialogue: 0,0:45:47.88,0:45:51.96,Default,,0000,0000,0000,,like – we have non-performing loans.\NBanks give out less loans because they
Dialogue: 0,0:45:51.96,0:45:59.40,Default,,0000,0000,0000,,have so many non-performing loans, and\Nthen the non-performing loans are written
Dialogue: 0,0:45:59.40,0:46:03.100,Default,,0000,0000,0000,,off, and suddenly banks give out more\Nloans. But you have this oscillating
Dialogue: 0,0:46:03.100,0:46:09.24,Default,,0000,0000,0000,,financial system, where you give out a lot\Nof loans, a lot of them are non-
Dialogue: 0,0:46:09.24,0:46:13.02,Default,,0000,0000,0000,,performing, then you give out a lot of\Nloans again. And this is very, very
Dialogue: 0,0:46:13.02,0:46:19.15,Default,,0000,0000,0000,,vulnerable to crisis. If you have a real\Neconomic crisis during the time where non-
Dialogue: 0,0:46:19.15,0:46:24.57,Default,,0000,0000,0000,,performing loans are high, then a lot of\Nbanks will actually default, which is
Dialogue: 0,0:46:24.57,0:46:29.92,Default,,0000,0000,0000,,very, very dangerous for a financial\Nsystem as nationed as the Chinese one.
Dialogue: 0,0:46:29.92,0:46:36.80,Default,,0000,0000,0000,,Now, what are some possible corrections?\NYou could create a score that basically is
Dialogue: 0,0:46:36.80,0:46:41.26,Default,,0000,0000,0000,,the same as the Schufa score. So that it\Nlooks only at credit decisions, but
Dialogue: 0,0:46:41.26,0:46:45.19,Default,,0000,0000,0000,,suddenly, you lose a lot of incentives for\Nthe social credit score, if the social
Dialogue: 0,0:46:45.19,0:46:48.43,Default,,0000,0000,0000,,credit score doesn't matter for credit\Ndistribution anymore.
Dialogue: 0,0:46:48.43,0:46:52.04,Default,,0000,0000,0000,,Another thing, and this is, I\Nthink, the more likely one,
Dialogue: 0,0:46:52.04,0:46:55.96,Default,,0000,0000,0000,,is that you have a blacklist for people\Nthat have not repaid a loan
Dialogue: 0,0:46:55.96,0:46:59.40,Default,,0000,0000,0000,,in the past. So you basically\Nget one freebie, and afterwards
Dialogue: 0,0:46:59.40,0:47:04.15,Default,,0000,0000,0000,,if you didn't repay your loan in the past\Nthen you will not get a loan in the
Dialogue: 0,0:47:04.15,0:47:08.40,Default,,0000,0000,0000,,future. You will still be part of the\Nsocial credit system, and your social
Dialogue: 0,0:47:08.40,0:47:12.36,Default,,0000,0000,0000,,credit score will still be important for\Nall of these other access issues, but it
Dialogue: 0,0:47:12.36,0:47:15.85,Default,,0000,0000,0000,,won't be important for access to loans\Nanymore, once you've been on this
Dialogue: 0,0:47:15.85,0:47:21.67,Default,,0000,0000,0000,,blacklist. Which is probably something\Nthat the Chinese government could go
Dialogue: 0,0:47:21.67,0:47:30.18,Default,,0000,0000,0000,,behind, but it's also more effort to take\Ncare of it; then you have to think about,
Dialogue: 0,0:47:30.18,0:47:33.82,Default,,0000,0000,0000,,"well, you can't leave them on the\Nblacklist forever, so how long do you
Dialogue: 0,0:47:33.82,0:47:37.60,Default,,0000,0000,0000,,leave them on the black list? Do they have\Nto pay back the loan and then they get off
Dialogue: 0,0:47:37.60,0:47:45.67,Default,,0000,0000,0000,,the blacklist? Or do they have to pay back\Nthe loan and then stay not in default
