Jillian C. York: Hello. So when we thought about this theme… We were both at re:publica in Dublin last year. We got to see the announce­ment of this them of re:publica, and the first thought that we had was how can you love out loud in a time of ubiq­ui­tous sur­veil­lance? And so as we’ve cre­at­ed this talk we’ve set out to answer that ques­tion. I am going to say I’m not rep­re­sent­ing my orga­ni­za­tion today, I’m just speak­ing from my per­son­al per­spec­tive. And I think yeah, let’s take it away.

Mathana: We have a lot of con­tent we’re going to cov­er quick­ly. If any­body wants to dis­cuss things after­wards, we’re more than hap­py to talk. We built kind of a long pre­sen­ta­tion so we’re going to try to fly through, be as infor­ma­tive and delib­er­ate as possible.

So one of the things that we want­ed to dis­cuss is the way in which facial recog­ni­tion tech­nolo­gies are sur­round­ing us more and more. It’s a num­ber of things. One, retro­fitting of old, exist­ing sys­tems like CCTV sys­tems in sub­ways and oth­er trans­porta­tion sys­tems. But as well as new and minia­tur­ized cam­era sys­tems that are now prop­a­gat­ed to a back­end of machine learn­ing neur­al net­works and new AI technologies. 

York: There’s a com­bi­na­tion of data right now. Data that we’re hand­ing over vol­un­tar­i­ly to the social net­works that we take part in, to the dif­fer­ent struc­tures in which we par­tic­i­pate, be it med­ical records, any­thing across that spec­trum. And then there’s data, or cap­ture, that’s being tak­en from us with­out our consent.

Mathana: So our actions cap­tured on film paint a star­tling­ly com­plete pic­ture of our lives. To give you an exam­ple, in 2015 the New York metro sys­tem, the MTA, start­ed installing around 1,000 new video cam­eras. And we think about what this looks like on a dai­ly basis. The things that can be gleaned from this. Transportation pat­terns. What sta­tions you get on and off at. If you are able to be unique­ly iden­ti­fied by some­thing like a metro CCTV inte­grat­ed sys­tem, even when you wake up; the stops that you go to; if you are going to a new stop that you may not nor­mal­ly go to. That we’re actu­al­ly even in this one sys­tem, a star­tling­ly clear pic­ture starts to emerge of our lives in the pas­sive actions that we take on a dai­ly basis, that we’re not even real­ly aware that are being sur­veilled, but they are.

York: And then when that data’s com­bined with all of the oth­er data, both that we’re hand­ing over and that’s being cap­tured about us, it all comes togeth­er to cre­ate this pic­ture that’s both dis­tort­ed but also com­pre­hen­sive in a way.

So I think that we have to ask who’s cre­at­ing this tech­nol­o­gy and who ben­e­fits from it. Who should have the right to col­lect and use infor­ma­tion about our faces and our bod­ies? What are the mech­a­nisms of con­trol? We have gov­ern­ment con­trol on the one hand, cap­i­tal­ism on the oth­er hand, and this murky grey zone between who’s build­ing the tech­nol­o­gy, who’s cap­tur­ing, and who’s ben­e­fit­ing from it.

Mathana: This is going to be one of the focus­es of our talk today, kind of this inter­play between government-driven tech­nol­o­gy and corporate-driven technology—capitalism-driven tech­nol­o­gy. And one of the kind of inter­est­ing crossovers to bring up now is the poach­ing of uni­ver­si­ties and oth­er pub­lic research insti­tu­tions into the pri­vate sec­tor. Carnegie Mellon had around thir­ty of its top self-driving car engi­neers poached by Uber to start their AI depart­ment. And we’re see­ing this more and more, in which this knowl­edge capac­i­ty from the uni­ver­si­ty and resource field is being cap­tured by the cor­po­rate sec­tor. And so when the new advances in tech­nol­o­gy hap­pen it’s real­ly for prof­it com­pa­nies that are the ones that are kind of the tip of the spear now.

York: And one of those things that they do is they suck us in with all of the cool fea­tures. So just raise your hand real quick if you’ve ever par­tic­i­pat­ed in some sort of web site or app or meme that asked you to hand over your pho­to­graph and in exchange get some sort of insight into who you are. I know there’s one going around right now where it kind of changes the gen­der appear­ance of a per­son. Has any­one par­tic­i­pat­ed in these? Okay, excel­lent. I have, too. I’m guilty. 

