1 00:00:15,076 --> 00:00:24,356 Speaker 1: Pushkin. I'm Maybe Higgins, and this is solvable Interviews with 2 00:00:24,436 --> 00:00:27,876 Speaker 1: the world's most innovative thinkers who are working to solve 3 00:00:27,956 --> 00:00:33,436 Speaker 1: the world's biggest problems. My solvable is that every frontline 4 00:00:33,556 --> 00:00:36,796 Speaker 1: social organization is the ability to use data and AI 5 00:00:37,196 --> 00:00:40,876 Speaker 1: same way, same capacity that the big tech companies do today. 6 00:00:41,036 --> 00:00:43,196 Speaker 1: I want to see a world where the same algorithms 7 00:00:43,196 --> 00:00:45,876 Speaker 1: that are routing your packages to you your house coming 8 00:00:45,916 --> 00:00:48,076 Speaker 1: so efficiently because an AI figured out the best way 9 00:00:48,116 --> 00:00:50,676 Speaker 1: to avoid traffic and weather are just as equally being 10 00:00:50,716 --> 00:00:54,436 Speaker 1: applied to delivering a vaccine through an area before it spoils. 11 00:00:55,036 --> 00:00:59,036 Speaker 1: That's Jake poor Away, the founder and CEO of nonprofit 12 00:00:59,156 --> 00:01:02,796 Speaker 1: Data Kind. He's talking to Jacob Weisberg about how he's 13 00:01:02,876 --> 00:01:07,636 Speaker 1: working to make that world a reality. The Rockefeller Foundation 14 00:01:07,756 --> 00:01:10,716 Speaker 1: has thought about this too. More than two point five 15 00:01:10,916 --> 00:01:15,196 Speaker 1: quintillion bytes of data are produced every day. That's one 16 00:01:15,316 --> 00:01:20,196 Speaker 1: hundred trillion bytes. This abundance of data, combined with rapidly 17 00:01:20,276 --> 00:01:25,396 Speaker 1: advancing analytics capabilities, could really improve the lives of billions 18 00:01:25,436 --> 00:01:28,916 Speaker 1: of people around the world, but it's only living up 19 00:01:28,996 --> 00:01:33,316 Speaker 1: to a fraction of that potential. While private sector businesses 20 00:01:33,356 --> 00:01:36,916 Speaker 1: have been building and deploying data science capabilities for many 21 00:01:36,996 --> 00:01:41,236 Speaker 1: years now. Most organizations in the nonprofit and civic and 22 00:01:41,356 --> 00:01:45,716 Speaker 1: public sectors are way behind. Of course, they want to 23 00:01:45,836 --> 00:01:48,436 Speaker 1: use the applied data to make their work go farther 24 00:01:48,596 --> 00:01:51,956 Speaker 1: and faster and to help more people, but they don't 25 00:01:52,036 --> 00:01:56,156 Speaker 1: often have the resources. I mean, put yourself in the 26 00:01:56,196 --> 00:02:00,156 Speaker 1: shoes of a newly minted graduate. They're probably wearing tivas. 27 00:02:00,836 --> 00:02:04,436 Speaker 1: Last year, the San Francisco Chronicle analyze glass door data 28 00:02:04,476 --> 00:02:07,316 Speaker 1: of the starting salaries of some of the biggest tech 29 00:02:07,356 --> 00:02:10,836 Speaker 1: companies in the Bay Area. They found out that tech 30 00:02:11,036 --> 00:02:15,356 Speaker 1: pays even for the young and inexperienced. The average starting 31 00:02:15,356 --> 00:02:19,556 Speaker 1: salary for a software engineer was almost ninety two thousand dollars. 32 00:02:20,516 --> 00:02:23,956 Speaker 1: So there's the workers and then there's the technology itself. 33 00:02:24,596 --> 00:02:27,356 Speaker 1: We know the power data science can have for social 34 00:02:27,436 --> 00:02:30,836 Speaker 1: good because we've seen it in action. When mission driven 35 00:02:30,916 --> 00:02:34,756 Speaker 1: organizations have the right talent and tools and knowledge, data 36 00:02:34,756 --> 00:02:39,636 Speaker 1: science can generate real human impact, helping vulnerable families access 37 00:02:39,676 --> 00:02:44,316 Speaker 1: public benefits, saving water and money during droughts, and saving 38 00:02:44,356 --> 00:02:48,236 Speaker 1: time in resettling refugees so that they can find homes 39 00:02:48,276 --> 00:02:53,516 Speaker 1: and jobs faster. Jake Borway works on this stuff every day. 40 00:02:53,916 --> 00:02:57,356 Speaker 1: He's a machine learning and technology enthusiast who loves nothing 41 00:02:57,396 --> 00:03:00,996 Speaker 1: more than seeing good values in data. In twenty eleven, 42 00:03:01,036 --> 00:03:04,756 Speaker 1: he found a data Kind, bringing together leading data scientists 43 00:03:04,756 --> 00:03:09,436 Speaker 1: with high impact social organizations to better collect, analyze, and 44 00:03:09,596 --> 00:03:13,716 Speaker 1: visualized data in the service of humanity. Jake works to 45 00:03:13,996 --> 00:03:18,436 Speaker 1: ensure organizations like the Red Cross have access to AI 46 00:03:18,516 --> 00:03:21,796 Speaker 1: and data science that's as good as the access enjoyed 47 00:03:21,836 --> 00:03:25,796 Speaker 1: by huge companies like Facebook. Data Kind has twenty thousand 48 00:03:25,836 --> 00:03:29,196 Speaker 1: volunteers around the world, who he likens to mets on 49 00:03:29,276 --> 00:03:33,476 Speaker 1: San Frontier, the doctors without Borders, except their data scientists 50 00:03:33,636 --> 00:03:37,956 Speaker 1: working pro bono with leading social change organizations on all 51 00:03:38,076 --> 00:03:41,996 Speaker 1: kinds of projects, including one that has data scientists from 52 00:03:42,036 --> 00:03:47,196 Speaker 1: Netflix predicting water usage in a California neighborhood. It's fascinating, 53 00:03:47,316 --> 00:03:50,156 Speaker 1: So enjoy this conversation and I'll talk to you after. 54 00:03:55,716 --> 00:03:59,076 Speaker 1: What's the problem. In a nutshell, the problem is that 55 00:03:59,316 --> 00:04:04,436 Speaker 1: digital technology and artificial intelligence have exploded over the last 56 00:04:04,476 --> 00:04:08,516 Speaker 1: ten or fifteen years, which have created huge opportunities in 57 00:04:08,676 --> 00:04:11,916 Speaker 1: the corporate space or in building new apps for society, 58 00:04:12,356 --> 00:04:17,436 Speaker 1: but there's very little application of that to social sector causes. 59 00:04:17,756 --> 00:04:20,636 Speaker 1: So we have this huge opportunity to use a revolutionary 60 00:04:20,676 --> 00:04:24,796 Speaker 1: technology to predict the future of things, to understand our 61 00:04:24,836 --> 00:04:27,396 Speaker 1: society better, to automate things that we either don't want 62 00:04:27,396 --> 00:04:30,596 Speaker 1: to or couldn't do, And yet there's a huge potential 63 00:04:30,676 --> 00:04:33,156 Speaker 1: loss in that it's very difficult to get that applied 64 00:04:33,356 --> 00:04:36,436 Speaker 1: to pro social causes that we need. Jake is a 65 00:04:36,556 --> 00:04:40,076 Speaker 1: data scientist. When did you start to see some of 66 00:04:40,116 --> 00:04:44,476 Speaker 1: the downsides around big data? Really? The article that I 67 00:04:44,556 --> 00:04:46,676 Speaker 1: used to point to is like the beginnings of the 68 00:04:47,276 --> 00:04:50,716 Speaker 1: tide turning to the negative. Was the article that was 69 00:04:50,756 --> 00:04:54,756 Speaker 1: titled very salaciously, Target Knows You're pregnant, And if you 70 00:04:54,796 --> 00:04:57,636 Speaker 1: remember this one from twenty thirteen, but the basic idea 71 00:04:57,756 --> 00:05:01,396 Speaker 1: was that someone had their daughter, that maybe sixteen seventeen 72 00:05:01,436 --> 00:05:04,516 Speaker 1: year old daughter was receiving mailers from Target that said, Hey, 73 00:05:04,596 --> 00:05:07,716 Speaker 1: we think you need to buy kupons for baby diapers 74 00:05:07,996 --> 00:05:11,396 Speaker 1: or formula, and the dad called up, you know Target, all, Matt, 75 00:05:11,436 --> 00:05:13,236 Speaker 1: So what are you sending me all my daughter all 76 00:05:13,316 --> 00:05:16,116 Speaker 1: these deals for having babies. She's not pregnant, Like, why 77 00:05:16,116 --> 00:05:18,956 Speaker 1: are you trying to get her to become pregnant? And 78 00:05:19,236 --> 00:05:20,596 Speaker 1: the person on the other end of the line, of 79 00:05:20,596 --> 00:05:22,876 Speaker 1: course didn't know what was happening, because you know, the 80 00:05:22,916 --> 00:05:25,316 Speaker 1: algorithms just send you what they think you're going to 81 00:05:25,396 --> 00:05:28,636 Speaker 1: buy based on other stuff you've bought, and it's He 82 00:05:28,716 --> 00:05:30,636 Speaker 1: called back later, kind of shame facedly and said, you know, 83 00:05:30,716 --> 00:05:32,956 Speaker 1: I talked to my daughter and actually she is pregnant, 84 00:05:33,476 --> 00:05:35,156 Speaker 1: and you know, the data had picked up on that 85 00:05:35,196 --> 00:05:36,996 Speaker 1: simply because you know, it watched what she bought and 86 00:05:36,996 --> 00:05:40,396 Speaker 1: she was probably buying you know, prenatal care, vitamins and stuff. 