0:00:00.512,0:00:03.068 I'd like to tell you about two games of chess. 0:00:03.068,0:00:06.932 The first happened in 1997, in which Garry Kasparov, 0:00:06.932,0:00:10.648 a human, lost to Deep Blue, a machine. 0:00:10.648,0:00:12.888 To many, this was the dawn of a new era, 0:00:12.888,0:00:15.667 one where man would be dominated by machine. 0:00:15.667,0:00:19.001 But here we are, 20 years on, and the greatest change 0:00:19.001,0:00:21.691 in how we relate to computers is the iPad, 0:00:21.691,0:00:23.736 not HAL. 0:00:23.736,0:00:26.384 The second game was a freestyle chess tournament 0:00:26.384,0:00:29.353 in 2005, in which man and machine could enter together 0:00:29.353,0:00:34.019 as partners, rather than adversaries, if they so chose. 0:00:34.019,0:00:35.870 At first, the results were predictable. 0:00:35.870,0:00:38.367 Even a supercomputer was beaten by a grandmaster 0:00:38.367,0:00:40.679 with a relatively weak laptop. 0:00:40.679,0:00:43.664 The surprise came at the end. Who won? 0:00:43.664,0:00:46.440 Not a grandmaster with a supercomputer, 0:00:46.440,0:00:47.933 but actually two American amateurs 0:00:47.933,0:00:51.755 using three relatively weak laptops. 0:00:51.755,0:00:54.351 Their ability to coach and manipulate their computers 0:00:54.351,0:00:56.786 to deeply explore specific positions 0:00:56.786,0:00:59.176 effectively counteracted the superior chess knowledge 0:00:59.176,0:01:01.785 of the grandmasters and the superior computational power 0:01:01.785,0:01:03.694 of other adversaries. 0:01:03.694,0:01:06.599 This is an astonishing result: average men, 0:01:06.599,0:01:10.680 average machines beating the best man, the best machine. 0:01:10.680,0:01:13.879 And anyways, isn't it supposed to be man versus machine? 0:01:13.879,0:01:18.031 Instead, it's about cooperation, and the right type of cooperation. 0:01:18.031,0:01:20.888 We've been paying a lot of attention to Marvin Minsky's 0:01:20.888,0:01:24.130 vision for artificial intelligence over the last 50 years. 0:01:24.130,0:01:26.392 It's a sexy vision, for sure. Many have embraced it. 0:01:26.392,0:01:29.145 It's become the dominant school of thought in computer science. 0:01:29.145,0:01:32.217 But as we enter the era of big data, of network systems, 0:01:32.217,0:01:34.915 of open platforms, and embedded technology, 0:01:34.915,0:01:38.307 I'd like to suggest it's time to reevaluate an alternative vision 0:01:38.307,0:01:41.377 that was actually developed around the same time. 0:01:41.377,0:01:44.709 I'm talking about J.C.R. Licklider's human-computer symbiosis, 0:01:44.709,0:01:48.517 perhaps better termed "intelligence augmentation," I.A. 0:01:48.517,0:01:51.157 Licklider was a computer science titan who had a profound 0:01:51.157,0:01:54.163 effect on the development of technology and the Internet. 0:01:54.163,0:01:57.031 His vision was to enable man and machine to cooperate 0:01:57.031,0:02:00.621 in making decisions, controlling complex situations 0:02:00.621,0:02:02.391 without the inflexible dependence 0:02:02.391,0:02:04.924 on predetermined programs. 0:02:04.924,0:02:07.422 Note that word "cooperate." 0:02:07.422,0:02:10.169 Licklider encourages us not to take a toaster 0:02:10.169,0:02:12.453 and make it Data from "Star Trek," 0:02:12.453,0:02:15.988 but to take a human and make her more capable. 0:02:15.988,0:02:17.899 Humans are so amazing -- how we think, 0:02:17.899,0:02:20.517 our non-linear approaches, our creativity, 0:02:20.517,0:02:22.648 iterative hypotheses, all very difficult if possible at all 0:02:22.648,0:02:23.993 for computers to do. 0:02:23.993,0:02:26.445 Licklider intuitively realized this, contemplating humans 0:02:26.445,0:02:28.772 setting the goals, formulating the hypotheses, 0:02:28.772,0:02:32.148 determining the criteria, and performing the evaluation. 0:02:32.148,0:02:33.923 Of course, in other ways, humans are so limited. 0:02:33.923,0:02:37.158 We're terrible at scale, computation and volume. 