1 00:00:00,300 --> 00:00:04,500 This is epicenter episode 275 with guest Ben goertzel. 2 00:00:19,300 --> 00:00:23,000 This episode of epicenter is brought you by Cosmos Cosmos, is 3 00:00:23,000 --> 00:00:25,500 building the internal blockchains an ecosystem. 4 00:00:25,500 --> 00:00:28,400 Where thousands of blockchains can interoperate creating the 5 00:00:28,400 --> 00:00:30,300 foundation for a new token economy. 6 00:00:30,800 --> 00:00:34,400 If you have an idea for adapt, visit Cosmos, dot Network, / 7 00:00:34,400 --> 00:00:37,300 epicenter, to learn more and to get in touch with the cosmos 8 00:00:37,300 --> 00:00:41,800 team and by top-down experience a new way of hiring. 9 00:00:42,000 --> 00:00:43,700 Stop. Tell delivers only the top, 10 00:00:43,700 --> 00:00:46,300 three percent of applicants including highly skilled 11 00:00:46,300 --> 00:00:48,600 blockchain Engineers. If you're looking to scale your 12 00:00:48,600 --> 00:00:52,900 team with the very best talent, is it top.com / epicenter. 13 00:00:57,600 --> 00:01:01,000 Hi, welcome to epicenter. My name is Sebastian Cujo and my 14 00:01:01,000 --> 00:01:01,600 name is fine. Bob. 15 00:01:01,600 --> 00:01:05,900 And Kay today, we have an interview with Banker. 16 00:01:05,900 --> 00:01:10,200 It's all who is the CEO of Singularity that he also works 17 00:01:10,200 --> 00:01:11,700 for a company called Hanson robotics. 18 00:01:11,900 --> 00:01:15,100 Yes, you may have seen in the news and so Hanson robotics 19 00:01:15,100 --> 00:01:18,400 makes this robot. This is going to humanoid robot 20 00:01:18,400 --> 00:01:22,500 named Sofia. She's been featured and you know 21 00:01:22,500 --> 00:01:26,300 on TV shows and Silicon Valley the sitcom and sort of has 22 00:01:26,300 --> 00:01:28,100 become an ambassador for robotics. 23 00:01:28,400 --> 00:01:31,800 It's a band is an expert in artificial intelligence and 24 00:01:31,800 --> 00:01:35,000 Robotics and single learning. That is a company that is 25 00:01:35,000 --> 00:01:39,300 building as the marketplace for AI on boshane. 26 00:01:39,600 --> 00:01:41,800 So we talk to Ben about all kinds of interesting. 27 00:01:41,900 --> 00:01:45,100 Topics that we don't usually get a chance to discuss on the show 28 00:01:45,100 --> 00:01:49,600 since we primarily focus on watching but talking about AI in 29 00:01:49,600 --> 00:01:54,100 a more General sense and what future is there, and how that 30 00:01:54,100 --> 00:01:56,700 ties into block, shame is a really fascinating conversation 31 00:01:56,700 --> 00:02:02,100 and that is is a great speaker on these topics and there's a 32 00:02:02,107 --> 00:02:05,800 lot of thinking at a high level, so it's really fascinating to 33 00:02:05,800 --> 00:02:08,199 get to interview him. The first, we got a couple of 34 00:02:08,199 --> 00:02:10,900 announcements. So one one thing that you 35 00:02:10,908 --> 00:02:12,900 mentioned is that We're going to be at least. 36 00:02:12,900 --> 00:02:17,000 I will be at ECC week of March 4th. 37 00:02:17,100 --> 00:02:21,200 So, this first box chain week, happening in Paris the whole 38 00:02:21,200 --> 00:02:23,900 week. And so, I'll be a VC will also 39 00:02:23,900 --> 00:02:27,900 be having a meet-up and I mean Up is on March 6th on the 40 00:02:27,900 --> 00:02:31,400 Wednesday around 6:00. It's just going to be a casual 41 00:02:31,900 --> 00:02:34,500 get-together and drinks Meetup venue. 42 00:02:34,500 --> 00:02:37,100 Isn't totally figured out yet but it will be announced soon 43 00:02:37,600 --> 00:02:41,000 and hope you hope to see you. There is going to be I'll be 44 00:02:41,000 --> 00:02:43,300 there son. You might be there as well. 45 00:02:43,900 --> 00:02:47,300 And you will have some guests and other listeners. 46 00:02:47,300 --> 00:02:49,900 So happy, happy to have you with happy to have. 47 00:02:49,900 --> 00:02:53,800 You join us for that meet up and you can sign up and register at 48 00:02:54,100 --> 00:02:57,700 the center. Dot rocks, / EC, so that's 49 00:02:57,700 --> 00:03:01,400 epicenter, not rocks /e. See if you register their will 50 00:03:01,400 --> 00:03:05,600 send you the address for the venue when when it's a nest and 51 00:03:05,600 --> 00:03:08,000 Brian, I think you had an update on course one. 52 00:03:09,100 --> 00:03:11,000 Yeah, but first of all, that sounds great. 53 00:03:11,000 --> 00:03:13,300 I'm so jealous. I wish I could be in Paris, too. 54 00:03:13,900 --> 00:03:15,900 I think it's gonna be a really great conference. 55 00:03:15,900 --> 00:03:18,400 I mean the first one that they had a couple years ago, we were 56 00:03:18,400 --> 00:03:22,000 there together was terrific and I wasn't there for the 57 00:03:22,000 --> 00:03:25,700 subsequent ones, but it's now grown to the organizers of told 58 00:03:25,700 --> 00:03:29,400 me they're waiting for their were hoping to have 1,500 people 59 00:03:29,400 --> 00:03:30,700 there. It's going to be at the canal 60 00:03:30,700 --> 00:03:32,800 which is this really fantastic venue. 61 00:03:32,800 --> 00:03:37,900 And yeah, 300 speakers. So it's it's turning out to be 62 00:03:37,900 --> 00:03:41,800 quite a quite a Conference. Cool, fantastic. 63 00:03:41,800 --> 00:03:43,300 But yeah. So I did want to give a brief 64 00:03:43,300 --> 00:03:46,700 update on correspond so you know, as many listeners know, 65 00:03:46,700 --> 00:03:49,000 most less than 0. So we've been mayor and I 66 00:03:49,000 --> 00:03:52,000 started this company together. It's just over a year ago to 67 00:03:52,000 --> 00:03:55,200 work on kind of building valid as proof State networks. 68 00:03:55,800 --> 00:03:59,700 And finally, now, we are life in the first Network, which is 69 00:03:59,700 --> 00:04:01,800 Project called Loom. So it's in the theorem 70 00:04:01,800 --> 00:04:06,300 sidechain, sort of a plasma cash chain, you know, we did we did 71 00:04:06,300 --> 00:04:07,900 an interview on this before as well. 72 00:04:09,400 --> 00:04:14,200 And Using tenement as consensus. So we just launched share last 73 00:04:14,200 --> 00:04:18,700 Friday and so you know, anyone who has Loom can delegate to us. 74 00:04:19,000 --> 00:04:22,400 We also wrote this in death, kind of research report on the 75 00:04:22,400 --> 00:04:26,800 room so if people want to check that out that's also available 76 00:04:27,200 --> 00:04:32,000 and the Rue I will put the link to our website in the show know 77 00:04:32,000 --> 00:04:36,800 so people can find everything there and then Cosmos is also 78 00:04:36,800 --> 00:04:39,000 about to go live. So which is also something where 79 00:04:39,000 --> 00:04:43,000 we'll have a valid. Later on and that brings us to 80 00:04:43,000 --> 00:04:45,900 the other thing. We wanted to speak about which 81 00:04:45,900 --> 00:04:50,900 is a first of all, we have a cosmos is starting to sponsor 82 00:04:50,900 --> 00:04:53,700 epicenter. So that's very exciting and it's 83 00:04:53,700 --> 00:04:57,100 starting to do. So if this episode now we do 84 00:04:57,100 --> 00:04:59,400 need to make a disclaimer here which is that. 85 00:04:59,400 --> 00:05:02,200 First of all, I have some atoms and save us. 86 00:05:02,200 --> 00:05:07,600 He has some items to Sonny has atoms and met her as items to, 87 00:05:07,600 --> 00:05:10,300 right? So, the epicenter team is fairly 88 00:05:10,700 --> 00:05:14,400 Heavily. You know, has some atom 89 00:05:14,400 --> 00:05:17,200 positions. And in addition to that, of 90 00:05:17,200 --> 00:05:20,800 course only works for the cosmos team and I used to work for 91 00:05:20,800 --> 00:05:24,800 Cosmo Steam and then mayor and I have Nelson building it Cosmos 92 00:05:24,800 --> 00:05:27,100 validator. So we just wanted to give that 93 00:05:27,100 --> 00:05:29,700 disclaimer upfront to people know that and fully aware of 94 00:05:29,700 --> 00:05:33,300 that. And and that ties into a larger 95 00:05:33,300 --> 00:05:36,300 topic which we've had a bunch of discussions about. 