Dialogue: 0,0:47:45.67,0:47:52.86,Default,,0000,0000,0000,,for a year, or for five years?" There's a\Nlot of small decisions that, in my
Dialogue: 0,0:47:52.86,0:47:57.55,Default,,0000,0000,0000,,opinion, the Chinese government hasn't\Nreally thought about, up until now,
Dialogue: 0,0:47:57.55,0:48:01.17,Default,,0000,0000,0000,,because they're basically doing all these\Npilot studies, and all of these regional
Dialogue: 0,0:48:01.17,0:48:05.16,Default,,0000,0000,0000,,governments are thinking of all these\Nsmall things, but they're not documenting
Dialogue: 0,0:48:05.16,0:48:10.35,Default,,0000,0000,0000,,everything that they're doing. So, once\Nthey – they want to roll it out by 2020,
Dialogue: 0,0:48:10.35,0:48:15.20,Default,,0000,0000,0000,,by the way, nationwide – once they've\Nrolled it out there's a pretty big chance,
Dialogue: 0,0:48:15.20,0:48:18.63,Default,,0000,0000,0000,,in my opinion, that they'll have a lot of\Nunintended consequences. A lot of things
Dialogue: 0,0:48:18.63,0:48:28.52,Default,,0000,0000,0000,,that they haven't thought about, and that\Nthey will then have to look at. So, I
Dialogue: 0,0:48:28.52,0:48:33.27,Default,,0000,0000,0000,,believe that some sort of system is likely\Nto come, just in terms of how much energy
Dialogue: 0,0:48:33.27,0:48:37.29,Default,,0000,0000,0000,,they've expended into this one, and for\Nthe Chinese government at this point, for
Dialogue: 0,0:48:37.29,0:48:41.90,Default,,0000,0000,0000,,the party, it would be losing face if they\Ndid not include any such system, because
Dialogue: 0,0:48:41.90,0:48:45.97,Default,,0000,0000,0000,,they've been talking about this for a\Nwhile. But most likely, it would be a kind
Dialogue: 0,0:48:45.97,0:48:53.02,Default,,0000,0000,0000,,of decentralized data sharing system. And\Nwhen I ran my simulation... By the way I
Dialogue: 0,0:48:53.02,0:48:59.70,Default,,0000,0000,0000,,will make public my code, I still need\Nsome, basically, I used some proprietary
Dialogue: 0,0:48:59.70,0:49:06.46,Default,,0000,0000,0000,,data for my model, and I still need the\Npermission to publish this. Once I publish
Dialogue: 0,0:49:06.46,0:49:11.29,Default,,0000,0000,0000,,this one I will also tweet it, and we'll\Nput it on GitHub for everyone to play
Dialogue: 0,0:49:11.29,0:49:16.20,Default,,0000,0000,0000,,around with, if you want to. And some of\Nthese implementation details that were
Dialogue: 0,0:49:16.20,0:49:20.45,Default,,0000,0000,0000,,very important in determining model\Noutcomes where "do we have a relative or
Dialogue: 0,0:49:20.45,0:49:25.29,Default,,0000,0000,0000,,absolute ranking?" So far, all of the\Nsystems I looked at had absolute rankings,
Dialogue: 0,0:49:25.29,0:49:30.70,Default,,0000,0000,0000,,but there's a point to be made for\Nrelative rankings. Do we have one score,
Dialogue: 0,0:49:30.70,0:49:35.09,Default,,0000,0000,0000,,where, basically, if you're a Chinese\Nperson you get one score? Or do we have
Dialogue: 0,0:49:35.09,0:49:40.88,Default,,0000,0000,0000,,different sub-scores in different fields?\NDo we have people reporting behavior, or
Dialogue: 0,0:49:40.88,0:49:46.37,Default,,0000,0000,0000,,do we have automatic behavior recording?\NHow do you access other people's scores?