Does any­one remem­ber this one? So this was this fun lit­tle tool for end­point users, for us to inter­act with this cool fea­ture. And basi­cal­ly what hap­pened was that Microsoft, about two years ago they unveiled this exper­i­ment in machine learn­ing. It was a web site that could guess your age and your gen­der based on a photograph. 

He thought that this was kind of sil­ly to include, like this isn’t a very com­mon exam­ple. But I was actu­al­ly remind­ed of it on Facebook a cou­ple days ago, when it said oh, two years ago today this is what you were doing. And what I was doing with this, and what it was telling me was that my 25 year-old self was like 40 years old. 

So it was­n’t par­tic­u­lar­ly accu­rate tech­nol­o­gy but nev­er­the­less the cre­ators of this par­tic­u­lar demon­stra­tion had been hop­ing ​“opti­misti­cal­ly” to lure in fifty or so peo­ple from the pub­lic to test their prod­uct. But instead with­in hours the site was almost strug­gling to stay online because it was so pop­u­lar, with thou­sands of vis­i­tors from around the world and par­tic­u­lar­ly Turkey for some reason.

There was anoth­er web­site called Face My Age that launched more recent­ly. It does­n’t try to just guess at age and gen­der, but it also asks users to sup­ply infor­ma­tion like their age, like their gen­der, but also oth­er things like their mar­i­tal sta­tus, their edu­ca­tion­al back­ground, whether they’re a smok­er or not. And then to upload pho­tos of their face with no make­up on, unsmil­ing, so that they can try to basi­cal­ly cre­ate a data­base that would help machines to be able to guess your age even better.

And so they say that okay well, smok­ing for exam­ple ages peo­ple’s faces. So we need to have that data so that our machines can get bet­ter and bet­ter learn­ing this, and then bet­ter and bet­ter at guess­ing. And of course because this is pre­sent­ed as a fun exper­i­ment for peo­ple, they will­ing­ly upload their infor­ma­tion with­out think­ing about the ways in which that tech­nol­o­gy may or may not be even­tu­al­ly used.

So Matthew, I’m going to hand it over to you for this next exam­ple because it makes me so angry that I need to cry in a cor­ner for a minute.

Mathana: So, FindFace. VKontakte, VK, is the Russian kin­da Facebook clone. It’s a large plat­form with hun­dreds of mil­lions of users—

York: 410 mil­lion, I think.

Mathana: What some­one did was basi­cal­ly— How’d it go? The sto­ry goes they got access to the pho­to API. So they had kin­da the fire­hose API. They were able to have access to all the pho­tos on on this rather large social media plat­form. Two engi­neers from Moscow/St. Petersburg wrote a facial recog­ni­tion script that basi­cal­ly to date is one of the top facial recog­ni­tion pro­grams. And they were basi­cal­ly able to con­nect this on the fron­tend, which is an app used by users, to be able to query the entire VK pho­to data­base and in a mat­ter of sec­onds return results. So it gives you as a user the pow­er to take a pho­to of some­body on the street, query against the entire social media pho­to data­base, and get match­es that are either the match of a per­son or you can find peo­ple that look sim­i­lar to the per­son that you’re try­ing to identify.

York: So NTechLab, which builds this, won the MegaFace Benchmark (love that name), which is the world cham­pi­onship in face recog­ni­tion orga­nized by the University of Washington. Do you see the inter­play already hap­pen­ing between acad­e­mia and cor­po­ra­tions? The chal­lenge was to rec­og­nize the largest num­ber of peo­ple in this data­base of more than a mil­lion pho­tos. And with their accu­ra­cy rate of 73.3%, they bypassed more than a hun­dred com­peti­tors, includ­ing Google.

Now, here’s the thing about FindFace. It also as Matthew point­ed out looks for ​“sim­i­lar” peo­ple. So this is one of FindFace’s founders (that is a mouth­ful of words), Alexander Kabakov. And he said that you could just upload a pho­to of a movie star you like, or your ex, and then find ten girls who look sim­i­lar to her and send the mes­sages. I’m…real­ly not okay with this. 