96 00:05:36,300 --> 00:05:39,800 But we haven't really taken the necessary actions on, and it's 97 00:05:39,800 --> 00:05:42,900 long. Long overdue for us to do that, 98 00:05:43,100 --> 00:05:48,700 which is to have a better way of disclosing, those kind of 99 00:05:48,707 --> 00:05:53,000 things. So what we will start doing and 100 00:05:53,000 --> 00:05:57,500 we'll probably have to up within the next week or so is it's just 101 00:05:57,500 --> 00:06:02,500 a page on a website where, you know, you'll be able to see all 102 00:06:02,500 --> 00:06:08,300 of the hosts and and they will list all of the tokens or you 103 00:06:08,300 --> 00:06:12,700 know, other kind of Investments they have in the blocked in 104 00:06:12,700 --> 00:06:18,900 space and then I think the other thing will start doing this, 105 00:06:18,900 --> 00:06:22,000 maybe in the show notes for every episode. 106 00:06:22,000 --> 00:06:24,500 I mean, this is something we have been doing generally. 107 00:06:24,700 --> 00:06:28,200 Like let's say there's an episode and like somebody has 108 00:06:29,200 --> 00:06:34,100 this token, then we've generally been mentioning that probably 109 00:06:34,100 --> 00:06:39,500 with the exception of Bitcoin and hearing episodes but but 110 00:06:39,500 --> 00:06:41,700 they are we Just want to be more consistent. 111 00:06:41,700 --> 00:06:44,700 They're really make sure you mention it every time and also 112 00:06:44,700 --> 00:06:47,200 write it in the show notes. So that I think that's mentioned 113 00:06:48,100 --> 00:06:50,900 the fact that we hold atoms. I guess, at least for me. 114 00:06:50,900 --> 00:06:54,000 I've always been interested in Cosmos and I'm actually 115 00:06:54,000 --> 00:06:56,400 generally quite excited that they're launching and so having 116 00:06:56,400 --> 00:06:58,800 them become a sponsor. Just felt like a really natural 117 00:06:58,800 --> 00:07:00,900 fit. And so I think you'll see the 118 00:07:00,900 --> 00:07:08,700 add their intention to sponsor. The show is sort of benefit for 119 00:07:08,700 --> 00:07:12,500 both one because You think the cosmos is a great platform and 120 00:07:12,800 --> 00:07:15,800 people should generally have interested in and also the 121 00:07:15,800 --> 00:07:18,200 cosmos team has always been very closely. 122 00:07:18,600 --> 00:07:21,000 We're we've always been very close to that team and I've 123 00:07:21,000 --> 00:07:23,400 always sort of appreciate what we done and they would say was 124 00:07:23,500 --> 00:07:25,400 one of our early guest sort of things. 125 00:07:25,900 --> 00:07:27,200 Yeah. So that I think you've said 126 00:07:27,200 --> 00:07:30,900 enough on that and we'll have that page on our website within 127 00:07:31,100 --> 00:07:34,400 about a week or so. And we'll make a point of also 128 00:07:34,400 --> 00:07:36,000 mentioning in the show notes as you mentioned. 129 00:07:36,400 --> 00:07:38,800 So without further delay, here's our interview with an earth 130 00:07:38,800 --> 00:07:42,800 Style. Alright, so we're here with Ben 131 00:07:42,800 --> 00:07:46,600 goertzel and Ben is the founder and CEO of Singularity. 132 00:07:46,600 --> 00:07:51,500 Net is also Chief scientist and at Hanson, Robotics and holds a 133 00:07:51,508 --> 00:07:54,400 number of positions in other organizations. 134 00:07:54,400 --> 00:07:57,300 But we'll get to that in these interviews, been so high debt. 135 00:07:57,400 --> 00:08:00,500 Thanks for joining us. Hey, it's a pleasure to be here. 136 00:08:01,400 --> 00:08:03,700 Well, thank you for joining us. So, yeah, you have a very 137 00:08:03,700 --> 00:08:06,700 impressive resume. So, as I mentioned, your the 138 00:08:07,100 --> 00:08:10,600 background CEO of Singularity at your also Chief scientist is 139 00:08:10,600 --> 00:08:13,500 handsome. Robot Two PhD in mathematics. 140 00:08:14,300 --> 00:08:17,200 You've started a whole bunch of companies and lots of different 141 00:08:17,200 --> 00:08:21,500 areas and you're involved in some nonprofits and Foundations 142 00:08:21,500 --> 00:08:25,000 as well. So how did you, how did you get 143 00:08:25,000 --> 00:08:28,200 here? And what, what is your 144 00:08:28,200 --> 00:08:30,600 trajectory looks like? And how did you get involved in 145 00:08:30,600 --> 00:08:35,299 Ai and Robotics? I've been interested in AI 146 00:08:35,299 --> 00:08:40,400 robotics life, extension, nanotech, femto Tech time 147 00:08:40,400 --> 00:08:44,400 travel, all these things since, you know, the early 1970s. 148 00:08:44,400 --> 00:08:48,800 When I, when I was a little kid reading science fiction books 149 00:08:48,800 --> 00:08:52,700 and you know, now a few decades have passed. 150 00:08:52,700 --> 00:08:57,400 And I found myself in a world where many of these apparently 151 00:08:57,400 --> 00:09:01,900 science fictional Technologies are are gradually becoming 152 00:09:01,900 --> 00:09:04,800 reality. These and so, it's really, 153 00:09:04,800 --> 00:09:09,800 really exciting to me to actually be every day, you know, 154 00:09:09,800 --> 00:09:15,100 concretely, working on building thinking machines and networking 155 00:09:15,100 --> 00:09:19,400 people and computers together into a, into a global brain and 156 00:09:19,400 --> 00:09:22,400 applying AI to longevity and nanotechnology. 157 00:09:22,400 --> 00:09:27,200 It's astounding that we live in a time when these things are 158 00:09:27,500 --> 00:09:29,700 realities. And of course, it's also a bit. 159 00:09:30,800 --> 00:09:33,500 Scary and sobering at times because these things could go 160 00:09:33,500 --> 00:09:37,500 badly wrong or they could go. Amazingly. 161 00:09:37,600 --> 00:09:41,200 Right? And you know, why I've been 162 00:09:41,200 --> 00:09:43,500 involved in a lot of different aspects of all these 163 00:09:44,100 --> 00:09:45,900 Technologies. Many of which are converging 164 00:09:45,900 --> 00:09:49,200 together now. So, I did a PhD in in 165 00:09:49,200 --> 00:09:52,200 mathematics but even at that time, I was very interested in 166 00:09:53,300 --> 00:09:55,800 AI biotechnology and a bunch of other things. 167 00:09:55,800 --> 00:10:00,000 I just figured mathematics. You know, as Bitcoin says, in 168 00:10:00,000 --> 00:10:04,300 math, We Trust right mathematics underlie underlies, everything, 169 00:10:04,300 --> 00:10:07,200 that's the foundation of all modern science and technology so 170 00:10:07,200 --> 00:10:08,700 I figured learning a bunch of math. 171 00:10:09,900 --> 00:10:12,600 Couldn't be bad. But since shortly after getting 172 00:10:12,600 --> 00:10:16,200 my PhD, I've been really, which was 89. 173 00:10:16,200 --> 00:10:18,300 I got my degree. I mean, since I've been working 174 00:10:18,300 --> 00:10:22,600 on AI and in various dimensions and aspects. 175 00:10:22,600 --> 00:10:27,100 And now in the last few years, that's really taken off along 176 00:10:27,100 --> 00:10:30,000 with a bunch of other other Technologies and of course, 177 00:10:30,500 --> 00:10:33,200 blockchain and cryptocurrency. Would you guys know a lot about? 178 00:10:33,200 --> 00:10:36,600 This is all part of the mix. So, right now, there's a insane 179 00:10:36,600 --> 00:10:38,900 number of different Advanced Technologies. 180 00:10:39,000 --> 00:10:41,600 He's for manipulating creating different kinds of information 181 00:10:41,600 --> 00:10:46,800 that are all intersecting and, and pushing each other forward. 182 00:10:46,800 --> 00:10:49,200 And you could talk about these for hundreds of hours without 183 00:10:49,200 --> 00:10:52,100 exhausting at all. Yeah, that's definitely true. 184 00:10:52,600 --> 00:10:56,700 So, let's, let's spend a little bit of time first on the topic 185 00:10:56,700 --> 00:11:00,100 of AI, which is something that I think, you know, if tangentially 186 00:11:00,100 --> 00:11:04,000 talked about a bunch of times, but still, it's, I guess, like, 187 00:11:04,000 --> 00:11:06,600 like probably for many outside of the Block Chain spaces, you 188 00:11:06,600 --> 00:11:08,800 know, big scary term that's a little bit. 189 00:11:09,100 --> 00:11:13,900 To kind of demystify. So how do you define a? 190 00:11:13,900 --> 00:11:19,100 I And what's the difference between Ai and, you know, terms 191 00:11:19,100 --> 00:11:22,000 that people use like machine learning and deep learning? 192 00:11:23,000 --> 00:11:27,200 I don't think any of these terms are worth too much in in the 193 00:11:27,200 --> 00:11:29,700 end. I mean AI. 194 00:11:29,700 --> 00:11:32,400 I'm in what sense is it really artificial? 