0:02:37.158,0:02:38.994 We require high-end talent management 0:02:38.994,0:02:41.058 to keep the rock band together and playing. 0:02:41.058,0:02:43.262 Licklider foresaw computers doing all the routinizable work 0:02:43.262,0:02:46.538 that was required to prepare the way for insights and decision making. 0:02:46.538,0:02:48.762 Silently, without much fanfare, 0:02:48.762,0:02:52.116 this approach has been compiling victories beyond chess. 0:02:52.116,0:02:55.472 Protein folding, a topic that shares the incredible expansiveness of chess — 0:02:55.472,0:02:58.514 there are more ways of folding a protein than there are atoms in the universe. 0:02:58.514,0:03:00.867 This is a world-changing problem with huge implications 0:03:00.867,0:03:03.175 for our ability to understand and treat disease. 0:03:03.175,0:03:07.423 And for this task, supercomputer field brute force simply isn't enough. 0:03:07.423,0:03:09.807 Foldit, a game created by computer scientists, 0:03:09.807,0:03:12.309 illustrates the value of the approach. 0:03:12.309,0:03:15.350 Non-technical, non-biologist amateurs play a video game 0:03:15.350,0:03:18.423 in which they visually rearrange the structure of the protein, 0:03:18.423,0:03:19.922 allowing the computer to manage the atomic forces 0:03:19.922,0:03:22.879 and interactions and identify structural issues. 0:03:22.879,0:03:25.902 This approach beat supercomputers 50 percent of the time 0:03:25.902,0:03:28.486 and tied 30 percent of the time. 0:03:28.486,0:03:31.623 Foldit recently made a notable and major scientific discovery 0:03:31.623,0:03:34.783 by deciphering the structure of the Mason-Pfizer monkey virus. 0:03:34.783,0:03:37.798 A protease that had eluded determination for over 10 years 0:03:37.798,0:03:40.424 was solved was by three players in a matter of days, 0:03:40.424,0:03:42.449 perhaps the first major scientific advance 0:03:42.449,0:03:44.772 to come from playing a video game. 0:03:44.772,0:03:46.953 Last year, on the site of the Twin Towers, 0:03:46.953,0:03:48.426 the 9/11 memorial opened. 0:03:48.426,0:03:51.147 It displays the names of the thousands of victims 0:03:51.147,0:03:54.210 using a beautiful concept called "meaningful adjacency." 0:03:54.210,0:03:56.376 It places the names next to each other based on their 0:03:56.376,0:03:58.589 relationships to one another: friends, families, coworkers. 0:03:58.589,0:04:01.617 When you put it all together, it's quite a computational 0:04:01.617,0:04:05.840 challenge: 3,500 victims, 1,800 adjacency requests, 0:04:05.840,0:04:08.932 the importance of the overall physical specifications 0:04:08.932,0:04:11.069 and the final aesthetics. 0:04:11.069,0:04:13.684 When first reported by the media, full credit for such a feat 0:04:13.684,0:04:15.576 was given to an algorithm from the New York City 0:04:15.576,0:04:19.577 design firm Local Projects. The truth is a bit more nuanced. 0:04:19.577,0:04:22.448 While an algorithm was used to develop the underlying framework, 0:04:22.448,0:04:25.456 humans used that framework to design the final result. 0:04:25.456,0:04:27.681 So in this case, a computer had evaluated millions 0:04:27.681,0:04:31.016 of possible layouts, managed a complex relational system, 0:04:31.016,0:04:33.430 and kept track of a very large set of measurements 0:04:33.430,0:04:35.840 and variables, allowing the humans to focus 0:04:35.840,0:04:38.642 on design and compositional choices. 0:04:38.642,0:04:39.678 So the more you look around you, 0:04:39.678,0:04:41.640 the more you see Licklider's vision everywhere. 0:04:41.640,0:04:44.944 Whether it's augmented reality in your iPhone or GPS in your car, 0:04:44.944,0:04:47.914 human-computer symbiosis is making us more capable. 0:04:47.914,0:04:49.569 So if you want to improve human-computer symbiosis, 0:04:49.569,0:04:50.998 what can you do? 0:04:50.998,0:04:53.450 You can start by designing the human into the process. 