87 00:05:40,836 --> 00:05:44,596 Speaker 1: But that article got shared around as the sign that 88 00:05:44,676 --> 00:05:48,436 Speaker 1: big data was going to be negative. Target knows you're pregnant. 89 00:05:48,516 --> 00:05:50,996 Speaker 1: What a horrible invasion of privacy. That title alone should, 90 00:05:51,036 --> 00:05:54,076 Speaker 1: you know, make everyone's skin crawl. But that's the problem 91 00:05:54,196 --> 00:05:56,476 Speaker 1: is that that shouldn't be the case. We think of 92 00:05:56,916 --> 00:05:59,876 Speaker 1: there are so many opportunities to be using data and 93 00:05:59,956 --> 00:06:04,236 Speaker 1: algorithms to see where disease outbreaks are going to occur 94 00:06:04,516 --> 00:06:06,476 Speaker 1: or predict in the same way as what kind of 95 00:06:06,796 --> 00:06:09,276 Speaker 1: conditions you might have so you can live a healthier life. 96 00:06:09,476 --> 00:06:12,516 Speaker 1: And so I think it was then that we really thought, Okay, 97 00:06:12,636 --> 00:06:15,796 Speaker 1: we need to come out and show the positive sides 98 00:06:15,796 --> 00:06:17,996 Speaker 1: of this. Otherwise everyone's going to just run to the 99 00:06:18,036 --> 00:06:22,756 Speaker 1: fear around what data science can do. We're interested on 100 00:06:22,796 --> 00:06:26,076 Speaker 1: this podcast and people who've taken this leap to become 101 00:06:26,276 --> 00:06:29,876 Speaker 1: problem solvers and to take on the biggest problems in 102 00:06:29,916 --> 00:06:34,636 Speaker 1: the world. What made you take a leap to leave 103 00:06:34,716 --> 00:06:40,356 Speaker 1: the private sector to start an organization with an ambitious goal. Well, 104 00:06:40,396 --> 00:06:41,876 Speaker 1: I have to say it was a bit of an accident. 105 00:06:41,956 --> 00:06:45,516 Speaker 1: Actually it was maybe twenty ten or eleven, and I 106 00:06:45,556 --> 00:06:48,716 Speaker 1: had just coincidentally come out of school with a computer 107 00:06:48,796 --> 00:06:51,476 Speaker 1: science and a statistics degree, which little did I know 108 00:06:51,596 --> 00:06:53,636 Speaker 1: was going to become what would lead to the title 109 00:06:53,716 --> 00:06:56,436 Speaker 1: data scientist. And I was working at the New York 110 00:06:56,436 --> 00:07:00,116 Speaker 1: Times R and D Lab, and really what seemed obvious 111 00:07:00,396 --> 00:07:02,316 Speaker 1: was the fact that we had all of this new 112 00:07:02,356 --> 00:07:05,756 Speaker 1: digital technology, from cell phones that people were carrying around 113 00:07:05,756 --> 00:07:09,396 Speaker 1: with them, to satellites launching in the air, to sends 114 00:07:09,436 --> 00:07:12,156 Speaker 1: being put around the world, that we were digitizing our 115 00:07:12,476 --> 00:07:15,996 Speaker 1: very existence. We were becoming a digital species. There was 116 00:07:16,036 --> 00:07:18,796 Speaker 1: almost like a central nervous system to the world, and 117 00:07:18,876 --> 00:07:21,716 Speaker 1: that meant that were these huge opportunities to learn from 118 00:07:21,756 --> 00:07:25,556 Speaker 1: that to you know, have algorithms drive maybe our greatest 119 00:07:25,636 --> 00:07:28,236 Speaker 1: human values. But the folks who really knew how to 120 00:07:28,236 --> 00:07:32,396 Speaker 1: convert data into those actions. The data scientists were largely 121 00:07:32,476 --> 00:07:35,716 Speaker 1: locked up in tech companies, and you know, I would 122 00:07:35,836 --> 00:07:38,956 Speaker 1: actually go to hackathons, which are you know, like weekend 123 00:07:38,956 --> 00:07:41,796 Speaker 1: events where technologists would get together and just work on 124 00:07:41,836 --> 00:07:44,196 Speaker 1: whatever they thought was cool. And I would sit there 125 00:07:44,196 --> 00:07:46,116 Speaker 1: and think, this is so interesting because you know, we're 126 00:07:46,116 --> 00:07:48,556 Speaker 1: not at a company, we're not at our jobs. We're 127 00:07:48,596 --> 00:07:50,516 Speaker 1: here on the weekend. You know, I'm sitting next to 128 00:07:50,556 --> 00:07:53,676 Speaker 1: some machine learning engineer from Google and NASA scientist, and 129 00:07:53,756 --> 00:07:56,556 Speaker 1: I'm like, this is great. We can make whatever we want. 130 00:07:56,636 --> 00:07:59,956 Speaker 1: Like the world has just become so ripe for what's possible. 131 00:08:00,556 --> 00:08:01,956 Speaker 1: And at the end of the day, the stuff that 132 00:08:01,996 --> 00:08:05,956 Speaker 1: people made was just so unfulfilling. You know that someone 133 00:08:05,996 --> 00:08:08,796 Speaker 1: had made like Twitter for pets, or had improved how 134 00:08:08,836 --> 00:08:12,356 Speaker 1: you'd find local deals in your neighborhood, and so I 135 00:08:12,396 --> 00:08:14,396 Speaker 1: just said, man, there's got to be something more we 136 00:08:14,436 --> 00:08:17,516 Speaker 1: can do for society, or something more fulfilling really than this, 137 00:08:17,676 --> 00:08:20,676 Speaker 1: as opposed to solving the problems of very well paid 138 00:08:20,756 --> 00:08:23,796 Speaker 1: twenty somethings in the Bay Area, right, which is the parody, 139 00:08:23,836 --> 00:08:26,236 Speaker 1: but that is a lot of the new companies you 140 00:08:26,316 --> 00:08:29,476 Speaker 1: hear about are solving problems like how do you get 141 00:08:29,516 --> 00:08:32,476 Speaker 1: your food delivered or god knows how to get cannabis delivered? 142 00:08:32,596 --> 00:08:34,476 Speaker 1: You know when you when you could already buy it 143 00:08:34,516 --> 00:08:37,236 Speaker 1: by walking around the corner. You're exactly right. We solve 144 00:08:37,276 --> 00:08:39,796 Speaker 1: the problems that we ourselves have. And as you've pointed out, 145 00:08:39,836 --> 00:08:45,196 Speaker 1: the tech community for better for worse, excused young male US. So, yeah, 146 00:08:45,196 --> 00:08:46,396 Speaker 1: I just thought, you know, what would it take for 147 00:08:46,516 --> 00:08:48,396 Speaker 1: to be applied to the social sector. Where are the 148 00:08:48,436 --> 00:08:50,476 Speaker 1: people who are on the front lines of getting people 149 00:08:50,516 --> 00:08:52,836 Speaker 1: food or clean water? And how could you apply it there? 150 00:08:53,276 --> 00:08:57,036 Speaker 1: And so I just wanted that job myself. What didn't exist? 