WEBVTT - The Dog Will See You Now 0:00:15.356 --> 0:00:26.236 Bushkin. What's her name? 0:00:27.796 --> 0:00:29.436 Cooly? Sorry there you got some fun names. 0:00:29.636 --> 0:00:30.396 K U L. 0:00:32.036 --> 0:00:36.436 Cooley. It's a Belgian melanoa sleek, lean, dappled gray and black, 0:00:36.836 --> 0:00:39.756 panting happily on a patch of grass way up in 0:00:39.796 --> 0:00:44.556 the hills of northeastern Alabama. This dog is probably from hungry, hungry, hungry, 0:00:44.636 --> 0:00:45.276 and she's. 0:00:45.076 --> 0:00:50.436 Held ill exactly have her information? 0:00:50.676 --> 0:00:51.916 Twenty four months or below? 0:00:52.156 --> 0:00:54.676 Twenty four months or below. We have all that paperwork 0:00:54.756 --> 0:00:56.636 at the office, but right her own. 0:00:56.796 --> 0:01:00.476 That is the biggest tongue I've ever seen. We're all 0:01:00.476 --> 0:01:03.476 standing in front of a cluster of concrete buildings spread 0:01:03.516 --> 0:01:07.236 over a large, rolling, wooded campus an hour or so 0:01:07.316 --> 0:01:11.476 from Birmingham, the headcore of an organization called three sixty 0:01:11.556 --> 0:01:14.876 Canine Group. The whole operation looks like a scaled down 0:01:15.076 --> 0:01:18.556 army base, only instead of soldiers marching to and fro, 0:01:19.236 --> 0:01:24.596 there are dogs, dozens and dozens of them, Poppies stumbling playfully, 0:01:24.836 --> 0:01:29.236 chocolate labs chasing things, German shepherds looking for a job 0:01:29.276 --> 0:01:36.516 to do and cooling. One of the operation's newest recruits, 0:01:36.676 --> 0:01:39.916 a teenager in dog years, being put through her paces 0:01:40.556 --> 0:01:43.236 a robber toy for which she has acquired. A certain 0:01:43.316 --> 0:01:47.276 affection was taken from her and her job is to 0:01:47.316 --> 0:01:47.716 find it. 0:01:48.236 --> 0:01:50.276 You can see, like, where is she slaughtered? Right here, 0:01:50.596 --> 0:01:53.556 so she called the over here. Turned. Once she turned, 0:01:53.676 --> 0:01:54.756 she can see. 0:01:54.596 --> 0:01:55.636 That wokake Ovisually. 0:01:56.476 --> 0:01:59.636 She has never done an exercise like this before, and 0:01:59.716 --> 0:02:01.996 the assignment has filled her with enthusiasm. 0:02:03.516 --> 0:02:04.076 Us. 0:02:09.596 --> 0:02:14.396 Yes, Coolie is standing in front of one of the cabinets. 0:02:14.676 --> 0:02:20.756 Tail waggings were but she never toldly was in there. 0:02:21.556 --> 0:02:24.716 Cooley's a good girl, all right. She follows her nose 0:02:25.396 --> 0:02:30.836 and we have a lot to learn from her. My 0:02:30.956 --> 0:02:34.356 name is Malcolm Gladwell. You're listening to the final episode 0:02:34.476 --> 0:02:38.756 of season six of Revisions History, my podcast about things 0:02:38.876 --> 0:02:43.236 overlooked and misunderstood. This has been a season devoted to 0:02:43.356 --> 0:02:50.476 fixing things, college rankings, laundry, the Little Mermaid. We have 0:02:50.516 --> 0:02:54.316 looked at many burning issues and said we can do better. 0:02:54.956 --> 0:02:57.756 But there are some problems that we cannot fix on 0:02:57.836 --> 0:03:02.116 our own. Some problems are so big they can only 0:03:02.196 --> 0:03:05.236 be solved with the help of another species. 0:03:13.516 --> 0:03:15.396 Okay, so this is our dark room test. 0:03:18.516 --> 0:03:22.156 Cooley's undergoing yet another test with the same goal as before, 0:03:23.196 --> 0:03:26.636 find the hidden toy, only now the lights are off 0:03:27.116 --> 0:03:30.036 and the toy is not in the first room COOLi enters, 0:03:30.436 --> 0:03:33.236 but in a second, smaller room off to the side. 0:03:33.956 --> 0:03:36.236 This also tells us if it takes her a little 0:03:36.276 --> 0:03:39.156 while to try to find it, how much hunt draft 0:03:39.196 --> 0:03:43.036 she actually has, so it's not just immediate to make sure, 0:03:43.236 --> 0:03:45.796 even after she gets tired, even after she's hot, but 0:03:45.916 --> 0:03:47.756 she still wants that toy enough to be. 0:03:47.756 --> 0:03:48.556 Looking for it. 0:03:49.076 --> 0:03:52.996 Cooley bounds in sniff's the first room, everything all four walls. 