1 00:00:02,720 --> 00:00:05,320 Speaker 1: A few years ago, around the time of Twitter's I 2 00:00:05,480 --> 00:00:08,640 Speaker 1: p O, I noticed what I thought was an odd coincidence. 3 00:00:09,320 --> 00:00:13,720 Speaker 1: It's chief engineer had studied birds. Specifically, he'd studied the 4 00:00:13,760 --> 00:00:17,360 Speaker 1: auditory cortex of zebra finches. I thought that was pretty 5 00:00:17,360 --> 00:00:21,599 Speaker 1: funny for an engineer, especially because Twitter's mascot is that 6 00:00:21,680 --> 00:00:24,480 Speaker 1: little blue bird. I forgot about it until a couple 7 00:00:24,480 --> 00:00:27,360 Speaker 1: of years ago when I noticed another bird brain scholar 8 00:00:27,400 --> 00:00:30,680 Speaker 1: in the top echelons of tech. This person had been 9 00:00:30,800 --> 00:00:34,600 Speaker 1: hired by Elon Musk, the entrepreneur behind Tesla and SpaceX, 10 00:00:34,960 --> 00:00:38,880 Speaker 1: to join his new company, Neuralink. Neuralink is a very 11 00:00:38,920 --> 00:00:43,280 Speaker 1: secretive futuristic company which is trying to supercharge the human brain. 12 00:00:43,880 --> 00:00:47,440 Speaker 1: This all sounds pretty obscure. You'd think studying bird brains 13 00:00:47,479 --> 00:00:50,800 Speaker 1: wouldn't be too relevant to studying human brains or figuring 14 00:00:50,800 --> 00:00:54,200 Speaker 1: out social media, right, That's why I remembered it. It 15 00:00:54,280 --> 00:00:57,279 Speaker 1: just felt so random, But now here were two. So 16 00:00:57,360 --> 00:00:59,280 Speaker 1: one day, when I was a little bored, I just 17 00:00:59,360 --> 00:01:02,320 Speaker 1: tried type being zebra finch and the names of several 18 00:01:02,400 --> 00:01:05,120 Speaker 1: big tech companies into Google. And what did you find? 19 00:01:05,680 --> 00:01:08,160 Speaker 1: I found quite a few employees who knew a lot 20 00:01:08,240 --> 00:01:12,200 Speaker 1: about zebra finches. That was that companies like Intelent, Apple 21 00:01:12,319 --> 00:01:15,320 Speaker 1: and Google too. I was surprised. Okay, hang on, Sarah, 22 00:01:15,360 --> 00:01:18,319 Speaker 1: I'm googling this. The zebra finch is the most common 23 00:01:18,400 --> 00:01:21,520 Speaker 1: astrill did finch of central Australia and ranges over most 24 00:01:21,520 --> 00:01:24,440 Speaker 1: of the continent, avoiding only the cool, moist south and 25 00:01:24,520 --> 00:01:28,920 Speaker 1: tropical far North. Zebrafinches are loud and boisterous singers. So 26 00:01:28,959 --> 00:01:31,600 Speaker 1: what's the connection here, Well, that's where I jumped into 27 00:01:31,600 --> 00:01:35,000 Speaker 1: the reporting process. Our goal is to figure out why 28 00:01:35,080 --> 00:01:38,160 Speaker 1: tech companies are hiring bird brain experts, and this little 29 00:01:38,240 --> 00:01:41,679 Speaker 1: quest has taken us to college campuses around the country. 30 00:01:41,840 --> 00:01:45,440 Speaker 1: We visited several university labs. Some smelled better than others. 31 00:01:46,040 --> 00:01:48,360 Speaker 1: We heard a lot of birds on So this is 32 00:01:48,400 --> 00:01:53,160 Speaker 1: our guy, and so they're they're very light, um, but 33 00:01:53,280 --> 00:01:58,560 Speaker 1: super active. How big would you say that thing is? 34 00:01:58,600 --> 00:02:01,840 Speaker 1: It's like, so they're about fifty in Graham's total weight. 35 00:02:01,960 --> 00:02:06,640 Speaker 1: They're mostly feathers. That's Tim Achi, a neuroscientist and assistant 36 00:02:06,640 --> 00:02:10,280 Speaker 1: professor at Boston University, showing us czebra finch in his 37 00:02:10,400 --> 00:02:13,880 Speaker 1: bird lab and telling us about his research on bird brains. 38 00:02:14,360 --> 00:02:18,079 Speaker 1: He's running some pretty extraordinary experiments. Oh no, but the 39 00:02:18,480 --> 00:02:22,280 Speaker 1: the implant. So this is actually just a lens. And 40 00:02:22,320 --> 00:02:24,079 Speaker 1: so if you look at the top, so you see 41 00:02:24,080 --> 00:02:26,079 Speaker 1: that little black thing that's sticking up, So that's a lens. 42 00:02:26,760 --> 00:02:28,480 Speaker 1: If you look down and you couldn't see the brain, 43 00:02:28,480 --> 00:02:30,360 Speaker 1: but there's not probably not enough light getting down there 44 00:02:30,400 --> 00:02:33,519 Speaker 1: that you could actually see it. And it turns out 45 00:02:33,560 --> 00:02:36,120 Speaker 1: what Tim's learning about the brains of these tiny birds 46 00:02:36,200 --> 00:02:39,120 Speaker 1: is of great interest to the world's largest tech companies. 47 00:02:39,720 --> 00:02:43,800 Speaker 1: I'm brad Stone and Sarah mcbrad that, I'm Ashley Vance, 48 00:02:44,120 --> 00:02:52,320 Speaker 1: and you're listening to decrypt it. So, guys, tell me 49 00:02:52,360 --> 00:02:54,840 Speaker 1: a little bit more about Tim Acchi and his I'm 50 00:02:54,880 --> 00:02:58,800 Speaker 1: sure wonderfully smelling lab out in Boston. Yeah, well, Tim's 51 00:02:58,880 --> 00:03:00,919 Speaker 1: lab was one of the key st ups on our trip. 52 00:03:01,280 --> 00:03:04,680 Speaker 1: Tim took over a Boston University lab from Tim Gardner, 53 00:03:04,880 --> 00:03:07,560 Speaker 1: the guy who left to work for Elon Musk and Neuralink. 