195 00:11:32,400 --> 00:11:35,900 It's all part of Nature and to some extent, these systems are 196 00:11:36,300 --> 00:11:40,500 evolving and emerging. Instead of being purely 197 00:11:40,800 --> 00:11:44,500 artificially created and intelligence, we don't even have 198 00:11:44,500 --> 00:11:47,400 a good definition for among humans. 199 00:11:47,400 --> 00:11:50,600 Like, there's not an IQ test that works across different 200 00:11:50,600 --> 00:11:54,900 cultures or ages of people, let alone Girls different kinds of 201 00:11:54,900 --> 00:11:56,600 Minds. None of these are very rigorous 202 00:11:56,600 --> 00:11:59,400 terms. I mean, machine learning again, 203 00:12:00,200 --> 00:12:03,300 in essence, all AI really is about machines that learned. 204 00:12:03,300 --> 00:12:08,500 And, and reason, and think that term has lately, come to be used 205 00:12:08,500 --> 00:12:12,800 to describe particular types of AI algorithms that are trained 206 00:12:12,800 --> 00:12:19,400 on on large amounts of data, but then the term is also used more 207 00:12:19,400 --> 00:12:21,300 Loosely. So, you always reinforcement 208 00:12:21,300 --> 00:12:24,200 learning a kind of machine. learning or not, it's not, 209 00:12:24,200 --> 00:12:28,600 especially not, especially well, defined and I mean, deep deep 210 00:12:28,600 --> 00:12:32,200 learning again, in cognitive science, you know, a guy named 211 00:12:32,200 --> 00:12:36,800 Stella Nelson wrote a book on deep learning what 15 years ago, 212 00:12:36,800 --> 00:12:41,000 which encompassed neural networks logic, systems 213 00:12:41,000 --> 00:12:45,600 production systems, many kinds of AI algorithms but now now the 214 00:12:45,600 --> 00:12:48,500 term seems to be used for what used to be called multi-layer 215 00:12:48,500 --> 00:12:53,100 perceptrons multi-layer neural networks, which is really only 216 00:12:53,100 --> 00:12:57,500 one special kind of Deep learning system and then in the 217 00:12:57,500 --> 00:12:59,800 broader said. So I mean, what, what deep 218 00:12:59,800 --> 00:13:02,500 learning originally meant was any system that just does 219 00:13:02,500 --> 00:13:05,100 hierarchical pattern recognition, like recognizes 220 00:13:05,500 --> 00:13:08,200 patterns within patterns within patterns within patterns in the 221 00:13:08,208 --> 00:13:10,300 world and uses those to take some action. 222 00:13:10,900 --> 00:13:14,800 The Deep Learning Systems being talked about, mostly now are 223 00:13:14,800 --> 00:13:18,000 hierarchical neural networks, which is one special kind of 224 00:13:18,000 --> 00:13:19,400 deep learning system in the brother. 225 00:13:19,400 --> 00:13:23,400 Kay so I mean what we have is a lot of words with confusing 226 00:13:23,400 --> 00:13:28,100 definitions that that that Shift over time and don't necessarily 227 00:13:28,100 --> 00:13:32,300 mean what they sound like, they mean which, which comes back to 228 00:13:32,300 --> 00:13:33,700 in math, We Trust. Right? 229 00:13:33,700 --> 00:13:37,100 Because the thing is the algorithms are doing are doing 230 00:13:37,700 --> 00:13:40,500 what they're doing and that there is a real mathematical 231 00:13:40,500 --> 00:13:42,900 description to them which and they carry out practical 232 00:13:42,900 --> 00:13:47,200 functions but the buzz words associated when them serve 233 00:13:47,200 --> 00:13:50,300 mostly to sell things rather than to convey useful 234 00:13:50,300 --> 00:13:52,100 information. Okay. 235 00:13:52,100 --> 00:13:56,200 Okay that is helpful but then let's speak about About one 236 00:13:56,200 --> 00:13:58,300 term. And I think that's the term that 237 00:13:58,300 --> 00:14:02,500 maybe you have more relationship to which is a GI. 238 00:14:02,500 --> 00:14:05,000 So artificial general intelligence. 239 00:14:05,900 --> 00:14:13,000 Yeah, well let me let me try to go. to what I think are the 240 00:14:14,400 --> 00:14:19,300 foundations here because I mean there, It is possible to 241 00:14:19,300 --> 00:14:21,700 describe these things in a way that makes sense. 242 00:14:21,700 --> 00:14:26,100 It's just that Things become marketing buzzwords, and then, 243 00:14:26,100 --> 00:14:29,000 and then them become become confusing. 244 00:14:29,000 --> 00:14:34,400 So I think Fundamentally. You can think about a mind or an 245 00:14:34,400 --> 00:14:40,200 intelligent system as something that's recognizing patterns in 246 00:14:40,200 --> 00:14:45,000 itself and in the world around it and then the system may have 247 00:14:45,000 --> 00:14:47,200 some goals which doesn't mean everything. 248 00:14:47,200 --> 00:14:51,000 It does is goal-directed, but it may have some goals and then 249 00:14:51,000 --> 00:14:55,900 recognizes patterns regarding which actions will achieve which 250 00:14:55,900 --> 00:14:58,900 goals in which contexts, right? So you have a pattern 251 00:14:58,900 --> 00:15:02,800 recognition system. And it has goals among other 252 00:15:02,800 --> 00:15:05,000 Dynamics and its trying to learn. 253 00:15:05,000 --> 00:15:07,300 It starting to recognize patterns of how to achieve, what 254 00:15:07,300 --> 00:15:10,900 goals and what situations. And you know babies do that, 255 00:15:10,900 --> 00:15:12,600 right? Babies are recognizing patterns 256 00:15:12,600 --> 00:15:16,700 in the world around them all the time and they have some goals 257 00:15:16,700 --> 00:15:19,600 that they want to get some milk, some food they want to run 258 00:15:19,600 --> 00:15:24,000 around, and they try to figure out what patterns of activity 259 00:15:24,000 --> 00:15:27,100 will let them achieve their goals and what in what 260 00:15:27,300 --> 00:15:30,500 situations and then where? I mean, we're deep learning. 261 00:15:30,600 --> 00:15:35,400 Comes in is the world we live. In seems to be made largely of 262 00:15:35,400 --> 00:15:37,700 hard rock, quickly, composed patterns, where you have 263 00:15:37,700 --> 00:15:40,900 patterns that build up in the more complex ones going up into 264 00:15:40,900 --> 00:15:43,700 more complex ones. I mean, just like physics builds 265 00:15:43,700 --> 00:15:46,600 into chemistry builds into biology builds into psychology, 266 00:15:46,600 --> 00:15:49,400 builds into sociology. So we have a hierarchically 267 00:15:49,400 --> 00:15:52,700 composed patterns, which means if you have a learning engine 268 00:15:52,700 --> 00:15:57,300 that is trying to recognize, Patterns in the hierarchy in May 269 00:15:57,300 --> 00:16:00,400 well succeed. Because our world seems to be 270 00:16:00,600 --> 00:16:05,600 seems to be built that way. Now it happens that Most of the 271 00:16:05,600 --> 00:16:09,300 AI is out there in the world. Now are able to recognize 272 00:16:09,300 --> 00:16:12,600 patterns in a very narrowly defined context and to achieve, 273 00:16:12,600 --> 00:16:18,200 only a very narrow set of goals, like, Say the original alphago, 274 00:16:18,400 --> 00:16:22,800 could recognize patterns in go games and it could achieve the 275 00:16:22,800 --> 00:16:24,700 goal of winning, a go game, right? 276 00:16:24,700 --> 00:16:29,400 And that was it now, Alpha zero with a step beyond that. 277 00:16:30,300 --> 00:16:33,000 These programs are all by Google deepmind, which is one of the 278 00:16:33,000 --> 00:16:35,200 more interesting AI organizations out. 279 00:16:35,200 --> 00:16:38,800 There are 40, won't be on that because it can play a lot of 280 00:16:38,800 --> 00:16:42,200 different kinds of board games. So it can recognize patterns in 281 00:16:42,200 --> 00:16:45,100 a broader scope of environments and many, many different types 282 00:16:45,100 --> 00:16:48,400 of board games and it can achieve more types of goals 283 00:16:48,400 --> 00:16:52,900 because the ways of winning chess or go or shogi, whatever 284 00:16:52,900 --> 00:16:56,800 are different, right? It's still not nearly as general 285 00:16:56,800 --> 00:16:59,200 as a human being, though, because we can only play board 286 00:16:59,200 --> 00:17:02,700 games We can recognize patterns in a huge number of other kinds 287 00:17:02,700 --> 00:17:06,000 of environments and we can achieve many, many different 288 00:17:06,099 --> 00:17:07,900 types of goals, like we can prove math. 289 00:17:07,900 --> 00:17:10,700 Terms become blow people off. We can, we can chase girls. 