Dialogue: 0,0:49:46.37,0:49:50.34,Default,,0000,0000,0000,,How much information can you get from\Nother people's scores? Currently, if
Dialogue: 0,0:49:50.34,0:49:55.53,Default,,0000,0000,0000,,someone is on a blacklist, for example, if\Nyou have their ID number, again, you can
Dialogue: 0,0:49:55.53,0:49:59.65,Default,,0000,0000,0000,,put it into this blacklist, and then they\Nwill say "oh, this person is on this
Dialogue: 0,0:49:59.65,0:50:04.63,Default,,0000,0000,0000,,blacklist for not following this judge's\Norder," and then it says what kind of
Dialogue: 0,0:50:04.63,0:50:10.66,Default,,0000,0000,0000,,judge's order it was. So, most likely, it\Nwill be something like this. The idea is
Dialogue: 0,0:50:10.66,0:50:16.05,Default,,0000,0000,0000,,that the Social Credit system isn't only\Nfor individuals, but also for firms and
Dialogue: 0,0:50:16.05,0:50:22.22,Default,,0000,0000,0000,,for NGOs. So, what kind of roles will\Nfirms play in the system? I haven't looked
Dialogue: 0,0:50:22.22,0:50:28.32,Default,,0000,0000,0000,,at that, in detail, at this point, but it\Nwould be very interesting. Another idea
Dialogue: 0,0:50:28.32,0:50:34.39,Default,,0000,0000,0000,,that western people often talk about is,\Ndo people also rank each other? Currently,
Dialogue: 0,0:50:34.39,0:50:39.39,Default,,0000,0000,0000,,that's not part of the system in China,\Nbut it might be at one point. And lastly,
Dialogue: 0,0:50:39.39,0:50:44.84,Default,,0000,0000,0000,,where does the aggregation happen? So I've\Nsaid that a lot of it is actually data
Dialogue: 0,0:50:44.84,0:50:53.75,Default,,0000,0000,0000,,sharing in China. So what kind of data is\Nshared? Is the raw data shared? "Person A
Dialogue: 0,0:50:53.75,0:51:03.90,Default,,0000,0000,0000,,did something." Or is the aggregated data\Nshared? "Person A got this score." At this
Dialogue: 0,0:51:03.90,0:51:07.89,Default,,0000,0000,0000,,point, most of the time, it is actually\Nthe raw data that is shared, but that also
Dialogue: 0,0:51:07.89,0:51:12.81,Default,,0000,0000,0000,,has sort of these data privacy issues, of\Ncourse, that I've talked about. OK,
Dialogue: 0,0:51:12.81,0:51:18.95,Default,,0000,0000,0000,,perfect! No there's 10 more minutes. Thank\Nyou for your attention! If you have
Dialogue: 0,0:51:18.95,0:51:24.27,Default,,0000,0000,0000,,questions, remarks you can ask them now or\Nyou can catch me up later. You can tweet
Dialogue: 0,0:51:24.27,0:51:29.83,Default,,0000,0000,0000,,to me or send me an e-mail, whatever\Nyou're interested in. Thank you very much!
Dialogue: 0,0:51:29.83,0:51:37.33,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,0:51:37.33,0:51:42.04,Default,,0000,0000,0000,,Herald Angel: Hello! As Toni said, we have\N10 minutes left for questions. If you have
Dialogue: 0,0:51:42.04,0:51:46.98,Default,,0000,0000,0000,,a question in the room, please go crouch in\Nfront of our five microphones. If you're
Dialogue: 0,0:51:46.98,0:51:49.70,Default,,0000,0000,0000,,watching the stream, please ask your\Nquestions through IRC or Twitter, and
Dialogue: 0,0:51:49.70,0:51:53.85,Default,,0000,0000,0000,,we'll also try to make sure to get to\Nthose. Let's just go ahead and start with
Dialogue: 0,0:51:53.85,0:51:56.34,Default,,0000,0000,0000,,mic one.\NQuestion: Good! Thank you very much for
Dialogue: 0,0:51:56.34,0:52:02.81,Default,,0000,0000,0000,,this beautiful talk. I was wondering how\Ndid the Chinese government, companies, and
Dialogue: 0,0:52:02.81,0:52:07.47,Default,,0000,0000,0000,,most of all, the citizens themselves,\Nrespond to you doing this research, or,
Dialogue: 0,0:52:07.47,0:52:11.52,Default,,0000,0000,0000,,let's put it differently, if you would\Nhave been in the system yourself,
Dialogue: 0,0:52:11.52,0:52:14.44,Default,,0000,0000,0000,,how would your research affect your\Nsocial credit score?