54 00:03:07,919 --> 00:03:10,560 Speaker 1: Tim Acchi is a forty year old guy with black 55 00:03:10,600 --> 00:03:14,280 Speaker 1: frame glasses who throws in plenty of references to Portlandia 56 00:03:14,440 --> 00:03:17,360 Speaker 1: and other TV shows when he check chats on his 57 00:03:17,440 --> 00:03:21,400 Speaker 1: office wall. He's got a print of voltage chases called 58 00:03:21,639 --> 00:03:24,959 Speaker 1: five Seconds of Donkey Kong. Now we're talking about something 59 00:03:25,000 --> 00:03:27,600 Speaker 1: I'm familiar with, So what's the connection here. It turns 60 00:03:27,600 --> 00:03:30,520 Speaker 1: out that that old eighties video game became the subject 61 00:03:30,639 --> 00:03:34,400 Speaker 1: of a famous neuroscience research paper. There's one other thing 62 00:03:34,440 --> 00:03:37,200 Speaker 1: you should know about Tim. He has a giant zebra 63 00:03:37,280 --> 00:03:41,440 Speaker 1: finch tattoo on his right forearm. Brad, how familiar are 64 00:03:41,480 --> 00:03:43,839 Speaker 1: you with zebra finch? Well, actually, other than my very 65 00:03:43,840 --> 00:03:47,000 Speaker 1: brief Wikipedia search just now, other than small birds that 66 00:03:47,120 --> 00:03:49,440 Speaker 1: sing a lot, I would say not very familiar. They 67 00:03:49,480 --> 00:03:53,080 Speaker 1: are very cute little birds. Indeed, they're about four inches long. 68 00:03:53,280 --> 00:03:56,080 Speaker 1: The males have orange cheeks and black and white striped 69 00:03:56,080 --> 00:03:59,320 Speaker 1: feathers across their chests, hence the name. They're super easy 70 00:03:59,360 --> 00:04:02,760 Speaker 1: to breed, and they chirp a lot. Here's Tim's interpretation. 71 00:04:04,760 --> 00:04:07,360 Speaker 1: So how did Tim get into studying bird brains? He 72 00:04:07,440 --> 00:04:10,880 Speaker 1: took a circuitous route into neuroscience. One thing I found 73 00:04:10,920 --> 00:04:13,720 Speaker 1: interesting is he started as an engineer working at a 74 00:04:13,760 --> 00:04:17,440 Speaker 1: company that helped factories to automate. His job was teaching 75 00:04:17,520 --> 00:04:20,640 Speaker 1: robots how to sort stuff, everything from car parts to 76 00:04:20,680 --> 00:04:23,680 Speaker 1: gives moos for circuit boards. It was just astounding to 77 00:04:23,680 --> 00:04:25,599 Speaker 1: me how difficult it was to get things to do this, 78 00:04:25,680 --> 00:04:29,640 Speaker 1: and these were tasks that you know, children do. It 79 00:04:29,680 --> 00:04:33,359 Speaker 1: really put in my mind the idea that, you know, 80 00:04:33,480 --> 00:04:37,000 Speaker 1: a lot of the things that that children can do 81 00:04:37,200 --> 00:04:43,080 Speaker 1: almost effortlessly and without almost any training, are incredibly impossible 82 00:04:43,400 --> 00:04:46,720 Speaker 1: to get artificial systems to do or take an enormous 83 00:04:46,720 --> 00:04:49,720 Speaker 1: amount of thought. Okay, so I think I understand Tim 84 00:04:49,800 --> 00:04:52,599 Speaker 1: was curious about why a certain task is so easy 85 00:04:52,680 --> 00:04:55,400 Speaker 1: for a child but so difficult for a robot. Yeah, 86 00:04:55,480 --> 00:04:58,640 Speaker 1: that's exactly it. So after a detour to Tim ended 87 00:04:58,720 --> 00:05:02,040 Speaker 1: up studying neuroscience at Harvard, and that's where he discovered 88 00:05:02,120 --> 00:05:05,200 Speaker 1: zebra finches. Now he's teaching it be you and doing 89 00:05:05,200 --> 00:05:08,440 Speaker 1: his own research. One thing Tim focuses on is how 90 00:05:08,560 --> 00:05:11,680 Speaker 1: zebra finches learned to sing. Tim described his study of 91 00:05:11,760 --> 00:05:14,280 Speaker 1: zebra finches is more of a means to an end. 92 00:05:14,880 --> 00:05:17,560 Speaker 1: I don't, you know, really think of myself or or 93 00:05:17,640 --> 00:05:24,080 Speaker 1: care too much about, you know, songbird neuroscience specifically, I 94 00:05:24,200 --> 00:05:27,080 Speaker 1: take it as a as a way to investigate general 95 00:05:27,080 --> 00:05:31,640 Speaker 1: principles and mechanisms in neuroscience and how brains function generally. 96 00:05:31,880 --> 00:05:34,280 Speaker 1: So Tim is saying that he studies these teeny tiny 97 00:05:34,360 --> 00:05:36,600 Speaker 1: zebra finch brains because he thinks it will give us 98 00:05:36,640 --> 00:05:40,040 Speaker 1: insights into the way human brains function. That's right. Researchers 99 00:05:40,040 --> 00:05:42,480 Speaker 1: study all different kinds of animals for all different kinds 100 00:05:42,480 --> 00:05:44,640 Speaker 1: of purposes, but in this case they're trying to learn 101 00:05:44,640 --> 00:05:47,599 Speaker 1: more about the human brain. Seeing how birds learn to sing, 102 00:05:47,680 --> 00:05:51,280 Speaker 1: for example, can provide insights into how we learn things. 103 00:05:51,360 --> 00:05:55,080 Speaker 1: So I think that you know, the songbird is one 104 00:05:55,120 --> 00:05:59,480 Speaker 1: of those systems that we probably understand the best in 105 00:05:59,560 --> 00:06:02,840 Speaker 1: terms of different brain regions that are involved, in terms 106 00:06:02,839 --> 00:06:05,600 Speaker 1: of the roles of those different brain regions, and so 107 00:06:05,720 --> 00:06:08,799 Speaker 1: I think that we can ask very very precise questions 108 00:06:09,000 --> 00:06:14,840 Speaker 1: in the songbird about the interaction between brain activity and behavior. Sarah, 109 00:06:14,880 --> 00:06:17,080 Speaker 1: you started out by saying that studying the brains of 110 00:06:17,160 --> 00:06:21,480 Speaker 1: zebra finches are somehow interesting for tech companies. So if 111 00:06:21,520 --> 00:06:24,200 Speaker 1: Tim studies the birds because there are connections between bird 112 00:06:24,200 --> 00:06:27,440 Speaker 1: brains and human brains, are there also connections between bird 113 00:06:27,440 --> 00:06:31,600 Speaker 1: brains and computers. Absolutely, But to understand it, I need 114 00:06:31,640 --> 00:06:34,800 Speaker 1: to explain a little bit more about Tim's research. When 115 00:06:34,800 --> 00:06:37,080 Speaker 1: we met him in Boston, before we went into the 116 00:06:37,160 --> 00:06:40,160 Speaker 1: room with the real life zebra finches, he played us 117 00:06:40,160 --> 00:06:43,680 Speaker 1: one of his best songbird clips, and so that's what 118 00:06:43,760 --> 00:06:47,359 Speaker 1: the song sounds like. Okay, so that's one. Yeah, so 119 00:06:47,440 --> 00:06:50,760 Speaker 1: that's what the that's what this this particular zebra finch sings. 