And in fact I’ve got to go back a slide to tell you this oth­er sto­ry, which made me even more sad, which is that a cou­ple years ago an artist called Igor Tsvetkov high­light­ed how inva­sive this tech­nol­o­gy could be. He went through St. Petersburg and he pho­tographed ran­dom pas­sen­gers on the sub­way and then matched the pic­tures to the indi­vid­u­als VKontakte pages using FindFace. So, ​“In the­o­ry,” he said, ​“the ser­vice could be used by a ser­i­al killer or col­lec­tor try­ing to hunt down a debtor.”

Well he was not wrong, because what hap­pened was that after his exper­i­ment went live and the media cov­ered it—there was a lot of cov­er­age of it in the media as an art project—another group launched a cam­paign to basi­cal­ly demo­nize pornog­ra­phy actors in Russia by using this to iden­ti­fy them from their porn and then harass them. And so this has already been used in this kind of way that we’re point­ing out as a poten­tial. This is already hap­pen­ing. Stalkertech. 

But you know, I think that one of the real­ly inter­est­ing thing that Matthew found is that as we were look­ing through these dif­fer­ent com­pa­nies that cre­ate facial analy­sis tech­nol­o­gy and emo­tion­al analy­sis tech­nol­o­gy, the way that they’re brand­ed and mar­ket­ed is real­ly interesting. 

Mathana: We can go through some of these exam­ples, even just slide to slide and see. These are some of the top facial recog­ni­tion tech­nol­o­gy com­pa­nies out there. And what’s inter­est­ing, we’re not going to be talk­ing so much about Google, about Facebook, about Amazon, although these com­pa­nies are impor­tant. But we’re here high­light­ing one of the kind of under­stat­ed facts in the facial recog­ni­tion world, is that there are small com­pa­nies pop­ping up that’re build­ing incred­i­bly pow­er­ful and sophis­ti­cat­ed algo­rithms to be able to find facial recog­ni­tion match­es, even in low qual­i­ty, low light, reren­der­ing, and con­vert­ing from 2D to 3D. These sort of things whose names we don’t know. And we can sit back and kind of demo­nize the large tech­nol­o­gy com­pa­nies, but there is a lot to be done to hold small com­pa­nies accountable.

York: And I think you can see the famil­iar thread through all of these, which is what? Anyone want to wager a guess? Smiling hap­py faces. Usually beau­ti­ful women, smil­ing and hap­py, as we saw back on VKontakte’s page.

So, rather than focus­ing on the bad guys, they’re focused on this ​“Oh look! Everyone’s real­ly hap­py when we use these facial recog­ni­tion tech­nolo­gies.” But what a lot of these tech­nolo­gies are doing is in my opin­ion dan­ger­ous. So for exam­ple Kairos, which is a Miami-based facial recog­ni­tion tech­nol­o­gy com­pa­ny, they also own an emo­tion­al analy­sis tech­nol­o­gy com­pa­ny that they acquired a cou­ple of years ago and they they wrapped into their core services. 

And their CEO has said that the impe­tus for that came from their cus­tomers. Some of their cus­tomers are banks. Specifically that a bank teller, when you go to the bank, maybe a bank teller could use facial recog­ni­tion tech­nol­o­gy to iden­ti­fy you. And that would be a bet­ter way than you show­ing your ID or sign­ing some­thing or maybe even enter­ing a PIN. 

But, some­times you could have some­one maybe that day who comes in to rob the bank and their face is kind of show­ing it. And so with that emo­tion­al analy­sis tech­nol­o­gy, the bank teller could have it indi­cat­ed to them that today is a day that they will refuse you service.

But my imme­di­ate thought when I read that was what about peo­ple who live with anx­i­ety? What about peo­ple who are just in a hur­ry that day? So you could lit­er­al­ly be shut out of your own mon­ey because some algo­rithm says that you’re anx­ious or you’re too emo­tion­al that day to be able to do that.

Another exam­ple that I found real­ly trou­bling was this one, Cognitec, which is I think a Dutch com­pa­ny. Theirs does ​“gen­der detec­tion.” So this is used by casi­nos in Macau as well as in oth­er places. I thought gen­der detec­tion was a real­ly fun­ny con­cept, because as our soci­ety’s become more enlight­ened about gen­der and the fact that gen­der is not always a bina­ry thing, these tech­nolo­gies are basi­cal­ly using your facial fea­tures to place you in a gender.