290 00:17:10,700 --> 00:17:14,000 We can, we can make art, we can try this, save starving kids, 291 00:17:14,200 --> 00:17:17,400 there's a lot of goals. We can work toward in a fairly 292 00:17:18,000 --> 00:17:21,099 Rich collection of environments but we're still not like 293 00:17:21,099 --> 00:17:22,800 infinitely. General intelligence has you 294 00:17:22,808 --> 00:17:26,500 could you can imagine a mind that could recognize patterns in 295 00:17:26,500 --> 00:17:29,600 407 dimensional space, where very bad at that? 296 00:17:29,800 --> 00:17:32,300 We're much better like two or three or four dimensions. 297 00:17:32,700 --> 00:17:35,700 So we're still somewhat restricted. 298 00:17:35,700 --> 00:17:38,200 We're good at recognizing patterns and some kinds of 299 00:17:38,200 --> 00:17:40,600 environments. Achieving some kinds of goals 300 00:17:40,900 --> 00:17:44,400 better than Alpha 0 or existing, AI programs. 301 00:17:44,800 --> 00:17:48,400 But you can imagine some kind of mind that could recognize 302 00:17:48,400 --> 00:17:52,700 patterns in like a space of any dimensions and in things that 303 00:17:52,700 --> 00:17:55,400 just look like noise to human beings and they could achieve 304 00:17:55,400 --> 00:17:59,700 goals that. Humans can't even begin to 305 00:17:59,700 --> 00:18:03,400 understand to understand. So I think, you know, totally 306 00:18:03,400 --> 00:18:07,500 general intelligence, that could recognize any kind of pattern in 307 00:18:07,500 --> 00:18:11,600 any kind of world and could figure out how to achieve any 308 00:18:11,600 --> 00:18:15,100 kind of goal by recognizing patterns of other achieve that 309 00:18:15,100 --> 00:18:18,700 goal. That's probably not achievable 310 00:18:18,900 --> 00:18:22,700 in this physical Universe, like totally general intelligence, 311 00:18:23,300 --> 00:18:26,900 but we're much more General than any existing AI program like 312 00:18:26,900 --> 00:18:28,500 each of us. Deal with a lot of different 313 00:18:28,500 --> 00:18:30,000 problems. And if you give us something 314 00:18:30,000 --> 00:18:33,300 totally new to do with like the internet didn't exist. 315 00:18:33,300 --> 00:18:38,300 When I was born, let alone when the, my DNA evolved, that didn't 316 00:18:38,300 --> 00:18:41,700 exist when I went to school, but I like, everyone else was able 317 00:18:41,700 --> 00:18:44,800 to adapt to deal with this new thing, right? 318 00:18:44,800 --> 00:18:47,900 We don't yet have a eyes that can transfer the knowledge and 319 00:18:47,900 --> 00:18:51,400 adapt to deal. With some very new type of 320 00:18:51,400 --> 00:18:56,400 thing, they weren't programmed or Or trained for right and I 321 00:18:56,400 --> 00:18:59,100 think, I think we will but we're not there yet. 322 00:18:59,100 --> 00:19:03,500 So I think now the AI field is starting to begin the transition 323 00:19:04,200 --> 00:19:08,200 from narrow. A is that do highly specific 324 00:19:08,200 --> 00:19:10,800 things recognize patterns and Achieve goals and very specific 325 00:19:10,800 --> 00:19:15,800 domains toward more General II's that can just deal with a 326 00:19:15,800 --> 00:19:18,700 broader scope of knowledge and a broader variety of goals and can 327 00:19:18,700 --> 00:19:22,400 transfer. What they've learned so far, the 328 00:19:22,408 --> 00:19:24,500 very different conditions. Missions. 329 00:19:24,500 --> 00:19:26,500 And their, I mean this will be really important. 330 00:19:26,500 --> 00:19:29,400 You see that would like self-driving cars now are 331 00:19:29,400 --> 00:19:32,800 crashing into people because they're seeing situations that 332 00:19:32,800 --> 00:19:35,200 weren't in their training data. I mean that's a failure to 333 00:19:35,900 --> 00:19:40,200 failure to generalize right? And then in financial markets 334 00:19:40,400 --> 00:19:43,200 when you have what's called a regime change, oh, suddenly the 335 00:19:43,200 --> 00:19:46,000 markets acting totally different than it was acting before. 336 00:19:46,600 --> 00:19:51,400 Well again current quantitative Financial prediction systems and 337 00:19:51,400 --> 00:19:53,800 risk management systems have failed to generalize right 338 00:19:53,800 --> 00:19:55,600 there. They were trained on previous 339 00:19:55,600 --> 00:19:58,200 Market regimes. When you give them a new market 340 00:19:58,200 --> 00:20:02,000 regime. No, they're still acting on 341 00:20:02,000 --> 00:20:03,700 their on their previous knowledge. 342 00:20:03,700 --> 00:20:07,800 Now of course most people can't deal with the new market regime 343 00:20:07,800 --> 00:20:10,900 either. But at least foundationally we 344 00:20:10,900 --> 00:20:15,200 do have the ability to like go back to basics and deal with it, 345 00:20:15,600 --> 00:20:20,400 radically new new situation and that's that's a big A big 346 00:20:20,400 --> 00:20:26,600 challenge facing the a, I filled in the next phase, which I think 347 00:20:27,300 --> 00:20:30,100 we're going to meet but it says, still some research challenges 348 00:20:30,100 --> 00:20:31,600 there. That's interesting. 349 00:20:31,600 --> 00:20:33,500 I've never considered that way that. 350 00:20:34,400 --> 00:20:40,000 Yeah I guess that like humans and carbon-based beings are good 351 00:20:40,000 --> 00:20:42,800 at recognizing certain types of patterns and I guess you could 352 00:20:42,800 --> 00:20:46,400 maybe differentiate. So like humans are good at 353 00:20:47,000 --> 00:20:50,000 recognizing certain path types of patterns and acting on them 354 00:20:50,000 --> 00:20:52,100 and that might be different from. 355 00:20:52,500 --> 00:20:55,400 Like for instance the intelligence of a dolphin or 356 00:20:55,408 --> 00:20:59,600 another type of carbon-based being and then artificial or 357 00:20:59,600 --> 00:21:04,000 computer-based intelligence or like silicon-based intelligence 358 00:21:04,300 --> 00:21:07,500 Might be good at recognizing patterns and like you said, like 359 00:21:07,500 --> 00:21:12,100 you multiple hundreds of side engines and figuring out actions 360 00:21:12,100 --> 00:21:17,300 based on what see serums kind of subtle if you think about it? 361 00:21:17,300 --> 00:21:22,000 Because, I mean, we involved in this domain of like discrete 362 00:21:22,000 --> 00:21:25,100 solid objects, like bouncing off each other and so on, right. 363 00:21:25,100 --> 00:21:29,600 And this probably led us to ideas about causation, but if 364 00:21:29,600 --> 00:21:33,700 you're a dolphin in the water you are seeing things flow 365 00:21:33,700 --> 00:21:35,600 around. Sound and blend into each other. 366 00:21:35,600 --> 00:21:38,100 You're not seeing some of these solid objects, bouncing off each 367 00:21:38,100 --> 00:21:42,500 other, that probably leads to a quite different world view now. 368 00:21:42,800 --> 00:21:45,900 Also like, where each of our minds is stuck in an individual 369 00:21:45,900 --> 00:21:49,200 body, her like our entire life until we die, right? 370 00:21:49,200 --> 00:21:53,600 And I mean, into barring a reincarnation and other freaky 371 00:21:53,600 --> 00:21:56,800 things at least 2 or 2 or first degree of approximation right 372 00:21:56,900 --> 00:21:59,800 now, if you're an AI, that can put yourself between different 373 00:21:59,800 --> 00:22:03,400 bodies or occupy, a hundred different bodies of the time or 374 00:22:03,500 --> 00:22:07,800 like, for Self and roll back to your last version before a 375 00:22:07,800 --> 00:22:11,200 traumatic experience like it. How does that change? 376 00:22:11,700 --> 00:22:13,800 Your whole outlook? What kind of patterns you look 377 00:22:13,800 --> 00:22:16,400 for what goals you bother to achieve, what risks you're 378 00:22:16,400 --> 00:22:19,400 willing to take, right? I mean I mean there's yeah 379 00:22:19,400 --> 00:22:22,600 there's so many ways that we're over fit to the exact 380 00:22:22,600 --> 00:22:26,500 environment we evolved in and the problems that we're trying 381 00:22:26,500 --> 00:22:29,800 to solve and then the other thing to realize like realize is 382 00:22:29,800 --> 00:22:33,900 we were stuck without root access to our brains and bodies. 383 00:22:34,200 --> 00:22:36,200 Which is, which is pretty terrible, right? 384 00:22:36,200 --> 00:22:40,000 Because I mean, if you're an AI, you can have like root Superuser 385 00:22:40,000 --> 00:22:43,700 access your own brain and body. If you think, well, I don't 386 00:22:43,700 --> 00:22:46,100 like, the way I react in this situation. 