0:04:53.450,0:04:55.654 Instead of thinking about what a computer will do to solve the problem, 0:04:55.654,0:04:59.523 design the solution around what the human will do as well. 0:04:59.523,0:05:01.460 When you do this, you'll quickly realize that you spent 0:05:01.460,0:05:04.339 all of your time on the interface between man and machine, 0:05:04.339,0:05:07.438 specifically on designing away the friction in the interaction. 0:05:07.438,0:05:10.204 In fact, this friction is more important than the power 0:05:10.204,0:05:12.256 of the man or the power of the machine 0:05:12.256,0:05:14.187 in determining overall capability. 0:05:14.187,0:05:16.164 That's why two amateurs with a few laptops 0:05:16.164,0:05:18.620 handily beat a supercomputer and a grandmaster. 0:05:18.620,0:05:21.625 What Kasparov calls process is a byproduct of friction. 0:05:21.625,0:05:24.026 The better the process, the less the friction. 0:05:24.026,0:05:28.282 And minimizing friction turns out to be the decisive variable. 0:05:28.282,0:05:30.525 Or take another example: big data. 0:05:30.525,0:05:32.431 Every interaction we have in the world is recorded 0:05:32.431,0:05:35.490 by an ever growing array of sensors: your phone, 0:05:35.490,0:05:37.863 your credit card, your computer. The result is big data, 0:05:37.863,0:05:39.605 and it actually presents us with an opportunity 0:05:39.605,0:05:42.267 to more deeply understand the human condition. 0:05:42.267,0:05:44.572 The major emphasis of most approaches to big data 0:05:44.572,0:05:46.787 focus on, "How do I store this data? How do I search 0:05:46.787,0:05:49.063 this data? How do I process this data?" 0:05:49.063,0:05:51.267 These are necessary but insufficient questions. 0:05:51.267,0:05:53.738 The imperative is not to figure out how to compute, 0:05:53.738,0:05:55.922 but what to compute. How do you impose human intuition 0:05:55.922,0:05:57.713 on data at this scale? 0:05:57.713,0:06:01.212 Again, we start by designing the human into the process. 0:06:01.212,0:06:04.024 When PayPal was first starting as a business, their biggest 0:06:04.024,0:06:06.828 challenge was not, "How do I send money back and forth online?" 0:06:06.828,0:06:10.700 It was, "How do I do that without being defrauded by organized crime?" 0:06:10.700,0:06:12.788 Why so challenging? Because while computers can learn 0:06:12.788,0:06:15.932 to detect and identify fraud based on patterns, 0:06:15.932,0:06:17.411 they can't learn to do that based on patterns 0:06:17.411,0:06:19.527 they've never seen before, and organized crime 0:06:19.527,0:06:22.236 has a lot in common with this audience: brilliant people, 0:06:22.236,0:06:25.876 relentlessly resourceful, entrepreneurial spirit — (Laughter) — 0:06:25.876,0:06:28.588 and one huge and important difference: purpose. 0:06:28.588,0:06:31.420 And so while computers alone can catch all but the cleverest 0:06:31.420,0:06:33.673 fraudsters, catching the cleverest is the difference 0:06:33.673,0:06:36.218 between success and failure. 0:06:36.218,0:06:38.439 There's a whole class of problems like this, ones with 0:06:38.439,0:06:41.014 adaptive adversaries. They rarely if ever present with a 0:06:41.014,0:06:43.750 repeatable pattern that's discernable to computers. 0:06:43.750,0:06:47.743 Instead, there's some inherent component of innovation or disruption, 0:06:47.743,0:06:50.478 and increasingly these problems are buried in big data. 0:06:50.478,0:06:52.978 For example, terrorism. Terrorists are always adapting 0:06:52.978,0:06:55.030 in minor and major ways to new circumstances, and despite 0:06:55.030,0:06:58.124 what you might see on TV, these adaptations, 0:06:58.124,0:07:00.417 and the detection of them, are fundamentally human. 0:07:00.417,0:07:03.534 Computers don't detect novel patterns and new behaviors, 0:07:03.534,0:07:06.769 but humans do. Humans, using technology, testing hypotheses, 0:07:06.769,0:07:11.389 searching for insight by asking machines to do things for them. 0:07:11.389,0:07:13.709 Osama bin Laden was not caught by artificial intelligence. 