WEBVTT 00:00:01.229 --> 00:00:03.737 So whenever I visit a school and talk to students, 00:00:03.737 --> 00:00:06.341 I always ask them the same thing: 00:00:06.941 --> 00:00:08.891 Why do you Google? 00:00:08.891 --> 00:00:12.594 Why is Google the search engine of choice for you? 00:00:12.594 --> 00:00:15.277 Strangely enough, I always get the same three answers. 00:00:15.277 --> 00:00:17.807 One, "Because it works", 00:00:17.807 --> 00:00:20.478 which is a great answer, that's why I Google, too. 00:00:20.478 --> 00:00:22.637 Two, somebody will say, 00:00:22.637 --> 00:00:25.461 "I really don't know of any alternatives." 00:00:25.461 --> 00:00:27.320 It's not an equally great answer 00:00:27.320 --> 00:00:28.860 and my reply to that is usually, 00:00:28.860 --> 00:00:30.805 "Try to Google the word 'search engine', 00:00:30.805 --> 00:00:33.357 you may find a couple of interesting alternatives." 00:00:33.357 --> 00:00:35.524 And last but not least, thirdly, 00:00:35.524 --> 00:00:38.844 inevitably, one student will raise her or his hand and say, 00:00:38.844 --> 00:00:43.867 "With Google, I'm certain to always get the best, unbiased search result." 00:00:45.647 --> 00:00:52.032 Certain to always get the best, unbiased search result. 00:00:53.332 --> 00:00:55.654 Now, as a man of the humanities, 00:00:55.654 --> 00:00:57.976 albeit, a digital humanities man, 00:00:57.976 --> 00:00:59.741 that just makes my skin curl, 00:00:59.741 --> 00:01:04.617 even if I, too, realize that that trust, that idea of the unbiased search result 00:01:04.617 --> 00:01:08.936 is a cornerstone in our collective love for and appreciation of Google. 00:01:08.936 --> 00:01:13.022 I will show you why that, philosophically, is almost an impossibility. 00:01:13.022 --> 00:01:16.319 But let me first elaborate, just a little bit, on a basic principle 00:01:16.319 --> 00:01:19.331 behind the search query that we sometimes seem to forget. 00:01:19.691 --> 00:01:22.054 So whenever you set out to Google something, 00:01:22.054 --> 00:01:26.582 start by asking yourself this, "Am I looking for an isolated fact? 00:01:26.582 --> 00:01:29.519 What is the capital of France? 00:01:29.519 --> 00:01:32.131 What are the building blocks of a water molecule?" 00:01:32.131 --> 00:01:34.569 Great -- Google away. 00:01:34.569 --> 00:01:37.356 There's not a group of scientists who are this close 00:01:37.356 --> 00:01:39.564 to proving that it's actually London and H30. 00:01:39.564 --> 00:01:42.000 You don't see a big conspiracy among those things. 00:01:42.000 --> 00:01:44.786 We agree, on a global scale, what the answers are 00:01:44.786 --> 00:01:46.550 to these isolated facts. 00:01:46.550 --> 00:01:51.775 But if you complicate your question just a little bit and ask something like, 00:01:51.775 --> 00:01:54.674 "Why is there an Israeli-Palestine conflict?" 00:01:54.854 --> 00:01:57.579 You're not exactly looking for a singular fact anymore, 00:01:57.579 --> 00:01:59.739 you're looking for knowledge, 00:01:59.739 --> 00:02:02.873 which is something way more complicated and delicate. 00:02:02.873 --> 00:02:04.429 And to get to knowledge, 00:02:04.429 --> 00:02:07.564 you have to bring 10 or 20 or 100 facts to the table 00:02:07.564 --> 00:02:10.350 and acknowledge them and say, "Yes, these are all true." 00:02:10.350 --> 00:02:12.184 But because of who I am, 00:02:12.184 --> 00:02:14.367 young or old, black or white, gay or straight, 00:02:14.367 --> 00:02:15.855 I will value them differently. 