And I’ve test­ed some of these things before online where it tries to guess your gen­der, and it often gets them wrong. But in this case it’s actu­al­ly where does our gen­der auton­o­my even fit into this? Do we have gen­der auton­o­my if these sys­tems are try­ing to place us as one thing or another?

Mathana: So one of things that is real­ly quite star­tling about this is the cul­mi­na­tion of data points. When we’re talk­ing about the way in which young peo­ple are hav­ing their faces scanned ear­li­er, in which iden­ti­fi­ca­tion soft­ware is being used using facial recog­ni­tion tech­nol­o­gy, that no one action may be held in a sta­t­ic kind con­tain­er any­more. That we’re now deal­ing with dynam­ic datasets that are con­tin­u­ous­ly being built around us.

Algorithms learn by being fed cer­tain images, often cho­sen by engi­neers, and the sys­tem builds a mod­el of the world based on those images. If a sys­tem is trained on pho­tos of peo­ple who are over­whelm­ing­ly white, it will have a hard­er time rec­og­niz­ing non­white faces.
Kate Crawford, Artificial Intelligence’s White Guy Problem, New York Times [pre­sen­ta­tion slide]

And one of the issues is that there is an asym­me­try of the way in which we’re seen by these dif­fer­ent sys­tems has, as Kate Crawford has said— (She pre­sent­ed last year on stage one. Some of y’all may have caught that.) And a lot of her research has gone into kind of the engi­neer­ing side. Looking at the ways in which dis­crim­i­na­tion bias repli­cates inside of new tech­nolo­gies. And as facial recog­ni­tion tech­nolo­gies are just one vec­tor these dis­crim­i­na­to­ry algo­rithms, it’s more and more impor­tant that we take a step back and say well, what are the nec­es­sary require­ments and pro­tec­tions and safe­guards to make sure that we don’t end up with with things like this:

With a Google algo­rithm say­ing that a black cou­ple are goril­las. And there’s a lot of oth­er famous exam­ples of this. But if engi­neer­ing teams and if com­pa­nies are not think­ing about this holis­ti­cal­ly from the inside, then there may PR dis­as­ters like this, but the real-world impli­ca­tions of these tech­nolo­gies are very unset­tling and quite scary.

York: Now, Google was made aware of this and they apol­o­gized. And they said that ​“there is still clear­ly a lot of work to do with auto­mat­ic image label­ing, and we’re look­ing at how we can pre­vent these types of mis­takes from hap­pen­ing in the future.” And they have done great work on this but I still think that part of the prob­lem is that these com­pa­nies are not very diverse. And that’s not what this talks about but I can’t help but say it. It’s an impor­tant facet of why these com­pa­nies con­tin­ue to make the same mis­takes over and over again.

But let’s talk a lit­tle bit about what hap­pens when it’s not just a com­pa­ny mak­ing a mis­take in iden­ti­fy­ing peo­ple in an offen­sive way, but when the mis­take has real-world implications. 

Mathana: So, one of things that is real­ly quite startling-interesting— And we were talk­ing about some of the dif­fer­ent algo­rithms for VK. But here is anoth­er exam­ple, that now sci­en­tists around the world and tech­nol­o­gists are train­ing facial recog­ni­tion algo­rithms on data­bas­es. In this case a group of inno­cent peo­ple, a group of ​“guilty peo­ple,” and using machine learn­ing and neur­al net­works to try to dis­cern who is guilty out of the test data and who is [inno­cent].

York: Incidentally, their accu­ra­cy rate on this par­tic­u­lar task was 89.5%.

Mathana: So 89.5%. I mean, for some things almost 90%…not bad. But we’re talk­ing about a 10% rate of either false pos­i­tives or of just errors. And if we’re think­ing about a crim­i­nal jus­tice sys­tem in which one out of every ten peo­ple are sen­tenced incor­rect­ly, we’re talk­ing about a whole tenth of the pop­u­la­tion which at a time in the future may not be able to have access to due process because of auto­mat­ed sen­tenc­ing guide­lines and oth­er things.