120 00:06:50,800 --> 00:06:53,440 Speaker 1: All of them have slightly different songs. That sounded like 121 00:06:53,480 --> 00:06:56,800 Speaker 1: a cartoon. Yeah, it sounded a little like woody woodpecker. Right, 122 00:06:57,080 --> 00:06:59,440 Speaker 1: but that but that is actually, uh, that is actually 123 00:06:59,480 --> 00:07:04,920 Speaker 1: what it sounds. Yeah. Yeah, we had listened to these 124 00:07:04,920 --> 00:07:08,800 Speaker 1: all day long. Tim told us while male and female 125 00:07:08,880 --> 00:07:12,960 Speaker 1: zebra finches can chirp, only male zebra finches sing, even 126 00:07:13,000 --> 00:07:15,320 Speaker 1: if their song is so short it doesn't sound like 127 00:07:15,400 --> 00:07:18,040 Speaker 1: much of a song to us. Tim studies the brains 128 00:07:18,080 --> 00:07:20,760 Speaker 1: of the baby birds as they learn. Here's the baby 129 00:07:20,840 --> 00:07:24,400 Speaker 1: zebra finch trying to imitate his dad's song. He makes 130 00:07:24,440 --> 00:07:27,960 Speaker 1: an early effort, and then a better one a month later. Okay, 131 00:07:28,000 --> 00:07:39,280 Speaker 1: so this is the father this one, well, okay, and 132 00:07:39,320 --> 00:07:41,080 Speaker 1: then we can do the middle one where he's not 133 00:07:41,160 --> 00:07:54,080 Speaker 1: quite good. A right, Okay, this is the one. Yeah 134 00:07:55,000 --> 00:07:58,480 Speaker 1: that's pretty good. So Ashley, I'm kind of praying that 135 00:07:58,520 --> 00:08:03,640 Speaker 1: to these experiments, he's not hurting these beautiful, teeny tiny birds. Basically, 136 00:08:03,800 --> 00:08:06,040 Speaker 1: what he does is they take these birds that have 137 00:08:06,080 --> 00:08:09,520 Speaker 1: been injected with a benign virus. The virus makes their 138 00:08:09,520 --> 00:08:13,120 Speaker 1: brains produce a type of protein that causes individual neurons 139 00:08:13,160 --> 00:08:16,040 Speaker 1: to light up when they fire, they glow green and red. 140 00:08:16,760 --> 00:08:19,000 Speaker 1: To see them in action, Tim and the grad students 141 00:08:19,000 --> 00:08:22,120 Speaker 1: in his lab performed very delicate surgery on the zebra 142 00:08:22,200 --> 00:08:26,320 Speaker 1: finches to implant tiny, tiny microscopes in their brain. Tim 143 00:08:26,360 --> 00:08:28,640 Speaker 1: showed us one bird who had gone through this procedure, 144 00:08:29,600 --> 00:08:32,800 Speaker 1: male zebra finish. And so haven't we on his hiding 145 00:08:34,760 --> 00:08:37,800 Speaker 1: He's got one of these windows that I was telling 146 00:08:37,840 --> 00:08:43,480 Speaker 1: you about. And so this guy has already been through 147 00:08:43,480 --> 00:08:46,360 Speaker 1: the procedure in which we inject the viruses into the 148 00:08:46,400 --> 00:08:50,319 Speaker 1: brain and then we've put a window or a lens 149 00:08:50,400 --> 00:08:52,880 Speaker 1: on top so that we can actually image through it. 150 00:08:53,080 --> 00:08:55,720 Speaker 1: And it's going to get a microscope. He will eventually 151 00:08:55,760 --> 00:09:00,840 Speaker 1: get a microscope. Auch Sarah, that sounds completely crazy. Give 152 00:09:00,880 --> 00:09:03,600 Speaker 1: us a better sense for what this looks like. So 153 00:09:03,640 --> 00:09:07,199 Speaker 1: there are all these little birds hopping around with a 154 00:09:07,240 --> 00:09:11,240 Speaker 1: tiny bit of their skull missing, and instead they have 155 00:09:11,440 --> 00:09:14,240 Speaker 1: a microscope there and a little hole where you can 156 00:09:14,240 --> 00:09:16,400 Speaker 1: look in and see what's going on in their brain. 157 00:09:17,040 --> 00:09:20,480 Speaker 1: And the microscope sits there for days, weeks, months, at 158 00:09:20,480 --> 00:09:24,400 Speaker 1: a time, looking at their neurons fire in in real time, 159 00:09:24,440 --> 00:09:26,320 Speaker 1: and the birds are in their little cages, and then 160 00:09:26,400 --> 00:09:28,520 Speaker 1: when you go into their lab, it's actually it's pretty cool. 161 00:09:28,640 --> 00:09:32,200 Speaker 1: There's all these wires going off these cages straight into 162 00:09:32,240 --> 00:09:35,199 Speaker 1: like a data center where they store all of this information. 163 00:09:35,240 --> 00:09:37,440 Speaker 1: And so basically you just get to just get to 164 00:09:37,480 --> 00:09:41,360 Speaker 1: watch this bird's brain behave in real time and then 165 00:09:41,400 --> 00:09:43,600 Speaker 1: go back and look through all the information. So, Sarah, 166 00:09:43,640 --> 00:09:46,920 Speaker 1: should we feel sorry for these birds? I didn't. They 167 00:09:46,960 --> 00:09:50,359 Speaker 1: seemed really happy. They were hopping around, they were chirping, 168 00:09:50,600 --> 00:09:53,640 Speaker 1: they were acting normal as far as they could tell 169 00:09:53,720 --> 00:09:56,959 Speaker 1: based on the ones they saw without microscopes in their head. 170 00:09:57,000 --> 00:09:59,280 Speaker 1: They really didn't seem to be too much difference. Okay, 171 00:09:59,280 --> 00:10:01,920 Speaker 1: so how is all this useful for the neuroscientists. They 172 00:10:01,920 --> 00:10:04,440 Speaker 1: look deep inside each bird's brain and they look to 173 00:10:04,480 --> 00:10:07,240 Speaker 1: see which neurons are firing and for how long, when 174 00:10:07,320 --> 00:10:09,760 Speaker 1: say the bird is learning to sing, and they can 175 00:10:09,800 --> 00:10:14,080 Speaker 1: make hypothesis on the relationship between different neurons. So that's 176 00:10:14,120 --> 00:10:17,199 Speaker 1: actually a technique used pretty widely in science now and 177 00:10:17,320 --> 00:10:20,360 Speaker 1: all kinds of animals. While we were visiting Tim in Boston, 178 00:10:20,400 --> 00:10:23,240 Speaker 1: we also stopped by a mouse lab that does something similar, 179 00:10:23,320 --> 00:10:25,360 Speaker 1: and a fish lab. In each of the labs, the 180 00:10:25,400 --> 00:10:28,480 Speaker 1: neuroscientists there are studying other things as well, like how 181 00:10:28,600 --> 00:10:31,960 Speaker 1: animals move, what happens in their brains when they make decisions. 182 00:10:32,679 --> 00:10:35,480 Speaker 1: For example, at the Roland Institute in Cambridge, we saw 183 00:10:35,520 --> 00:10:38,400 Speaker 1: a couple of video games that mice play using tiny 184 00:10:38,480 --> 00:10:41,480 Speaker 1: joysticks size for a mouse paw Wow. So how do 185 00:10:41,520 --> 00:10:43,559 Speaker 1: they how do they mice get the quarters in the 186 00:10:43,679 --> 00:10:48,400 Speaker 1: video game machine. They're highly trained and and the video 187 00:10:48,440 --> 00:10:50,880 Speaker 1: games tells us what it tells us how they're making 188 00:10:50,880 --> 00:10:54,600 Speaker 1: decisions exactly. So the mice are a lot like the birds. 189 00:10:54,640 --> 00:10:57,559 Speaker 1: They have bits of their their skull have been removed, 190 00:10:57,600 --> 00:11:01,000 Speaker 1: and we're watching their brains and real time again and 191 00:11:01,160 --> 00:11:03,080 Speaker 1: and so you watch them play these video games, and 192 00:11:03,120 --> 00:11:05,920 Speaker 1: you see how they adapt. Sometimes they get different rules, 193 00:11:06,200 --> 00:11:08,760 Speaker 1: and you see how the mouse learns the rules of 194 00:11:08,760 --> 00:11:11,120 Speaker 1: the game, and they move this choice stick around and 195 00:11:11,160 --> 00:11:13,600 Speaker 1: one of them to find the edges of a box. 196 00:11:14,080 --> 00:11:16,720 Speaker 1: And if the mouse is successful, it gets a little 197 00:11:16,760 --> 00:11:19,520 Speaker 1: bit of sugar water. And and all this time the 198 00:11:19,559 --> 00:11:21,760 Speaker 1: scientists are sitting in there seeing which parts of the 199 00:11:21,800 --> 00:11:24,480 Speaker 1: brain light up and how the mouse reacts to these 200 00:11:24,520 --> 00:11:27,280 Speaker 1: different situations. So what is it about this research that 201 00:11:27,400 --> 00:11:31,040 Speaker 1: tech companies find so interesting? One reason is that these 202 00:11:31,040 --> 00:11:33,839 Speaker 1: scientists are working with tons and tons of data, and 203 00:11:33,880 --> 00:11:37,640 Speaker 1: that's something that every tech company needs to do. And 204 00:11:37,800 --> 00:11:42,240 Speaker 1: also it's to do with artificial intelligence, the field that 205 00:11:42,360 --> 00:11:46,959 Speaker 1: has computer systems mastering tasks that require human traits like 206 00:11:47,200 --> 00:11:51,800 Speaker 1: visual perception or decision making. Maybe how to identify a cat. 207 00:11:52,480 --> 00:11:55,800 Speaker 1: That sounds simple, but it's actually way more nuanced than 208 00:11:55,840 --> 00:11:59,520 Speaker 1: a traditional computer task and very hard for a computer 209 00:11:59,640 --> 00:12:02,600 Speaker 1: to masse term. So there's this one school of thought 210 00:12:02,760 --> 00:12:06,679 Speaker 1: that AI moving forward should loosely be modeled after the 211 00:12:06,760 --> 00:12:11,160 Speaker 1: human brain. The AI systems we have today are still 212 00:12:11,320 --> 00:12:15,480 Speaker 1: basically number crunching systems. They're doing tons of statistical calculations. 213 00:12:15,720 --> 00:12:18,000 Speaker 1: And if we want to get to this this future 214 00:12:18,040 --> 00:12:21,120 Speaker 1: that Sarah's talking about where you actually have decision making 215 00:12:21,640 --> 00:12:25,280 Speaker 1: and much more sophisticated thought, the ideas that we could 216 00:12:25,280 --> 00:12:29,360 Speaker 1: borrow from the human brain and and maybe I have 217 00:12:29,520 --> 00:12:32,000 Speaker 1: something that's way more flexible than what a computer could do. 218 00:12:32,160 --> 00:12:35,360 Speaker 1: And so presumably, since we can't cut open human skulls 219 00:12:35,360 --> 00:12:37,920 Speaker 1: and make people play video games against their will, the 220 00:12:38,280 --> 00:12:42,080 Speaker 1: zebra finch brains are first Yeah, exactly, human brains are 221 00:12:42,080 --> 00:12:44,720 Speaker 1: a little too big and complicated to study in the 222 00:12:44,800 --> 00:12:47,560 Speaker 1: kind of detail we can get from animals right now, 223 00:12:48,000 --> 00:12:52,280 Speaker 1: they're smaller, easier to study, and obviously the ethics of 224 00:12:52,400 --> 00:12:56,199 Speaker 1: studying a living human brain very tricky. So I'm so 225 00:12:56,240 --> 00:12:58,680 Speaker 1: curious about why zebra finches have become the bird to 226 00:12:58,720 --> 00:13:01,920 Speaker 1: study and what exactly tech companies are doing with all this. 227 00:13:02,200 --> 00:13:17,680 Speaker 1: Let's get to that after the break, Okay, Sarah n Ashley, 228 00:13:17,800 --> 00:13:21,560 Speaker 1: So before the break, you explain how neuroscience is starting 229 00:13:21,559 --> 00:13:25,400 Speaker 1: to inform the way tech companies design AI systems. So 230 00:13:25,440 --> 00:13:27,480 Speaker 1: I guess this means it's a very good time to 231 00:13:27,520 --> 00:13:30,120 Speaker 1: be a neuroscientist. Yeah. So these people that used to 232 00:13:30,160 --> 00:13:33,160 Speaker 1: be in academia or working at universities for their whole 233 00:13:33,200 --> 00:13:37,360 Speaker 1: career are now finding tons of job opportunities in Silicon Valley. 