0:00:06.440,0:00:18.430 35C3 Intro music 0:00:19.490,0:00:25.046 Herald Angel: We at the Congress, we not[br]only talk about technology, we also talk 0:00:25.046,0:00:30.990 about social and ethical responsibility.[br]About how we can change the world for 0:00:30.990,0:00:36.190 good. The Good Technology Collective[br]supports the development guidelines... 0:00:36.190,0:00:41.289 sorry, it supports the development process[br]of new technology with ethical engineering 0:00:41.289,0:00:47.829 guidelines, that offer a practical way to[br]take ethic and social impact into account. 0:00:47.829,0:00:54.235 Yannick Leretaille - and I hope this was[br]okay - will tell you more about it. 0:00:54.235,0:00:57.820 Please welcome on stage with a very warm applause[br]Yann Leretaille. 0:00:57.820,0:01:02.649 applause 0:01:02.649,0:01:07.060 Yannick Leretaille: Hi, thanks for the[br]introduction. So before we start, can you 0:01:07.060,0:01:11.360 kind of show me your hand if you, like,[br]work in tech building products as 0:01:11.360,0:01:17.720 designers, engineers, coders, product[br]management? OK, so it's like 95 percent, 0:01:17.720,0:01:27.820 90 percent. Great. Yeah. So, today we kind[br]of try to answer the question: What is 0:01:27.820,0:01:33.560 good technology and how can we build[br]better technology. Before that, shortly 0:01:33.560,0:01:39.810 something of me. So I am Yann. I'm French-[br]German. Kind of a hacker, among the CCC 0:01:39.810,0:01:44.140 for a long time, entrepreneur, like, co-[br]founder of a startup in Berlin. And I'm 0:01:44.140,0:01:48.110 also founding member of the Good[br]Technology Collective. The Good 0:01:48.110,0:01:55.920 Technology Collective was founded about a[br]year ago or almost over a year now actually 0:01:55.920,0:02:02.250 by a very diverse expert council[br]and we kinda have like 3 areas of work. 0:02:02.250,0:02:07.660 The first one is trying to educate the[br]public about current issues with 0:02:07.660,0:02:12.190 technology, then, to educate engineers or[br]to build better technology, and then 0:02:12.190,0:02:19.980 long-term hopefully one day we'll be[br]able to work like in legislation as well. 0:02:19.980,0:02:27.250 Here, it's a bit of what we achieved so[br]far. We've like 27 council members now. We 0:02:27.250,0:02:30.870 have several media partnerships and[br]published around 20 articles, that's kind 0:02:30.870,0:02:36.260 of the public education part. Then we[br]organized or participated in roughly 15 0:02:36.260,0:02:44.780 events already. And we are now publishing[br]one standard, well, kind of today 0:02:44.780,0:02:49.459 actually, and[br]applause 0:02:49.459,0:02:54.290 and if you're interested in what we do,[br]then, yeah, sign up for the newsletter and 0:02:54.290,0:03:00.290 we keep you up to date and you can join[br]events. So as I said the Expert Council is 0:03:00.290,0:03:08.290 really, really diverse. We have everything[br]from people in academia, to people in 0:03:08.290,0:03:13.380 government, to technology makers, to[br]philosophers, all sorts, journalists. 0:03:13.380,0:03:22.260 And the reason that is the case is that a year[br]ago we kind of noticed that in our own 0:03:22.260,0:03:27.840 circles, like, as technology makers or[br]academics, we were all talking about a lot 0:03:27.840,0:03:33.209 of, kind of, voice on development and[br]technology, but no one was really kind of 0:03:33.209,0:03:37.470 getting together and looking at it from[br]all angles. And there have been a lot of 0:03:37.470,0:03:43.580 very weird and troublesome developments in[br]the last two years. I think we really 0:03:43.580,0:03:49.010 finally feel, you know like, the impact of[br]filter bubbles. Something we have talked 0:03:49.010,0:03:54.310 for like five years, but now it's like,[br]really like, you know, deciding over 0:03:54.310,0:04:01.290 elections and people become politically[br]radicalized and society is, kind of, 0:04:01.290,0:04:06.360 polarized more because they only see a[br]certain opinion. And we have situations 0:04:06.360,0:04:11.310 that we only knew, like, from science[br]fiction, just kind of, you know, pre-crime, 0:04:11.310,0:04:16.629 like, governments, kind of, over-arching[br]and trying to use machine learning to make 0:04:16.629,0:04:22.820 decisions on whether or not you should go[br]to jail. We have more and more machine 0:04:22.820,0:04:26.820 learning and big data and automization[br]going into basically every single aspect 0:04:26.820,0:04:31.620 of our lives and not all of it been has[br]been positive. You know, like, literally 0:04:31.620,0:04:36.510 everything from e-commerce to banking to[br]navigating to moving to the vault now goes 0:04:36.510,0:04:45.889 through these interfaces. That present us[br]the data and a slice of the world at a time. 0:04:45.889,0:04:49.699 And then at the same time we have[br]really positive developments. Right? We have 0:04:49.699,0:04:54.180 things like this, you know, like space[br]travel, finally something's happening. 0:04:54.180,0:05:00.850 And we have huge advances in medicine. Maybe[br]soon we'll have, like, self-driving cars 0:05:00.850,0:05:10.870 and great renewable technology. And it kind[br]of begs the question: How can it be that 0:05:10.870,0:05:16.889 good and bad use of technology are kind of[br]showing up at such an increasing rate in 0:05:16.889,0:05:25.130 this, like, on such extremes, right? And[br]maybe the reason is just that everything 0:05:25.130,0:05:30.330 got so complicated, right? Data is[br]basically doubling every couple of years, 0:05:30.330,0:05:35.620 so no human can possibly process anymore.[br]So we had to build more and more complex 0:05:35.620,0:05:40.979 algorithms to process it, connecting more[br]and more parts together. And no one really 0:05:40.979,0:05:46.020 seems to understand it anymore, it seems.[br]And that leads to unintended consequences. 0:05:46.020,0:05:50.270 I've an example here: So, Google Photos –[br]this is actually only two years ago – 0:05:50.270,0:05:56.960 launched a classifier to automatically go[br]through all of your pictures and tell you 0:05:56.960,0:06:01.254 what it is. You could say "Show me the[br]picture of the bird in summer at this 0:06:01.254,0:06:06.330 location" and it would find it for you.[br]Kind of really cool technology, and they 0:06:06.330,0:06:11.180 released it to, like, a planetary user[br]base until someone figured out that people 0:06:11.180,0:06:17.090 of color were always marked as gorillas.[br]Of course it was a huge PR disaster, why 0:06:17.090,0:06:21.680 somehow no one found out about this before[br]it came out... But now the interesting thing 0:06:21.680,0:06:27.240 is: In two years they didn't even manage[br]to fix it! Their solution was to just 0:06:27.240,0:06:32.853 block all kind of apes, so they're just[br]not found anymore. And that's how they 0:06:32.853,0:06:37.645 solved it, right? But if even Google can't [br]solve this... what does it mean? 0:06:38.795,0:06:42.245 And then, at the same time, you know, [br]sometimes we seem to have, kind of, 0:06:42.245,0:06:44.130 [br]intended consequences? 