Perpetual Line Up

And it’s hap­pen­ing now. Nearly one half of American cit­i­zens have their face in a data­base that’s acces­si­ble by some lev­el of law enforce­ment. And that’s mas­sive. That means that one out of every two adults in the US, their face has been tak­en from them. That their like­ness now resides in a data­base which is able to be used for crim­i­nal jus­tice inves­ti­ga­tions and oth­er things. And so you may not even know that your face is in one of a num­ber of dif­fer­ent data­bas­es, and yet on a dai­ly basis these data­bas­es maybe crawled to look for new match­es for guilty peo­ple or sus­pects. But we’re not aware of this a lot of times. 

York: And it’s not just our faces, it’s also oth­er iden­ti­fy­ing mark­ers about us. It’s our tat­toos, which I’m cov­ered with which and which now I know— I did­n’t know when I got them, but now I know that that’s a way that I can be iden­ti­fied by police so I’m going to have to come up with some sort of thing that cov­ers them with infrared—I don’t even know.

Image: OU-ISIR Biometric Database

It’s also our gait. It’s the way that we walk. And one of the real­ly scary things about this is that while facial recog­ni­tion usu­al­ly requires high-quality images, gait recog­ni­tion does not. If you’re walk­ing on the sub­way plat­form and the CCTV cam­era picks you up—and you know that the U‑Bahn sta­tions are cov­ered with them—if you’re walk­ing that way, a low-bandwidth image is enough to rec­og­nize you by your gait. Yesterday, my boss from across the room rec­og­nized me by my gait. Even our eyes can do this. So it’s possible.

The machines have eyes, and in some ways minds. And as more of these sys­tems become auto­mat­ed, we believe that humans will be increas­ing­ly place out­side of the loop. 

Mathana: So just anoth­er exam­ple, going back to New York, to show how these things real­ly don’t just… We’re now in this inter­con­nect­ed world. I don’t know if you all are famil­iar with stop-and-frisk. It was an unpop­u­lar law in New York City which allowed police offi­cers to essen­tial­ly go up to any­one they might be sus­pi­cious about and ask them for ID and pat them down.

What hap­pened, why this was even­tu­al­ly pulled by the police depart­ment, there were some chal­lenges in court say­ing it was uncon­sti­tu­tion­al. The pol­i­cy was was rescind­ed before it went to court. But the idea of…there are now records from the time that was in place of the peo­ple who were charged under this pol­i­cy. It was found to be very dis­crim­i­na­to­ry in the sense that young men of col­or were dis­pro­por­tion­ate­ly tar­get­ed by this program. 

So if we’re look­ing at crime sta­tis­tics, let’s just say a read­out of the num­ber of arrests in New York City. And if we were to put demo­graph­ic data with that and then feed this into a machine learn­ing algo­rithm, if the machine learn­ing algo­rithms sees that a large per­cent­age of indi­vid­u­als are young black and brown men, what is the machine learn­ing algo­rithm to think except that these indi­vid­u­als have a high­er like­li­hood of com­mit­ting crimes?

In the real world, it was real-world bias by police offi­cers that were tar­get­ing minor­i­ty com­mu­ni­ties. But a machine algorithm…if we’re not weight­ing this sort of infor­ma­tion in the test and train­ing data, there’s no way for a machine to log­i­cal­ly or intu­itive­ly see a causal rela­tion­ship between seg­re­ga­tion and bias in the real world, and the crime sta­tis­tics that are the result from that.

York: So with that in mind, we want to talk a lit­tle bit about how we can love out loud in a future where ubiq­ui­tous cap­ture is even big­ger than it is now. So Matthew, tell me a lit­tle bit about this exam­ple from Sesame Credit. Has any­one heard of Sesame cred­it? Okay, so we’ve got a few peo­ple who are famil­iar with it.

Mathana: So Sesame Credit is a sys­tem that’s now been imple­ment­ed in China—there’s an ongo­ing roll­out. It’s a social cred­it rat­ing, essen­tial­ly. It uses a num­ber of dif­fer­ent fac­tors. It’s being pio­neered by Ant Financial, which is a large finan­cial insti­tu­tion in China. It’s a com­pa­ny, it’s not tech­ni­cal­ly a state-owned enter­prise, but it’s fuzzy when it comes to the rul­ing Chinese Communist Party.