234 00:13:37,520 --> 00:13:41,480 Speaker 1: Companies like Google and Apple and Amazon are all snatching 235 00:13:41,520 --> 00:13:44,520 Speaker 1: them up, including a lot of these zebra finch experts 236 00:13:44,520 --> 00:13:46,960 Speaker 1: that we've been talking about. It's how the company that 237 00:13:47,080 --> 00:13:50,040 Speaker 1: probably makes the chips in your laptop won't say exactly 238 00:13:50,040 --> 00:13:52,520 Speaker 1: how many people it has working in AI, but it's 239 00:13:52,559 --> 00:13:55,720 Speaker 1: a lot, and many of them have this expertise. Intel 240 00:13:55,760 --> 00:13:58,199 Speaker 1: helps its customers soup up machines to get them to 241 00:13:58,280 --> 00:14:01,800 Speaker 1: behave smarter and in more uman lake ways. That could 242 00:14:01,880 --> 00:14:05,120 Speaker 1: mean maybe working with a self driving carmaker, and that 243 00:14:05,240 --> 00:14:08,800 Speaker 1: carmaker needs its vehicles to make lightning fast decisions on 244 00:14:08,840 --> 00:14:11,800 Speaker 1: the road, and they have to be good decisions. A 245 00:14:11,840 --> 00:14:14,400 Speaker 1: promising way to do that is to build computer systems 246 00:14:14,440 --> 00:14:17,640 Speaker 1: that imitate how the human brain works. In a human brain, 247 00:14:17,880 --> 00:14:21,480 Speaker 1: the systems are called signapses and pathways. In a machine, 248 00:14:21,760 --> 00:14:25,720 Speaker 1: they're called neural networks. We talked to a mirror because Shahi, 249 00:14:26,000 --> 00:14:30,960 Speaker 1: the chief technology officer for Intel's AI products divisions of 250 00:14:31,040 --> 00:14:35,040 Speaker 1: architectures having wildly successful and uh sends around two thousand 251 00:14:35,120 --> 00:14:45,440 Speaker 1: eleven UM, and while wider and excreted vision, m navigation, 252 00:14:45,680 --> 00:14:50,520 Speaker 1: reforcement learning things that are car related to neuroscience becauses 253 00:14:50,520 --> 00:14:54,880 Speaker 1: are also tacit humans doing pretty well. A mirror is 254 00:14:54,880 --> 00:14:58,480 Speaker 1: a computational neuroscientists by training. He got his PhD at 255 00:14:58,560 --> 00:15:02,280 Speaker 1: UC Berkeley, and he ended up hiring another Berkeley PhD grad, 256 00:15:02,520 --> 00:15:06,320 Speaker 1: Tyler Lee, to work it into and helping virtual assistants 257 00:15:06,440 --> 00:15:09,840 Speaker 1: understand human speech. Tyler spends a lot of time thinking 258 00:15:09,880 --> 00:15:14,400 Speaker 1: about how speech works in different environments like cars. Knowing 259 00:15:14,480 --> 00:15:17,960 Speaker 1: context simple things like if the speakers the driver, or 260 00:15:18,000 --> 00:15:20,760 Speaker 1: the passenger makes it a lot easier to understand what 261 00:15:20,800 --> 00:15:23,800 Speaker 1: they're saying. For example, a driver is more likely to 262 00:15:23,840 --> 00:15:26,600 Speaker 1: ask about directions. Some of the work on context he 263 00:15:26,640 --> 00:15:29,000 Speaker 1: does actually reminds him a lot of his PhD work 264 00:15:29,200 --> 00:15:34,360 Speaker 1: studying you guessed it zebra fitches. The bird has to 265 00:15:34,400 --> 00:15:38,360 Speaker 1: recognize what type of call is being being omitted by 266 00:15:38,480 --> 00:15:41,600 Speaker 1: by the one it's hearing, uh, and then then it 267 00:15:41,640 --> 00:15:43,960 Speaker 1: can go and recognize who that bird is. Is that 268 00:15:44,400 --> 00:15:46,600 Speaker 1: is that a family member of mate? You know? Is 269 00:15:46,600 --> 00:15:48,600 Speaker 1: it another zuber fringe or a different type of bird? 270 00:15:48,960 --> 00:15:52,120 Speaker 1: Should I be concerned? The vocal identification is one thing 271 00:15:52,120 --> 00:15:54,800 Speaker 1: where it's it's context specific, and recognizing the context would 272 00:15:54,920 --> 00:15:59,160 Speaker 1: let you better understand the vocal signature. Mostly, Tyler says 273 00:15:59,160 --> 00:16:03,000 Speaker 1: his studies help him big picture stuff. Neuroscience teaches you 274 00:16:03,120 --> 00:16:06,960 Speaker 1: how to think about complex problems of signal processing. Where 275 00:16:07,000 --> 00:16:10,360 Speaker 1: I take some something from the world that comes in. 276 00:16:10,440 --> 00:16:15,040 Speaker 1: It's an image, it's uh, it's sound, and it's it's noisy, 277 00:16:15,320 --> 00:16:18,320 Speaker 1: and it's high very very high dimensional, and I have 278 00:16:18,400 --> 00:16:21,400 Speaker 1: to like break it down into into features that I 279 00:16:21,440 --> 00:16:25,560 Speaker 1: can then use to to like do something with, solve 280 00:16:25,600 --> 00:16:27,680 Speaker 1: a task with. That's what the brain does all the time, 281 00:16:27,720 --> 00:16:30,880 Speaker 1: and that's sort of the abstract level. Tyler's boss a 282 00:16:30,960 --> 00:16:34,160 Speaker 1: Mirror says he's no zebra finch chauvinist. He's got people 283 00:16:34,200 --> 00:16:38,640 Speaker 1: on his team who studied flies, rats, locus, even worms. 