0:06:45.640,0:06:49.650 I have an example... another example here:[br]Uber Greyball. I don't know if anyone 0:06:49.650,0:06:58.490 heard about it. So Uber was very eager to[br]change regulation and push the services 0:06:58.490,0:07:03.210 globally as much as possible, and kind of[br]starting a fight with, you know, all the 0:07:03.210,0:07:06.539 taxi laws and regulation, and taxi[br]drivers in the various countries around the 0:07:06.539,0:07:11.880 world. And what they realized, of course,[br]is that they didn't really want people to 0:07:11.880,0:07:18.620 be able to, like, investigate what they[br]were doing or, like, finding individual 0:07:18.620,0:07:22.170 drivers. So they built this absolutely[br]massive operation which was like following 0:07:22.170,0:07:27.120 data in social media profiles, linking,[br]like, your credit card and location data 0:07:27.120,0:07:31.039 to find out if you were working for the[br]government. And if you did, you would just 0:07:31.039,0:07:36.680 never find a car. It would just not show[br]up, right? And that was clearly 0:07:36.680,0:07:41.307 intentional, all right. So at the same[br]time they were pushing, like, on, like, 0:07:41.307,0:07:46.490 the lobbyism, political side to change[br]regulation, while heavily manipulating the 0:07:46.490,0:07:52.039 people that were pushing to change the[br]regulation, right? Which is really not a 0:07:52.039,0:07:54.754 very nice thing to do, I would say. 0:07:55.534,0:07:56.917 And... 0:07:58.417,0:08:02.711 The thing that I find, kind of... [br]worrisome about this: 0:08:02.711,0:08:08.207 No matter if it's intended or unintended, [br]is that it actually gets worse, right? 0:08:08.207,0:08:12.746 The more and more systems we[br]interconnect, the worse these consequences 0:08:12.746,0:08:17.300 can get. And I've an example here: So this[br]is a screenshot I took of Google Maps 0:08:17.300,0:08:24.100 yesterday and you notice there are, like,[br]certain locations... So they're kind of 0:08:24.100,0:08:28.160 highlighted on this map and I don't know[br]if you knew it but this map and the 0:08:28.160,0:08:32.510 locations that Google highlight look[br]different for every single person. 0:08:32.510,0:08:38.130 Actually, I went again and looked today[br]and it looked different again. So, Google 0:08:38.130,0:08:42.520 is already heavily filtering and kind of[br]highlighting certain places, like, maybe 0:08:42.520,0:08:48.850 this restaurant over there, if you can see[br]it. And I would say, like, from just 0:08:48.850,0:08:53.550 opening the map, that's not obvious to you[br]that it's doing that. Or that it's trying 0:08:53.550,0:08:59.130 to decide for you which place is[br]interesting for you. However, that's 0:08:59.130,0:09:05.620 probably not such a big issue. But the[br]same company, Google with Waymo, is also 0:09:05.620,0:09:11.360 developing this – and they just started[br]deploying them: self-driving cars. They're... 0:09:12.720,0:09:18.690 ...still a good couple of years away from[br]actually making it reality, but they are 0:09:18.690,0:09:22.547 really – in terms of, like, all the others[br]trying it at the moment – the farthest, I 0:09:22.547,0:09:28.915 would say, and in some cities they started[br]deploying self-driving cars. So now, just 0:09:28.915,0:09:33.892 think like 5, 10 years into the future[br]and you have signed up in your Google... 0:09:35.246,0:09:38.834 ...self-driving car. Probably you don't [br]have your own car, right? And you go in 0:09:38.834,0:09:44.381 the car and you are like: "Hey, Yann, where[br]do you want to go?" Do you want to go to 0:09:44.381,0:09:49.770 work? Because, I mean obviously that's why[br]I probably go most of the time. Do you 0:09:49.770,0:09:52.620 want to go to your favorite Asian[br]restaurant, like the one we just saw on the 0:09:52.620,0:09:56.700 map? Which is actually not my favorite,[br]but the first one I went to. So Google 0:09:56.700,0:10:00.300 just assumed it was. Do you want to go to[br]another Asian restaurant? Because, 0:10:00.300,0:10:06.960 obviously, that's all I like. And then[br]McDonald's. Because, everyone goes there. 0:10:06.960,0:10:11.420 And maybe the fifth entry is an[br]advertisement. And you would say: Well, 0:10:11.420,0:10:17.899 Yann, you know, that's still kind of fine,[br]but it's OK because I can still click on: 0:10:17.899,0:10:24.750 'No, I don't want these 5 options, give me,[br]like, the full map.' But now, we went back 0:10:24.750,0:10:31.050 here. So, even though you are seeing the[br]map, you're not actually not seeing all 0:10:31.050,0:10:35.650 the choices, right? Google is actually[br]filtering for you where it thinks you want 0:10:35.650,0:10:42.910 to go. So now we have, you know, the car[br]like this symbol of mobility and freedom. 0:10:42.910,0:10:49.790 It enables so much change in our society[br]that it's actually reducing the part of 0:10:49.790,0:10:53.910 the world that you see. And because, I[br]mean these days they call it AI, I think 0:10:53.910,0:10:58.660 it's just machine learning, because these[br]machine learning algorithms all do pattern 0:10:58.660,0:11:04.730 matching and basically just can recognize[br]similarities. When you open the map and 0:11:04.730,0:11:09.330 you zoom in and you select a random place,[br]it would only suggest places to you where 0:11:09.330,0:11:14.730 other people have been before. So now the[br]restaurant that opened around the corner 0:11:14.730,0:11:19.010 you'll probably not even discover it[br]anymore. And no one will. And it will 0:11:19.010,0:11:23.080 probably close. And the only ones that[br]will stay are the ones that are already 0:11:23.080,0:11:32.410 established now. And all of that without[br]being really obvious to anyone who would 0:11:32.410,0:11:40.200 use the technology. Because it has become[br]like kind of a black box. So, I do want 0:11:40.200,0:11:47.740 self-driving cars, I really do. I don't[br]want a future like this. Right. And if we 0:11:47.740,0:11:52.850 want to prevent that future, I think we[br]have to first ask a very simple question, 0:11:52.850,0:11:59.519 which is: Who is responsible for designing[br]these products? So, do you know the 0:11:59.519,0:12:01.519 answer?[br]audience: inaudible 0:12:01.519,0:12:05.200 Yann: Say it louder.[br]audience: We are. 0:12:05.200,0:12:10.220 Yann: Yeah, we are. Right. That's a really[br]frustrating thing about it that actually 0:12:10.220,0:12:15.490 gets us, right, as engineers and[br]developers. You know we are always driven 0:12:15.490,0:12:20.230 by perfection. We want to create, like,[br]the perfect code sources. One problem, 0:12:20.230,0:12:25.420 really, really nice. You know. Chasing the[br]next challenge over and over trying to be 0:12:25.420,0:12:31.950 first. But we have to realize that at the[br]same time we are kind of working on 0:12:31.950,0:12:37.040 frontier technologies, right, on things,[br]technology, that are really kind of on the 0:12:37.040,0:12:42.230 edge of values and norms we have in[br]society. And if we are not careful and 0:12:42.230,0:12:46.450 just, like, focus on our small problem and[br]don't look at the big picture, then we 0:12:46.450,0:12:52.270 have no say in on which side of the coin[br]the technology will fall. And probably it 0:12:52.270,0:12:58.680 will take a couple of years, or by that[br]time we alreaday moved on, I guess. So. 0:12:58.680,0:13:06.620 It's just that technology has become so[br]powerful and interconnected and impactful, 0:13:06.620,0:13:10.720 because we are not building stuff that[br]it's not affecting like 10 or 100 people 0:13:10.720,0:13:14.520 or a city but literally millions of[br]people, that we really have to take a step 0:13:14.520,0:13:20.950 back and not only look at the individual[br]problem as the challenge but also the big 0:13:20.950,0:13:27.120 picture. And I think if you want to do[br]that we have to start by asking the right 0:13:27.120,0:13:33.510 questions. And the first question of[br]course is: What is good technology? So, 0:13:33.510,0:13:39.250 that's also the name of the talk.[br]Unfortunately, I don't have a perfect 0:13:39.250,0:13:45.680 answer for that. And probably we will[br]never find a perfect answer for that. So, 0:13:45.680,0:13:53.060 what I would like to propose is to[br]establish some guidelines and engineering 0:13:53.060,0:13:57.980 processes that help us to build better[br]technology. To kind of ensure the same 0:13:57.980,0:14:03.959 where we have quality insurance and[br]project management systems and processes 0:14:03.959,0:14:09.480 to, like, kind of, this you were tasked[br]with. And companies that what we build is 0:14:09.480,0:14:15.589 actually, has a net positive outcome for[br]society. And we call it the good 0:14:15.589,0:14:22.130 technology standard. We've kind of been[br]working that over, the last year, and we 0:14:22.130,0:14:26.840 really wanted to make it really practical.