387 00:22:46,200 --> 00:22:48,400 Just go in and fix the damn bug, right? 388 00:22:48,400 --> 00:22:51,300 But we, we can. If you want to fix bugs in our 389 00:22:51,300 --> 00:22:55,800 self, it's like years and years of meditation, or therapy or 390 00:22:55,800 --> 00:22:59,700 reflection rather than just go in and change the Rogue piece of 391 00:22:59,700 --> 00:23:02,800 code, right? So, I mean, there there's a lot 392 00:23:02,800 --> 00:23:06,300 of A lot of things we take for granted. 393 00:23:06,300 --> 00:23:10,600 Now in terms of biases, we have our restrictions that we have 394 00:23:10,900 --> 00:23:15,400 which are not really intrinsic to being an intelligent mind, 395 00:23:15,400 --> 00:23:21,400 but but I mean they're just particularities of how we happen 396 00:23:21,400 --> 00:23:27,500 to evolve out of apes and in Africa, right? 397 00:23:27,500 --> 00:23:32,000 And I mean, that's in, in general, this is sort of why I 398 00:23:32,000 --> 00:23:35,900 like a uncle and conceptual view of things because how things 399 00:23:35,900 --> 00:23:39,700 evolved No, there's some fundamental reality to but 400 00:23:39,700 --> 00:23:41,500 there's a lot of historical contingency. 401 00:23:41,500 --> 00:23:45,400 I mean you see the same thing with exchange and money and so 402 00:23:45,400 --> 00:23:47,600 forth. I mean, people take so many 403 00:23:47,600 --> 00:23:50,200 things for granted about how economies work, which aren't 404 00:23:50,200 --> 00:23:54,300 necessarily intrinsic to the nature of, like, exchanging 405 00:23:54,300 --> 00:23:58,000 value in the community. They're just how things happen 406 00:23:58,000 --> 00:24:01,500 to evolve for quasi, random combination of reasons. 407 00:24:03,900 --> 00:24:06,500 This episode of epicenter is brought to you by Cosmos. 408 00:24:06,900 --> 00:24:09,700 The internet of blockchains we couldn't be more excited about 409 00:24:09,700 --> 00:24:12,300 the upcoming maenette launch, and to see so many projects 410 00:24:12,300 --> 00:24:15,200 already building on it. Blockchain Technologies are 411 00:24:15,200 --> 00:24:18,400 evolving fast and development shouldn't be one-size-fits-all 412 00:24:18,900 --> 00:24:20,900 as a dab. Developer you need the tools 413 00:24:20,900 --> 00:24:24,300 that will allow your dad to scale grow and evolve over time. 414 00:24:25,000 --> 00:24:28,200 The cosmos SDK is a user-friendly modular framework, 415 00:24:28,200 --> 00:24:30,200 which allows you to customize your dap to best suit your 416 00:24:30,200 --> 00:24:32,000 needs. It's powered by 10 or Marine 417 00:24:32,000 --> 00:24:35,500 Corps in advanced implementation of the bft proof of State 418 00:24:35,500 --> 00:24:38,000 protocol. Cosmos takes care of networking 419 00:24:38,000 --> 00:24:40,300 and consensus, and allows you to focus on building your 420 00:24:40,300 --> 00:24:42,100 application in your language of choice. 421 00:24:42,700 --> 00:24:45,800 If you're a smart contracts will be supported soon and the SDK 422 00:24:45,800 --> 00:24:48,200 makes it simple for you to connect to other blockchains in 423 00:24:48,200 --> 00:24:51,200 the cosmos network. If you have an idea for a dab, I 424 00:24:51,208 --> 00:24:54,300 would like to learn more about the cosmos SDK or if you'd like 425 00:24:54,300 --> 00:24:56,100 to connect your existing that the cosmos. 426 00:24:56,500 --> 00:25:00,100 Is it cause most dot Network / epicenter for epicenter 427 00:25:00,100 --> 00:25:01,700 listeners. The cosmos team will reach out 428 00:25:01,700 --> 00:25:03,900 to answer your questions and help you get started. 429 00:25:04,500 --> 00:25:07,000 We'd like to thank Cosmos for their support of epicenter. 430 00:25:08,100 --> 00:25:12,200 So let's talk about about about Ai and data. 431 00:25:12,700 --> 00:25:16,800 And this is a topic that has been brought a lot in the 432 00:25:16,800 --> 00:25:21,200 conversation about Ai and the fact that the AI needs large 433 00:25:21,200 --> 00:25:24,800 quantities as if ETA to dries off. 434 00:25:24,800 --> 00:25:26,300 And we kind of talked about this. 435 00:25:27,000 --> 00:25:32,600 So you've to today that data is very centralized data is held 436 00:25:32,600 --> 00:25:36,900 and owned by by a small number of, very large companies. 437 00:25:37,100 --> 00:25:42,200 You see the The problem, you know, type of repercussions that 438 00:25:42,200 --> 00:25:45,600 were unintended there, or is there a better system that you 439 00:25:45,600 --> 00:25:49,900 think we could achieve? Yeah, I mean the situation with 440 00:25:51,000 --> 00:25:57,100 the collection storage and use of data regarding human beings 441 00:25:57,100 --> 00:26:01,500 on the planet Now is really pretty ridiculous. 442 00:26:01,800 --> 00:26:05,500 I mean, it's not that it's necessarily entirely bad or 443 00:26:05,500 --> 00:26:10,800 malevolent some of its really good and useful but overall the 444 00:26:11,900 --> 00:26:16,600 ownership and control of data from the various sensors, we 445 00:26:16,600 --> 00:26:23,700 have everywhere is It's centralized in a pretty bizarre 446 00:26:23,700 --> 00:26:26,000 way. I mean some of its good, of 447 00:26:26,000 --> 00:26:29,700 course I could Google Maps. It's pretty nice and it's 448 00:26:29,700 --> 00:26:32,500 collecting data on where everyone's driving to so you can 449 00:26:32,500 --> 00:26:34,900 like see where there where there's a traffic jam, right. 450 00:26:34,900 --> 00:26:39,100 So these are these are very these are very useful functions 451 00:26:39,400 --> 00:26:42,900 and I don't really mind sharing location anonymously of where 452 00:26:42,900 --> 00:26:45,800 I'm driving with Google Maps so we can tell everyone else where 453 00:26:45,800 --> 00:26:48,500 there's a traffic jam, right? I mean that That seems like a 454 00:26:48,500 --> 00:26:54,900 fair exchange but I mean in in the end the agency regarding use 455 00:26:54,900 --> 00:26:58,400 of people's data is in a very confused state. 456 00:26:58,400 --> 00:27:01,500 So like they're this phone, I carry with me everywhere, right? 457 00:27:01,500 --> 00:27:06,400 There's a tremendous amount of data Coming through this phone 458 00:27:06,400 --> 00:27:11,200 onto the internet and it's all innocence my data, right? 459 00:27:11,200 --> 00:27:14,300 It's data about when I'm talking about who I'm talking to like 460 00:27:14,300 --> 00:27:18,700 where I am when I'm taking pictures of, but all this data 461 00:27:18,700 --> 00:27:21,300 coming for me through this device that I bought and then I 462 00:27:21,308 --> 00:27:24,300 pay a subscription to connect to the internet each month, like 463 00:27:24,500 --> 00:27:29,100 this data is going sort of haphazardly into various 464 00:27:29,100 --> 00:27:34,000 databases owned by various large corporations probably passing 465 00:27:34,000 --> 00:27:38,300 along to Various governments along the way and then this data 466 00:27:38,300 --> 00:27:42,100 is then being used, you know, for some useful things, like 467 00:27:42,100 --> 00:27:44,700 telling me when there's when there's a traffic jam, right? 468 00:27:44,700 --> 00:27:50,200 And and then it's being used to Advertise things to me, which 469 00:27:50,200 --> 00:27:53,000 doesn't matter to me much since. I've never I've never clicked on 470 00:27:53,000 --> 00:27:59,300 an ad in my life, I think. But, I mean, it's, it's, it's 471 00:27:59,300 --> 00:28:04,200 being used by big companies to make themselves money and 472 00:28:04,200 --> 00:28:07,500 increase their ability to manipulate people as a whole. 473 00:28:07,500 --> 00:28:10,700 Like, even if I double click on their heads by studying me along 474 00:28:10,700 --> 00:28:13,100 with everyone else, so learning, how to manipulate the minds of 475 00:28:13,100 --> 00:28:15,700 other people who do click on their ads, and do read their 476 00:28:15,700 --> 00:28:17,600 fake news. So then yeah, you got to Ask 477 00:28:17,600 --> 00:28:22,100 like, okay this data that comes from me through this device and 478 00:28:22,100 --> 00:28:24,400 I'm paying for and paying to connect to the internet. 