284 00:16:38,680 --> 00:16:42,600 Speaker 1: These very simple organisms exhibit really complex behaviors that are 285 00:16:42,600 --> 00:16:46,680 Speaker 1: still a challenge for us to to simulate and silicon 286 00:16:46,800 --> 00:16:49,520 Speaker 1: using our neual networks and machine and learning. So even 287 00:16:49,640 --> 00:16:55,040 Speaker 1: a simple worm inchworm is a really complicated robotic machine 288 00:16:55,080 --> 00:17:00,000 Speaker 1: that's really miraculous. So we look for inspiration from simple 289 00:17:00,280 --> 00:17:03,880 Speaker 1: to humans. So I guess Intel and all these other 290 00:17:03,920 --> 00:17:07,240 Speaker 1: companies must be developing products which incorporate AI developed with 291 00:17:07,280 --> 00:17:10,520 Speaker 1: the help of these zebra finch experts. Besides commands to 292 00:17:10,560 --> 00:17:13,520 Speaker 1: self driving cars, is there any area where knowing a 293 00:17:13,520 --> 00:17:16,520 Speaker 1: lot about sound itself is helpful? And I guess when 294 00:17:16,560 --> 00:17:19,280 Speaker 1: when do all these years of studying zebra finches finally 295 00:17:19,359 --> 00:17:21,880 Speaker 1: pay off? You can think about features on your phone 296 00:17:21,960 --> 00:17:24,639 Speaker 1: or computer that let you unlock the device with your voice, 297 00:17:25,320 --> 00:17:28,000 Speaker 1: or stuff like noise reduction and phone calls and on 298 00:17:28,119 --> 00:17:30,119 Speaker 1: video calls. I can't wait till I can sing a 299 00:17:30,119 --> 00:17:33,680 Speaker 1: little Zebra Finch song and my phone unlocks. But anything 300 00:17:33,680 --> 00:17:37,520 Speaker 1: beyond gadgets, yeah, absolutely. Animal neuroscience connects to a lot 301 00:17:37,600 --> 00:17:41,600 Speaker 1: of fields, especially health. Somewhere could be relevant for Parkinson's 302 00:17:41,640 --> 00:17:45,120 Speaker 1: research because animals help researchers figure out how to stop tremors. 303 00:17:45,600 --> 00:17:48,560 Speaker 1: They also help in areas like how to handle prosthetic limbs. 304 00:17:48,920 --> 00:17:52,040 Speaker 1: Outside of medicine, the work might grow even more futuristic. 305 00:17:52,280 --> 00:17:55,040 Speaker 1: This is where we get into dystopian scenarios. I suspect 306 00:17:55,080 --> 00:17:58,960 Speaker 1: what what else exactly like Neuralink. I mentioned that company 307 00:17:59,000 --> 00:18:01,840 Speaker 1: earlier because Ellen Musk get hired a Zebra Finch scholar, 308 00:18:01,920 --> 00:18:04,080 Speaker 1: and it's one of the companies that we believe is 309 00:18:04,119 --> 00:18:08,320 Speaker 1: working on very futuristic technology. Ellen Musk keeps dropping hints 310 00:18:08,400 --> 00:18:10,560 Speaker 1: on Twitter that the company is about to announce a 311 00:18:10,600 --> 00:18:14,320 Speaker 1: big breakthrough. Nobody knows exactly what, but it's going to 312 00:18:14,359 --> 00:18:17,200 Speaker 1: have something to do with brain machine interfaces. So I'm 313 00:18:17,240 --> 00:18:19,760 Speaker 1: trying hard not to think about some Star Trek episodes 314 00:18:19,760 --> 00:18:22,680 Speaker 1: on this topic, which all ended quite badly. But help 315 00:18:22,720 --> 00:18:26,320 Speaker 1: me understand actually what a brain machine interfaces about. I mean, 316 00:18:26,480 --> 00:18:28,720 Speaker 1: at its most basic level, is this idea that you 317 00:18:28,800 --> 00:18:33,040 Speaker 1: have a two way interplay between humans and computers where 318 00:18:33,040 --> 00:18:36,639 Speaker 1: you could actually funnel information back and forth. We already 319 00:18:36,640 --> 00:18:38,960 Speaker 1: have examples of stuff like this with the implants that 320 00:18:39,000 --> 00:18:43,160 Speaker 1: help people here or stop Parkinson's tremors in this case, 321 00:18:43,359 --> 00:18:46,760 Speaker 1: I think people are looking at much more futuristic applications 322 00:18:46,800 --> 00:18:50,120 Speaker 1: where you might even have like a mesh that's attached 323 00:18:50,160 --> 00:18:53,440 Speaker 1: to your brain and you could full on download your 324 00:18:53,440 --> 00:18:57,040 Speaker 1: brain to a machine or learn Japanese in five seconds. 325 00:18:57,720 --> 00:19:00,639 Speaker 1: There's another company called Colonel that's in this same field, 326 00:19:01,000 --> 00:19:05,080 Speaker 1: and like Neuralalink, it's also very mysterious, and in some 327 00:19:05,119 --> 00:19:07,640 Speaker 1: ways that's kind of the best part. When we don't 328 00:19:07,760 --> 00:19:10,920 Speaker 1: know exactly what they're doing, we can imagine all kinds 329 00:19:10,920 --> 00:19:15,080 Speaker 1: of crazy stuff. Going back to Elon Musk, he's been 330 00:19:15,080 --> 00:19:19,040 Speaker 1: talking about where neurally could go, maybe allowing people to 331 00:19:19,119 --> 00:19:22,920 Speaker 1: have this kind of superhuman cognition where you could you 332 00:19:22,960 --> 00:19:25,280 Speaker 1: could think on par with a machine, are certainly much 333 00:19:25,320 --> 00:19:28,919 Speaker 1: better than we do today. That means basically stuff like 334 00:19:28,960 --> 00:19:32,439 Speaker 1: you could download an entire FOURGN language directly into your brain, 335 00:19:32,840 --> 00:19:37,399 Speaker 1: or maybe instantly grabbing encyclopedia. People like Timachi have been 336 00:19:37,400 --> 00:19:40,840 Speaker 1: thinking about exactly these scenarios for years and can really 337 00:19:40,880 --> 00:19:44,960 Speaker 1: nerd out on the possibilities. Trivia would be over, Jeopardy 338 00:19:45,000 --> 00:19:48,160 Speaker 1: would not be a thing anymore. Um, Alex Trebek would 339 00:19:48,160 --> 00:19:50,280 Speaker 1: be out of a job. He has some more serious 340 00:19:50,280 --> 00:19:53,439 Speaker 1: thoughts on the topic too. I find the idea that 341 00:19:53,480 --> 00:19:57,359 Speaker 1: we could, you know, pretend one day in the maybe 342 00:19:57,400 --> 00:20:01,000 Speaker 1: just in future, you know, really right information directly into 343 00:20:01,000 --> 00:20:05,320 Speaker 1: the brain, that we could actually have a high bandwidth way, 344 00:20:05,800 --> 00:20:08,359 Speaker 1: uh to get really sci fi about it, kind of 345 00:20:08,400 --> 00:20:11,880 Speaker 1: a matrix E like, Um, I think that would be amazing. 