[br]And what we kind of realized is that if you 0:14:26.840,0:14:32.250 want to make it practical you have to make[br]it very easy to use and also mostly, 0:14:32.250,0:14:39.000 actually what was surprising, just ask the[br]right questions. So, what is important 0:14:39.000,0:14:46.270 though, is that if you adapt the standard,[br]it has to be in all project phases. It has 0:14:46.270,0:14:50.200 to involve everyone. So, from, like, the[br]CTO to, like, the product managers to 0:14:50.200,0:14:55.820 actually legal. Today, legal has this[br]interesting role, where you develop 0:14:55.820,0:15:00.350 something and then you're like: Okay, now,[br]legal, make sure that we can actually ship it. 0:15:00.350,0:15:06.480 And that's what usually happens. And,[br]yeah, down to the individual engineer. And 0:15:06.480,0:15:09.940 if it's not applied globally and people[br]start making exceptions then of course it 0:15:09.940,0:15:17.970 won't be worth very much. Generally, we[br]kind of identified four main areas that we 0:15:17.970,0:15:22.730 think are important, kind of defining,[br]kind of an abstract way, if a product is 0:15:22.730,0:15:30.470 good. And the first one is empowerment. A[br]good product should empower its users. And 0:15:30.470,0:15:36.290 that's kind of a tricky thing. So, as[br]humans we have very limited decision 0:15:36.290,0:15:40.360 power. Right? And we are faced with, as I[br]said before, like, this huge amount of 0:15:40.360,0:15:46.339 data and choices. So it seems very natural[br]to build machines and interfaces that try 0:15:46.339,0:15:51.260 to make a lot of decisions for us. Like[br]the Google Maps one we saw before. But we 0:15:51.260,0:15:55.730 have to be careful because if we do that[br]too much then the machine ends up making 0:15:55.730,0:16:03.079 all decisions for us. So often, when you[br]develop something you should really ask 0:16:03.079,0:16:07.320 yourself, like, in the end if I take[br]everything together am I actually 0:16:07.320,0:16:12.880 empowering users, or am I taking[br]responsibility away from them? Do I 0:16:12.880,0:16:18.360 respect the individual choice? Why does he[br]say: I don't want this, or they give you 0:16:18.360,0:16:23.570 their preference, do we actually respect[br]it or do we still try to, you know, just 0:16:23.570,0:16:29.639 figure out what is better for them. Do my[br]users actually feel like they benefit from 0:16:29.639,0:16:34.330 using the product? So, I couldn't,[br]actually not a lot of people ask themselves, 0:16:34.330,0:16:39.720 because usually you think like in terms[br]of: Are you benefiting your company? And I 0:16:39.720,0:16:45.699 think what's really pressing in that[br]aspect: does it help the users, the humans 0:16:45.699,0:16:53.529 behind it, to grow in any way. If it helps[br]them to be more effective or faster or do 0:16:53.529,0:16:57.990 more things or be more relaxed or more[br]healthy, right, then it's probably positive. 0:16:57.990,0:17:02.100 But if you can't identify any of these,[br]then you really have to think about it. 0:17:02.100,0:17:08.789 And then, in terms of AI, in machine[br]learning, are we actually kind of 0:17:08.789,0:17:17.049 impacting their own reasoning so that they[br]can't make proper decisions anymore. The 0:17:17.049,0:17:22.560 second one is Purposeful Product Design.[br]That one is one that, it's been kind of a 0:17:22.560,0:17:26.789 pet peeve for me for a really long time.[br]So these days we have a lot of products 0:17:26.789,0:17:32.289 that are kind of like this. I don't have[br]something specifically against Philips 0:17:32.289,0:17:37.639 Hue, but there seems to be, like, this[br]trend that is kind of, making smart 0:17:37.639,0:17:42.899 things, right? You take a product, put a[br]Wi-Fi chip on it, just slap it on there. 0:17:42.899,0:17:47.860 Label it "smart", and then you make tons[br]of profit, right? And a lot of these new 0:17:47.860,0:17:50.309 products we've been seeing around us,[br]like, everyone is saying, like, oh yeah, 0:17:50.309,0:17:55.119 we will have this great interconnected[br]feature, but most of them are actually not 0:17:55.119,0:17:58.489 changing the actual product, right, like,[br]the Wi-Fi connected washing machine today 0:17:58.489,0:18:03.320 is still a boring washing machine that[br]breaks down after two years. But it has 0:18:03.320,0:18:08.970 Wi-Fi, so you can see what it's doing when[br]you're in the park. And we think we should 0:18:08.970,0:18:16.000 really think more in terms of intelligent[br]design. How can we design it in the first 0:18:16.000,0:18:21.769 place so it's intelligent, not smart. That[br]the different components interact in a 0:18:21.769,0:18:26.779 way, that it serves a purpose well, and[br]the kind of intelligent by design 0:18:26.779,0:18:33.950 philosophy is, when you start using your[br]product you kind of try to identify the 0:18:33.950,0:18:40.629 core purpose of it. And based on that, you[br]just use all the technologies available to 0:18:40.629,0:18:44.359 rebuild it from scratch. So, instead of[br]building a Wi-Fi connect washing machine 0:18:44.359,0:18:47.450 would actually try to build a better[br]washing machine. And if it ends up having 0:18:47.450,0:18:51.489 Wi-Fi, then that's good, but it doesn't[br]has to. And along each step actually try 0:18:51.489,0:18:58.309 to ask yourself: Am I actually improving[br]washing machines here? Or am I just 0:18:58.309,0:19:06.279 creating another data point? And yeah, a[br]good example for that is, kind of, a 0:19:06.279,0:19:09.960 watch. Of course it's very old and old[br]technology, it was invented a long time 0:19:09.960,0:19:14.149 ago. But back when it was invented it was[br]for something you could have on your arm 0:19:14.149,0:19:19.510 or in your pocket in the beginning and it[br]was kind of a natural extension of 0:19:19.510,0:19:25.489 yourself, right, that kind of enhances[br]your senses because it's never there, you 0:19:25.489,0:19:29.850 don't really feel it. But when you need it[br]it's always there and then you can just 0:19:29.850,0:19:33.940 look at it and you know the time. And that[br]profoundly changed how, like, we humans 0:19:33.940,0:19:37.570 actually worked in society because we[br]couldn't meet in the same place at the 0:19:37.570,0:19:42.779 same time. So, when you build a new[br]product try to ask yourself what is the 0:19:42.779,0:19:46.590 purpose of the product, who is it for.[br]Often I talk to people and they talk to me 0:19:46.590,0:19:50.750 for one hour, what like, literally the[br]details of how they solved the problem but 0:19:50.750,0:19:55.130 they can't tell me who their customer is.[br]Then does this product actually make 0:19:55.130,0:19:59.570 sense? Do I have features, and these[br]distract my users, that I maybe just don't 0:19:59.570,0:20:03.850 need. And can I find more intelligent[br]solutions by kind of thinking outside of 0:20:03.850,0:20:09.549 the box and focusing on the purpose of it.