151 00:08:57,636 --> 00:08:59,516 Speaker 1: So I just wrote to a couple of folks in 152 00:08:59,516 --> 00:09:02,196 Speaker 1: the community here in New York and said, hey, you know, 153 00:09:02,476 --> 00:09:06,076 Speaker 1: instead of going and building you know, a door dash competitor, 154 00:09:06,676 --> 00:09:09,116 Speaker 1: could we, I don't know, work with the Red Cross 155 00:09:09,196 --> 00:09:12,996 Speaker 1: US or Kiva who goes cash transfers to folks, and 156 00:09:13,036 --> 00:09:14,796 Speaker 1: say what could we do with their data? What could 157 00:09:14,796 --> 00:09:16,636 Speaker 1: we learn? What are the positive ways we could work 158 00:09:16,636 --> 00:09:19,516 Speaker 1: together with them? And I thought people would just say, yeah, 159 00:09:19,596 --> 00:09:22,276 Speaker 1: good idea, Jake, but no thanks. I kind of just 160 00:09:22,396 --> 00:09:26,676 Speaker 1: buried the little sign up link for folks, and I 161 00:09:26,716 --> 00:09:29,236 Speaker 1: was surprised to find that people started sharing around before 162 00:09:29,236 --> 00:09:31,356 Speaker 1: I knew it. I came back to work the next time, 163 00:09:31,596 --> 00:09:33,996 Speaker 1: hundreds of emails in my inbox from people not just 164 00:09:34,036 --> 00:09:36,316 Speaker 1: in the city but around the world saying, though this 165 00:09:36,396 --> 00:09:38,556 Speaker 1: is great, I want to get involved with data kind, 166 00:09:38,596 --> 00:09:41,516 Speaker 1: I want to do data kind France. At one point, 167 00:09:41,556 --> 00:09:44,036 Speaker 1: a few months into this, the White House called and said, hey, 168 00:09:44,036 --> 00:09:46,396 Speaker 1: we're interested in big data initiatives. What's this thing? And 169 00:09:46,636 --> 00:09:50,556 Speaker 1: you know, joke because I don't know, it's not really 170 00:09:50,596 --> 00:09:53,876 Speaker 1: a thing, But to me it really tapped into an 171 00:09:53,996 --> 00:09:56,996 Speaker 1: energy from both the technology side and the nonprofits and 172 00:09:57,076 --> 00:10:01,356 Speaker 1: governments who are writing, who said, we're energetic to take 173 00:10:01,356 --> 00:10:03,796 Speaker 1: on this new wave of this technology and figure out 174 00:10:03,796 --> 00:10:06,196 Speaker 1: how could be applied. And so our job ever since 175 00:10:06,276 --> 00:10:09,556 Speaker 1: has really just been trying to support that community, harness 176 00:10:09,556 --> 00:10:11,676 Speaker 1: its energy, and be helpful in any way we can. 177 00:10:12,316 --> 00:10:15,516 Speaker 1: Since you've been doing this, it's amazing how quickly attitudes 178 00:10:15,556 --> 00:10:19,956 Speaker 1: have shifted around big data and algorithms. I mean, just 179 00:10:19,996 --> 00:10:23,236 Speaker 1: think about Facebook, which even a few years ago was 180 00:10:23,396 --> 00:10:27,556 Speaker 1: thought as a socially positive company. That was why part 181 00:10:27,596 --> 00:10:30,836 Speaker 1: of why people went to work there, and in just 182 00:10:30,916 --> 00:10:34,956 Speaker 1: a couple of years it's become something that people think 183 00:10:35,116 --> 00:10:38,956 Speaker 1: is an overwhelmingly negative force. Are we're swinging too far 184 00:10:39,076 --> 00:10:43,396 Speaker 1: in the other direction in our skepticism about what data 185 00:10:43,516 --> 00:10:45,996 Speaker 1: is going to be used for? Well, I think there's 186 00:10:46,036 --> 00:10:49,076 Speaker 1: a healthy reckoning on how we've been using data and 187 00:10:49,076 --> 00:10:52,316 Speaker 1: technology in the past. You're right that in the last 188 00:10:52,356 --> 00:10:55,196 Speaker 1: couple of years there was sort of unfettered techno optimism 189 00:10:55,196 --> 00:10:57,516 Speaker 1: amongst a lot of the big companies and that this 190 00:10:57,556 --> 00:11:00,676 Speaker 1: would just change everything and nothing could ever go wrong 191 00:11:00,716 --> 00:11:03,636 Speaker 1: with social media and data. So I think there is 192 00:11:03,676 --> 00:11:05,796 Speaker 1: an obviously very healthy reckoning of this, and we're starting 193 00:11:05,796 --> 00:11:09,556 Speaker 1: to realize what the downsides could be. What your point 194 00:11:09,596 --> 00:11:11,676 Speaker 1: I think is missing and we really need to get 195 00:11:11,956 --> 00:11:15,596 Speaker 1: acclimated to, is where do we go from there? You know, 196 00:11:15,796 --> 00:11:17,876 Speaker 1: is the idea that we're just going to put the 197 00:11:17,916 --> 00:11:20,916 Speaker 1: genie back in the bottle, not use digital information in 198 00:11:20,956 --> 00:11:25,676 Speaker 1: these ways, regulate all companies into existence. I'm in favor of, 199 00:11:25,836 --> 00:11:28,956 Speaker 1: by the way, stronger regulation, for sure, But I think 200 00:11:28,996 --> 00:11:32,316 Speaker 1: what we need now is more examples and more of 201 00:11:32,356 --> 00:11:34,716 Speaker 1: a community of practice around what it looks like to 202 00:11:34,796 --> 00:11:38,276 Speaker 1: use these technologies ethically. That's a big conversation obviously, that's 203 00:11:38,276 --> 00:11:39,916 Speaker 1: in the space right now. You hear a lot about 204 00:11:39,916 --> 00:11:42,996 Speaker 1: the ethics of data use, ethics of AI, but even 205 00:11:42,996 --> 00:11:45,876 Speaker 1: then I find those conversations fairly academic. I think what 206 00:11:45,916 --> 00:11:48,036 Speaker 1: we need are some more positive examples of how it 207 00:11:48,076 --> 00:11:51,076 Speaker 1: can be applied and positive principles that we all agree 208 00:11:51,076 --> 00:11:53,916 Speaker 1: to adhere to. And so the data kind that's something 209 00:11:53,916 --> 00:11:58,156 Speaker 1: we're really working to try to demonstrate, is to say, yes, 210 00:11:58,276 --> 00:12:02,956 Speaker 1: we need to protect ourselves, uphold our civil liberties through data. 211 00:12:03,076 --> 00:12:05,716 Speaker 1: Make sure that we're not degrading human life with what's 212 00:12:05,716 --> 00:12:09,316 Speaker 1: going on with data in the business world? And what 213 00:12:09,356 --> 00:12:12,476 Speaker 1: does it look like when you want to use data 214 00:12:12,516 --> 00:12:16,596 Speaker 1: and algorithms to predict, say, inclement weather that could wipe 215 00:12:16,596 --> 00:12:20,236 Speaker 1: out a crop and that's critical to someone's sustenance in 216 00:12:20,356 --> 00:12:22,836 Speaker 1: another part of the world. What's the good version of this? 217 00:12:22,956 --> 00:12:25,036 Speaker 1: You know? How do you make sure that it's accountable 218 00:12:25,076 --> 00:12:27,276 Speaker 1: to those folks? How do we make sure that everyone 219 00:12:27,356 --> 00:12:30,196 Speaker 1: involved has some sense of what the algorithm is doing 220 00:12:30,196 --> 00:12:32,516 Speaker 1: and how their data is being used. And I don't 221 00:12:32,556 --> 00:12:34,716 Speaker 1: think we can move past that point just by talking 222 00:12:34,716 --> 00:12:37,956 Speaker 1: about it. I think we need real concrete examples of 223 00:12:38,356 --> 00:12:43,076 Speaker 1: data scientists, nonprofits, social organizations, constituents getting together to say, 224 00:12:43,116 --> 00:12:45,276 Speaker 1: what does the good version of this look like a 225 00:12:45,356 --> 00:12:48,116 Speaker 1: better version. I should say there was a positive example 226 00:12:48,156 --> 00:12:52,076 Speaker 1: in the news recently with the prediction of the cyclone 227 00:12:52,396 --> 00:12:57,156 Speaker 1: in South Asia that killed very few people, and in 228 00:12:57,196 --> 00:13:00,956 Speaker 1: the world before big data, that same storm might have 229 00:13:01,036 --> 00:13:04,836 Speaker 1: killed a lot of people through panic, through all sorts 230 00:13:04,876 --> 00:13:07,516 Speaker 1: of consequences because people wouldn't have known it was coming. 