0:03:53.156 --> 0:03:55.836 Then she zeros in on the closet. We can't see her, 0:03:56.116 --> 0:03:59.916 only hear her. Her trainers, Sean and Ashley are grading 0:03:59.956 --> 0:04:03.516 every step of her performance on a long, detailed checklist. 0:04:06.436 --> 0:04:09.516 You can hear the difference in a yes when she 0:04:09.516 --> 0:04:11.036 got it out him and she's gotten out for. 0:04:13.716 --> 0:04:14.036 Cans. 0:04:15.676 --> 0:04:18.876 Cooley walked across a mattress, didn't stop, jumped on one 0:04:18.956 --> 0:04:21.316 end of a makeshift teter totter, didn't phaze her. 0:04:22.116 --> 0:04:24.156 These are the dogs that people get as pets, and 0:04:24.236 --> 0:04:27.116 they eat their couch and they're the worst, terrible pets 0:04:27.276 --> 0:04:28.876 that make the really really good dogs. 0:04:28.916 --> 0:04:31.516 For the most part, you. 0:04:31.516 --> 0:04:32.276 Love all of that. 0:04:33.076 --> 0:04:42.036 She runs down a hallway, sniff sniff a large canvas 0:04:42.076 --> 0:04:45.636 bag is suspended from the ceiling, ready to drop and 0:04:45.676 --> 0:04:49.436 block her way, and pulled up to the ceiling, big 0:04:49.476 --> 0:04:49.996 heavy bag. 0:04:50.676 --> 0:04:51.276 Three guns. 0:04:51.796 --> 0:05:02.356 Here she comes. We all jumped when the bag dropped 0:05:02.356 --> 0:05:07.236 from the ceiling. Cooley didn't. She paused, looked at the bag, 0:05:07.356 --> 0:05:16.876 sniffed it, and carried on. Dogs noses are something like 0:05:16.996 --> 0:05:20.636 one hundred thousand times more sensitive than ours. A puppy 0:05:20.916 --> 0:05:23.636 fresh off the boat can find a hidden toy in 0:05:23.676 --> 0:05:28.036 a darkened room. Dogs can perform all manner of olfactory 0:05:28.076 --> 0:05:32.796 tasks using judgment, drive, and discipline. It's their superpower. That's 0:05:32.796 --> 0:05:35.516 why they are dogs at the airport sniffing luggage, or 0:05:35.516 --> 0:05:37.956 why it's always a dog who finds the hiker lost 0:05:37.996 --> 0:05:41.116 in the woods. Spend a little time at three point 0:05:41.076 --> 0:05:44.396 sixty k nine in the hills of Alabama, and every 0:05:44.396 --> 0:05:48.356 one of your preconceptions about the genius of dogs will 0:05:48.396 --> 0:05:49.116 be confirmed. 0:05:53.396 --> 0:05:57.116 We had a dog, a black lab that we retired, 0:05:57.476 --> 0:05:59.596 and we allowed the dog. You know, we had a 0:05:59.636 --> 0:06:03.716 ceremony and we bought we ceremony and we hired you know, 0:06:03.716 --> 0:06:06.556 there's a local bakery that makes dog friendly cakes, so 0:06:06.596 --> 0:06:07.796 we had a cake for the dog. 0:06:08.316 --> 0:06:11.436 We were sitting in the three sixty Knine conference room 0:06:11.476 --> 0:06:14.676 listening to Jerry Johnson and John Pierce, two of the 0:06:14.716 --> 0:06:19.956 company's senior executives, two hardcore dog guys, talking about the 0:06:19.996 --> 0:06:20.996 best they ever saw. 0:06:21.396 --> 0:06:23.236 And we had the dog in the room and run 0:06:23.276 --> 0:06:25.796 around and just you know, an off leash and sniffing 0:06:25.796 --> 0:06:28.956 everybody's super friendly dog. The lady comes in to deliver 0:06:29.036 --> 0:06:31.716 the cake and she's got on her boyfriend or husband 0:06:31.716 --> 0:06:33.636 with her and he's standing there hole in the cake 0:06:33.676 --> 0:06:37.356 and the dog key's running up behind the guy and 0:06:37.396 --> 0:06:40.556 going behind him and basically like sniffing his rear in 0:06:40.596 --> 0:06:43.996 and sniffing the back of his uh, you know, his shirt, 0:06:44.116 --> 0:06:46.676 and won't leave him alone. And I remember who asked him, 0:06:46.716 --> 0:06:49.436 but finally someone said, all right, yeah, John said, are 0:06:49.476 --> 0:06:52.676 you carrying a weapon? And he was like, no, I 0:06:53.476 --> 0:06:55.436 was carrying a weapon. I took it off and left 0:06:55.476 --> 0:06:57.156 it in the car, just prayer to walking in here. 0:06:57.356 --> 0:07:00.236 And the guy had, you know, three minutes before taking 0:07:00.276 --> 0:07:02.796 his gun off and left it in his car, and 0:07:02.836 --> 0:07:04.276 the dog still found it. And you know, we were 0:07:04.596 --> 0:07:05.196 having a party. 