00:02:15.855 --> 00:02:17.567 And I will say, "Yes, this is true, 00:02:17.567 --> 00:02:19.754 "but this is more important to me than that." 00:02:19.754 --> 00:02:21.809 And this is where it becomes interesting 00:02:21.809 --> 00:02:24.049 because this is where we become human. 00:02:24.049 --> 00:02:26.528 This is when we start to argue, to form society 00:02:26.528 --> 00:02:28.321 and to really get somewhere. 00:02:28.321 --> 00:02:30.290 We need to filter all our facts here 00:02:30.290 --> 00:02:32.872 through friends and neighbors and parents and children 00:02:32.872 --> 00:02:35.032 and coworkers and newspapers and magazines 00:02:35.032 --> 00:02:38.259 to finally be grounded in real knowledge, 00:02:38.259 --> 00:02:42.323 which is something that a search engine is a poor help to achieve. 00:02:43.553 --> 00:02:49.636 So I promised you an example just to show you why it's so hard 00:02:49.636 --> 00:02:53.282 to get to the point of true, clean, objective knowledge -- 00:02:53.282 --> 00:02:54.745 that's food for thought. 00:02:54.745 --> 00:02:58.385 I will conduct a couple of simple queries, search queries. 00:02:58.745 --> 00:03:02.006 We'll start by "Michelle Obama", 00:03:02.016 --> 00:03:04.706 the First Lady of the United States. 00:03:04.706 --> 00:03:06.303 And we'll click for pictures. 00:03:06.303 --> 00:03:09.303 It works really well, as you can see. 00:03:09.303 --> 00:03:12.530 It's a perfect search result, more or less. 00:03:12.530 --> 00:03:15.280 It's just her in the picture, not even the President. 00:03:15.920 --> 00:03:18.057 How does this work? 00:03:18.057 --> 00:03:19.233 Quite simple, 00:03:19.233 --> 00:03:22.472 Google uses a lot of smartness to achieve this, but quite simply 00:03:22.472 --> 00:03:24.556 they look at two things more than anything. 00:03:24.556 --> 00:03:27.058 First, what does it say in the caption, 00:03:27.058 --> 00:03:29.438 what does it say under the picture on each website? 00:03:29.438 --> 00:03:31.698 Does it say "Michelle Obama" under the picture? 00:03:31.748 --> 00:03:34.101 Pretty good indication it's actually her on there. 00:03:34.101 --> 00:03:37.300 Second, Google looks at the picture file, 00:03:37.300 --> 00:03:40.115 the name of the file as such uploaded to the website. 00:03:40.115 --> 00:03:42.813 Again, is it called "Michelle Obama.jpeg"? 00:03:42.813 --> 00:03:45.826 Pretty good indication it's not Clint Eastwood in the picture. 00:03:45.826 --> 00:03:50.261 So you got those two and you get a search result like this, almost. 00:03:50.261 --> 00:03:56.855 Now, in 2009, Michelle Obama was the victim of a racist campaign 00:03:56.855 --> 00:04:01.269 where people set out to insult her through her search results. 00:04:01.269 --> 00:04:04.262 There was a picture distributed widely over the Internet 00:04:04.262 --> 00:04:07.002 where her face was distorted to look like a monkey. 00:04:07.002 --> 00:04:10.230 And that picture was published all over. 00:04:10.230 --> 00:04:13.657 And people published it very, very purposefully 00:04:13.657 --> 00:04:15.686 to get it up there in the search results. 00:04:15.686 --> 00:04:18.890 They made sure to write "Michelle Obama" in the caption 00:04:18.890 --> 00:04:22.977 and they made sure to upload the picture as "Michelle Obama.jpeg" or the like. 00:04:22.977 --> 00:04:25.368 You get why -- to manipulate the search result. 00:04:25.368 --> 00:04:26.687 And it worked, too. 00:04:26.687 --> 00:04:29.501 So when you picture-Googled for "Michelle Obama" in 2009, 00:04:29.501 --> 00:04:33.054 that distorted money picture showed up among the first results. 