231 00:13:07,676 --> 00:13:09,796 Speaker 1: I mean, is that the kind of example we're talking 232 00:13:09,836 --> 00:13:13,156 Speaker 1: about here? Something positive? I think that's exactly right. So 233 00:13:13,756 --> 00:13:16,796 Speaker 1: at data Kind we team technologists like data scientists who 234 00:13:16,836 --> 00:13:20,396 Speaker 1: want to volunteer their time alongside social change organizations, be 235 00:13:20,476 --> 00:13:24,276 Speaker 1: they government agencies or nonprofits who have a pro social mission, 236 00:13:24,596 --> 00:13:26,436 Speaker 1: might be able to use data and algorithms to do 237 00:13:26,516 --> 00:13:29,956 Speaker 1: even more, and we together they collaborate and kind of 238 00:13:29,996 --> 00:13:33,716 Speaker 1: codesign the solutions that they might foster a better world. 239 00:13:34,116 --> 00:13:36,116 Speaker 1: So some examples that we've seen are exactly what you're 240 00:13:36,156 --> 00:13:38,836 Speaker 1: talking about. There was a project that a group did 241 00:13:38,836 --> 00:13:41,636 Speaker 1: as a water district in California, and the problem they 242 00:13:41,676 --> 00:13:44,476 Speaker 1: faced was when drought season comes, you know, it's really 243 00:13:44,476 --> 00:13:46,556 Speaker 1: hard to get water to folks. People don't have water. 244 00:13:47,316 --> 00:13:51,556 Speaker 1: That's obviously problematic. You need drinking water and water to bathe, etc. 245 00:13:52,356 --> 00:13:55,556 Speaker 1: But more than that, the cost of not getting them 246 00:13:55,556 --> 00:13:58,356 Speaker 1: water is really high because the only way that they 247 00:13:58,356 --> 00:13:59,996 Speaker 1: can get water to the places they don't have it 248 00:14:00,076 --> 00:14:03,276 Speaker 1: is to actually take a dump truck, drive it up 249 00:14:03,316 --> 00:14:06,516 Speaker 1: to some other reservoir, maybe over to Nevada, literally fill 250 00:14:06,556 --> 00:14:09,556 Speaker 1: it by hand and drive it back. So you're also 251 00:14:09,556 --> 00:14:13,436 Speaker 1: facing like huge environmental costs, huge energy costs. So they 252 00:14:13,436 --> 00:14:15,716 Speaker 1: ask the question, you know, could we figure out a 253 00:14:15,756 --> 00:14:18,196 Speaker 1: way to predict how much water demand there's going to 254 00:14:18,276 --> 00:14:20,516 Speaker 1: be at a more granular level so we can really 255 00:14:20,596 --> 00:14:23,956 Speaker 1: understand and ration more effectively. And so we team them 256 00:14:23,996 --> 00:14:27,036 Speaker 1: up with some data scientists that come from everywhere from 257 00:14:27,036 --> 00:14:31,596 Speaker 1: Netflix to environmental science organizations, and together they collected the 258 00:14:31,676 --> 00:14:34,316 Speaker 1: data at almost a block by block level, and they 259 00:14:34,396 --> 00:14:36,716 Speaker 1: built an algorithm that sort of takes that data in 260 00:14:36,756 --> 00:14:39,916 Speaker 1: and continually gives updates. Does water district to say, hey, 261 00:14:39,916 --> 00:14:41,596 Speaker 1: this is how much we think people are going to use. 262 00:14:41,636 --> 00:14:43,836 Speaker 1: Here's how much they've already used. Tomorrow, you're probably going 263 00:14:43,876 --> 00:14:45,756 Speaker 1: to see this, And they said, in the first year 264 00:14:45,756 --> 00:14:48,236 Speaker 1: of using this, they saved over twenty five million dollars 265 00:14:48,476 --> 00:14:50,876 Speaker 1: in addition to getting water to people much more effectively. 266 00:14:51,516 --> 00:14:53,276 Speaker 1: So I think when you hear about cases like that 267 00:14:53,716 --> 00:14:55,876 Speaker 1: those are the kinds of examples that we want to 268 00:14:56,596 --> 00:14:58,996 Speaker 1: kind of platform and see more even the world where 269 00:14:59,436 --> 00:15:03,556 Speaker 1: within the confines of social organization these data and algorithms 270 00:15:03,556 --> 00:15:07,356 Speaker 1: that can really drive real effectiveness. Now your people are 271 00:15:07,356 --> 00:15:11,116 Speaker 1: all doing this for good. We've all heard about the 272 00:15:11,236 --> 00:15:14,676 Speaker 1: kinds of bias issues that have started to turn up 273 00:15:14,716 --> 00:15:18,796 Speaker 1: with predictive algorithms of different kinds, and they seem to 274 00:15:18,796 --> 00:15:24,236 Speaker 1: get embedded just because of the inherited unconscious biases of 275 00:15:24,276 --> 00:15:27,076 Speaker 1: the people who write the algorithm. Absolutely, how do you 276 00:15:27,116 --> 00:15:32,556 Speaker 1: avoid recapitulating that problem again with the projects you're working on? 277 00:15:32,956 --> 00:15:35,796 Speaker 1: Such an awesome question, and I think just to comment 278 00:15:35,836 --> 00:15:39,836 Speaker 1: on the challenge generally, I think you really nailed it there. 279 00:15:39,876 --> 00:15:43,436 Speaker 1: That the challenge that we face is that humans have 280 00:15:43,476 --> 00:15:48,236 Speaker 1: been collecting data from our activities that incorporate unconscious bias, 281 00:15:48,316 --> 00:15:50,196 Speaker 1: and so if you then have a machine learn from 282 00:15:50,196 --> 00:15:53,836 Speaker 1: it or you analyze it, you write replicating that. So, 283 00:15:54,596 --> 00:15:56,636 Speaker 1: while I will not admit that we have a perfect solution, 284 00:15:56,636 --> 00:15:58,796 Speaker 1: because I mean we're sort of talking about the challenge 285 00:15:58,796 --> 00:16:02,436 Speaker 1: of bias and humanity, some of the things that we 286 00:16:02,476 --> 00:16:05,676 Speaker 1: really focus on is the technology to us that we're 287 00:16:05,676 --> 00:16:09,996 Speaker 1: building is secondary to the outcome for people. So, for example, 288 00:16:10,076 --> 00:16:13,516 Speaker 1: it's not exciting to us to build an algorithm that 289 00:16:13,956 --> 00:16:18,076 Speaker 1: helps a homeless shelter triage people to the right homeless 290 00:16:18,116 --> 00:16:21,556 Speaker 1: shelters correctly just because it's a cool algorithm. We only 291 00:16:21,596 --> 00:16:23,556 Speaker 1: care if at the end of the day, the ultimate 292 00:16:23,596 --> 00:16:27,196 Speaker 1: success metric that you know, a wide range of inclusive 293 00:16:27,236 --> 00:16:31,196 Speaker 1: folks are getting housing is achieved. So I want to 294 00:16:31,196 --> 00:16:33,036 Speaker 1: say that first because I think one of the reasons 295 00:16:33,036 --> 00:16:36,116 Speaker 1: we see some of these biased challenges rise up is 296 00:16:36,116 --> 00:16:39,116 Speaker 1: that folks say, hey, the algorithm is doing something. It's 297 00:16:39,116 --> 00:16:42,236 Speaker 1: doing a thing I want, like giving out sentences in 298 00:16:42,396 --> 00:16:45,476 Speaker 1: courts or you know, policing folks, but without a question 299 00:16:45,476 --> 00:16:48,676 Speaker 1: of and how is it biased? Towards the end, you know, 300 00:16:48,716 --> 00:16:50,836 Speaker 1: what's it achieving. But the other thing we do is 301 00:16:50,876 --> 00:16:54,356 Speaker 1: we work extremely closely with our NGEO partners who are 302 00:16:54,436 --> 00:16:58,116 Speaker 1: on the ground and who understand a lot of those challenges. 303 00:16:58,476 --> 00:17:00,516 Speaker 1: And so we'll actually do what we call a pre 304 00:17:00,596 --> 00:17:03,036 Speaker 1: mortem some other companies do, which is before we even 305 00:17:03,036 --> 00:17:05,116 Speaker 1: start a project, we'll say, okay, let's pretend we jump 306 00:17:05,196 --> 00:17:08,716 Speaker 1: to the end. Well, you know, basic questions like how 307 00:17:08,716 --> 00:17:11,276 Speaker 1: will this be maintained, who's actually going to use this 308 00:17:11,276 --> 00:17:12,956 Speaker 1: tool at the end of the day. But then we'll 309 00:17:12,996 --> 00:17:15,836 Speaker 1: also ask two questions, which is one, what's the worst 310 00:17:15,876 --> 00:17:18,236 Speaker 1: that happens if we fail? So if you're relying on 311 00:17:18,356 --> 00:17:21,436 Speaker 1: us to build, this is not something we would necessarily build. 