0:07:05.276 --> 0:07:06.276 What was that dog's name? 0:07:06.596 --> 0:07:11.236 That was Cole Cole coolly add them to the long 0:07:11.276 --> 0:07:14.676 list of Doggie All Stars. But you already knew dogs 0:07:14.716 --> 0:07:18.556 could do this. Right now, it's time to go further. 0:07:23.476 --> 0:07:27.156 A few years ago, a labrador named Florin flew in 0:07:27.236 --> 0:07:31.396 from London to Boston at the invitation of a research 0:07:31.476 --> 0:07:36.556 scientist at MIT named Andreas Mershen, coming in on a 0:07:36.596 --> 0:07:40.796 winter's morning for a special one dog exhibition. 0:07:42.436 --> 0:07:45.716 Flooring. The famous dog flew first class to MIT, so 0:07:45.756 --> 0:07:47.316 I didn't have time to go through all the hooks, 0:07:47.356 --> 0:07:49.836 so we had to rent the hotel room, the hotel 0:07:49.996 --> 0:07:53.116 ballroom right across the street at the Maria. 0:07:53.796 --> 0:07:57.156 Mershin is a kind of scientific jack of all trades. 0:07:57.756 --> 0:08:01.236 He's technically a physicist, but as he likes to say, 0:08:01.476 --> 0:08:07.436 he operates in blissful ignorance of disciplinary boundaries. He's interested 0:08:07.516 --> 0:08:12.836 in solar panels, quantum effects in molecular biology, not to 0:08:12.876 --> 0:08:18.036 mention something that I cannot begin to understand called cytoskeletal 0:08:18.276 --> 0:08:23.916 memory and coding. Mershan got interested in prostate cancer, one 0:08:23.916 --> 0:08:26.756 of the most common of all cancers. Mershan heard about 0:08:26.796 --> 0:08:29.836 the work being done by a British researcher named Claire Guest, 0:08:30.436 --> 0:08:34.236 who had been training dogs to detect diseases the same 0:08:34.236 --> 0:08:37.796 way we've been training dogs for years to detect explosives, 0:08:38.436 --> 0:08:41.236 and Merchan and Guest ended up working together, which is 0:08:41.276 --> 0:08:44.756 what led to Florin guests four year old labrador flying 0:08:44.796 --> 0:08:48.996 into Boston one cold winter's morning to perform a special 0:08:49.076 --> 0:08:54.636 sniffing demonstration for the senior leadership of the Prostate Cancer Foundation. 0:08:55.196 --> 0:08:56.516 So wait, described a scene. 0:08:56.516 --> 0:08:57.796 This is this is hilarious. 0:08:58.036 --> 0:09:00.596 You in the ballroom of the Marrion. Yes, how many people? 0:09:00.796 --> 0:09:03.196 Oh maybe maybe a dozen people maybe if you motif, 0:09:03.436 --> 0:09:05.836 But that's they're all by invitation. Okay, No. And there's 0:09:05.836 --> 0:09:08.396 a giant paris on the middle, about seven feet across, 0:09:08.676 --> 0:09:11.916 which has eight arms, and each arm is a jar 0:09:12.156 --> 0:09:14.796 glass jar of urine closed off by a metal threat. 0:09:14.876 --> 0:09:16.996 Everything is standless still. Everything is this in fact that 0:09:17.036 --> 0:09:17.956 everything is super clean. 0:09:18.236 --> 0:09:22.516 The jars have all been pre selected from a biological repository. 0:09:22.716 --> 0:09:24.956 Some are filled with urine from people known to have 0:09:24.996 --> 0:09:30.236 suffered from prostate cancer. Others are controls. Florin's job is 0:09:30.276 --> 0:09:33.756 to sniff each jar. If she detects what she believes 0:09:33.876 --> 0:09:37.636 is a positive sample, she's supposed to stop and stare 0:09:37.676 --> 0:09:41.236 at the jar before moving on. Were the people from 0:09:41.316 --> 0:09:43.916 the prostic Foundation skeptical. 