479 00:28:24,400 --> 00:28:29,600 Like why, why isn't there an easy way for me to observe, what 480 00:28:29,600 --> 00:28:33,200 this data is being used for and have some agency over? 481 00:28:33,200 --> 00:28:36,200 What this data is is being used for, like it. 482 00:28:36,400 --> 00:28:39,800 If my data is being used, you know, to provide data to fuel 483 00:28:39,800 --> 00:28:42,500 someone's political campaign, I'd rather have it be used only 484 00:28:42,500 --> 00:28:44,600 for a candidate, I agree with or something, right. 485 00:28:44,600 --> 00:28:51,000 And, you know, it's Quite within our reach technologically to put 486 00:28:51,000 --> 00:28:54,400 agency over use of our data. Whether for a I are four basic 487 00:28:54,400 --> 00:29:00,800 statistics in the hands of the human being who produces that 488 00:29:00,900 --> 00:29:04,800 data. On the other hand, it's not in 489 00:29:04,800 --> 00:29:07,500 accordance with the business model of the large corporations 490 00:29:07,500 --> 00:29:10,800 involved in the phone and the internet services behind it. 491 00:29:11,500 --> 00:29:13,600 It's not in the interest of the business model. 492 00:29:13,600 --> 00:29:17,600 These corporations to provide that agency to the user except 493 00:29:17,600 --> 00:29:22,100 in so far as like government Regulators, force them to but of 494 00:29:22,100 --> 00:29:26,500 course, government regulators. Even when well intention which 495 00:29:26,500 --> 00:29:29,900 is only a fraction of the time. They can't keep up with the 496 00:29:30,400 --> 00:29:37,800 advances of the of Technology. Now this I mean this right now 497 00:29:38,500 --> 00:29:44,900 is mostly An inconvenience in this sort of aesthetic and moral 498 00:29:44,900 --> 00:29:48,600 in Felicity, but I mean, as As you move from their, I eye 499 00:29:48,600 --> 00:29:52,700 toward AGI. If it turns out that these 500 00:29:52,700 --> 00:29:57,300 stores of data, you know are critical for giving some parties 501 00:29:57,500 --> 00:30:00,500 a boost toward a GI more so than others, right? 502 00:30:00,500 --> 00:30:05,400 Then then this hoarding of data could actually have a more a 503 00:30:05,400 --> 00:30:09,600 more critical importance and then in principle blockchain and 504 00:30:09,600 --> 00:30:14,500 related Technologies give a way to circumvent these issues by 505 00:30:14,500 --> 00:30:18,200 putting each individual's data in some, you know, online 506 00:30:18,200 --> 00:30:21,900 repository or distributed decentralized repository, which 507 00:30:21,900 --> 00:30:25,500 is encrypted by their private key and then giving that 508 00:30:25,500 --> 00:30:30,500 individual agency over how the data is used and then there are 509 00:30:30,500 --> 00:30:33,400 fancy tools, like homomorphic encryption and multi-party 510 00:30:33,400 --> 00:30:38,100 computation which can be used to, you know, that a person give 511 00:30:38,500 --> 00:30:42,500 certain aspects of their data To certain other parties to use for 512 00:30:42,500 --> 00:30:44,300 certain things without giving all over the way. 513 00:30:44,300 --> 00:30:47,200 So, and yeah, in theory, the blockchain based decentralized, 514 00:30:47,200 --> 00:30:51,500 ecosystem, provides the technical tools and the sort of 515 00:30:51,500 --> 00:30:55,300 cultural oomph to the solve these problems. 516 00:30:55,300 --> 00:31:00,400 And on the other hand, the centralized Ecosystem, 517 00:31:00,500 --> 00:31:05,000 underlying, you know, big data and mobile phones and computers 518 00:31:05,000 --> 00:31:08,900 and embedded devices is as multiple trillion dollar 519 00:31:08,900 --> 00:31:13,100 companies, pushing things, pushing things forward. 520 00:31:13,500 --> 00:31:18,100 So there's a, the decentralized world has the right tools but 521 00:31:18,100 --> 00:31:19,800 the Big Challenge on their heads here. 522 00:31:20,800 --> 00:31:26,200 I'd love to speak a little bit about the concept of AI safety. 523 00:31:26,200 --> 00:31:30,900 And, and just to take a step back here, I guess, like you 524 00:31:30,908 --> 00:31:33,400 know what, what are some of the fear scenarios here? 525 00:31:33,400 --> 00:31:38,100 I so, so, like, let's say, fear scenario today is okay, AI is 526 00:31:38,100 --> 00:31:42,500 replacing all of these jobs, so people become, you know, 527 00:31:42,500 --> 00:31:47,800 unemployed maybe it leads to accumulation of resources for a 528 00:31:47,800 --> 00:31:50,400 few people, more and more. And you have this like East, 529 00:31:50,600 --> 00:31:54,100 Mini quality, right? Like that's like one Pearson are 530 00:31:54,100 --> 00:31:56,000 about a. I may be a different one. 531 00:31:56,000 --> 00:32:00,600 Is that then this AI starts to have its own objectives, becomes 532 00:32:00,600 --> 00:32:03,600 more and more powerful, gets more, and more resources and its 533 00:32:03,600 --> 00:32:07,900 objectives. Maybe it's hostile to humans or 534 00:32:07,900 --> 00:32:10,700 maybe it's like indifferent to humans. 535 00:32:11,100 --> 00:32:14,200 And so you have these potentially like bad outcomes, 536 00:32:14,200 --> 00:32:15,800 right? In that may be extreme 537 00:32:15,800 --> 00:32:19,800 inequality, maybe just have like human beings, becoming a kind 538 00:32:19,800 --> 00:32:22,900 of, you know, in inferior species being exploited. 539 00:32:23,300 --> 00:32:26,400 And then, you know, this is this is feel right? 540 00:32:26,400 --> 00:32:29,100 Like this idea of AI. Safety is like, what are your 541 00:32:29,100 --> 00:32:31,200 thoughts on it? You think this is an important 542 00:32:31,200 --> 00:32:33,700 field. Do you think efforts around AI 543 00:32:33,700 --> 00:32:37,900 safety or needed? People are certainly right to 544 00:32:37,900 --> 00:32:47,900 be. thinking about AI safety and really about the impacts and 545 00:32:47,900 --> 00:32:54,300 implications of AI for the advance of technology in the 546 00:32:54,300 --> 00:32:56,800 growth of humanity in general. Because I mean, looking at AI 547 00:32:56,800 --> 00:33:01,000 separately from politics and from all the other Tech 548 00:33:01,000 --> 00:33:04,800 connected with a, I probably probably doesn't make sense. 549 00:33:04,800 --> 00:33:09,000 So people are certainly right to be thinking and worrying about 550 00:33:09,000 --> 00:33:10,800 it. Now whether the things that 551 00:33:11,100 --> 00:33:13,600 people will do about it will have a positive or negative 552 00:33:13,600 --> 00:33:19,100 impact is a different question. And right, like I'm in bioethics 553 00:33:19,100 --> 00:33:24,000 is somewhat similar thing. And in general, it's easy to 554 00:33:24,000 --> 00:33:27,100 agree. We should be thinking somewhat 555 00:33:27,100 --> 00:33:31,000 about the ethics of, you know, genetic engineering in the bio 556 00:33:31,000 --> 00:33:35,700 hacking and so forth, I can I don't want people to create, 557 00:33:35,700 --> 00:33:40,700 like, weird say, artificial babies that have like a 558 00:33:41,000 --> 00:33:44,100 hypertrophied pain cortex. So they're just suffering and 559 00:33:44,100 --> 00:33:47,300 screaming with the billion times the level of suffering, any 560 00:33:47,300 --> 00:33:50,400 normal human can have, right? So I mean, clearly clearly 561 00:33:50,400 --> 00:33:54,300 there, some things that as a society, we just don't want 562 00:33:54,300 --> 00:33:57,800 people to bioengineer because they're just plain old nasty, 563 00:33:57,800 --> 00:33:59,600 and you're just creating suffering. 564 00:33:59,900 --> 00:34:04,400 On the other hand, in practice, the role of most This seems to 565 00:34:04,400 --> 00:34:08,800 be just to say no genetic engineering is bad. 566 00:34:08,800 --> 00:34:11,900 Don't make crispr babies. Don't upgrade your intelligence, 567 00:34:11,900 --> 00:34:15,400 right. So while in theory, yes, there 568 00:34:15,400 --> 00:34:20,199 are things that are just morally bad to do by essentially any 569 00:34:20,199 --> 00:34:24,900 human standard and we want to reflect on what to actually do 570 00:34:24,900 --> 00:34:26,500 and what not to do. Not all possible. 571 00:34:26,500 --> 00:34:30,400 Things should be done on the other hand in practice. 