The inter­est­ing thing about this new sys­tem is that it uses things like your… It’ll look into things like your WeChat account and see who you are talk­ing with. And if peo­ple are dis­cussing sen­si­tive things, your social cred­it rat­ing can be docked. 

So now we’re see­ing this sys­tem devel­oped right now in China that brings in ele­ments from your social life, your net­work con­nec­tions, as well as things like your cred­it his­to­ry, to paint a pic­ture of how good of a cit­i­zen are you. Alipay is now launch­ing the US. I think it was announced today. And so they’re now pio­neer­ing some very sophis­ti­cat­ed tech­nol­o­gy like iris scans. And the con­tact­less mar­ket in China has explod­ed from a few hun­dred mil­lions of dol­lars now into the hun­dreds of bil­lions of dol­lars. And so this tech­nol­o­gy in the last even twenty-four months has become much more preva­lent and has become much more ubiq­ui­tous for its capac­i­ty for indi­vid­ual surveillance.

York: And this is where life begins to resem­ble an episode of Black Mirror. So what we’d like to remind you is that dig­i­tal images aren’t sta­t­ic. With each new devel­op­ment, each sweep of an algo­rithm, each time you put some­thing there you’ve left it there. I know that I’ve got hun­dreds, pos­si­bly thou­sands of images sit­ting on Flickr. With each new sweep of an algo­rithm, these images are being reassessed. They’re being recon­sid­ered and rei­de­al­ized to match oth­er data that this com­pa­ny or X Company or a gov­ern­ment might have on you. What you share today may mean some­thing else tomorrow.

So right now we feel that there’s no uni­ver­sal rea­son­able expec­ta­tion that exists between our­selves and our tech­nol­o­gy. The con­se­quence of data aggre­ga­tion is that increased cap­ture of our per­son­al infor­ma­tion results in this more robust, yet dis­tort­ed, pic­ture of who we are that we men­tioned at the beginning.

And so I think that we’ll take the last few min­utes, and we’ll try to leave a few min­utes for ques­tions, just to talk about that emerg­ing social con­tract that we would like to see exist, we would like to see forged between us and tech­nol­o­gy com­pa­nies and governments.

We can’t see behind the cur­tain. We have no way of know­ing how the col­lec­tion of our visu­al imagery is even being used, aggre­gat­ed, or repur­posed. And we want to remind you also that these tech­nolo­gies are mech­a­nisms of con­trol. And so the first ques­tion that I want to ask, per­son­al­ly, is what kind of world do we want? I think that’s the start­ing point, is ask­ing do we want a world where our faces are cap­tured all the time? Where I can walk down this hall­way and have dif­fer­ent cam­eras that are attached to dif­fer­ent com­pa­nies that have dif­fer­ent meth­ods and modes of analy­sis look­ing at me and try­ing to decide who I am and mak­ing deter­mi­na­tions about me.

But per­haps we’re past that point, and so we’ve decid­ed to be prag­mat­ic a lit­tle bit in try­ing to for­mu­late some things that we can do. So in terms of what we want, we want active life that’s free from pas­sive sur­veil­lance. We want more con­trol over our choic­es and over the images that we share. And we want a tech­nol­o­gy mar­ket that isn’t based on sell­ing us out to the high­est bid­der. Luckily there are some peo­ple work­ing on all of these things right now, not just us, and so we feel real­ly sup­port­ed in these choic­es. And we’ll turn to you for regulation.

Mathana: I’ve had the oppor­tu­ni­ty to sit in on a cou­ple of smart cities ses­sions yes­ter­day, talk­ing between devel­op­ment of smart cities in Barcelona and Berlin, as well as smart devices yes­ter­day. We are I think see­ing to some degree a devel­op­ment of a glob­al set of best prac­tices, but very piece­meal and frag­ment­ed. And I think that if we think about image recog­ni­tion tech­nol­o­gy as kind of a lay­er that fits on the many of the oth­er modes of tech­no­log­i­cal devel­op­ment, that it becomes clear that actu­al­ly we need to have some sort of best prac­tices as bio­met­ric data­bas­es con­tin­ue to be aggregated.

It’s very dif­fi­cult for one per­son in one coun­try and a dif­fer­ent per­son in a dif­fer­ent coun­try to have rea­son­able expec­ta­tions of what the best prac­tices are going for­ward. I mean, maybe today it’s pos­si­ble but twen­ty years from now what’s the world look like that we want to live in?