Dialogue: 0,0:52:14.44,0:52:17.27,Default,,0000,0000,0000,,{\i1}laughter{\i0}
Dialogue: 0,0:52:17.27,0:52:25.64,Default,,0000,0000,0000,,Answer: So, um... There's actually two\Ndifferent responses that I've seen. When I
Dialogue: 0,0:52:25.64,0:52:31.38,Default,,0000,0000,0000,,talk to the government themselves, because\NI was there on a government scholarship,
Dialogue: 0,0:52:31.38,0:52:34.87,Default,,0000,0000,0000,,and mentioned that I'm really interested\Nin this, they basically said oh well this
Dialogue: 0,0:52:34.87,0:52:38.93,Default,,0000,0000,0000,,is just a technical system. You don't\Nreally need to be concerned with this. It
Dialogue: 0,0:52:38.93,0:52:43.44,Default,,0000,0000,0000,,is not very important. Just, you know,\Nit's just a technicality. It's just for us
Dialogue: 0,0:52:43.44,0:52:49.30,Default,,0000,0000,0000,,to make life more efficient and better for\Neveryone. So I assume my score would
Dialogue: 0,0:52:49.30,0:52:54.66,Default,,0000,0000,0000,,actually go down from doing this research,\Nactually. But when I talk to a lot of
Dialogue: 0,0:52:54.66,0:53:00.89,Default,,0000,0000,0000,,people at universities, they were also\Nvery – they were very interested in my
Dialogue: 0,0:53:00.89,0:53:05.23,Default,,0000,0000,0000,,research, and a lot of them mentioned that\Nthey didn't even know that the system
Dialogue: 0,0:53:05.23,0:53:10.20,Default,,0000,0000,0000,,existed!\NHerald: Before we go to a question from
Dialogue: 0,0:53:10.20,0:53:14.73,Default,,0000,0000,0000,,our signal angel, a request for all the\Npeople leaving the room, please do so as
Dialogue: 0,0:53:14.73,0:53:20.50,Default,,0000,0000,0000,,quietly as possible, so we can continue\Nthis Q and A. The signal angel, please!
Dialogue: 0,0:53:20.50,0:53:26.25,Default,,0000,0000,0000,,Signal Angel: Jaenix wants to know, is\Nthis score actually influenced by
Dialogue: 0,0:53:26.25,0:53:31.55,Default,,0000,0000,0000,,association with people with a low score.\NMeaning that, is there any peer pressure
Dialogue: 0,0:53:31.55,0:53:36.20,Default,,0000,0000,0000,,to stay away from people with bad scores?\NAnswer: The Sesame credit score definitely
Dialogue: 0,0:53:36.20,0:53:42.85,Default,,0000,0000,0000,,is influenced by your friends' scores, the\NRongcheng score, so far, apparently, is
Dialogue: 0,0:53:42.85,0:53:47.62,Default,,0000,0000,0000,,not influenced, but it is definitely in\Nthe cards, and it is planned that it will
Dialogue: 0,0:53:47.62,0:53:53.80,Default,,0000,0000,0000,,be part of this. I think WeChat, which is\Nthe main platform – it's sort of like
Dialogue: 0,0:53:53.80,0:53:59.88,Default,,0000,0000,0000,,WhatsApp, except it can do a lot a lot\Nmore – WeChat is still not connected to
Dialogue: 0,0:53:59.88,0:54:05.04,Default,,0000,0000,0000,,the Social Credit Score in Rongcheng. Once\Nthey do that, it will most likely also
Dialogue: 0,0:54:05.04,0:54:09.85,Default,,0000,0000,0000,,reflect your score.\NHerald: All right, let's continue with
Dialogue: 0,0:54:09.85,0:54:15.43,Default,,0000,0000,0000,,mic 3.\NQ: I have a question about your models.
Dialogue: 0,0:54:15.43,0:54:19.79,Default,,0000,0000,0000,,I'm wondering, what kind of interactions\Nare you modeling? Or actions, like, what
Dialogue: 0,0:54:19.79,0:54:24.91,Default,,0000,0000,0000,,can the agents actually do? You mentioned\Nmoving somewhere else. And, what else?