[br]And then of course what is the long term 0:20:09.549,0:20:12.950 product vision like, where do we want this[br]to go? This kind of technology I'm 0:20:12.950,0:20:20.090 developing in the next years. The next one[br]is kind of, Societal Impact, that goes 0:20:20.090,0:20:27.659 into what I talked about in the beginning[br]with all the negative consequences we have 0:20:27.659,0:20:30.820 seen. A lot of people these days don't[br]realize that even if you're, like, in a 0:20:30.820,0:20:34.571 small start up and you're working on, I[br]don't know, a technology, or robots, or 0:20:34.571,0:20:39.669 whatever. You don't know if your[br]algorithm, or your mechanism, or whatever 0:20:39.669,0:20:45.190 you build, will be used by 100 million[br]people in five years. Because this has 0:20:45.190,0:20:49.849 happened a lot, right? So, only when[br]starting to build it you have to think: If 0:20:49.849,0:20:54.220 this product would be used by 10 million,[br]100, maybe even a billion people, like 0:20:54.220,0:20:58.149 Facebook, would it have negative[br]consequences? Right, because then you get 0:20:58.149,0:21:03.470 completely different effects in society,[br]completely different engagement cycles and 0:21:03.470,0:21:09.279 so on. Then, are we taking advantage of[br]human weaknesses? So this is arguably 0:21:09.279,0:21:16.249 something that's just their technology. A[br]lot of products these days kind of try to 0:21:16.249,0:21:19.789 hack your brain, what, we understand[br]really well how, like, engagement works 0:21:19.789,0:21:24.870 and addiction. So a lot of things, like[br]social networks, actually have been 0:21:24.870,0:21:28.070 focusing, you know, and also built by[br]engineers, you know, trying to get a 0:21:28.070,0:21:34.509 little number from 0.1% to 0.2%, can mean[br]that you just do extensive A/B testing, 0:21:34.509,0:21:38.460 create an interface that no one can stop[br]looking at. You just continue scrolling, 0:21:38.460,0:21:42.340 right? You just continue, and two hours[br]have passed and you haven't actually 0:21:42.340,0:21:48.830 talked to anyone. And this attention[br]grabbing is kind of an issue and we can 0:21:48.830,0:21:53.590 see that Apple actually now implemented[br]screen time and they actually tell you how 0:21:53.590,0:21:56.700 much time you spend on your phone. So[br]there's definitely ways to build 0:21:56.700,0:22:01.830 technology that even helps you to get away[br]from these. And then for everything that 0:22:01.830,0:22:06.269 involves AI and machine learning, you[br]really have to take a really deep look at 0:22:06.269,0:22:11.090 your data sets and your algorithms because[br]it's very, very easy to build in biases 0:22:11.090,0:22:16.700 and discrimination. And again, if you it[br]applied to all of society many people who 0:22:16.700,0:22:20.440 are less fortunate, or more fortunate, or[br]they're just different, you know they just 0:22:20.440,0:22:24.899 do different things, kind of fall out of[br]the grid and now suddenly they can't, 0:22:24.899,0:22:30.849 like, [unintelligible] anymore. Or use[br]Uber, or Air B'n'B, or just live a normal 0:22:30.849,0:22:35.440 life, or do financial transactions. And[br]then, kind of what I said in the 0:22:35.440,0:22:39.789 beginning, not only look at your product[br]but also, if you combine it with other 0:22:39.789,0:22:43.689 technologies that are upcoming, are there[br]certain combinations that are dangerous? 0:22:43.689,0:22:48.509 And for that I kind of recommend to do,[br]like, some techno or litmus test to just 0:22:48.509,0:22:59.359 try to come up with the craziest scenario[br]that your technology could entail. And if 0:22:59.359,0:23:05.450 it's not too bad then, probably good. The[br]next thing is, kind of, sustainability. I 0:23:05.450,0:23:11.460 think in today's world it really should be[br]part of a good product, right. The first 0:23:11.460,0:23:16.700 question is of course kind of obvious. Are[br]we limiting product lifetime? Do we maybe 0:23:16.700,0:23:20.049 have planned obsolescence, or if we[br]build something that is so dependent on so 0:23:20.049,0:23:23.950 many services and we're not only going to[br]support it for one year anyways, that 0:23:23.950,0:23:29.389 basically it will have to be thrown in the[br]trash afterwards. Maybe it would be 0:23:29.389,0:23:34.159 possible to add a standalone node or a[br]very basic fallback feature so that at 0:23:34.159,0:23:37.659 least the products continues to work.[br]Especially if you talk about things like 0:23:37.659,0:23:43.710 home appliances. Then, what is the[br]environmental impact? A good example here 0:23:43.710,0:23:49.009 would be, you know, crypto currencies who[br]are now using as much energy as certain 0:23:49.009,0:23:56.960 countries. And when you consider that just[br]think like is there maybe an alternative solution 0:23:56.960,0:24:00.950 that doesn't have such a big impact. And[br]of course we are still capitalism, it has 0:24:00.950,0:24:05.289 to be economically viable, but often there[br]aren't, often it's again just really small 0:24:05.289,0:24:12.809 tweaks. Then of course: Which other[br]services are you working with? But for 0:24:12.809,0:24:17.879 example I would say, like, as european[br]companies, we're in Europe here, maybe try 0:24:17.879,0:24:22.049 to work mostly with suppliers from Europe,[br]right, because you know they follow GDPR 0:24:22.049,0:24:28.309 and strict rules, and in a sense the US.[br]Or check your supply chain if you build 0:24:28.309,0:24:32.919 hardware. And then for hardware[br]specifically that's because also I have, 0:24:32.919,0:24:37.979 like, we also do hardware in my company, I[br]always found that interesting. We're kind 0:24:37.979,0:24:41.710 of in a world where everyone tries to[br]save, like, the last little bit of money 0:24:41.710,0:24:46.460 out of every device that is built and[br]often makes the difference between plastic 0:24:46.460,0:24:51.690 and metal screws like half a cent, right.[br]And at that point it doesn't really change 0:24:51.690,0:24:57.039 your margins much. And maybe as an[br]engineer, you know, just say no and say: 0:24:57.039,0:25:00.620 You know, we don't have to do that. The[br]savings are too small to redesign 0:25:00.620,0:25:05.509 everything and it will impact upon our[br]quality so much that it just breaks 0:25:05.509,0:25:13.159 earlier. These are kind of the main four[br]points. I hope that makes sense. Then we 0:25:13.159,0:25:17.230 have two more, kind of, additional[br]checklists. The first one is data 0:25:17.230,0:25:24.490 collection. So really, just if, especially[br]like in terms of like IOT, you know, 0:25:24.490,0:25:29.419 everyone focuses on kind of collecting as[br]much data as possible without actually 0:25:29.419,0:25:33.629 having an application. And I think we[br]really have to start seeing that as a 0:25:33.629,0:25:39.889 liability. And instead try to really[br]define the application first, define which 0:25:39.889,0:25:44.279 data we need for it, and then really just[br]collect that. And we can start collecting 0:25:44.279,0:25:49.350 more data later on. And that can really[br]prevent a lot of these negative cycles we 0:25:49.350,0:25:53.029 have seen. By just having machine learning[br]organisms run on of it kind of 0:25:53.029,0:25:59.349 unsupervised and seeing what comes out.[br]Then also kind of really interesting I 0:25:59.349,0:26:02.820 found that, many times, like, a lot of[br]people are so fascinated by the amount of 0:26:02.820,0:26:08.649 data, right, just try to have as many data[br]points as possible. But very often you can 0:26:08.649,0:26:13.539 realize exactly the same application with a[br]fraction of data points. Because what you 0:26:13.539,0:26:17.879 really need is, like, trends. And that[br]usually also makes the product more 0:26:17.879,0:26:24.230 efficient. Then how privacy intrusive is[br]the data we collect? Right. There's a big 0:26:24.230,0:26:27.619 difference between, let's say, the[br]temperature in this building and 0:26:27.619,0:26:32.070 everyone's individual movements here. And[br]if it is privacy intrusive then we should 0:26:32.070,0:26:35.759 really, really think hard if we want to[br]collect it. Because we don't know how it 0:26:35.759,0:26:44.389 might be used at a later point. And then,[br]are we actually collecting data without 0:26:44.389,0:26:47.610 people realizing that they do it, right,[br]especially if we look at Facebook and 0:26:47.610,0:26:52.720 Google. They're collecting a lot of data[br]without really implicit consent. But of 0:26:52.720,0:26:58.659 course at some point you like all agreed[br]to the privacy policy. But it's often not 0:26:58.659,0:27:03.860 clear to you when and which data is[br]collected. And that's kind of dangerous 0:27:03.860,0:27:12.149 and kind of in the same way if you kind of[br]build dark patterns into your app. They 0:27:12.149,0:27:17.669 kind of fool you into sharing even more[br]data. I had, like, an example that someone 0:27:17.669,0:27:24.830 told me yesterday. I don't if you know[br]Venmo which is this American system where 0:27:24.830,0:27:27.381 you pay each other with your smartphone.