York: So these are some of the areas that we think gov­ern­ments, and par­tic­u­lar­ly local gov­ern­ments, can inter­vene. We also think that we can have bet­ter pri­va­cy stan­dards for tech­nol­o­gy. And we know that there are a lot of pri­va­cy advo­cates here and that those things are already being worked on, too. So we want to acknowl­edge the great work of all the orga­ni­za­tions that are fight­ing for this. But we see one way to this is seg­ment pri­va­cy by feature—location, visu­al rep­re­sen­ta­tion, search, social, and move­ment, all of these dif­fer­ent areas in which our pri­va­cy is being violated. 

We also think user-centric pri­va­cy con­trols… Right now, most pri­va­cy con­trols on dif­fer­ent plat­forms that you use are not real­ly user-friendly. And trust me, I spend a lot of time on these platforms. 

And then anoth­er thing, too, that’s real­ly impor­tant to me—and the rest of my life I work on censorship—but I think for­ward con­sent. I don’t feel that I am con­sent­ing to these fifteen-page terms and con­di­tions doc­u­ments that com­pa­nies try to make as con­fus­ing for me as pos­si­ble. And so I think that if com­pa­nies keep in mind for­ward con­sent every time that you use their ser­vice, that’s one way that they can man­age this problem.

But also, we think that you have to keep lov­ing out loud. That you can’t hide. That you can’t live in fear. That just because these sys­tems are out there, yes of course we have to take pre­cau­tions. I talk a lot in my day job about dig­i­tal secu­ri­ty. And I think that this is that same area, where we have to con­tin­ue liv­ing the way that we want to live. We can take pre­cau­tions, but we can’t sac­ri­fice our lives out of fear.

And so one thing is pho­to aware­ness. We’re real­ly glad to see that at a lot of con­fer­ences recent­ly there’ve been ways of iden­ti­fy­ing your­self if you don’t want to be pho­tographed. But also in clubs in Berlin if any­one’s ever gone club­bing (and if you haven’t you prob­a­bly should), usu­al­ly they put a stick­er over your cam­era when you walk in the front door. And that to me… The first time I saw that I was so elat­ed that I think I danced til 7AM. I mean, that’s not normal.

We also think that you know, reg­u­late your own spaces. I don’t want to get into this divi­sion of public/private prop­er­ty. That’s not what I’m here for. But I think that we have our own spaces and we decide inside of those what’s accept­able and what isn’t. And that includes con­fer­ences like this. So if we feel in these spaces—I don’t real­ly know what’s going on with cam­eras here, if they exist or not. But if we feel in these spaces that that’s some­thing we want to take con­trol of, then we should band togeth­er and do that. 

Put sta­t­ic in the sys­tem? [Gestures to Mathana.]

Mathana: This is a gen­er­al point but in any sys­tem that is being focused on, how can we find ways to put sta­t­ic in the sys­tem, whether this is tag­ging a pho­to that’s not you with your name on Facebook, or whether that’s… Yeah, these sort of strate­gies, right. So it’s ways in which, how can we think a lit­tle more out­side to actu­al­ly con­fuse the algo­rithms or to make their jobs a lit­tle more dif­fi­cult. And there’s dif­fer­ent ways to do this, whether it’s wear­ing reflec­tive cloth­ing, anti-paparazzi cloth­ing in pub­lic, or tag­ging things under one label that may not be that.

York: It also means you know, wear­ing those flash-proof gar­ments, cov­er­ing your face, going to buy your burn­er phone in the store and wear­ing a Halloween cos­tume. I’m not say­ing you should do that [whis­per­ing] but you should total­ly do that. 

And con­tin­u­ing to love out loud and not live in fear. I can see that we’ve com­plete­ly run out of time because I think the sched­ule’s a lit­tle bit behind. But there’s some good news. If you want to keep talk­ing to us about this, we’re both pret­ty eas­i­ly acces­si­ble. But we’re also going to take advan­tage of that sun­light that did­n’t exist yes­ter­day and go out back for a cel­e­bra­to­ry beer. So if you want to keep talk­ing about this sub­ject you’re wel­come to join us out there. Thank you so much.