346 00:20:11,960 --> 00:20:15,320 Speaker 1: We are nowhere near knowing anything about how to get there. 347 00:20:15,840 --> 00:20:19,119 Speaker 1: We can barely even scratch the surface of what that 348 00:20:19,119 --> 00:20:22,480 Speaker 1: would be like. But you know, in terms of fantasy, 349 00:20:24,119 --> 00:20:26,200 Speaker 1: what would I like to do one day? I would 350 00:20:26,240 --> 00:20:29,080 Speaker 1: love to be able to contribute even a small way 351 00:20:29,200 --> 00:20:32,800 Speaker 1: to figuring out how we can have this sort of 352 00:20:33,400 --> 00:20:37,240 Speaker 1: bi directional interface with the brain. Oh my god. So yeah, 353 00:20:37,280 --> 00:20:39,760 Speaker 1: this is While the evocation of the matrix does not 354 00:20:39,880 --> 00:20:42,199 Speaker 1: make me feel more comfortable about this, what do the 355 00:20:42,200 --> 00:20:45,240 Speaker 1: skeptics say about about it all? Well, there are plenty 356 00:20:45,280 --> 00:20:48,880 Speaker 1: of skeptics out there. I checked in with one scholar 357 00:20:48,960 --> 00:20:53,040 Speaker 1: at the University of Chicago, Dan Margolias. The idea that 358 00:20:53,080 --> 00:20:59,440 Speaker 1: we're going to reverse engineers, not reverse engineer, uh, forward 359 00:20:59,520 --> 00:21:03,800 Speaker 1: engineer or the human brain, so we can download tons 360 00:21:03,840 --> 00:21:07,479 Speaker 1: of material into it very rapidly. And I don't know 361 00:21:07,520 --> 00:21:12,400 Speaker 1: what pick up a language overnight or something. I it's 362 00:21:12,440 --> 00:21:15,879 Speaker 1: it's the way people make progresses to dream and so 363 00:21:16,000 --> 00:21:21,919 Speaker 1: I'm I'm a scientist. I'm for that. But that really 364 00:21:21,960 --> 00:21:30,159 Speaker 1: sounds more fantastical than realistic. It ignores it ignores the 365 00:21:30,240 --> 00:21:35,520 Speaker 1: remarkable ways we learn, and it ignores our evolutionary history. Um, 366 00:21:35,840 --> 00:21:39,399 Speaker 1: so I would, I would. Uh, it'll be interesting to 367 00:21:39,440 --> 00:21:46,000 Speaker 1: see what progress they make for sure. So guys, the 368 00:21:46,040 --> 00:21:50,840 Speaker 1: possibility of super human cognition sounds appealing. Um, you know, 369 00:21:50,880 --> 00:21:53,679 Speaker 1: if we're up to you two and and somebody was 370 00:21:53,720 --> 00:21:56,240 Speaker 1: offering to put a chip or an apparatus and in 371 00:21:56,280 --> 00:21:58,359 Speaker 1: the year brains like they're doing to the to the 372 00:21:58,400 --> 00:22:00,720 Speaker 1: poor little zebra finch, would you do it? I mean, 373 00:22:00,880 --> 00:22:03,800 Speaker 1: you know, in some sence, we're already doing this stuff today. 374 00:22:03,920 --> 00:22:07,159 Speaker 1: If you have an implant to help you here, or 375 00:22:07,480 --> 00:22:10,520 Speaker 1: things to stop Parkinson's tremors, yeah, if I had one 376 00:22:10,520 --> 00:22:13,560 Speaker 1: of those conditions, I would absolutely get one of these 377 00:22:13,560 --> 00:22:16,919 Speaker 1: implants and supercharge myself. When you start going into this 378 00:22:17,080 --> 00:22:19,640 Speaker 1: this next wave of stuff. It gets I think, far 379 00:22:19,680 --> 00:22:25,040 Speaker 1: more philosophical and complicated, because you're talking about changing humans 380 00:22:25,040 --> 00:22:27,639 Speaker 1: from what they are, some sort of weird next step 381 00:22:27,720 --> 00:22:32,320 Speaker 1: of evolution where we're kind of half man half machine. Um. 382 00:22:32,359 --> 00:22:34,719 Speaker 1: You know, in some ways there's people I talked to, 383 00:22:35,359 --> 00:22:38,399 Speaker 1: like the guys at Colonel, who argue that this is 384 00:22:38,440 --> 00:22:40,280 Speaker 1: the only way humans will be able to keep up 385 00:22:40,359 --> 00:22:44,280 Speaker 1: with machines. And we always hear about losing jobs artificial 386 00:22:44,280 --> 00:22:48,240 Speaker 1: intelligence and and seeing what humans can do going away. 387 00:22:48,320 --> 00:22:50,240 Speaker 1: And so you know, if you're half and half, you 388 00:22:50,280 --> 00:22:52,280 Speaker 1: can you can keep up, but maybe keep some of 389 00:22:52,320 --> 00:22:56,320 Speaker 1: your humanness as well. Sarah, what about cyborg? Sarah McBride, 390 00:22:56,400 --> 00:22:58,679 Speaker 1: will we ever see? Will we ever see that? I 391 00:22:58,720 --> 00:23:03,320 Speaker 1: already hate myself, so I doubt it. But um, what 392 00:23:03,359 --> 00:23:07,080 Speaker 1: about this idea of keeping up with robots? And and yeah, 393 00:23:07,359 --> 00:23:09,879 Speaker 1: there are people who think that AI could help us 394 00:23:09,960 --> 00:23:14,640 Speaker 1: solve problems like climate change. I think that's too optimistic. 