[br]Basically to split the bill in a 0:27:27.381,0:27:32.570 restaurant. By default, all transactions[br]are public. So, like 200 million public 0:27:32.570,0:27:38.659 transactions which everyone can see,[br]including the description of it. So for 0:27:38.659,0:27:45.249 some of the more maybe not so legal[br]payments that was also very obvious, 0:27:45.249,0:27:50.720 right? And it's totally un-obvious when[br]you use the app that that is happening. So 0:27:50.720,0:27:56.809 that's definitely a dark pattern that[br]they're employing here. And then the next 0:27:56.809,0:28:02.809 point is User Product Education and[br]Transparency. Is a user able to understand 0:28:02.809,0:28:08.320 how the product works? And, of course, we[br]can't really ever have a perfect 0:28:08.320,0:28:15.609 explanation of all the intricacies of the[br]technology. But these days for most people 0:28:15.609,0:28:20.889 almost all of the apps, the interfaces,[br]the building technology and tech. This is 0:28:20.889,0:28:25.340 a complete black box and no one is really[br]doing an effort to explain it to them why 0:28:25.340,0:28:30.460 most companies advertise it like this[br]magical thing. But that just leads to kind 0:28:30.460,0:28:35.950 of this immunization where you just look at[br]it and you don't even try to understand 0:28:35.950,0:28:46.399 it. I'm pretty sure that no one ever,[br]like, these days is still opening up a PC 0:28:46.399,0:28:50.029 and trying looking at the components,[br]right, because everything is in tablet and 0:28:50.029,0:28:56.639 it's integrated and it's sold to us like[br]this magical media consumption machine. 0:28:56.639,0:29:01.809 Then, are users informed when decisions[br]are made for them? So we had that in 0:29:01.809,0:29:07.950 Empowerment, that we should try to reduce[br]the amount of decisions we make for the 0:29:07.950,0:29:12.149 user. But sometimes, that's a good thing[br]to do. But then, is it transparently 0:29:12.149,0:29:17.789 communicated? I would be totally fine with[br]Google Maps filtering out for me the 0:29:17.789,0:29:21.629 points of interest if it would actually[br]tell me that it's doing that. And if you 0:29:21.629,0:29:25.779 can't understand why it made that decision[br]and why it showed me this place. And maybe 0:29:25.779,0:29:29.589 also have a way to switch it off if I[br]want. But today we seem to kind of assume 0:29:29.589,0:29:34.039 that we know better for the people why[br]it's, so we found the perfect algorithm 0:29:34.039,0:29:37.869 that has a perfect answer. So we don't[br]even have to explain how it works, right? 0:29:37.869,0:29:42.440 We just do it and people will be happy.[br]But then we end up with is very negative 0:29:42.440,0:29:48.999 consequences. And then, that's more like a[br]marketing thing, how is it actually 0:29:48.999,0:29:54.820 advertised? I find it, for example, quite[br]worrisome that things like Siri and 0:29:54.820,0:30:00.210 Alexa and Google home are, like, sold as[br]these magical AI machines that make your 0:30:00.210,0:30:03.499 life better, and are you personal[br]assistant. When in reality they are 0:30:03.499,0:30:10.119 actually still pretty dumb, pattern[br]matching. And that also creates a big 0:30:10.119,0:30:14.379 disconnect. Because now we have children[br]growing up who actually think that Alexa 0:30:14.379,0:30:20.529 is a person. And that's kind of dangerous.[br]And I think we should try to prevent that 0:30:20.529,0:30:27.450 because for these children, basically, it[br]kind of creates this veil and it's 0:30:27.450,0:30:33.340 humanized. And that's especially dangerous[br]if then the machine starts to make 0:30:33.340,0:30:37.359 decisions for them. And suggestions[br]because they will take them as if a human 0:30:37.359,0:30:48.539 did it for them. So, what is that? So,[br]these are kind of the main areas. Of course 0:30:48.539,0:30:54.929 it's a bit more complicated. So we just[br]published the standard today in the first 0:30:54.929,0:31:00.990 draft version. And it's basically three[br]parts of science introduction, kind of the 0:31:00.990,0:31:04.809 questions and checklists that you just saw.[br]And then actually how to implement it in 0:31:04.809,0:31:10.239 your company, which processes to have, at[br]which point you basically should have 0:31:10.239,0:31:16.289 kind of a feature gate. And I would kind of[br]ask everyone to go there, look at it, 0:31:16.289,0:31:22.749 contribute, shared it with people. We hope[br]that we'll have a final version ready kind 0:31:22.749,0:31:40.499 of in Q1 and that by then people can start[br]to implement it. Oh, yeah. So, even though 0:31:40.499,0:31:45.330 we have this standard, right, I want to[br]make it clear having such a standard and 0:31:45.330,0:31:50.539 implementing it in your organization or[br]for yourself or your product is great. It 0:31:50.539,0:31:55.859 actually doesn't remove your[br]responsibility, right? This can only be 0:31:55.859,0:32:01.899 successful if we actually all accept that[br]we are responsible. Right? If today I 0:32:01.899,0:32:06.719 build a bridge as a structural engineer[br]and the bridge breaks down because I 0:32:06.719,0:32:10.479 miscalculated, I am responsible. And I[br]think, equally, we have to accept that if 0:32:10.479,0:32:18.589 we build technology like this we also have[br]to, kind of, assume that responsibility. 0:32:18.589,0:32:25.289 And before we kind of move to Q&A, I'd[br]like to kind of take the citations. This 0:32:25.289,0:32:30.839 is Chamath Palihapitiya, former Facebook[br]executive, from the really early times. 0:32:30.839,0:32:35.340 And also, around a year ago when we[br]actually saw the GTC he said this in a 0:32:35.340,0:32:40.190 conference: "I feel tremendous guilt. I[br]think in the back in the deep restlessness 0:32:40.190,0:32:44.269 of our mind we knew something bad could[br]happen. But I think the way we defined it 0:32:44.269,0:32:48.490 is not like this. It is now literally at a[br]point where I think we have created 0:32:48.490,0:32:54.169 tools that are ripping apart the social[br]fabric of how society works." And 0:32:54.169,0:33:02.769 personally, and I hope the same for you, I[br]do not want to be that person that five 0:33:02.769,0:33:07.979 years down the line realizes that they[br]built that technology. So if there is one 0:33:07.979,0:33:14.190 take-away that you can take home from this[br]talk, then to just start asking yourself: 0:33:14.190,0:33:18.720 What is good technology, what does it mean[br]for you? What does it mean for the 0:33:18.720,0:33:24.779 products you build and what does it mean[br]for your organization? Thanks. 0:33:24.779,0:33:29.970 applause 0:33:29.970,0:33:37.739 Herald: Thank you. Yann Leretaille. Do we[br]have questions in the room? There are 0:33:37.739,0:33:44.539 microphones, microphones number 1, 2, 3,[br]4, 5. If you have a question please speak 0:33:44.539,0:33:49.269 loud into the microphone, as the people in[br]the stream want to hear you as well. 0:33:49.269,0:33:52.616 I think microphone number 1 was the fastest. [br]So please. 