572 00:34:31,300 --> 00:34:39,800 bioethics seems very one-sidedly inclined to just push against 573 00:34:40,300 --> 00:34:44,100 advancing if humanity and new directions, and to push against 574 00:34:44,100 --> 00:34:47,699 reduction of suffering in favor of Maintenance of the status 575 00:34:47,699 --> 00:34:52,300 quo, when I would say most, most people who talk a lot about AI 576 00:34:52,300 --> 00:35:00,300 safety, Are not really thinking about how to maximize the odds 577 00:35:00,300 --> 00:35:04,000 of a beneficial outcome. But Humanity, All Things 578 00:35:04,000 --> 00:35:07,900 Considered, they more are thinking, like, how do we slow 579 00:35:07,900 --> 00:35:10,800 down AI development? Because we don't understand that 580 00:35:10,800 --> 00:35:14,300 and we're scared about it. So, I found myself disagreeing 581 00:35:14,300 --> 00:35:16,600 with almost everyone who's putting themselves out there is 582 00:35:16,600 --> 00:35:20,900 an AI safety Condit, but that doesn't mean I don't think AI 583 00:35:20,900 --> 00:35:23,200 safety is important. Like I don't want the Terminator 584 00:35:23,200 --> 00:35:25,100 2. Be roaming, roaming the streets, 585 00:35:25,100 --> 00:35:27,100 like a. I mean, I have four kids. 586 00:35:27,100 --> 00:35:30,900 I don't want an AI to be turning them into fuel or something, 587 00:35:30,900 --> 00:35:32,900 right? So a lot of us talk that about 588 00:35:32,900 --> 00:35:36,300 about, I think I sent the Hanson robotics quite quite well 589 00:35:36,300 --> 00:35:40,400 because with regards to has to Robotics and you guys have built 590 00:35:40,400 --> 00:35:45,200 this robot named Sophia that I'm sure most most of our listeners 591 00:35:45,200 --> 00:35:48,800 have seen, at least, at least once on the internet because 592 00:35:49,600 --> 00:35:53,100 she's had quite a few media appearances, she's been on Jimmy 593 00:35:53,100 --> 00:35:56,900 Fallon. It was a bunch of different 594 00:35:56,900 --> 00:35:59,700 conferences and what's the purpose of this robot? 595 00:35:59,700 --> 00:36:06,400 And how does it maybe its genetic Sophia indeed was partly 596 00:36:07,800 --> 00:36:12,600 Created and envisioned, as sort of an ambassador ambassador of a 597 00:36:12,600 --> 00:36:19,000 I love and love and compassion. And I think that's been 598 00:36:19,000 --> 00:36:22,900 interesting to see because David Henson, who's a good friend of 599 00:36:22,908 --> 00:36:26,400 mine. I mean, he, I've known them for 600 00:36:26,400 --> 00:36:29,700 a long time and I came to Hong Kong where I'm living now in 601 00:36:29,700 --> 00:36:32,900 2011, and he visited me here once. 602 00:36:32,900 --> 00:36:36,200 And I ended up convincing him to come here and move his company 603 00:36:36,200 --> 00:36:40,600 or introducing him to some. To help inject funding into his 604 00:36:40,600 --> 00:36:42,200 company here. So we've been talking about 605 00:36:42,200 --> 00:36:46,000 these things a long time and I think what's interesting is 606 00:36:46,600 --> 00:36:49,300 David is really a warm loving good-hearted person. 607 00:36:49,300 --> 00:36:52,900 He wanted to create a robot that would emanate love and 608 00:36:52,900 --> 00:36:56,500 compassion make people love it. So that it would, you know, 609 00:36:57,600 --> 00:37:01,100 build a positive relation between humans and robots, like 610 00:37:01,100 --> 00:37:03,600 proactively even before we have human level AGI. 611 00:37:03,600 --> 00:37:06,800 So that as a eyes and robots, get more and more generally 612 00:37:06,800 --> 00:37:10,000 intelligent That positive relationship is there on the 613 00:37:10,000 --> 00:37:16,500 other hand, David, he's an artist and as such he can't help 614 00:37:16,500 --> 00:37:22,600 himself from poking people and, and provoking controversy a 615 00:37:22,600 --> 00:37:25,500 little bit and making making things a little bit creepy. 616 00:37:25,500 --> 00:37:29,700 Sometimes just because he thinks that that looks that looks 617 00:37:29,700 --> 00:37:33,400 coolest. So I mean, I would say Sophia 618 00:37:33,400 --> 00:37:38,900 and all the Hanson robot sir, they're driven by David's desire 619 00:37:39,500 --> 00:37:41,900 to build a sort of compassionate, loving bond, 620 00:37:41,900 --> 00:37:47,500 between human and Ai, and robot. And also at some level driven by 621 00:37:47,500 --> 00:37:52,700 David's Sunny conscious artistic desire to poke at people a 622 00:37:52,707 --> 00:37:55,300 little bit and make and make them a little a little 623 00:37:55,300 --> 00:37:58,100 uncomfortable, right? And the these come together in 624 00:37:58,500 --> 00:38:01,600 an interesting way and I think that's good because you know my 625 00:38:02,100 --> 00:38:06,900 my emotional orientation is optimistic and positive. 626 00:38:06,900 --> 00:38:09,800 So I mean that my intuition and feeling is that the 627 00:38:09,800 --> 00:38:13,800 technological singularity is going to come out awesome and 628 00:38:13,800 --> 00:38:20,400 closer to utopic then dystopic. But I also think There's a 629 00:38:20,400 --> 00:38:22,700 fundamental uncertainty to all this. 630 00:38:22,700 --> 00:38:26,300 So people are are certainly Justified to feel a little bit 631 00:38:27,000 --> 00:38:31,000 uncomfortable and Confused. Like in in the end none of us 632 00:38:31,000 --> 00:38:33,500 knows what's going to happen. We're on the verge of creating 633 00:38:33,500 --> 00:38:37,600 machines that are, you know, 10 20 100 billion times more 634 00:38:37,600 --> 00:38:39,500 intelligent and capable than we are. 635 00:38:39,500 --> 00:38:43,900 And we'd be idiotic to believe we could predict and detail how 636 00:38:43,900 --> 00:38:46,500 this is going to Is going to come out. 637 00:38:46,500 --> 00:38:49,600 I mean, I, I find this irreducible uncertainty, 638 00:38:49,600 --> 00:38:54,100 beautiful, and exciting. And I see it as what Humanity 639 00:38:54,100 --> 00:38:55,600 has been doing since the beginning. 640 00:38:55,600 --> 00:38:58,200 Like, this is why we're less boring than cows and sheep, 641 00:38:58,200 --> 00:39:00,700 right? I mean, we decided not to remain 642 00:39:00,700 --> 00:39:05,000 monkeys and we invented language and fire and wheels and, and 643 00:39:05,000 --> 00:39:09,200 machines, and money, and Bitcoin and and Ai. 644 00:39:09,200 --> 00:39:13,900 And and a.g., I mean, that's the trajectory we're revolutionizing 645 00:39:13,900 --> 00:39:16,800 ourselves. Over and over and we never know 646 00:39:16,800 --> 00:39:18,500 what's. We never know what's going to 647 00:39:18,500 --> 00:39:21,100 happen next, right? And that's, that's part of the 648 00:39:21,100 --> 00:39:24,600 essence of what it is, totally human. 649 00:39:24,600 --> 00:39:29,400 And I think David, he baked some of that into Sophia, as an 650 00:39:29,400 --> 00:39:32,900 artist along with the love and compassion, which is quite cool. 651 00:39:34,200 --> 00:39:41,000 Yeah, and so probably many listeners have seen the TV show, 652 00:39:41,000 --> 00:39:46,600 Silicon Valley, so there is this kind of basically inspired by by 653 00:39:46,600 --> 00:39:49,000 you. And by Sofia the part there were 654 00:39:49,000 --> 00:39:53,100 basically, you know, someone Meant To Be. 655 00:39:53,100 --> 00:39:56,900 You is kind of playing this role but let's move to block chain. 656 00:39:56,900 --> 00:39:59,700 Now, when did you get interested in blockchain? 657 00:39:59,700 --> 00:40:03,300 And why did you think that blockchain had, you know, kind 658 00:40:03,300 --> 00:40:07,200 of It's to the Future course of AI. 659 00:40:09,000 --> 00:40:10,700 I've been. So I've been interested in 660 00:40:11,600 --> 00:40:15,400 crypto for a long time, like, since the early 90s knows doing 661 00:40:15,400 --> 00:40:20,300 that with finite fields and cryptography Tech, and that that 662 00:40:20,300 --> 00:40:22,500 seemed like it could potentially be important. 663 00:40:22,500 --> 00:40:26,500 Just politically in terms of stopping governments from having 664 00:40:26,500 --> 00:40:30,700 like, the ability to spy on everyone's information, and keep 665 00:40:30,700 --> 00:40:32,700 it uniquely uniquely for themselves. 666 00:40:33,500 --> 00:40:36,700 Bitcoin. I didn't like because it just 667 00:40:36,700 --> 00:40:39,900 because proof-of-work annoyed, me, just heats up the 668 00:40:39,900 --> 00:40:42,200 environment and waste energy unnecessarily. 