0:09:44.076 --> 0:09:46.116 Oh yes, of course, and as rightly they should be. 0:09:46.596 --> 0:09:49.316 You must see this to believe it. Nobody should take 0:09:49.356 --> 0:09:52.676 these kinds of things on just somebody's, especially when we're 0:09:52.676 --> 0:09:54.476 talking about people's lives here, right. 0:09:54.796 --> 0:09:57.796 It was early in the morning. Merchen looked at Florin 0:09:58.196 --> 0:10:00.796 and worried she wasn't up to the task, so they. 0:10:00.676 --> 0:10:02.916 Had brought flooring the dog on first class. Then the 0:10:02.956 --> 0:10:06.716 American customs had to by lod dewarm her, give her medicine, 0:10:06.716 --> 0:10:08.876 so she had to be vomiting. And also some other 0:10:09.276 --> 0:10:12.796 unpleasantless happened to the poor dog. And there's coffee smells everywhere, 0:10:12.836 --> 0:10:14.876 by the way, in the myriad people who are serving breakfast. 0:10:14.916 --> 0:10:16.316 I'm like, oh my god, we're going to confuse this 0:10:16.316 --> 0:10:18.556 dog now with all the coffee. And she's, you know, 0:10:18.636 --> 0:10:21.356 jet lagged and dewormed, and this poor dog is not 0:10:21.436 --> 0:10:23.436 going to perform, and she kicks it out of the park. 0:10:24.276 --> 0:10:27.596 The people from the Prostate Cancer Foundation are so blown 0:10:27.596 --> 0:10:30.796 away they end up funding guests and merchants research project 0:10:31.316 --> 0:10:33.956 to explore the idea that smell might be the best 0:10:33.956 --> 0:10:36.116 way to diagnose a dangerous cancer. 0:10:36.036 --> 0:10:37.076 To a layman like me. 0:10:37.196 --> 0:10:42.316 The idea that a cancer or a virus would have 0:10:42.356 --> 0:10:44.476 a smell is not intuitive. 0:10:44.556 --> 0:10:47.196 It's not the disease itself that has the smell. What 0:10:47.236 --> 0:10:49.996 happens is it changes your body's metabolic pathways, of which 0:10:49.996 --> 0:10:52.436 you have thousands going on in every cell. There's different 0:10:52.436 --> 0:10:55.836 things happening. They change, and many of these pathways, many 0:10:55.876 --> 0:10:59.356 of these processes inside your cells, they have as a 0:10:59.356 --> 0:11:02.916 byproduct odorance. Odorance are volletilor gant and compounds that tend 0:11:02.996 --> 0:11:05.876 to fly around and come out of your body. So cancer, 0:11:05.916 --> 0:11:10.236 for instance, likely leaves an imprint on every emission that 0:11:10.236 --> 0:11:13.436 you have, from sweat to urine, to saliva to tears, 0:11:13.436 --> 0:11:14.076 so you name it. 0:11:14.916 --> 0:11:19.636 Human beings are a site species. Sight is our superpower. 0:11:20.036 --> 0:11:22.956 The biggest web of sensory connections in our brains is 0:11:22.996 --> 0:11:27.436 between our cortex and our eyes. We can differentiate colors 0:11:27.436 --> 0:11:30.276 in a way that few other species can. But as 0:11:30.316 --> 0:11:33.876 our eyesight evolved to get better, our noses got worse. 0:11:34.676 --> 0:11:37.636 As the biologist John Bradshaw puts it, there appear to 0:11:37.636 --> 0:11:40.836 be limits on how much information any brain can process. 0:11:41.516 --> 0:11:45.236 So if you optimize for seeing, you compromise on smelling. 0:11:46.276 --> 0:11:50.516 Dogs make the opposite trade off. Their eyes are mediocre, 0:11:50.796 --> 0:11:56.476 their noses are amazing. To Florin, prostate cancer wasn't a 0:11:56.516 --> 0:11:59.476 clump of cells. It was an odor. An odor is 0:11:59.516 --> 0:12:03.956 so distinctive that it made her stop and stare. Let's 0:12:04.036 --> 0:12:08.716 imagine that you and I both have crostake cancer at 0:12:08.716 --> 0:12:09.716 the same time. Stage. 