0:33:52.616,0:33:57.659 Question: Thank you for your talk. I just[br]want to make a short comment first and 0:33:57.659,0:34:01.750 then ask a question. I think this last[br]thing you mentioned about offering users 0:34:01.750,0:34:06.999 the options to have more control of the[br]interface there is also a problem that 0:34:06.999,0:34:11.330 users don't want it. Because when you look[br]at the statistics of how people use online 0:34:11.330,0:34:17.250 web tools, only maybe 5 percent of them[br]actually use that option. So companies 0:34:17.250,0:34:22.260 remove them because for them it seems like[br]it's something not so efficient for user 0:34:22.260,0:34:26.409 experience. This was just one thing to[br]mention and maybe you can respond to that. 0:34:26.409,0:34:33.589 But what I wanted to ask you was, that all[br]these principles that you presented, they 0:34:33.589,0:34:40.079 seem to be very sound and interesting and[br]good. We can all accept them as 0:34:40.079,0:34:45.329 developers. But how would you propose to[br]actually sell them to companies. Because 0:34:45.329,0:34:50.090 if you adopt a principle like this as an[br]individual based on your ideology or the 0:34:50.090,0:34:53.680 way that you think, okay, it's great it[br]will work, but how would you convince a 0:34:53.680,0:34:58.590 company which is driven by profits to[br]adopt these practices? Have you thought of 0:34:58.590,0:35:04.589 this and what's your idea about this?[br]Thank you. 0:35:04.589,0:35:11.310 Yann: Yeah. Maybe to the first part.[br]First, that giving people choice is 0:35:11.310,0:35:16.760 something that people do not want and[br]that's why companies removed it. I think 0:35:16.760,0:35:21.970 if you look at the development process[br]it's basically like a huge cycle of 0:35:21.970,0:35:26.359 optimization and user testing geared[br]towards a very specific goal, right, which 0:35:26.359,0:35:31.060 is usually set by leadership which is,[br]like, bringing engagement up or increase 0:35:31.060,0:35:37.670 user amount by 200 percent. So I would say[br]the goals were, or are today, mostly 0:35:37.670,0:35:41.849 misaligned. And that's why we end up with[br]interfaces that are in a very certain way, 0:35:41.849,0:35:46.049 right? If we set the goals[br]differently, and I mean that's why we have 0:35:46.049,0:35:51.130 like UI and UX research. I'm very sure we[br]can find ways to build interfaces that are 0:35:51.130,0:35:59.370 just different. And still engaging, but[br]also give that choice. To the second 0:35:59.370,0:36:06.289 question. I mean it's kind of interesting.[br]So I wouldn't expect a company like Google 0:36:06.289,0:36:10.730 to implement something like this, because[br]it's a bit against that. This is more by 0:36:10.730,0:36:16.359 that point probably but I've met a lot of,[br]like, also high level executives already, 0:36:16.359,0:36:23.250 who were actually very aware of kind of[br]the issues of technology that they built. 0:36:23.250,0:36:28.480 And there is definitely interest there.[br]Also, more like industrial side, and so 0:36:28.480,0:36:34.250 on, especially, it seems like self-driving[br]cars to actually adopt that. And in the 0:36:34.250,0:36:39.530 end I think, you know, if everyone[br]actually demands it, then there's a pretty 0:36:39.530,0:36:44.069 high probability that it might actually[br]happen. Especially, as workers in the tech 0:36:44.069,0:36:50.760 field, we are quite flexible in the[br]selection of our employer. So I think if 0:36:50.760,0:36:56.340 you give it some time, that's definitely[br]something that's very possible. The second 0:36:56.340,0:37:01.930 aspect is that, actually, if we looked at[br]something like Facebook, I think they 0:37:01.930,0:37:08.730 overdid it. Say, optimize that so far and[br]push the engagement machine and kind of 0:37:08.730,0:37:13.150 triggering like your brain cells to[br]never stop being on the site and keeps 0:37:13.150,0:37:17.960 scrolling, that people got too much of it.[br]And now they're leaving the platform in 0:37:17.960,0:37:21.950 droves. And of course Facebook would not[br]go down, they own all these other social 0:37:21.950,0:37:27.420 networks. But for the product itself. as[br]you can see, that, long term it's not even 0:37:27.420,0:37:33.519 necessarily a positive business outcome.[br]And everything we are advertising here 0:37:33.519,0:37:38.610 still also to have very profitable businesses,[br]right, just tweaking the right screws. 0:37:38.610,0:37:42.730 Herald: Thank you. We have a question from[br]the interweb. 0:37:42.730,0:37:48.480 Signal Angel: Yes. Fly asks a question 0:37:48.480,0:37:54.700 that goes into a similar direction. In[br]recent months we had numerous reports 0:37:54.700,0:37:59.040 about social media executives forbidding[br]their children to use the products they 0:37:59.040,0:38:05.289 create at work. I think these people know[br]that their products are made addictive 0:38:05.289,0:38:11.079 deliberately. Do you think your work is[br]somewhat superfluous because big companies 0:38:11.079,0:38:16.400 are doing the opposite on purpose.[br]Yann: Right. I think that's why you have 0:38:16.400,0:38:23.119 to draw the line between intentional and[br]unintentional. If we go to intentional 0:38:23.119,0:38:27.220 things like what Uber did and so on. At[br]some point it should probably become a 0:38:27.220,0:38:32.289 legal issue. Unfortunately we are not[br]there yet and usually regulation is kind 0:38:32.289,0:38:39.019 of lagging way behind. So I think for now[br]we should focus on, you know, the more 0:38:39.019,0:38:45.190 unintentional consequences, of which there[br]are plentiful and kind of appeal to the 0:38:45.190,0:38:52.329 good in humans.[br]Herald: Okay. Microphone number 2 please. 0:38:52.329,0:38:59.619 Q: Thank you for sharing your ideas about[br]educating the engineer. What about 0:38:59.619,0:39:05.440 educating the customer, the consumer who[br]purchases the product. 0:39:05.440,0:39:12.390 Yann: Yeah. So that's a really valid[br]point. Right. As I said I think 0:39:12.390,0:39:19.609 [unintelligible] like part of your product[br]development. And the way you build a 0:39:19.609,0:39:24.619 product should also be how you educate[br]your users on how it works. Generally, we 0:39:24.619,0:39:30.940 have a really big kind of technology[br]illiteracy problem. Things have been 0:39:30.940,0:39:34.990 moving so fast in the last year that most[br]people haven't really caught up and they 0:39:34.990,0:39:39.849 just don't understand things anymore. And[br]I think again that's like a shared 0:39:39.849,0:39:44.500 responsibility, right? You can't just do[br]that in the tech field. You have to talk 0:39:44.500,0:39:48.670 to your relatives, to people. That's why[br]we're doing, like, this series of articles 0:39:48.670,0:39:54.860 and media partnerships to kind of explain[br]and make these things transparent. One 0:39:54.860,0:40:01.510 thing we just started working on is a[br]children's book. Because for children, 0:40:01.510,0:40:06.510 like, the entire world just exists with[br]this shiny glass surfaces and they don't 0:40:06.510,0:40:11.180 understand at all what is happening. But[br]it's also primetime to explain to them, 0:40:11.180,0:40:15.420 like, really simple machine learning[br]algorithms. How they work, how like, 0:40:15.420,0:40:19.280 filterbubbles work, how decisions are[br]made. And if you understand that from an 0:40:19.280,0:40:24.740 early age on, then maybe you'll be able to[br]deal with what is happening. In a way 0:40:24.740,0:40:31.730 better, an educated way. But I do think[br]that is a very long process and so only if 0:40:31.730,0:40:37.349 we start and the more work we invest in[br]that, the earlier people will be better 0:40:37.349,0:40:40.730 educated.[br]Herald: Thank you. Microphone number 1 0:40:40.730,0:40:45.090 please.