312 00:17:21,436 --> 00:17:23,676 Speaker 1: But let's say someone said, hey, we want a tool 313 00:17:23,716 --> 00:17:26,396 Speaker 1: that predicts whether you have cancer or not. Okay, well 314 00:17:26,436 --> 00:17:29,236 Speaker 1: that's pretty serious. And if we don't succeed, are you 315 00:17:29,996 --> 00:17:32,316 Speaker 1: stuck because you really needed that and now your organization 316 00:17:32,356 --> 00:17:34,756 Speaker 1: can't proceed. That's important to know. But then we also 317 00:17:34,796 --> 00:17:37,516 Speaker 1: ask what's the worst that happens if we succeed? So 318 00:17:38,196 --> 00:17:39,956 Speaker 1: who is this going to affect? How would you know 319 00:17:40,036 --> 00:17:42,236 Speaker 1: that it's wrong? Right? Like, how would you know just 320 00:17:42,236 --> 00:17:45,036 Speaker 1: because it's chugging away making predictions? Is it doing the 321 00:17:45,116 --> 00:17:48,636 Speaker 1: right thing? Is it disenfranchising certain groups? Could somebody use 322 00:17:48,636 --> 00:17:51,956 Speaker 1: it to intentionally target people who have cancer? We ask 323 00:17:51,956 --> 00:17:54,356 Speaker 1: a lot of those questions, and what's really important us 324 00:17:54,356 --> 00:17:58,156 Speaker 1: in that questioning is who has the power and agency 325 00:17:58,236 --> 00:18:02,116 Speaker 1: to both understand the algorithm and change the algorithm Because 326 00:18:02,156 --> 00:18:05,036 Speaker 1: in the current landscape, when tech companies build algorithms, it's 327 00:18:05,076 --> 00:18:06,516 Speaker 1: not much you can do. But you know, I don't 328 00:18:06,516 --> 00:18:10,116 Speaker 1: have enough agency to know how Facebook's news feed algorithm works, 329 00:18:10,116 --> 00:18:13,156 Speaker 1: nor can I really affect it much? But that's not 330 00:18:13,196 --> 00:18:16,116 Speaker 1: acceptable to me when you're bringing algorithms into the public 331 00:18:16,396 --> 00:18:19,116 Speaker 1: good space and this is actually affecting folks lives. So 332 00:18:19,156 --> 00:18:20,716 Speaker 1: those are some of the questions we ask up front 333 00:18:20,756 --> 00:18:22,596 Speaker 1: and really try to be rigorous with our partners around 334 00:18:22,676 --> 00:18:25,316 Speaker 1: oversight of and oftentimes that's enough for us to not 335 00:18:25,396 --> 00:18:28,356 Speaker 1: take on a project. It's great that you're thinking steps 336 00:18:28,356 --> 00:18:33,716 Speaker 1: ahead about these projects, and your own solvable is, ironically, 337 00:18:33,756 --> 00:18:35,876 Speaker 1: to put yourself out of business is to create a 338 00:18:35,916 --> 00:18:38,436 Speaker 1: world in which you don't need a data kind to 339 00:18:38,476 --> 00:18:41,716 Speaker 1: point people towards positive uses of data. That's right, What 340 00:18:41,716 --> 00:18:44,556 Speaker 1: would it take to make that happen? And I guess 341 00:18:44,716 --> 00:18:48,556 Speaker 1: playing your chess game. What happens when that happens. The 342 00:18:48,636 --> 00:18:51,116 Speaker 1: day we close our doors is the data. Every frontline 343 00:18:51,236 --> 00:18:54,356 Speaker 1: social change organization has the capabilities to use data and 344 00:18:54,396 --> 00:18:57,156 Speaker 1: AI the same way the big tech companies do ethically 345 00:18:57,236 --> 00:19:00,636 Speaker 1: and capably. And so you know, our little slice of 346 00:19:00,676 --> 00:19:03,796 Speaker 1: that today is to bridge the gap in getting the 347 00:19:03,876 --> 00:19:06,836 Speaker 1: human capital, the talent, the data scientists AI engineers to 348 00:19:07,036 --> 00:19:10,436 Speaker 1: social organizations. That sort of step one is to show 349 00:19:10,436 --> 00:19:12,876 Speaker 1: people the art of the possible and really get some 350 00:19:12,876 --> 00:19:15,076 Speaker 1: of those challenges solved. But what do it take to 351 00:19:15,076 --> 00:19:16,636 Speaker 1: do that? Long runs to think about what are the 352 00:19:16,676 --> 00:19:19,356 Speaker 1: problems and hurdles we're trying to overcome with that model today, 353 00:19:19,756 --> 00:19:23,076 Speaker 1: and they are that in the social sector there isn't 354 00:19:23,156 --> 00:19:25,516 Speaker 1: enough awareness about what the technology could do or where 355 00:19:25,516 --> 00:19:27,476 Speaker 1: it would be applied. So we have to start with that, 356 00:19:27,556 --> 00:19:30,956 Speaker 1: and I think now increasingly you're seeing more of more 357 00:19:30,996 --> 00:19:35,276 Speaker 1: folks understanding that, more companies talking about doing data and 358 00:19:35,396 --> 00:19:38,636 Speaker 1: AI for good. So I feel like there's some progress there, 359 00:19:38,876 --> 00:19:40,796 Speaker 1: But if you go further, you have to think, well, 360 00:19:40,796 --> 00:19:44,036 Speaker 1: how would a government or nonprofit get access to these 361 00:19:44,076 --> 00:19:47,316 Speaker 1: resources in the long term, And there I think there's 362 00:19:47,356 --> 00:19:50,036 Speaker 1: going to be a long term shift in getting funding 363 00:19:50,316 --> 00:19:54,876 Speaker 1: to move towards nonprofits for things like data science and AI. 364 00:19:55,036 --> 00:19:59,196 Speaker 1: You're going to need maybe consultancies that actually provide this 365 00:19:59,316 --> 00:20:02,436 Speaker 1: service in the social sector. There's lots of different models 366 00:20:02,436 --> 00:20:04,836 Speaker 1: for where that capacity could come from, but I think 367 00:20:04,836 --> 00:20:06,836 Speaker 1: the biggest things that we need right now are that 368 00:20:06,876 --> 00:20:09,476 Speaker 1: awareness of how could be used and then the I say, 369 00:20:09,476 --> 00:20:12,276 Speaker 1: the funding for ngox to be able to hire a 370 00:20:12,356 --> 00:20:15,916 Speaker 1: data sciences and incorporate them into the work they do. Now. 371 00:20:16,356 --> 00:20:19,476 Speaker 1: When that happens, what happens. Oh, I mean, I'd love 372 00:20:19,476 --> 00:20:23,676 Speaker 1: to say that all challenges that are stymied by not 373 00:20:23,756 --> 00:20:27,196 Speaker 1: having data science and AI are solved live apply ever after. 374 00:20:27,796 --> 00:20:31,116 Speaker 1: But actually, what I think my most ambitious hope for 375 00:20:31,156 --> 00:20:34,516 Speaker 1: the world is that we could actually tip the balance 376 00:20:34,556 --> 00:20:38,716 Speaker 1: a little bit to where the social sector is paving 377 00:20:38,756 --> 00:20:42,036 Speaker 1: the path for how machine learning and AI could be used. 378 00:20:42,516 --> 00:20:44,876 Speaker 1: I think we're so built into this default model that 379 00:20:45,356 --> 00:20:48,996 Speaker 1: business and wealthy countries set the agenda and everyone else 380 00:20:49,076 --> 00:20:52,676 Speaker 1: kind of struggles to catch up and imitate. We're talking 381 00:20:52,796 --> 00:20:55,956 Speaker 1: about a technology that is so fundamental to humanity because 382 00:20:55,996 --> 00:20:59,036 Speaker 1: it relies on data about us. When we talk about AI, 383 00:20:59,156 --> 00:21:01,916 Speaker 1: it is like automating human processes that I don't think 384 00:21:01,916 --> 00:21:04,276 Speaker 1: that's something that should be just a business application that 385 00:21:04,396 --> 00:21:07,276 Speaker 1: is ported to the world. There should be a place 386 00:21:07,636 --> 00:21:10,036 Speaker 1: for us to say, what does it look like when 387 00:21:10,036 --> 00:21:12,516 Speaker 1: we apply the technology to the better angels of our nature? 