0:12:10.276 --> 0:12:11.556 Is it fair to assume that we. 0:12:12.996 --> 0:12:16.596 The way in which our bodies smell is our odor 0:12:16.636 --> 0:12:19.036 is altered by that cancer? Correct might not be the 0:12:19.076 --> 0:12:23.516 same exactly. So why can the dog generalize for. 0:12:23.436 --> 0:12:25.716 The same reason that you can generalize the gait You 0:12:25.756 --> 0:12:29.316 can track my footprints if I teach you what footprints 0:12:29.316 --> 0:12:32.716 look like on the website and how they're distributed. So 0:12:32.756 --> 0:12:34.836 the dog learns to identify the footprints on the web 0:12:34.836 --> 0:12:36.996 site where it's easy to see and easy to see 0:12:36.756 --> 0:12:40.836 the gate, and then it generalizes, It extrapolates and says, oh, 0:12:40.956 --> 0:12:43.156 this smells like the stuff that they used to want 0:12:43.196 --> 0:12:43.676 me to find. 0:12:44.596 --> 0:12:47.956 Human beings try to screen for prostate cancer by looking 0:12:48.036 --> 0:12:50.596 for it. Of course we do. We're a species that's 0:12:50.636 --> 0:12:53.636 in love with sight. We ask men over a certain 0:12:53.676 --> 0:12:56.196 age to take a blood test to look for something 0:12:56.236 --> 0:13:00.156 called the prostate specific antigen. And if we see a 0:13:00.156 --> 0:13:02.316 lot of that antigen. We take a closer look at 0:13:02.316 --> 0:13:05.516 the prostate itself. We take half a dozen or more 0:13:05.556 --> 0:13:08.116 little slices of the prostate and look at them under 0:13:08.116 --> 0:13:12.236 the microscope. We look, and we look again. The problem 0:13:12.436 --> 0:13:14.636 is that only around a quarter of men who have 0:13:14.716 --> 0:13:19.556 elevated PSA levels actually end up having prostate cancer. You 0:13:19.596 --> 0:13:22.676 could have high PSA levels for some other reason. You 0:13:22.716 --> 0:13:26.236 could also have normal PSA levels and still have prostate cancer. 0:13:26.396 --> 0:13:28.636 And so the result of all that looking is an 0:13:28.756 --> 0:13:31.796 enormous amount of error. Lots of men are told they 0:13:31.796 --> 0:13:34.716 could have a dangerous cancer when they don't, and lots 0:13:34.716 --> 0:13:38.116 of other men have cancer that gets missed. The prostate 0:13:38.316 --> 0:13:41.916 screening air rate is so bad that many men avoid 0:13:41.916 --> 0:13:45.076 it altogether. I'm not getting tested, Are you kidding me? 0:13:45.396 --> 0:13:47.116 All you have to do is spend a few hours 0:13:47.116 --> 0:13:50.156 reading the available literature, and you realize that looking for 0:13:50.236 --> 0:13:54.196 early signs of prostate cancer is a fool's game. It's 0:13:54.276 --> 0:14:00.356 really really hard. But Florin the dog didn't look. Florin 0:14:00.476 --> 0:14:04.956 bypassed all that nonsense. Florina just sniffed and did Florin 0:14:05.036 --> 0:14:06.076 make any mistake? Oh? 0:14:06.156 --> 0:14:08.996 Sure, of course, if something is one hundred percent right, 0:14:09.116 --> 0:14:11.956 it's one hundred percent wrong. So yes, there's definitely errors 0:14:11.996 --> 0:14:14.716 in the samples, and the dog can definitely make errors. However, 0:14:14.876 --> 0:14:18.396 you should remember this, so do all of our tests. 0:14:19.196 --> 0:14:21.756 And currently the dogs are better than any test. 0:14:22.356 --> 0:14:25.916 Notice how he said dogs are better. It's not like 0:14:26.036 --> 0:14:28.916 Flowing is some kind of super dog the Usain Bolt 0:14:28.916 --> 0:14:32.556 of the canine world. Lots of dogs could do just 0:14:32.636 --> 0:14:33.236 what she did. 0:14:34.756 --> 0:14:37.276 A dog can do better than the tests you trust, 0:14:37.836 --> 0:14:39.356 and you don't trust the dog, And I'm glad you 0:14:39.356 --> 0:14:41.716 don't trust the dog. We have to not trust the dog. 0:14:41.796 --> 0:14:45.436 The dog is teaching us Now. Dogs are definitely error prone, 0:14:45.476 --> 0:14:48.476 same as everything else. But currently, and this is a 0:14:48.516 --> 0:14:51.836 mind blowing statistic, of all the diseases that have ever 0:14:51.876 --> 0:14:54.316 been tried to be identified by trained dog, all of 0:14:54.356 --> 0:14:57.916 them have succeeded. That is mind blowing. It should be 0:14:58.236 --> 0:15:00.076 we should be paying attention to this. What the hell 0:15:00.156 --> 0:15:02.636 is happening where we can't make this thing not work? 0:15:03.236 --> 0:15:06.396 I mean, shouldn't you be angry at this? I mean, 0:15:06.436 --> 0:15:07.476 come on, what are we doing? 