[br]Q: Thanks for sharing your insights. I 0:40:45.090,0:40:50.829 feel like, while you presented these rules[br]along with their meaning, the specific 0:40:50.829,0:40:55.710 selection might seem a bit arbitrary. And[br]for my personal acceptance and willingness 0:40:55.710,0:41:01.840 to implement them it would be interesting[br]to know the reasoning, besides common 0:41:01.840,0:41:07.280 sense, that justifies this specific[br]selection of rules. So, it would be 0:41:07.280,0:41:12.779 interesting to know if you looked at[br]examples from history, or if you just sat 0:41:12.779,0:41:19.299 down and discussed things, or if you just[br]grabbed some rules out of the air. And so 0:41:19.299,0:41:26.230 my question is: What influenced you for[br]the development of these specific rules? 0:41:26.230,0:41:34.130 Yann: It's a very complicated question. So[br]how did we come up this specific selection 0:41:34.130,0:41:39.470 of rules and also, like, the main building[br]blocks of what we think should good 0:41:39.470,0:41:47.099 technology be. Well, let's say first what[br]we didn't want to do, right. We didn't 0:41:47.099,0:41:51.119 want to create like a value framework and[br]say, like, this is good, this is bad, 0:41:51.119,0:41:55.290 don't do this kind of research or[br]technology. Because this would also be 0:41:55.290,0:41:59.960 outdated, it doesn't apply to everyone. We[br]probably couldn't even agree in the expert 0:41:59.960,0:42:05.300 council on that because it's very diverse.[br]Generally, we try to get everyone on the 0:42:05.300,0:42:12.200 table. And we talked about issues we had,[br]like, for example me as an entrepreneur. And when 0:42:12.200,0:42:18.890 I was, like, in developing products with[br]our own engineers. Issues we've seen in terms 0:42:18.890,0:42:26.790 of public perception. Issues we've seen,[br]like, on a more governmental level. Then 0:42:26.790,0:42:32.349 we also have, like, cryptologists in[br]there. So we looked at that as well and 0:42:32.349,0:42:42.509 then we made a really, really long list[br]and kind of started clustering it. And a 0:42:42.509,0:42:49.589 couple of things did get cut off. But[br]generally, based on the clustering, these 0:42:49.589,0:42:57.569 were kind of the main themes that we saw.[br]And again, it's really more of a tool for 0:42:57.569,0:43:03.680 yourself as a company that developers,[br]designers and engineers to really 0:43:03.680,0:43:08.690 understand the impact and evaluate it. Right.[br]This is what these questions are 0:43:08.690,0:43:13.369 aimed at. And we think that for that they[br]do a very good job. 0:43:13.369,0:43:18.559 From microphone 1: Thank you.[br]Herald: Thank you. And I think. Microphone 0:43:18.559,0:43:22.359 number 2 has a question again.[br]Q: Hi. I was just wondering how you've 0:43:22.359,0:43:26.730 gone about engaging with other standards[br]bodies, that perhaps have a wider 0:43:26.730,0:43:32.540 representation. It looks largely like from[br]your team of the council currently that 0:43:32.540,0:43:36.740 there's not necessarily a lot of[br]engagement outside of Europe. So how do 0:43:36.740,0:43:41.540 you go about getting representation from[br]Asia. For example. 0:43:41.540,0:43:52.369 Yann: No, at the moment you're correct the[br]GTC is mostly a European initaitive. We 0:43:52.369,0:43:57.710 are in talks with other organizations who[br]work on similar issues and regularly 0:43:57.710,0:44:04.220 exchange ideas. But, yeah, we thought we[br]should probably start somewhere. In Europe 0:44:04.220,0:44:09.250 is actually a really good place to start.[br]Like a societal discourse about technology 0:44:09.250,0:44:14.049 and the impact it has and also to to have[br]change. But I think if for example 0:44:14.049,0:44:19.599 compared to things like Asia or the US[br]where is a very different perception of 0:44:19.599,0:44:25.029 privacy and technology and progress and[br]like the rights of the individual Europe 0:44:25.029,0:44:29.400 is actually a really good place to do[br]that. And we can also see things like GDPR 0:44:29.400,0:44:35.920 regulation, that actually, ... It's kind[br]of complicated. It's also kind of a big 0:44:35.920,0:44:40.790 step forward in terms of protecting the[br]individual from exactly these kind of 0:44:40.790,0:44:47.150 consequences. Of course though, long term[br]we would like to expand this globally. 0:44:47.150,0:44:52.640 Herald: Thank you. Microphone number 1[br]again. 0:44:52.640,0:44:57.270 Q: Hello. Just a short question. I[br]couldn't find a donate button on your 0:44:57.270,0:45:03.549 website. Do you accept donations? Is money[br]a problem? Like, do you need it? 0:45:03.549,0:45:12.960 Yann: Yes, we do need money. However it's[br]a bit complicated because we want to stay 0:45:12.960,0:45:19.750 as independent as possible. So we are not[br]accepting project related money. So you can't 0:45:19.750,0:45:22.470 like say we want to do certain research[br]product with you, it has to be 0:45:22.470,0:45:29.800 unconditional. And the second thing we do[br]is for the events we organize. We usually 0:45:29.800,0:45:33.690 have sponsors that provide, like, venue[br]and food and logistics and things like 0:45:33.690,0:45:39.140 that. But that's an, ... for the event.[br]And again, I can't, like, change the 0:45:39.140,0:45:44.249 program of it. So if you want to do that[br]you can come into contact with us. We 0:45:44.249,0:45:48.509 don't have a mechanism yet for individuals[br]to donate. We might add that. 0:45:48.509,0:45:54.109 Herald: Thank you. Did you think about[br]Patreon or something like that? 0:45:54.109,0:46:03.509 Yann: We thought about quite a few[br]options. Yeah, but it's actually not so 0:46:03.509,0:46:09.470 easy to not fall into the trap that,[br]like, as organizations in space have been, 0:46:09.470,0:46:15.190 like, Google at some point sweeps in and[br]it's like: Hey, do you want all this cash. 0:46:15.190,0:46:18.840 And then very quickly you have a big[br]conflict of interest. Even if you don't 0:46:18.840,0:46:25.660 want that to happen it starts happening.[br]Herald: Yeah right. Number 1 please. 0:46:25.660,0:46:32.730 Q: I was wondering how do you unite the[br]second and third points in your checklist. 0:46:32.730,0:46:37.960 Because the second one is intelligence by[br]design. The third one is to take into 0:46:37.960,0:46:43.080 account future technologies. But companies[br]do not want to push back their 0:46:43.080,0:46:48.519 technologies endlessly to take into[br]account future technologies. And on the 0:46:48.519,0:46:52.000 other hand they don't want to compromise[br]their own design too much. 0:46:52.000,0:47:00.160 Yann: Yeah. Okay. Okay. Got it. So you[br]were saying if we should always stop 0:47:00.160,0:47:04.109 these, like, future scenarios and the[br]worst case and everything and incorporate 0:47:04.109,0:47:07.869 every possible thing that might happen in[br]the future we might end up doing nothing 0:47:07.869,0:47:14.210 because everything looks horrible. For[br]that I would say, like, we are not like 0:47:14.210,0:47:21.079 technology haters. We are all from areas[br]working in tech. So of course the idea is 0:47:21.079,0:47:25.859 that you can just take a look at what is[br]there today and try to make an assessment 0:47:25.859,0:47:30.470 based on that. And the idea is if you look[br]it up and meet the standards that over 0:47:30.470,0:47:35.289 time actually you try to,... When you add[br]new major features to look back at your 0:47:35.289,0:47:40.079 assessment from before and see if it[br]changed. So the idea is you kind of create 0:47:40.079,0:47:46.819 a snapshot of how it is now. And this kind[br]of document that you end up as part of 0:47:46.819,0:47:50.549 your documentation kind of evolved over[br]time as your product changes and the 0:47:50.549,0:47:57.390 technology around it changes as well.[br]Herald: Thank you. Microphone number 2. 