388 00:21:12,756 --> 00:21:16,036 Speaker 1: What is human based AI? What are the things we 389 00:21:16,076 --> 00:21:18,316 Speaker 1: care about? And I can't think of any other place 390 00:21:18,316 --> 00:21:21,356 Speaker 1: besides the social sector whose sole mandate is to look 391 00:21:21,356 --> 00:21:24,756 Speaker 1: out for humanity. So my dream is when you bridge 392 00:21:24,836 --> 00:21:27,476 Speaker 1: that gap, when that's there. You could actually have this 393 00:21:27,596 --> 00:21:30,836 Speaker 1: voice from the social sector itself saying what it looks 394 00:21:30,876 --> 00:21:33,236 Speaker 1: like to have human based ai Jick. Do you think 395 00:21:33,276 --> 00:21:37,716 Speaker 1: about the training of data scientists. I sometimes think we're 396 00:21:37,796 --> 00:21:43,996 Speaker 1: just missing the intersection between moral philosophy and computer science. 397 00:21:44,076 --> 00:21:46,476 Speaker 1: You know, the people who are majoring in college and 398 00:21:46,676 --> 00:21:50,396 Speaker 1: electronic engineering aren't reading much Kant, and the people who 399 00:21:50,396 --> 00:21:54,756 Speaker 1: are reading Kant don't understand much about computer programming, you know, 400 00:21:54,796 --> 00:21:57,476 Speaker 1: And in a way, the problem is that the people 401 00:21:57,516 --> 00:22:00,716 Speaker 1: at these tech companies don't have a different kind of 402 00:22:00,756 --> 00:22:05,356 Speaker 1: background in literature and philosophy and history to think through 403 00:22:05,436 --> 00:22:08,836 Speaker 1: the implications of what they're building the way you clearly 404 00:22:08,876 --> 00:22:12,076 Speaker 1: are thinking through those implications. I think it's a really 405 00:22:12,116 --> 00:22:16,716 Speaker 1: great point that when wielding the technology, it's really important 406 00:22:16,756 --> 00:22:21,476 Speaker 1: to have a very varied sense of skills somewhere in 407 00:22:21,476 --> 00:22:24,556 Speaker 1: the conversation. And increasingly you're seeing data science and tech 408 00:22:25,236 --> 00:22:29,196 Speaker 1: curricula incorporate ethics training into their courses, which I think 409 00:22:29,316 --> 00:22:31,636 Speaker 1: is great. In the same way that I'm not a 410 00:22:31,716 --> 00:22:35,076 Speaker 1: historian myself, I feel like physics went through this reckoning 411 00:22:36,076 --> 00:22:38,276 Speaker 1: with the ethics of what was being built when they 412 00:22:38,316 --> 00:22:41,116 Speaker 1: went from the joy of all energy and nuclear power 413 00:22:41,156 --> 00:22:43,516 Speaker 1: to the realizations of the downsides of the nuclear bomb 414 00:22:43,636 --> 00:22:46,156 Speaker 1: nuclear weapons. So I think you're going to see that 415 00:22:46,196 --> 00:22:48,996 Speaker 1: similar shift, which is which is great, But you know, 416 00:22:49,076 --> 00:22:52,036 Speaker 1: I think what your question raises actually a bigger point 417 00:22:52,076 --> 00:22:56,316 Speaker 1: to me, which is who holds the responsibility for the 418 00:22:56,356 --> 00:23:00,076 Speaker 1: ethical applications of this technology? And I'll just say, while 419 00:23:00,316 --> 00:23:03,276 Speaker 1: I would love to see, you know, ethical code around 420 00:23:03,316 --> 00:23:07,516 Speaker 1: data science, it's a lot of responsibility to say that 421 00:23:08,236 --> 00:23:11,676 Speaker 1: engineer x it has come out of college engineering college 422 00:23:11,716 --> 00:23:13,876 Speaker 1: for two years and is working at big tech company 423 00:23:14,476 --> 00:23:18,276 Speaker 1: and gets asked by their boss to build something fairly benign, 424 00:23:18,556 --> 00:23:23,436 Speaker 1: like I upgrade to their their GPS system that recommends 425 00:23:23,796 --> 00:23:26,956 Speaker 1: routes you can walk that avoid crime ridden areas. I say, 426 00:23:26,996 --> 00:23:30,716 Speaker 1: here's an algorith build that. Well, number one, that's not 427 00:23:30,756 --> 00:23:33,076 Speaker 1: necessarily a bad thing to builds not like you know, 428 00:23:33,116 --> 00:23:34,996 Speaker 1: it's not as black and white as some people may 429 00:23:34,996 --> 00:23:37,996 Speaker 1: feel about building a weapon or something. But of course, 430 00:23:38,036 --> 00:23:40,876 Speaker 1: if you sort of play the game through, if everyone 431 00:23:40,916 --> 00:23:43,556 Speaker 1: were using an app that avoided crime ridden areas, probably 432 00:23:43,636 --> 00:23:46,796 Speaker 1: end up with some sort of digital segregation. So number one, 433 00:23:46,836 --> 00:23:49,676 Speaker 1: there's already long range effects that you'd have to anticipate. 434 00:23:49,756 --> 00:23:52,156 Speaker 1: But more than that, It also relies on that, you know, 435 00:23:52,276 --> 00:23:55,716 Speaker 1: second year engineer to say, hey boss, yeah, I'm not 436 00:23:55,796 --> 00:23:59,316 Speaker 1: doing that. You know this is I'm quitting, which, given 437 00:23:59,676 --> 00:24:02,956 Speaker 1: you know people's career paths and the money associate with 438 00:24:03,036 --> 00:24:06,276 Speaker 1: these jobs, is a big ask. So I would say 439 00:24:06,396 --> 00:24:09,356 Speaker 1: it's not just about the technologies. I think the question is, 440 00:24:09,436 --> 00:24:12,116 Speaker 1: you know, how do we share that responsibility? Is it 441 00:24:12,156 --> 00:24:14,436 Speaker 1: the technologist to make this call? Was it the manager 442 00:24:14,556 --> 00:24:16,756 Speaker 1: said we want to build this feature? Was it the 443 00:24:16,756 --> 00:24:19,876 Speaker 1: constituents would be affected by that? Is a government to 444 00:24:19,876 --> 00:24:22,436 Speaker 1: come regulate. I don't think there's any one answer, but 445 00:24:22,556 --> 00:24:25,076 Speaker 1: I do think the frame that people have I'm hearing 446 00:24:25,076 --> 00:24:28,636 Speaker 1: more in the public right now around technologists need to 447 00:24:28,676 --> 00:24:31,196 Speaker 1: know the ethics, I think is missing the bigger picture 448 00:24:31,236 --> 00:24:34,156 Speaker 1: that that alone isn't the right responsibility model. In my mind. 449 00:24:34,596 --> 00:24:37,796 Speaker 1: You have two very different ideas of capitalism, right. I mean, 450 00:24:37,836 --> 00:24:42,516 Speaker 1: there's an older idea that government sets the rules, tells 451 00:24:42,556 --> 00:24:44,836 Speaker 1: you what you can and can't do, and that businesses 452 00:24:44,876 --> 00:24:48,636 Speaker 1: should obey the law and regulation but go be very 453 00:24:48,676 --> 00:24:52,036 Speaker 1: free to do what they want. Within that, the newer 454 00:24:52,116 --> 00:24:56,516 Speaker 1: model suggests that the businesses themselves have a higher degree 455 00:24:56,596 --> 00:25:00,676 Speaker 1: of social responsibility, and it's not enough to follow the 456 00:25:00,756 --> 00:25:04,236 Speaker 1: rules that they have to be thinking about outcomes. Look, 457 00:25:04,316 --> 00:25:07,356 Speaker 1: I would love to live in a world where business 458 00:25:07,396 --> 00:25:12,236 Speaker 1: and social outcome were somehow linked, where the fact that 459 00:25:12,316 --> 00:25:16,196 Speaker 1: businesses were accountable somehow to at least not doing harm, 460 00:25:16,236 --> 00:25:18,316 Speaker 1: if not improving human life. That would be a really 461 00:25:18,356 --> 00:25:22,436 Speaker 1: great intersection. Call me a cynic, but we're not really 462 00:25:22,436 --> 00:25:25,796 Speaker 1: currently set up for that. The incentives aren't there. In 463 00:25:25,836 --> 00:25:29,196 Speaker 1: my mind, businesses