0:07:13.709,0:07:16.262 He was caught by dedicated, resourceful, brilliant people 0:07:16.262,0:07:20.531 in partnerships with various technologies. 0:07:20.531,0:07:23.349 As appealing as it might sound, you cannot algorithmically 0:07:23.349,0:07:24.950 data mine your way to the answer. 0:07:24.950,0:07:27.805 There is no "Find Terrorist" button, and the more data 0:07:27.805,0:07:30.107 we integrate from a vast variety of sources 0:07:30.107,0:07:32.240 across a wide variety of data formats from very 0:07:32.240,0:07:35.549 disparate systems, the less effective data mining can be. 0:07:35.549,0:07:37.573 Instead, people will have to look at data 0:07:37.573,0:07:41.029 and search for insight, and as Licklider foresaw long ago, 0:07:41.029,0:07:43.714 the key to great results here is the right type of cooperation, 0:07:43.714,0:07:45.238 and as Kasparov realized, 0:07:45.238,0:07:48.269 that means minimizing friction at the interface. 0:07:48.269,0:07:51.027 Now this approach makes possible things like combing 0:07:51.027,0:07:54.413 through all available data from very different sources, 0:07:54.413,0:07:57.205 identifying key relationships and putting them in one place, 0:07:57.205,0:08:00.133 something that's been nearly impossible to do before. 0:08:00.133,0:08:02.075 To some, this has terrifying privacy and civil liberties 0:08:02.075,0:08:05.485 implications. To others it foretells of an era of greater 0:08:05.485,0:08:07.394 privacy and civil liberties protections, 0:08:07.394,0:08:10.330 but privacy and civil liberties are of fundamental importance. 0:08:10.330,0:08:12.523 That must be acknowledged, and they can't be swept aside, 0:08:12.523,0:08:15.053 even with the best of intents. 0:08:15.053,0:08:17.571 So let's explore, through a couple of examples, the impact 0:08:17.571,0:08:19.977 that technologies built to drive human-computer symbiosis 0:08:19.977,0:08:22.896 have had in recent time. 0:08:22.896,0:08:26.312 In October, 2007, U.S. and coalition forces raided 0:08:26.312,0:08:28.728 an al Qaeda safe house in the city of Sinjar 0:08:28.728,0:08:30.662 on the Syrian border of Iraq. 0:08:30.662,0:08:33.038 They found a treasure trove of documents: 0:08:33.038,0:08:35.373 700 biographical sketches of foreign fighters. 0:08:35.373,0:08:37.957 These foreign fighters had left their families in the Gulf, 0:08:37.957,0:08:41.103 the Levant and North Africa to join al Qaeda in Iraq. 0:08:41.103,0:08:42.719 These records were human resource forms. 0:08:42.719,0:08:45.574 The foreign fighters filled them out as they joined the organization. 0:08:45.574,0:08:46.785 It turns out that al Qaeda, too, 0:08:46.785,0:08:49.382 is not without its bureaucracy. (Laughter) 0:08:49.382,0:08:51.480 They answered questions like, "Who recruited you?" 0:08:51.480,0:08:54.334 "What's your hometown?" "What occupation do you seek?" 0:08:54.334,0:08:57.503 In that last question, a surprising insight was revealed. 0:08:57.503,0:08:59.903 The vast majority of foreign fighters 0:08:59.903,0:09:02.303 were seeking to become suicide bombers for martyrdom -- 0:09:02.303,0:09:06.641 hugely important, since between 2003 and 2007, Iraq 0:09:06.641,0:09:10.885 had 1,382 suicide bombings, a major source of instability. 0:09:10.885,0:09:12.943 Analyzing this data was hard. The originals were sheets 0:09:12.943,0:09:15.685 of paper in Arabic that had to be scanned and translated. 0:09:15.685,0:09:17.877 The friction in the process did not allow for meaningful 0:09:17.877,0:09:21.227 results in an operational time frame using humans, PDFs 0:09:21.227,0:09:23.445 and tenacity alone. 0:09:23.445,0:09:25.398 The researchers had to lever up their human minds 0:09:25.398,0:09:27.743 with technology to dive deeper, to explore non-obvious 0:09:27.743,0:09:30.961 hypotheses, and in fact, insights emerged. 