0:15:09.116 --> 0:15:09.196 Well? 0:15:09.556 --> 0:15:28.636 Exactly what are we doing? The chief research scientist for 0:15:28.716 --> 0:15:32.276 three sixty K nine is a man named Bill Schneider, doctor, 0:15:32.516 --> 0:15:36.116 William Schneider. Schneider used to work at Fort Dietrich, the 0:15:36.236 --> 0:15:39.076 army garrison in Maryland where the Pentagon does its top 0:15:39.156 --> 0:15:45.116 secret biodefense work. Schneider's specialty was plant virology, diseases that 0:15:45.156 --> 0:15:47.476 affect things like peach or cherry trees. 0:15:47.676 --> 0:15:52.436 I come from a plant background, but being at Fort Dietrich, 0:15:52.516 --> 0:15:55.396 I had access to a lot of tools and toys 0:15:55.436 --> 0:15:59.356 that typical scientists didn't have in their repertoire, and I'd 0:15:59.396 --> 0:16:00.476 been playing around with those. 0:16:00.916 --> 0:16:05.356 But then, just like Andreas Mershin over at MIT, Schneider 0:16:05.436 --> 0:16:09.076 began to hear of the supposedly miraculous powers of dogs, 0:16:09.596 --> 0:16:12.676 so he decided to hold a sniff off. Fort Dietrich 0:16:12.796 --> 0:16:15.796 has a special containment facility where they could do an 0:16:15.836 --> 0:16:20.116 actual head to head challenge comparing detection results between a 0:16:20.156 --> 0:16:25.396 dog and one of modern science's highest tech tests. Schneider 0:16:25.476 --> 0:16:28.916 chose as the test case a plant virus called plum 0:16:29.036 --> 0:16:32.516 pux and bet against the dogs. He was a lab guy, 0:16:33.036 --> 0:16:33.916 not a dog guy. 0:16:34.716 --> 0:16:37.916 Yeah, you name it. We had access to it. Gas chromatography, 0:16:38.076 --> 0:16:43.956 mass spectrometry, all sorts of very deep sequencing, heavy duty 0:16:44.036 --> 0:16:49.356 nucleic acid analysis. Things that are pretty darn cool actually, 0:16:49.396 --> 0:16:52.636 and not one single one of them can beat the dogs. 0:16:52.996 --> 0:16:57.636 In fact, the US Army invested about nineteen billion dollars 0:16:57.676 --> 0:17:02.356 in trying to find a machine that could detect explosives 0:17:02.356 --> 0:17:06.156 better than dogs, and in the end, all these projects 0:17:06.236 --> 0:17:08.236 nothing could come close to what the dogs can do. 0:17:08.716 --> 0:17:10.756 So I had to convert myself. 0:17:11.196 --> 0:17:15.796 Bill Schneider, the convert, decided to change careers. He left 0:17:15.836 --> 0:17:20.836 government work started making sophisticated canine training aids. I come 0:17:20.876 --> 0:17:23.196 to you, Bill, and I say, Bill, we have a new, 0:17:23.956 --> 0:17:25.556 highly infectious virus out there. 0:17:25.596 --> 0:17:28.236 We know almost nothing about it. What do you do next? 0:17:29.076 --> 0:17:32.516 I say, I need the genomic information. So that's a 0:17:32.636 --> 0:17:36.556 key step in practically any diagnostics these days, and with 0:17:37.116 --> 0:17:41.276 good purpose, because when you have genomic information, you can 0:17:41.316 --> 0:17:44.436 design assays, or in my case, you can design a 0:17:44.516 --> 0:17:48.356 training aid that will teach a dog to directly detect 0:17:48.356 --> 0:17:52.956 that you're making a cent. Yes exactly, You're. 0:17:52.916 --> 0:17:56.676 You're looking at a blueprint and you're saying, ah, ye, gene, 0:17:56.756 --> 0:18:00.436 this protein, this protein, this protein, this protein. I can 0:18:00.956 --> 0:18:02.716 construct that in a lab. 0:18:03.396 --> 0:18:03.996 Yeah. 0:18:04.116 --> 0:18:08.116 Florin the prostate cancer dog was trained on samples taken 0:18:08.276 --> 0:18:12.756 directly from people with prostate cancer. Schneider was taking the 0:18:12.796 --> 0:18:16.076 next step, figuring out a way to extract the scent 0:18:16.236 --> 0:18:22.196 of the disease itself, refining the process creating customized disease 0:18:22.436 --> 0:18:26.516 fragrances that could be shipped overnight to dogs anywhere in 0:18:26.556 --> 0:18:30.316 the world. I'm assuming you're avoiding everything that has to 0:18:30.356 --> 0:18:33.596 do with actual virulence, contagiousness. 