are still held mostly to the bottom line, 464 00:25:29,236 --> 00:25:33,236 Speaker 1: even though we are seeing some increased interest in social entrepreneurship, 465 00:25:33,276 --> 00:25:36,236 Speaker 1: where businesses may have a double bottom line, one that's 466 00:25:36,276 --> 00:25:39,956 Speaker 1: monetary and one that's social, or new structures like b 467 00:25:40,116 --> 00:25:42,196 Speaker 1: corps that actually say, hey, we are committed to some 468 00:25:42,236 --> 00:25:45,436 Speaker 1: social cause. But I think it's a lot to ask 469 00:25:46,076 --> 00:25:48,436 Speaker 1: of a company. And as much as it's a nice 470 00:25:48,476 --> 00:25:51,476 Speaker 1: idea of a future of capitalism, it's certainly not the 471 00:25:51,556 --> 00:25:55,316 Speaker 1: rule or the law. And so I don't think that's 472 00:25:55,316 --> 00:25:56,996 Speaker 1: going to be the sole model that brings us to 473 00:25:57,036 --> 00:26:00,236 Speaker 1: a world of pro social technology and AI. If for 474 00:26:00,316 --> 00:26:06,676 Speaker 1: no other reason then certain human needs are inherently cost ineffective, 475 00:26:06,676 --> 00:26:08,716 Speaker 1: I would say to solve at least currently if people 476 00:26:08,716 --> 00:26:10,996 Speaker 1: could cry those if every social problem were able to 477 00:26:11,036 --> 00:26:13,996 Speaker 1: align perfectly with a business needs, be in great shape. 478 00:26:13,996 --> 00:26:17,836 Speaker 1: But when it comes to housing the homeless or making 479 00:26:17,836 --> 00:26:21,516 Speaker 1: sure that people have food to eat, that is a 480 00:26:21,556 --> 00:26:25,236 Speaker 1: difficult challenge that I don't see an immediate market solution too, 481 00:26:25,356 --> 00:26:27,636 Speaker 1: and so I don't think even the best intention companies 482 00:26:27,636 --> 00:26:30,516 Speaker 1: could survive in a market based world trying to solve 483 00:26:30,556 --> 00:26:33,556 Speaker 1: that problem. I mean, Google, which is still the first 484 00:26:33,596 --> 00:26:37,996 Speaker 1: and best known data company essentially has held out this 485 00:26:38,036 --> 00:26:40,916 Speaker 1: promise that we're going to make a lot of money 486 00:26:41,036 --> 00:26:45,996 Speaker 1: using data commercially targeting advertising, but we're going to use 487 00:26:46,276 --> 00:26:48,076 Speaker 1: a lot of what we make, or at least some 488 00:26:48,156 --> 00:26:51,036 Speaker 1: of it in a kind of philanthropy. We're going to 489 00:26:51,116 --> 00:26:54,436 Speaker 1: try to create some of the kinds of solutions you're 490 00:26:54,476 --> 00:26:57,716 Speaker 1: talking about that aren't driven by the profit motive. Does 491 00:26:57,756 --> 00:27:00,476 Speaker 1: that work look like? I said, One of the big 492 00:27:00,556 --> 00:27:03,876 Speaker 1: challenges we face, I think in the social sector right 493 00:27:03,916 --> 00:27:07,156 Speaker 1: now is the lack of funding for innovation for your technology. 494 00:27:07,716 --> 00:27:12,556 Speaker 1: And so if company are going to offer that great netwin, 495 00:27:13,316 --> 00:27:17,316 Speaker 1: do I believe that the world's biggest challenges will be 496 00:27:17,396 --> 00:27:21,156 Speaker 1: solved on the you know, philanthropic efforts of large companies 497 00:27:22,396 --> 00:27:25,636 Speaker 1: that I'm not so hopeful. I think there. I still 498 00:27:25,676 --> 00:27:28,356 Speaker 1: wonder where are the folks for whom the mandate is 499 00:27:28,436 --> 00:27:32,516 Speaker 1: solely pro social, you know, for governments or again nonprofits 500 00:27:32,596 --> 00:27:35,876 Speaker 1: or civic organizations whose very guiding mission is to make 501 00:27:35,916 --> 00:27:39,756 Speaker 1: sure that human prosperity is enhanced. There's a little bit 502 00:27:39,796 --> 00:27:41,636 Speaker 1: more of a direct line there. And so that's why 503 00:27:41,676 --> 00:27:43,996 Speaker 1: I think it has to be a combination of the two, 504 00:27:44,276 --> 00:27:46,916 Speaker 1: and why we focus so much on saying instead of 505 00:27:46,956 --> 00:27:50,676 Speaker 1: trying to bend the Googles of the world to you know, 506 00:27:50,796 --> 00:27:52,756 Speaker 1: being in charge of clean water, which frankly I think 507 00:27:52,876 --> 00:27:54,356 Speaker 1: is really not not the way you want to go. 508 00:27:54,916 --> 00:27:57,116 Speaker 1: Where the you know, the clean water organizations of the 509 00:27:57,116 --> 00:27:59,156 Speaker 1: world who just need that same technology to be ten 510 00:27:59,356 --> 00:28:03,156 Speaker 1: hundred times more effective. What are some things listeners to 511 00:28:03,196 --> 00:28:06,476 Speaker 1: this podcast might be able to do to work towards 512 00:28:06,476 --> 00:28:09,436 Speaker 1: the kinds of solutions you're thinking about. Well, the great 513 00:28:09,476 --> 00:28:12,356 Speaker 1: thing about this cross cutting technology is that everyone has 514 00:28:12,356 --> 00:28:15,196 Speaker 1: a role to play in creating this future vision of 515 00:28:15,516 --> 00:28:19,036 Speaker 1: more social and positive AI. Well, first, I would say, 516 00:28:19,036 --> 00:28:21,636 Speaker 1: if you're a technologist who works with data and you 517 00:28:21,716 --> 00:28:24,156 Speaker 1: want to give your time and energy back, come aboard. 518 00:28:24,476 --> 00:28:26,716 Speaker 1: There's a whole movement of folks doing this work. Whether 519 00:28:26,756 --> 00:28:28,276 Speaker 1: you want to come work with us at Data Kind 520 00:28:28,276 --> 00:28:30,876 Speaker 1: and work on projects pro bono, or with many of 521 00:28:30,876 --> 00:28:34,916 Speaker 1: the other organizations like Driven Data, Data Science for Social Good, 522 00:28:35,156 --> 00:28:38,236 Speaker 1: CODE for America who take technologists and apply them to 523 00:28:38,276 --> 00:28:41,516 Speaker 1: social problems, come aboard. There's no reason to wait. And 524 00:28:41,636 --> 00:28:44,756 Speaker 1: increasing Link asked the company you work for if there's 525 00:28:44,916 --> 00:28:47,236 Speaker 1: opportunities to give back, because we see more tech companies 526 00:28:47,276 --> 00:28:49,876 Speaker 1: do that. But if you're not a data scientist, non 527 00:28:49,956 --> 00:28:53,316 Speaker 1: data scientist, I would say, yeah, I have to first 528 00:28:53,316 --> 00:28:55,116 Speaker 1: give a shout out to anyone of the funder or 529 00:28:55,196 --> 00:28:58,036 Speaker 1: donor world. One of the big gaps here is that 530 00:28:58,076 --> 00:29:00,556 Speaker 1: there isn't enough funding for technology and innovation in the 531 00:29:00,596 --> 00:29:03,636 Speaker 1: social sector. So I've been very impressed with the efforts 532 00:29:03,636 --> 00:29:08,076 Speaker 1: of Rockefeller Foundation and MasterCard Impact Fund and others who 533 00:29:08,076 --> 00:29:11,476 Speaker 1: are giving big amounts of funding to data and AI 534 00:29:11,516 --> 00:29:13,716 Speaker 1: and social good to bring it on. We need more 535 00:29:13,716 --> 00:29:17,156 Speaker 1: of that for this happen. But very lastly, if not 536 00:29:17,196 --> 00:29:19,636 Speaker 1: a data scientist and you're not a funder, I would 537 00:29:19,676 --> 00:29:22,716 Speaker 1: say there's a huge opportunity to get involved in just 538 00:29:22,876 --> 00:29:26,396 Speaker 1: understanding what this new technology can do. Ciicero had a 539 00:29:26,436 --> 00:29:28,956 Speaker 1: quote that you should take an interest in politics, because 540 00:29:29,236 --> 00:29:31,316 Speaker 1: politics is definitely going to take an interest in you. 541 00:29:31,916 --> 00:29:34,396 Speaker 1: And I feel exactly the same about data and algorithms. 