395 00:23:15,119 --> 00:23:17,199 Speaker 1: The funny part about doing the story for me was 396 00:23:17,280 --> 00:23:21,080 Speaker 1: that that in the AI camp you tend to have 397 00:23:21,280 --> 00:23:23,640 Speaker 1: I feel like people who think the technology is really 398 00:23:23,640 --> 00:23:26,199 Speaker 1: far along. You know, if you're talking about specifically computer 399 00:23:26,280 --> 00:23:30,240 Speaker 1: scientists and the Silicon Valley kind of crew. Um, they're 400 00:23:30,359 --> 00:23:32,720 Speaker 1: very impressed what they've come up with over the last 401 00:23:32,760 --> 00:23:34,360 Speaker 1: few years. When we went to talk to all these 402 00:23:34,400 --> 00:23:38,080 Speaker 1: brain researchers, the ones who are down in the box, 403 00:23:38,119 --> 00:23:41,199 Speaker 1: down at the neurons, they seemed, on the whole to 404 00:23:41,280 --> 00:23:46,240 Speaker 1: me much more skeptical about when we would see huge breakthroughs. 405 00:23:46,280 --> 00:23:48,600 Speaker 1: They seem to think that a lot of this stuff 406 00:23:48,640 --> 00:23:51,320 Speaker 1: was years and years away, and and I felt like 407 00:23:51,359 --> 00:23:55,480 Speaker 1: they had this sense of how complicated the brain really 408 00:23:55,560 --> 00:23:58,600 Speaker 1: is and that unlocking its secrets is going to take along. 409 00:23:58,680 --> 00:24:00,440 Speaker 1: But given the fact that they are taking the first 410 00:24:00,480 --> 00:24:04,919 Speaker 1: steps to what will ultimately be very transformative and challenging, 411 00:24:05,200 --> 00:24:07,800 Speaker 1: controversial technology. Sarah, did you get the sense that they 412 00:24:07,800 --> 00:24:11,040 Speaker 1: were wrestling with the ethical complications of their work? You know, 413 00:24:11,080 --> 00:24:13,240 Speaker 1: a lot of them actually turned out to have studied 414 00:24:13,280 --> 00:24:16,160 Speaker 1: philosophy at some point in their careers, which I thought 415 00:24:16,320 --> 00:24:20,400 Speaker 1: was pretty interesting. And yeah, they talk about the decisions 416 00:24:20,480 --> 00:24:23,000 Speaker 1: that a carmaker might have to make. Is it more 417 00:24:23,040 --> 00:24:27,040 Speaker 1: important to preserve the life of a passenger or a pedestrian, 418 00:24:27,160 --> 00:24:30,800 Speaker 1: stuff like that. So, yeah, they're thinking about these big problems. 419 00:24:30,840 --> 00:24:33,160 Speaker 1: That doesn't mean they know how to answer them though, 420 00:24:33,240 --> 00:24:35,520 Speaker 1: any more than we would. But they say that it 421 00:24:35,560 --> 00:24:38,159 Speaker 1: means these systems will be very human one day. Do 422 00:24:38,240 --> 00:24:40,719 Speaker 1: they understand that they have now found themselves right at 423 00:24:40,760 --> 00:24:43,639 Speaker 1: the center of this next wave in in computing. I 424 00:24:43,680 --> 00:24:46,000 Speaker 1: mean they do to a degree. They're definitely excited. I mean, 425 00:24:46,600 --> 00:24:49,760 Speaker 1: just in very crass terms. A lot of these people 426 00:24:50,040 --> 00:24:52,639 Speaker 1: get much better job offers than they would have in 427 00:24:52,680 --> 00:24:55,520 Speaker 1: the past. You know, there used to be far less 428 00:24:55,640 --> 00:24:59,560 Speaker 1: neuroscience graduates. It wasn't that appealing of a field. We 429 00:24:59,560 --> 00:25:02,480 Speaker 1: didn't know much about the brain. All these promised breakthroughs 430 00:25:02,480 --> 00:25:05,320 Speaker 1: weren't happening at all. In Now you can go to 431 00:25:05,400 --> 00:25:07,920 Speaker 1: a university do this amazing work, or if you kind 432 00:25:07,960 --> 00:25:09,800 Speaker 1: of get tired of that, or you want to poke 433 00:25:09,840 --> 00:25:11,879 Speaker 1: around somewhere else, you can go work for one of 434 00:25:11,880 --> 00:25:13,960 Speaker 1: these tech companies and get paid I don't know, like 435 00:25:14,040 --> 00:25:16,159 Speaker 1: ten times what you need to make at one of 436 00:25:16,200 --> 00:25:18,280 Speaker 1: these labs. Well, to bring this all the way back 437 00:25:18,320 --> 00:25:21,320 Speaker 1: to the beginning, Sarah, is it too early? When I 438 00:25:21,359 --> 00:25:25,000 Speaker 1: talked to Alexa or Sirie or Google Voice to thank 439 00:25:25,359 --> 00:25:28,880 Speaker 1: the little zebra finch? Can we can we see any 440 00:25:28,960 --> 00:25:31,840 Speaker 1: of the zebra finch and that research in today's AI. 441 00:25:32,320 --> 00:25:35,000 Speaker 1: We see some of it already and things like voice 442 00:25:35,040 --> 00:25:40,159 Speaker 1: recognition and making better audio quality, but I think some 443 00:25:40,240 --> 00:25:44,320 Speaker 1: of the biggest stuff is uh yet to come, and 444 00:25:44,400 --> 00:25:51,119 Speaker 1: we can always check in with Timachi. And that's it 445 00:25:51,200 --> 00:25:54,240 Speaker 1: for this week's episode of Decrypted. Thanks for listening. We 446 00:25:54,320 --> 00:25:56,160 Speaker 1: always like to know what you think of the show. 447 00:25:56,320 --> 00:25:59,399 Speaker 1: You can write to us at Decrypted at Bloomberg dot 448 00:25:59,400 --> 00:26:04,240 Speaker 1: net or I'm on Twitter at McBride s G, I'm 449 00:26:04,240 --> 00:26:08,000 Speaker 1: at Bradstone, and I'm at Valley Hack. And please help 450 00:26:08,080 --> 00:26:10,199 Speaker 1: us spread the word about our show by leaving us 451 00:26:10,200 --> 00:26:13,760 Speaker 1: a rating or review in your favorite podcast app. This 452 00:26:13,880 --> 00:26:18,560 Speaker 1: episode was produced by Pierre Gadkari and Lindsay Cradowell. Our 453 00:26:18,600 --> 00:26:22,400 Speaker 1: story editor was Aki Ito. Thank you also to Ann 454 00:26:22,520 --> 00:26:26,280 Speaker 1: vander May and Emily Buso. Francesca Levi is head of 455 00:26:26,320 --> 00:26:29,160 Speaker 1: Bloomberg Podcasts. We'll see you next week.