00:04:33.054 --> 00:04:36.746 Now, the results are self-cleansing, 00:04:36.746 --> 00:04:38.069 and that's sort of the beauty of it 00:04:38.069 --> 00:04:42.016 because Google measures relevance every hour every day. 00:04:42.016 --> 00:04:44.547 However, Google didn't settle for that this time, 00:04:44.547 --> 00:04:47.879 they just thought, "That's racist and it's a bad search result 00:04:47.879 --> 00:04:50.956 and we're going to go back and clean that up, manually. 00:04:50.956 --> 00:04:53.812 We are going to write some code and fix it", 00:04:53.812 --> 00:04:55.623 which they did, 00:04:55.623 --> 00:04:59.028 and I don't think anyone in this room thinks that was a bad idea. 00:04:59.028 --> 00:05:00.708 Me neither. 00:05:02.648 --> 00:05:06.071 But then, a couple years go by, 00:05:06.071 --> 00:05:08.997 and the world's most googled Anders, 00:05:08.997 --> 00:05:11.899 Anders Behring Breivik, 00:05:11.899 --> 00:05:13.246 did what he did. 00:05:13.246 --> 00:05:17.774 This is July 22 in 2011 and a terrible day in Norwegian history. 00:05:17.774 --> 00:05:21.181 This man, a terrorist, blew up a couple of government buildings, 00:05:21.181 --> 00:05:24.172 walking distance from where we are right now in Oslo, Norway 00:05:24.172 --> 00:05:26.390 and then he traveled to the island of Utøya 00:05:26.390 --> 00:05:29.360 and shot and killed a group of kids. 00:05:29.360 --> 00:05:32.657 Almost 80 people died that day. 00:05:32.657 --> 00:05:36.980 And a lot of people would describe this act of terror as two steps, 00:05:36.980 --> 00:05:38.613 that he did two things: 00:05:38.613 --> 00:05:40.956 he blew up the buildings and he shot those kids. 00:05:40.956 --> 00:05:42.548 It's not true. 00:05:42.548 --> 00:05:44.548 It was three steps. 00:05:44.548 --> 00:05:45.892 He blew up those buildings, 00:05:45.892 --> 00:05:46.844 he shot those kids, 00:05:46.844 --> 00:05:51.481 and he sat down and waited for the world to Google him. 00:05:51.481 --> 00:05:53.878 And he prepared all three steps equally well. 00:05:53.878 --> 00:05:56.921 And if there was somebody who immediately understood this, 00:05:56.921 --> 00:06:00.706 it was a Swedish web developer, or search engine optimization expert 00:06:00.706 --> 00:06:02.679 in Stockholm named Nikke Lindqvist, 00:06:02.679 --> 00:06:04.165 he's also a very political guy 00:06:04.165 --> 00:06:07.465 and he was right out there in social media, on his blog and Facebook. 00:06:07.465 --> 00:06:08.695 And he told everybody, 00:06:08.695 --> 00:06:11.346 "If there's something that this guy wants right now, 00:06:11.346 --> 00:06:14.179 it's to control the image of himself. 00:06:14.869 --> 00:06:17.841 Let's see if we can distort that. 00:06:17.841 --> 00:06:21.626 Let's see if we, in the civilized world, can protest against what he did 00:06:21.626 --> 00:06:25.132 through insulting him in his search results." 00:06:25.132 --> 00:06:26.966 And how? 00:06:26.966 --> 00:06:29.056 He told all of his readers the following, 00:06:29.056 --> 00:06:30.656 "Go out there on the Internet, 00:06:30.656 --> 00:06:33.340 find pictures of dog poop on sidewalks -- 00:06:33.934 --> 00:06:36.360 find pictures of dog poop on sidewalks -- 00:06:36.360 --> 00:06:40.133 publish them in your feeds, on your websites, on your blogs/ 00:06:40.133 --> 00:06:43.429 Make sure to write the terrorist's name in the caption, 00:06:43.429 --> 00:06:48.014 make sure to name the picture file "Breivik.jpeg", 00:06:48.014 --> 00:06:52.278 let's teach Google that that's the face of the terrorist." 