Dialogue: 0,0:54:24.91,0:54:31.19,Default,,0000,0000,0000,,A: Okay so the way I set up my model was,\NI set up a multilevel model. So I looked
Dialogue: 0,0:54:31.19,0:54:38.34,Default,,0000,0000,0000,,at different kinds of levels. I started\Nout with, basically, they can evade taxes,
Dialogue: 0,0:54:38.34,0:54:46.89,Default,,0000,0000,0000,,they can get loans and repay loans, they\Ncan choose where to live, and they can
Dialogue: 0,0:54:46.89,0:54:54.48,Default,,0000,0000,0000,,follow traffic rules or not follow traffic\Nrules. And because these were, sort of,
Dialogue: 0,0:54:54.48,0:54:58.66,Default,,0000,0000,0000,,four big issues that were mentioned in all\Nof the different systems, so I started out
Dialogue: 0,0:54:58.66,0:55:04.81,Default,,0000,0000,0000,,with these issues, and looked at, what\Nkind of behavior do I see? I used some
Dialogue: 0,0:55:04.81,0:55:11.03,Default,,0000,0000,0000,,research that – some friends of mine\Nactually sent out surveys to people and
Dialogue: 0,0:55:11.03,0:55:16.30,Default,,0000,0000,0000,,asked them "well, you're now part of the\Nsystem. Did your behavior change, and how
Dialogue: 0,0:55:16.30,0:55:23.11,Default,,0000,0000,0000,,did it change depending on your responses,\Ndepending on your score, and depending on
Dialogue: 0,0:55:23.11,0:55:27.73,Default,,0000,0000,0000,,the score system that exists?" And I,\Nbasically, used that, and some other
Dialogue: 0,0:55:27.73,0:55:34.43,Default,,0000,0000,0000,,research on nudging and on behavioral\Nadaptation, to look at how likely is it
Dialogue: 0,0:55:34.43,0:55:39.16,Default,,0000,0000,0000,,that someone would change their behavior\Nbased on the score.
Dialogue: 0,0:55:39.16,0:55:42.11,Default,,0000,0000,0000,,Herald: All right let's do another\Nquestion from the interwebs.
Dialogue: 0,0:55:42.11,0:55:48.49,Default,,0000,0000,0000,,Q: Yeah, it's actually two questions in\None. How does this system work for Chinese
Dialogue: 0,0:55:48.49,0:55:53.86,Default,,0000,0000,0000,,people living abroad, or for noncitizens\Nthat do business in China?
Dialogue: 0,0:55:53.86,0:56:00.53,Default,,0000,0000,0000,,A: Currently the system does not work for\Nnoncitizens that do business in China,
Dialogue: 0,0:56:00.53,0:56:04.70,Default,,0000,0000,0000,,because it works through the {\i1}shēnfènzhèng{\i0}.\NYou only get a {\i1}shēnfènzhèng{\i0} if you're a
Dialogue: 0,0:56:04.70,0:56:10.79,Default,,0000,0000,0000,,Chinese citizen or you live in China for\N10 or more years. So everyone who is not
Dialogue: 0,0:56:10.79,0:56:16.03,Default,,0000,0000,0000,,Chinese is currently excluded. Chinese\Npeople not living in China, if they have a
Dialogue: 0,0:56:16.03,0:56:21.10,Default,,0000,0000,0000,,{\i1}shēnfènzhèng{\i0}, are on this system, but\Nthere's not a lot of information.