542 00:29:34,516 --> 00:29:36,196 Speaker 1: They're going to take an interest in all of us. 543 00:29:36,316 --> 00:29:39,596 Speaker 1: In fact, they're shaping our lives already today. Maybe the 544 00:29:39,636 --> 00:29:42,116 Speaker 1: reason you're listening to this podcast is because an algorithm 545 00:29:42,196 --> 00:29:44,956 Speaker 1: recommended it to you based on your previous listening habits. 546 00:29:45,236 --> 00:29:48,116 Speaker 1: And so if these tools are going to be shaping 547 00:29:48,196 --> 00:29:52,036 Speaker 1: and visibly shaping our decisions, then it's all the more 548 00:29:52,076 --> 00:29:57,556 Speaker 1: incumbent on us as society to understand what the ramifications are, 549 00:29:57,916 --> 00:30:00,796 Speaker 1: where it's showing up in society, and how we might 550 00:30:00,836 --> 00:30:03,196 Speaker 1: have some agency over the role we want it to play. 551 00:30:03,676 --> 00:30:05,396 Speaker 1: I think so much of the reason you hear so 552 00:30:05,436 --> 00:30:09,236 Speaker 1: much negativity today is because we don't understand it well 553 00:30:09,356 --> 00:30:11,196 Speaker 1: enough and we don't have any agency to change it. 554 00:30:11,276 --> 00:30:13,476 Speaker 1: So our only options are to shrug and say, well, 555 00:30:13,476 --> 00:30:14,916 Speaker 1: I guess that's going to be the way it is, 556 00:30:15,396 --> 00:30:17,356 Speaker 1: or to rail against it and say this is bad. 557 00:30:17,796 --> 00:30:19,716 Speaker 1: But if we could get to a place where we 558 00:30:19,756 --> 00:30:23,796 Speaker 1: had call it algorithmic literacy. Not everyone needs to code, 559 00:30:23,836 --> 00:30:25,676 Speaker 1: but if you just understand a little more about it, 560 00:30:25,996 --> 00:30:28,636 Speaker 1: then I think we'd progress towards a society where we 561 00:30:28,716 --> 00:30:31,916 Speaker 1: felt like we had a more control agency over how 562 00:30:31,956 --> 00:30:34,396 Speaker 1: we work with the machines instead of against them. That's 563 00:30:34,396 --> 00:30:36,596 Speaker 1: a great point. And I have to ask you for 564 00:30:36,676 --> 00:30:40,116 Speaker 1: a reading recommendation. If people need to get educated, what 565 00:30:40,196 --> 00:30:42,476 Speaker 1: should they read. What's a thing or two they should 566 00:30:42,516 --> 00:30:46,516 Speaker 1: read to get more sophisticated about data. So the best 567 00:30:46,556 --> 00:30:48,636 Speaker 1: thing I think you can read are some of the 568 00:30:48,676 --> 00:30:51,276 Speaker 1: blogs that actually talk about the state of the space today, 569 00:30:51,956 --> 00:30:54,516 Speaker 1: because it's changing so much that you know there's no 570 00:30:54,556 --> 00:30:56,356 Speaker 1: one book that's going to capture it. Yeah. So some 571 00:30:56,396 --> 00:31:00,036 Speaker 1: of the ones I love are the company O'Reilly O'Reilly 572 00:31:00,116 --> 00:31:03,836 Speaker 1: dot com. They have a feature on data and AI 573 00:31:03,916 --> 00:31:06,636 Speaker 1: that's a weekly newsletter that comes out talking about everything 574 00:31:06,676 --> 00:31:10,636 Speaker 1: from the interesting innovations and AI to what kind of 575 00:31:10,956 --> 00:31:13,836 Speaker 1: privacy concerns are in the space today, and it's very 576 00:31:13,916 --> 00:31:15,796 Speaker 1: readable for a common audience. I think that's one of 577 00:31:15,796 --> 00:31:18,676 Speaker 1: the most interesting ones. I would also read Data and 578 00:31:18,796 --> 00:31:21,476 Speaker 1: Society's newsletter. They are a group here in New York 579 00:31:21,516 --> 00:31:24,276 Speaker 1: who are really tackling the question of what does it 580 00:31:24,276 --> 00:31:26,956 Speaker 1: mean to have data and algorithms in society. They have 581 00:31:27,036 --> 00:31:30,076 Speaker 1: some really great accessible writing there The other thing I 582 00:31:30,076 --> 00:31:33,036 Speaker 1: would say is if you have the privilege of living 583 00:31:33,116 --> 00:31:36,316 Speaker 1: near a medium, miss or big city that has a 584 00:31:36,356 --> 00:31:40,676 Speaker 1: meetup community. There are tons of data science AI meetups 585 00:31:40,716 --> 00:31:42,796 Speaker 1: where people go and just talk about what's going on 586 00:31:42,836 --> 00:31:45,516 Speaker 1: in the space. And I always recommend that people drop 587 00:31:45,556 --> 00:31:47,916 Speaker 1: by at least one because if you see it and 588 00:31:48,116 --> 00:31:50,676 Speaker 1: feel it and here people are talking about you don't 589 00:31:50,716 --> 00:31:52,676 Speaker 1: have to understand, you know, if there's any math on 590 00:31:52,716 --> 00:31:56,356 Speaker 1: the board, but just you almost immediately, it creates a 591 00:31:56,436 --> 00:31:59,356 Speaker 1: states where people walk and go, oh, I actually see 592 00:31:59,396 --> 00:32:01,516 Speaker 1: what this is all about. So I would say if 593 00:32:01,516 --> 00:32:03,756 Speaker 1: you happen to be a checkout meetup, dot com or 594 00:32:03,796 --> 00:32:07,116 Speaker 1: any of those communities. The data scientists AI folks are 595 00:32:07,236 --> 00:32:09,476 Speaker 1: very friendly and I know you'll have a great time, 596 00:32:09,516 --> 00:32:12,876 Speaker 1: if not an educational one. Terrific. Well, Jake Probi, thanks 597 00:32:12,876 --> 00:32:15,636 Speaker 1: for joining us Unsolvable My pleasure. Thanks so much for 598 00:32:15,636 --> 00:32:20,396 Speaker 1: having me reasons for hope all of this potential being 599 00:32:20,436 --> 00:32:25,196 Speaker 1: harnessed to improve people's lives, the really big stuff. Although 600 00:32:25,356 --> 00:32:28,036 Speaker 1: my ears certainly did prick up when Jake mentioned Twitter 601 00:32:28,076 --> 00:32:31,156 Speaker 1: for pets, as did my dog's ears. She has been 602 00:32:31,236 --> 00:32:35,476 Speaker 1: dying to get online and really drag other dogs anonymously, 603 00:32:35,596 --> 00:32:39,556 Speaker 1: of course, but both myself and my dog are pleased 604 00:32:39,596 --> 00:32:42,396 Speaker 1: to see what data Kind has actually managed to do 605 00:32:42,596 --> 00:32:46,596 Speaker 1: so far, creating algorithms that have helped transport clean water 606 00:32:46,716 --> 00:32:51,516 Speaker 1: more effectively, informed government policy that protects communities from corruption, 607 00:32:52,076 --> 00:32:56,236 Speaker 1: and detected crop disease using satellite imagery. Jake and his 608 00:32:56,356 --> 00:32:59,636 Speaker 1: team and all those volunteers are leveling the playing fields 609 00:32:59,636 --> 00:33:02,836 Speaker 1: and you can help too. Read more about data Kind 610 00:33:02,916 --> 00:33:06,596 Speaker 1: and how to get involved at Rockefella Foundation dot org. 611 00:33:06,756 --> 00:33:12,516 Speaker 1: Slash solvable. Solvable is a collaboration between Pushkin Industries and 612 00:33:12,516 --> 00:33:17,076 Speaker 1: the Rockefella Foundation, with production by Chalk and Blade. Pushkin's 613 00:33:17,076 --> 00:33:21,516 Speaker 1: executive producer is Mia LaBelle. Engineering by Jason Gambrell and 614 00:33:21,596 --> 00:33:26,036 Speaker 1: the fine folks at GSI Studios. Original music composed by 615 00:33:26,076 --> 00:33:31,036 Speaker 1: Pascal Wise. Special thanks to Maggie Taylor, Heather Faine, Julia Barton, 616 00:33:31,316 --> 00:33:36,436 Speaker 1: Carlie Migliori, Sheriff Vincent, Jacob Weisberg, and Malcolm Gladwell. You 617 00:33:36,476 --> 00:33:40,236 Speaker 1: can learn more about solving today's biggest problems at Rockefella 618 00:33:40,316 --> 00:33:45,156 Speaker 1: Foundation dot org. Slash Solvable. I'm Mave Higgins, Now go 619 00:33:45,476 --> 00:33:46,036 Speaker 1: solve Itt