0:18:33.956 --> 0:18:36.156 It's an inert it's a harmless substance. 0:18:36.316 --> 0:18:39.036 Yeah, you could eat it if you wanted to. I 0:18:39.036 --> 0:18:41.916 wouldn't taste all that great, but yeah, it's completely safe, 0:18:42.236 --> 0:18:44.996 it's completely stable. You can make as much of it 0:18:45.036 --> 0:18:45.556 as you want. 0:18:45.756 --> 0:18:49.996 Yeah, now we have a little vial of something that 0:18:50.116 --> 0:18:57.436 has a characteristic odorit You and I can't smell that. 0:18:58.276 --> 0:19:01.396 Nope. No, In fact, I would have bet you a 0:19:01.436 --> 0:19:04.316 lot of money that nothing could smell it, because in 0:19:04.396 --> 0:19:07.116 our mind, none of these things had sent. 0:19:09.196 --> 0:19:13.316 The world. Scientists were studying and refining the diagnostic power 0:19:13.316 --> 0:19:16.636 of dogs. Then COVID nineteen. 0:19:16.316 --> 0:19:20.756 Happened, breaking news tonight, the coronavirus forcing millions more Americans 0:19:20.796 --> 0:19:25.516 into virtual lockdown. Over seventy five million people in New York, California, Illinois, 0:19:25.516 --> 0:19:27.716 and Connecticut ordered to stay at home. 0:19:28.196 --> 0:19:31.196 Bill Schneider realized that his technique of making these little 0:19:31.236 --> 0:19:35.276 vials of disease fragrance might prove really useful. 0:19:36.396 --> 0:19:38.356 There's really reason to believe that COVID might be the 0:19:38.396 --> 0:19:39.796 exception to this pattern. 0:19:40.956 --> 0:19:44.956 Not in my mind, no, No. Once I crossed that 0:19:45.076 --> 0:19:47.756 threshold with the original virus that we've worked on, which 0:19:47.796 --> 0:19:51.396 was that plumphaks virus, I knew you could translate this, 0:19:51.796 --> 0:19:55.556 not just to another virus in plants, I can take 0:19:55.716 --> 0:19:59.756 and transfer that to humans or cattle or swine or 0:19:59.916 --> 0:20:03.276 chickens or anything, and off you go. 0:20:04.476 --> 0:20:08.516 COVID first services at the end of twenty nineteen. The 0:20:08.596 --> 0:20:13.116 novel Coronavirus Genome is published. In early January of twenty twenty, 0:20:13.476 --> 0:20:16.476 the Who declares a pandemic that spring. 0:20:16.796 --> 0:20:19.836 And across the country. There's demand for COVID nineteen testing 0:20:20.076 --> 0:20:22.116 as positive cases are on the rise. 0:20:22.556 --> 0:20:25.236 Some people have had to wait longer than two weeks 0:20:25.396 --> 0:20:26.716 to receive results. 0:20:28.476 --> 0:20:30.956 While the world is in lockdown. In the early summer 0:20:30.996 --> 0:20:34.836 of twenty twenty, Bill Schneider's lab gets going synthesizing the 0:20:34.876 --> 0:20:37.516 proteins that make up the smell of SARS COB two. 0:20:38.036 --> 0:20:42.756 The trainers start training dogs on it. Hardworking, focused, unflappable 0:20:42.796 --> 0:20:46.236 dogs who really loved the idea of sniffing nine to five. 0:20:46.596 --> 0:20:49.756 By the time we got to September fifteenth, we were 0:20:49.916 --> 0:20:53.716 starting to collect clinical samples from people who were confirmed positive, 0:20:54.076 --> 0:20:56.476 and we were checking the dogs on those clinical samples, 0:20:56.556 --> 0:20:58.756 and they didn't miss a beat. They were ninety nine 0:20:58.756 --> 0:21:02.876 percent accurate. And at that point I was very confident 0:21:02.956 --> 0:21:07.076 that the dogs we were training could detect COVID in 0:21:07.236 --> 0:21:14.476 people's masks, on their sacks, in sweat and saliva labs. 0:21:14.756 --> 0:21:18.876 Shepherds Belgian Malona was floppy year pointy year. Doesn't much matter. 