0:09:30.961,0:09:33.605 Twenty percent of the foreign fighters were from Libya, 0:09:33.605,0:09:36.573 50 percent of those from a single town in Libya, 0:09:36.573,0:09:39.023 hugely important since prior statistics put that figure at 0:09:39.023,0:09:41.406 three percent. It also helped to hone in on a figure 0:09:41.406,0:09:44.383 of rising importance in al Qaeda, Abu Yahya al-Libi, 0:09:44.383,0:09:47.014 a senior cleric in the Libyan Islamic fighting group. 0:09:47.014,0:09:49.678 In March of 2007, he gave a speech, after which there was 0:09:49.678,0:09:53.144 a surge in participation amongst Libyan foreign fighters. 0:09:53.144,0:09:56.250 Perhaps most clever of all, though, and least obvious, 0:09:56.250,0:09:58.323 by flipping the data on its head, the researchers were 0:09:58.323,0:10:01.223 able to deeply explore the coordination networks in Syria 0:10:01.223,0:10:03.740 that were ultimately responsible for receiving and 0:10:03.740,0:10:06.204 transporting the foreign fighters to the border. 0:10:06.204,0:10:08.837 These were networks of mercenaries, not ideologues, 0:10:08.837,0:10:11.235 who were in the coordination business for profit. 0:10:11.235,0:10:13.139 For example, they charged Saudi foreign fighters 0:10:13.139,0:10:15.338 substantially more than Libyans, money that would have 0:10:15.338,0:10:17.658 otherwise gone to al Qaeda. 0:10:17.658,0:10:19.703 Perhaps the adversary would disrupt their own network 0:10:19.703,0:10:22.738 if they knew they cheating would-be jihadists. 0:10:22.738,0:10:26.483 In January, 2010, a devastating 7.0 earthquake struck Haiti, 0:10:26.483,0:10:29.399 third deadliest earthquake of all time, left one million people, 0:10:29.399,0:10:31.983 10 percent of the population, homeless. 0:10:31.983,0:10:35.120 One seemingly small aspect of the overall relief effort 0:10:35.120,0:10:37.296 became increasingly important as the delivery of food 0:10:37.296,0:10:39.456 and water started rolling. 0:10:39.456,0:10:40.914 January and February are the dry months in Haiti, 0:10:40.914,0:10:43.856 yet many of the camps had developed standing water. 0:10:43.856,0:10:45.978 The only institution with detailed knowledge of Haiti's 0:10:45.978,0:10:47.275 floodplains had been leveled 0:10:47.275,0:10:50.283 in the earthquake, leadership inside. 0:10:50.283,0:10:52.858 So the question is, which camps are at risk, 0:10:52.858,0:10:54.779 how many people are in these camps, what's the 0:10:54.779,0:10:57.090 timeline for flooding, and given very limited resources 0:10:57.090,0:11:00.474 and infrastructure, how do we prioritize the relocation? 0:11:00.474,0:11:02.818 The data was incredibly disparate. The U.S. Army had 0:11:02.818,0:11:05.747 detailed knowledge for only a small section of the country. 0:11:05.747,0:11:08.258 There was data online from a 2006 environmental risk 0:11:08.258,0:11:10.922 conference, other geospatial data, none of it integrated. 0:11:10.922,0:11:13.880 The human goal here was to identify camps for relocation 0:11:13.880,0:11:16.275 based on priority need. 0:11:16.275,0:11:18.715 The computer had to integrate a vast amount of geospacial 0:11:18.715,0:11:21.299 information, social media data and relief organization 0:11:21.299,0:11:24.779 information to answer this question. 0:11:24.779,0:11:27.194 By implementing a superior process, what was otherwise 0:11:27.194,0:11:29.802 a task for 40 people over three months became 0:11:29.802,0:11:32.978 a simple job for three people in 40 hours, 0:11:32.978,0:11:35.606 all victories for human-computer symbiosis. 0:11:35.606,0:11:37.660 We're more than 50 years into Licklider's vision 0:11:37.660,0:11:39.902 for the future, and the data suggests that we should be 0:11:39.902,0:11:42.932 quite excited about tackling this century's hardest problems, 0:11:42.932,0:11:45.879 man and machine in cooperation together. 0:11:45.879,0:11:48.076 Thank you. (Applause) 0:11:48.076,0:11:50.581 (Applause)