Dialogue: 0,0:56:21.10,0:56:28.86,Default,,0000,0000,0000,,Herald: All right, mic 4.\NQ: Well, we've come a long way since the
Dialogue: 0,0:56:28.86,0:56:34.03,Default,,0000,0000,0000,,{\i1}Volkszählungsurteil{\i0}. Can you\Ntell us anything about the dynamic in the
Dialogue: 0,0:56:34.03,0:56:43.70,Default,,0000,0000,0000,,time dimension? How quickly can I regain\Ncredit that was lost? Do you have any
Dialogue: 0,0:56:43.70,0:56:47.49,Default,,0000,0000,0000,,observations there?\NA: So in the Suining system what they
Dialogue: 0,0:56:47.49,0:56:53.52,Default,,0000,0000,0000,,actually did was they had a very, very\Nstrict period. So if you evaded taxes your
Dialogue: 0,0:56:53.52,0:56:58.60,Default,,0000,0000,0000,,score would be down for two years and then\Nit would rebounce. In the Rongcheng
Dialogue: 0,0:56:58.60,0:57:03.36,Default,,0000,0000,0000,,system, they did not publish this kind of\Nperiod. So, my assumption is that it's
Dialogue: 0,0:57:03.36,0:57:08.76,Default,,0000,0000,0000,,going to be more on a case by case basis.\NBecause, I looked at the Chinese data, I
Dialogue: 0,0:57:08.76,0:57:13.85,Default,,0000,0000,0000,,looked at the Chinese policy documents,\Nand they didn't really, for most of the
Dialogue: 0,0:57:13.85,0:57:20.02,Default,,0000,0000,0000,,stuff, they didn't say how long it would\Ncount. For the blacklists, which was kind
Dialogue: 0,0:57:20.02,0:57:24.63,Default,,0000,0000,0000,,of the predecessor that we look at\Ncurrently, the way it works is you stay on
Dialogue: 0,0:57:24.63,0:57:29.31,Default,,0000,0000,0000,,there until whatever the blacklist –\Nuntil whatever the reason for the
Dialogue: 0,0:57:29.31,0:57:34.51,Default,,0000,0000,0000,,blacklist is has been resolved. So, you\Nstay on there until you send off this
Dialogue: 0,0:57:34.51,0:57:40.04,Default,,0000,0000,0000,,apology that the judge ordered you to. And\Nthen, usually, you still needed to apply to
Dialogue: 0,0:57:40.04,0:57:44.98,Default,,0000,0000,0000,,get off. So it doesn't – for blacklists,\Nit does not work that you automatically
Dialogue: 0,0:57:44.98,0:57:49.41,Default,,0000,0000,0000,,get off. You need to apply, you need to\Nshow that you've done what they've asked
Dialogue: 0,0:57:49.41,0:57:53.09,Default,,0000,0000,0000,,you to do, and then you can get off this\Nblacklist. And I assume it will be a
Dialogue: 0,0:57:53.09,0:57:57.02,Default,,0000,0000,0000,,similar sort of appeals procedure for the\Nsystem.
Dialogue: 0,0:57:57.02,0:58:04.07,Default,,0000,0000,0000,,Herald: All right. Let's go to mic 2.\NQ: Thank you. I just wanted to if looking
Dialogue: 0,0:58:04.07,0:58:08.91,Default,,0000,0000,0000,,up someone else's data in details, like\Nposition et cetera, does affect your own
Dialogue: 0,0:58:08.91,0:58:12.05,Default,,0000,0000,0000,,score?\NA: Currently, it apparently does not, or
Dialogue: 0,0:58:12.05,0:58:17.37,Default,,0000,0000,0000,,at least they haven't published that it\Ndoes. It might in the future, but most
Dialogue: 0,0:58:17.37,0:58:22.55,Default,,0000,0000,0000,,likely it's actually behavior that they\Nwant. So they want you to look up other
Dialogue: 0,0:58:22.55,0:58:26.48,Default,,0000,0000,0000,,people's scores before doing business with\Nthem. They want you to, basically, use
Dialogue: 0,0:58:26.48,0:58:30.07,Default,,0000,0000,0000,,this to decide who you're going to\Nassociate with.
Dialogue: 0,0:58:30.07,0:58:33.20,Default,,0000,0000,0000,,Q: Thank you!\NHerald: All right, do we have another
Dialogue: 0,0:58:33.20,0:58:38.75,Default,,0000,0000,0000,,question from the Internet, maybe?\NSignal: Yes, I do! Standby... The question
Dialogue: 0,0:58:38.75,0:58:47.57,Default,,0000,0000,0000,,is, how is this actually implemented for\Nthe offline rural population in China?