0:21:19.036 --> 0:21:21.796 What matters is that they're a dog. They have a big, 0:21:21.996 --> 0:21:25.556 powerful nose, and they're perfectly happy to put that magnificent 0:21:25.636 --> 0:21:28.676 nose to work in the service of helping all of us. 0:21:29.356 --> 0:21:40.476 They're best friends. Let's think about some of the options 0:21:40.476 --> 0:21:44.236 we have when detecting a deadly new pathogen as it 0:21:44.316 --> 0:21:48.396 moves through a human population. One option is what we 0:21:48.516 --> 0:21:51.076 ended up trying with COVID in the United States. We 0:21:51.196 --> 0:21:53.996 make you come to a testing site stand in line 0:21:54.036 --> 0:21:56.196 with a lot of people who also think they might 0:21:56.196 --> 0:21:59.076 have COVID. Really good idea, by the way, stick a 0:21:59.156 --> 0:22:02.196 nasty swab up your nose, charge you a couple hundred dollars, 0:22:02.556 --> 0:22:06.516 use an insanely expensive and complicated and high tech system 0:22:06.716 --> 0:22:09.996 called PCR to give you an answer that may or 0:22:09.996 --> 0:22:12.356 may not be useful because it didn't always get to 0:22:12.356 --> 0:22:15.316 you in time. Remember, sometimes it would take forever to 0:22:15.316 --> 0:22:18.396 get your test results, sometimes two weeks, and maybe in 0:22:18.396 --> 0:22:21.276 those two weeks while you were waiting, you were infecting 0:22:21.316 --> 0:22:24.556 everyone you med. The result of all the cost and 0:22:24.596 --> 0:22:29.036 inconvenience and hassle and imperfection of Option one was that 0:22:29.156 --> 0:22:32.716 we've never done enough testing at any stage of the pandemic. 0:22:32.996 --> 0:22:36.396 And because we didn't do enough testing, the pandemic sword 0:22:36.476 --> 0:22:40.316 out of control. Someone has a wedding or a Thanksgiving dinner, 0:22:40.396 --> 0:22:43.436 and in a perfect world everyone would get tested before coming, 0:22:43.756 --> 0:22:46.356 but of course they don't, and one person ends up 0:22:46.396 --> 0:22:49.356 infecting ten other people, and maybe one or two of 0:22:49.396 --> 0:22:54.236 those ten will die. All these horror stories have at 0:22:54.316 --> 0:23:03.236 their core a failure of disease detection. I emailed Michael 0:23:03.276 --> 0:23:06.076 Minna at the Harvard School of Public Health, who was 0:23:06.116 --> 0:23:07.996 one of the big critics of the way we tested 0:23:07.996 --> 0:23:10.876 for COVID, and this is what he wrote back about 0:23:10.916 --> 0:23:13.596 the way testing works. It has to do with what 0:23:13.676 --> 0:23:17.596 epidemiologists called are not, which is the average number of 0:23:17.676 --> 0:23:22.236 people infected by an infected person. If an epidemic has 0:23:22.316 --> 0:23:26.716 are not above one, it grows exponentially. If are not 0:23:26.956 --> 0:23:32.156 falls below one, an epidemic dies out. Here's what Minna wrote. 0:23:32.596 --> 0:23:36.316 This whole pandemic and all of the massive outbreaks we've 0:23:36.316 --> 0:23:39.036 seen have been with an or not of about one 0:23:39.036 --> 0:23:43.996 point three. That means that every ten people infected went 0:23:44.076 --> 0:23:48.636 on to infect on average thirteen people. Well, in that case, 0:23:48.916 --> 0:23:52.276 if you have one hundred people infected on day one, 0:23:52.396 --> 0:23:55.716 then thirty days later you have about six hundred new 0:23:55.756 --> 0:24:01.916 infections exponential growth end quote. All we had to do 0:24:02.476 --> 0:24:06.036 to stop the pandemic was to test just enough people 0:24:06.236 --> 0:24:10.596 and prevent just enough new infections so that those ten 0:24:10.916 --> 0:24:15.676 newly infected people only infected on average nine other people. 0:24:16.636 --> 0:24:19.636