0:47:57.390,0:48:02.789 Q: So thanks for the talk and especially[br]the effort. Just to echo back the 0:48:02.789,0:48:07.430 question that was asked a bit before on[br]starting with Europe. I do think it's a 0:48:07.430,0:48:14.010 good option. What I'm a little bit worried[br]is it might be the only option. It might 0:48:14.010,0:48:19.569 become irrelevant rather quickly because[br]it's easy to do, it's less hard to 0:48:19.569,0:48:26.220 implement. Maybe in Europe now. Okay. The[br]question is. It might work in Europe now 0:48:26.220,0:48:30.960 but if Europe doesn't have the same[br]economical power it cannot bog in as much 0:48:30.960,0:48:36.549 politically with let's say China or the US[br]in Silicon Valley. So will it still be 0:48:36.549,0:48:41.390 possible and relevant if the economical[br]balance shifts? 0:48:41.390,0:48:52.329 Yann: Yes, I mean we have to start[br]somewhere, right? Just saying "Oh, 0:48:52.329,0:48:59.040 economical balance will shift anyway,[br]Google will invent singularity, and that's 0:48:59.040,0:49:02.039 why we shouldn't do anything" is, I think,[br]one of the reasons why we actually got 0:49:02.039,0:49:07.730 here, why it kind of is this assumption[br]that there is like this really big picture 0:49:07.730,0:49:14.490 that is kind of working against us, so we[br]all do our small part to fulfill that 0:49:14.490,0:49:20.779 kind of evil vision by not doing anything.[br]I think we have to start somewhere and I 0:49:20.779,0:49:26.780 think for having operated for one year, we[br]have been actually quite successful so far 0:49:26.780,0:49:31.690 and we have a good progress. And I'm[br]totally looking forward to make it a bit 0:49:31.690,0:49:35.769 more global and to start traveling more, I[br]think that like one event outside Europe 0:49:35.769,0:49:40.330 last year in the US and that will[br]definitely increase over time, and we're 0:49:40.330,0:49:46.450 also working on making kind of our[br]ambassadors more mobile and kind of expand 0:49:46.450,0:49:50.310 to other locations. So it's definitely on[br]the roadmap but it's not like yeah, just 0:49:50.310,0:49:54.030 staying here. But yeah, you have to start[br]somewhere and that's what we did. 0:49:54.030,0:50:01.809 Herald: Nice, thank you. Number 1 please.[br]Mic 1: Yeah. One thing I haven't found was 0:50:01.809,0:50:08.420 – all those general rules you formulated[br]fit into the more general rules of 0:50:08.420,0:50:16.390 society, like the constitutional rules.[br]Have you considered that and it's just not 0:50:16.390,0:50:25.319 clearly stated and will be stated, or did[br]you develop them more from the bottom up? 0:50:25.319,0:50:33.470 Yann: Yes, you are completely right. So we[br]are defining the process and the questions 0:50:33.470,0:50:39.330 to ask yourself, but we are actually not[br]defining a value framework. The reason for 0:50:39.330,0:50:42.809 that is that societies are different, as I[br]said they are widely different 0:50:42.809,0:50:48.260 expectations towards technology, privacy,[br]how society should work, all the ones 0:50:48.260,0:50:53.799 about. The second one is that every[br]company is also different, right, every 0:50:53.799,0:50:58.240 company has their own company culture and[br]things they want to do and they don't want 0:50:58.240,0:51:04.640 to do. If I would say, for example, we[br]would have put in there "You should not 0:51:04.640,0:51:08.220 build weapons or something like that",[br]right, that would mean that all these 0:51:08.220,0:51:12.950 companies that work in that field couldn't[br]try to adapt it. And while I don't want 0:51:12.950,0:51:17.029 them to build weapons maybe in their value[br]framework that's OK and we don't want to 0:51:17.029,0:51:21.069 impose that, right. That's why I said in[br]the beginning we actually, we're called 0:51:21.069,0:51:24.730 the Good Technology Collective, we are not[br]defining what it is and I think that's 0:51:24.730,0:51:28.780 really important. We are not trying to[br]impose our opinion here. We want others to 0:51:28.780,0:51:33.750 decide for themselves what is good and[br]cannot support them and guide them in 0:51:33.750,0:51:36.299 building products that they believe are[br]good. 0:51:36.299,0:51:44.599 Herald: Thank you. Number two .[br]Mic 2: Hello, thanks for sharing. As 0:51:44.599,0:51:51.710 engineer we always want users to spend[br]more time to use our product, right? But 0:51:51.710,0:51:58.990 I'm working at mobile game company. Yep.[br]We are making, we are making a world that 0:51:58.990,0:52:05.539 users love our product. So we want users[br]spend more time in our game. So we may 0:52:05.539,0:52:13.510 make a lot of money, yeah, but when users[br]spend time to play our game they may lose 0:52:13.510,0:52:19.549 something. Yeah. You know. So how do we[br]think about the balance in a game, mobile 0:52:19.549,0:52:24.910 game. Yeah.[br]Yann: Hmm. It's a really difficult 0:52:24.910,0:52:32.470 question. So the question was like[br]specifically for mobile gaming. Where's 0:52:32.470,0:52:38.490 kind of the balance between trying to[br]engage people more and, yeah, basically 0:52:38.490,0:52:44.510 making them addicted and having them spend[br]all their money, I guess. I personally 0:52:44.510,0:52:53.880 would say it's about intent, right? It's[br]totally fine with a business model where 0:52:53.880,0:52:58.119 you make money with a game. I mean that's[br]kind of good and people do want 0:52:58.119,0:53:08.750 entertainment. But if you actively use,[br]like, research in how, like, you know, 0:53:08.750,0:53:14.750 like the brain actually works and how it[br]get super engaged, and if you basically 0:53:14.750,0:53:18.540 build in, like, gamification and[br]lotteries, which a lot of, I think, have 0:53:18.540,0:53:21.829 done, where basically your game becomes a[br]slot machine, right, you always want to 0:53:21.829,0:53:28.270 see the next opening of a crate[br]and see what you got. Kind of making it a 0:53:28.270,0:53:32.651 luck based game, actually. I think if you[br]go too far into that direction, at some 0:53:32.651,0:53:36.280 point you cross the line. Where that line[br]is you have to decide yourself, right, 0:53:36.280,0:53:40.060 some of it could be a good game and[br]dynamic but there definitely some games 0:53:40.060,0:53:44.700 out there, I would say with quite a reason[br]to say that they pushed to the limit quite 0:53:44.700,0:53:48.099 a bit too far. And if you actually look[br]how they did it because they wrote about 0:53:48.099,0:53:52.730 it, they actually did use very modern[br]research and very extensive testing to 0:53:52.730,0:53:58.180 really find out these, all these patterns[br]that make you addicted. And then it's not 0:53:58.180,0:54:02.260 much better than an actual slot machine.[br]And that probably we don't want. 0:54:02.260,0:54:08.140 Herald: So it's also an ethical question[br]for each and every one of us, right? 0:54:08.140,0:54:10.750 Yann: Yes.[br]Herald: I think there is a light and I 0:54:10.750,0:54:13.500 think this light means the interwebs has a[br]question. 0:54:13.500,0:54:21.589 Signal angel: I, there's another question[br]from ploy about practical usage, I guess. 0:54:21.589,0:54:25.199 Are you putting your guidelines at work in[br]your company? You said you're an 0:54:25.199,0:54:29.880 entrepeneur.[br]Yann: That's a great question. Yes, we 0:54:29.880,0:54:37.569 will. So we kind of just completed some[br]and there was kind of a lot of work to get 0:54:37.569,0:54:41.740 there. Once they are finished and released[br]we will definitely be one of the first 0:54:41.740,0:54:47.910 adopters.[br]Herald: Nice. And with this I think we're 0:54:47.910,0:54:50.440 done for today.[br]Yann: Perfect. 0:54:50.440,0:54:54.049 Herald: Yann, people, warm applause! 0:54:54.049,0:54:55.549 applause 0:54:55.549,0:54:57.049 postroll music 0:54:57.049,0:55:19.000 subtitles created by c3subtitles.de[br]in the year 2020. Join, and help us!