00:06:53.738 --> 00:06:56.060 And it worked. 00:06:56.060 --> 00:06:58.775 Two years after that campaign against Michelle Obama, 00:06:58.775 --> 00:07:02.065 this manipulation campaign against Anders Behring Breivik worked. 00:07:02.065 --> 00:07:06.717 If you picture-Googled for him weeks after the July 22 events from Sweden, 00:07:06.717 --> 00:07:10.637 you'd see that picture of dog poop high up in your search result, 00:07:10.637 --> 00:07:12.747 as a little protest 00:07:13.637 --> 00:07:17.778 Strangely enough, Google didn't intervene this time. 00:07:18.768 --> 00:07:24.178 They did not step in and manually clean those search results up. 00:07:24.178 --> 00:07:25.943 So the million-dollar question is, 00:07:25.943 --> 00:07:29.096 is there anything different between these two happenings here? 00:07:29.096 --> 00:07:32.371 Is there anything different between what happened to Michelle Obama 00:07:32.371 --> 00:07:34.626 and what happened to Anders Behring Breivik? 00:07:34.626 --> 00:07:36.972 Of course not. 00:07:36.972 --> 00:07:38.713 It's the exact same thing, 00:07:38.713 --> 00:07:41.523 yet Google intervened in one case and not in the other. 00:07:41.523 --> 00:07:42.987 Why? 00:07:43.607 --> 00:07:46.607 Because Michelle Obama is an honorable person, that's why, 00:07:46.607 --> 00:07:50.601 and Anders Behring Breivik is a despicable person. 00:07:50.601 --> 00:07:51.948 See what happens there? 00:07:51.948 --> 00:07:54.149 An evaluation of a person takes place 00:07:54.149 --> 00:07:56.199 and there's only one player in the world, 00:07:56.199 --> 00:07:57.915 one power-player in the world 00:07:57.915 --> 00:08:00.487 with the authority to say who's who, 00:08:00.487 --> 00:08:02.999 "We like you, we dislike you. 00:08:02.999 --> 00:08:05.047 We believe in you, we don't believe in you. 00:08:05.047 --> 00:08:06.560 You're right, you're wrong. 00:08:06.560 --> 00:08:08.007 You're true, you're false. 00:08:08.007 --> 00:08:11.057 You're Obama, and you're Breivik. 00:08:11.057 --> 00:08:13.932 That's power if I ever saw it. 00:08:14.862 --> 00:08:18.941 So I'm asking you to remember that behind every algorithm 00:08:18.941 --> 00:08:20.925 there's always a person, 00:08:20.925 --> 00:08:22.964 a person with a set of personal beliefs 00:08:22.964 --> 00:08:25.964 that no code could ever completely eradicate. 00:08:25.964 --> 00:08:28.115 and my message goes out to not only Google, 00:08:28.115 --> 00:08:30.956 but to all believers in the faith of code around the world. 00:08:30.956 --> 00:08:33.916 You need to identify your own personal bias. 00:08:33.916 --> 00:08:36.157 You need to understand that you are human 00:08:36.157 --> 00:08:38.954 and take responsibility accordingly. 00:08:39.564 --> 00:08:42.774 And I say this because I believe that we've reach a point in time 00:08:42.774 --> 00:08:44.609 when it's absolutely imperative 00:08:44.609 --> 00:08:47.673 that we tie those bonds together again, tighter: 00:08:47.673 --> 00:08:50.065 the humanities and the technology. 00:08:50.065 --> 00:08:52.302 Tighter than ever. 00:08:52.312 --> 00:08:55.675 And, if nothing else, to remind us that that wonderfully seductive idea 00:08:55.675 --> 00:08:58.882 of the unbiased, clean search result 00:08:58.882 --> 00:09:01.592 is, and is likely to remain, a myth. NOTE Paragraph 00:09:01.592 --> 00:09:02.859 Thank you for your time. NOTE Paragraph 00:09:02.859 --> 00:09:05.599 (Applause)