Dialogue: 0,0:58:47.57,0:58:53.12,Default,,0000,0000,0000,,A: Quite easily; not at all at this point.\NThe idea is, by 2020, that they will
Dialogue: 0,0:58:53.12,0:58:59.09,Default,,0000,0000,0000,,actually have all of this is implemented.\NBut even for the offline – or let's say
Dialogue: 0,0:58:59.09,0:59:05.87,Default,,0000,0000,0000,,offline rural population in China is\Ngetting smaller and smaller. Even in rural
Dialogue: 0,0:59:05.87,0:59:13.04,Default,,0000,0000,0000,,villages you have about 50-60% of people\Nthat are online. And most of them are
Dialogue: 0,0:59:13.04,0:59:16.05,Default,,0000,0000,0000,,online via smartphone, and their\Nsmartphone is connected to the
Dialogue: 0,0:59:16.05,0:59:21.31,Default,,0000,0000,0000,,{\i1}shēnfènzhèng{\i0}. So it's not very complicated\Nto do that for everyone who is online. For
Dialogue: 0,0:59:21.31,0:59:25.99,Default,,0000,0000,0000,,everyone who's offline, off course, this\Nis more problematic, but I think the end
Dialogue: 0,0:59:25.99,0:59:31.80,Default,,0000,0000,0000,,goal is to not have people offline at all.\NHerald: All right. Let's jump right back
Dialogue: 0,0:59:31.80,0:59:38.06,Default,,0000,0000,0000,,to microphone 2, please.\NQ: Thank you for the very good and
Dialogue: 0,0:59:38.06,0:59:44.94,Default,,0000,0000,0000,,frightening talk, so far. At first I have\Nto correct you in one point. In Germany we
Dialogue: 0,0:59:44.94,0:59:50.28,Default,,0000,0000,0000,,have a similar system because we have this\Ntax I.D., which is set from birth on and
Dialogue: 0,0:59:50.28,0:59:58.77,Default,,0000,0000,0000,,rests 30 years after a person's dead.\NYeah. So we have a lifelong I.D.
Dialogue: 0,0:59:58.77,1:00:02.25,Default,,0000,0000,0000,,A: You're right. I just... I don't know\Nmine, so I figured… {\i1}dismissive sound.{\i0}
Dialogue: 0,1:00:02.25,1:00:08.13,Default,,0000,0000,0000,,Q: No problem! But, at least we could\Nestablish a similar system, if we have a
Dialogue: 0,1:00:08.13,1:00:16.15,Default,,0000,0000,0000,,government which would want it. A question\Nfor you: you mentioned this "{\i1}guanxi{\i0}." Is
Dialogue: 0,1:00:16.15,1:00:20.79,Default,,0000,0000,0000,,it a kind of a social network? I didn't\Nunderstand it, really.
Dialogue: 0,1:00:20.79,1:00:26.00,Default,,0000,0000,0000,,A: Yes, it is a kind of social network,\Nbut one that is a lot more based on
Dialogue: 0,1:00:26.00,1:00:31.93,Default,,0000,0000,0000,,hierarchies than it is in the West. So you\Nhave people that are above you and people
Dialogue: 0,1:00:31.93,1:00:36.52,Default,,0000,0000,0000,,that are below you. And the expectation is\Nthat, while it's a quid pro quo, people
Dialogue: 0,1:00:36.52,1:00:41.54,Default,,0000,0000,0000,,that are above you in the hierarchy will\Ngive you less than you will give to them.
Dialogue: 0,1:00:41.54,1:00:46.53,Default,,0000,0000,0000,,Q: Aha, okay.\NHerald: OK, all right. Unfortunately, we
Dialogue: 0,1:00:46.53,1:00:52.10,Default,,0000,0000,0000,,are out of time, so, please give another\Nhuge applause for Toni!
Dialogue: 0,1:00:52.10,1:00:54.39,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,1:00:54.39,1:00:57.84,Default,,0000,0000,0000,,{\i1}postroll music{\i0}
Dialogue: 0,1:00:57.84,1:01:17.00,Default,,0000,0000,0000,,subtitles created by c3subtitles.de\Nin the year 2019. Join, and help us!