669 00:40:42,200 --> 00:40:46,100 So I didn't get involved in that one if barium came out. 670 00:40:46,100 --> 00:40:49,300 That's the first thing where I thought well this this is 671 00:40:49,300 --> 00:40:53,500 actually cool like it did it did use proof of work but you can 672 00:40:53,500 --> 00:40:57,200 see there was a will and a path to going beyond that and then 673 00:40:57,200 --> 00:41:00,500 you had celebrity I mean you have a scripting language which 674 00:41:00,500 --> 00:41:03,100 basically lets you create this, you know? 675 00:41:03,300 --> 00:41:08,300 Secure and encrypted decentralized World computer and 676 00:41:08,300 --> 00:41:12,700 I thought I thought it fairy. Mm was a vision in the right 677 00:41:12,700 --> 00:41:15,600 direction and it was a reasonable software tool set 678 00:41:15,600 --> 00:41:19,100 although obviously immature at first and not not that mature 679 00:41:19,100 --> 00:41:21,600 still. So, once in theory, mm came out, 680 00:41:21,600 --> 00:41:26,700 I started really thinking like, how do we use this to create 681 00:41:26,900 --> 00:41:30,100 like a decentralized, global AI Network? 682 00:41:30,100 --> 00:41:33,400 Because that in 2001, I published a book called Eating 683 00:41:33,400 --> 00:41:36,500 internet intelligence, which Envision the, you know, a 684 00:41:36,500 --> 00:41:40,500 decentralized, Global Network of AI is coming together as a 685 00:41:40,500 --> 00:41:47,700 society of mind and before that in in 95, I posted some web 686 00:41:47,700 --> 00:41:50,500 pages claiming I was going to run for US, president on the 687 00:41:50,500 --> 00:41:55,100 decentralization party platform, which I ended up not doing 688 00:41:55,100 --> 00:41:58,000 because I realized in time with a terrible job, it would be to 689 00:41:58,000 --> 00:42:02,400 be president anyway, but I'm in the these ideas were interesting 690 00:42:02,400 --> 00:42:05,000 me for a long time. Book decentralized control 691 00:42:05,000 --> 00:42:09,000 politically because I always had a sort of anarchist socialist 692 00:42:09,000 --> 00:42:13,300 bent and then the idea of making a decentralized, global AI 693 00:42:13,300 --> 00:42:17,600 Network, like Marvin Minsky's Society of minds, but an economy 694 00:42:17,600 --> 00:42:20,300 of Minds when they eyes are paying each other for work and 695 00:42:20,300 --> 00:42:23,400 there's collective intelligence coming out of the of the whole 696 00:42:23,400 --> 00:42:25,400 network beyond the intelligence in the parts. 697 00:42:25,400 --> 00:42:29,900 But it Syrian seem like a critical step forward toward 698 00:42:29,900 --> 00:42:35,800 having a tool set that that would Let you do this and so 699 00:42:35,800 --> 00:42:40,000 that then as soon as if their aim was there, yet the idea of 700 00:42:40,800 --> 00:42:45,000 dows decentralized autonomous organizations which again 701 00:42:45,000 --> 00:42:47,400 they've been spelled out in science fiction like in Charlie 702 00:42:47,400 --> 00:42:50,000 straws his book accelerometer and a bunch of others. 703 00:42:50,000 --> 00:42:53,000 But with the solidity programming language, like, wow, 704 00:42:53,000 --> 00:42:55,600 you can script a dow in a short script, right? 705 00:42:55,600 --> 00:43:00,300 That's that's similar feeling to how when I first learned Java in 706 00:43:00,300 --> 00:43:03,100 like 1995. It's like, wow, you can create 707 00:43:03,300 --> 00:43:05,600 Web page or send an email with this much code. 708 00:43:05,700 --> 00:43:07,800 That's power, right? Solidity was like that. 709 00:43:07,800 --> 00:43:10,800 Like, you can, you can create a decentralized corporation and 710 00:43:10,800 --> 00:43:14,100 just a little bit of code. It's not the perfect language, 711 00:43:14,100 --> 00:43:19,300 just like Java wasn't but I mean it it really opened the door and 712 00:43:19,500 --> 00:43:22,500 once I saw how a fairy, mm worked, I started thinking, 713 00:43:22,500 --> 00:43:26,000 well, how do we put this together with, for example, 714 00:43:26,000 --> 00:43:31,100 opencog which is my open source, AI platform aimed at general 715 00:43:31,100 --> 00:43:33,600 intelligence? Or, you know, Rooted neural 716 00:43:33,600 --> 00:43:36,600 networks and genetic algorithms or whatever other type of AI. 717 00:43:36,600 --> 00:43:40,200 It seemed clear you could use ethereum as a basis for 718 00:43:40,200 --> 00:43:43,300 connecting together, many different a inodes and to some 719 00:43:43,300 --> 00:43:47,200 sort of decentralized Hey, I mind. 720 00:43:47,200 --> 00:43:50,600 And then this logically, this should be able to kick the asses 721 00:43:50,600 --> 00:43:55,100 of Google Amazon, 10 cents in the IBM and all these big 722 00:43:55,100 --> 00:43:58,400 companies by, you know, the power of decentralized 723 00:43:58,900 --> 00:44:02,100 community. And then when when I met Simone 724 00:44:02,100 --> 00:44:08,000 a jackal Mellie who was later to co-found Singularity net with me 725 00:44:08,000 --> 00:44:11,100 and he had, he had a blockchain development team in Italy and 726 00:44:11,100 --> 00:44:14,100 he'd been helping out a host of different blockchain project. 727 00:44:14,100 --> 00:44:18,400 So when I met Simona, Hey, who was really conversant with the 728 00:44:18,400 --> 00:44:20,900 blockchain world, both technically and on the business 729 00:44:20,900 --> 00:44:23,200 level, then we sort of put our heads together. 730 00:44:23,900 --> 00:44:28,600 And we like roughed out what became the singular unit design 731 00:44:28,600 --> 00:44:33,400 and then So in moving toward the initial token cell, and then 732 00:44:33,400 --> 00:44:36,700 David Henson already was a close friend. 733 00:44:36,800 --> 00:44:39,500 I mean, he he saw the vision immediately. 734 00:44:39,500 --> 00:44:42,000 Our first meetings on this were in the Hanson, robotics office, 735 00:44:42,000 --> 00:44:45,800 in Hong Kong and David saw this as a way to get like a 736 00:44:45,800 --> 00:44:49,700 decentralized global robot mind Cloud behind his robots because 737 00:44:49,700 --> 00:44:52,500 you always knew the intelligence is going to be in Sofia, right? 738 00:44:52,500 --> 00:44:56,600 I mean some is about seeing and moving but the cognitive Parts, 739 00:44:56,600 --> 00:44:59,300 the long-term knowledge, you're going to be in the cloud, but 740 00:44:59,300 --> 00:45:00,600 what? Cloud, right? 741 00:45:00,700 --> 00:45:03,300 Do you have a million robots around the world and all the 742 00:45:03,300 --> 00:45:06,200 intelligence is running an Amazon's cloud or a teasing like 743 00:45:06,200 --> 00:45:10,700 Microsoft Azure API or do you have like a decentralized mind 744 00:45:10,700 --> 00:45:13,200 Cloud that's owned and controlled by all the people who 745 00:45:13,200 --> 00:45:14,600 are buying these robots. Right. 746 00:45:14,600 --> 00:45:18,700 So that was David was seeing it as a robot mind Club, but it was 747 00:45:18,700 --> 00:45:21,100 really the same thing that Simona and I were seeing with a 748 00:45:21,100 --> 00:45:24,000 decentralized blockchain based AI mind. 749 00:45:25,400 --> 00:45:26,400 Okay. Okay. 750 00:45:27,000 --> 00:45:31,800 So, would it be fair to kind of characterize this, as, you know, 751 00:45:31,800 --> 00:45:35,600 you see this to reject or uses AI coming, but then the question 752 00:45:35,600 --> 00:45:38,700 is, yeah, where do those are? Is coordinate, where do they 753 00:45:38,700 --> 00:45:41,000 share information? What kind of substrate with a 754 00:45:41,000 --> 00:45:44,000 run on? And, of course, if you get it 755 00:45:44,000 --> 00:45:47,800 today, but it will be mostly controlled by companies like 756 00:45:47,800 --> 00:45:52,800 Google and Facebook and then with something like Singularity 757 00:45:52,800 --> 00:45:57,300 net, they could be Be, you know, kind of open decentralized 758 00:45:57,900 --> 00:46:03,300 transparent accessible Democratic platform where, you 759 00:46:03,300 --> 00:46:07,000 know, a ice could coordinate a ice can share data, actually 760 00:46:07,000 --> 00:46:10,700 evolved. That's kind of the yeah. 761 00:46:11,600 --> 00:46:14,200 That's right. So there I mean as I've said 762 00:46:14,200 --> 00:46:18,100 before, what really excited me about the singularity that 763 00:46:18,100 --> 00:46:23,500 design inversion is seeing that two different goals which are 764 00:46:23,500 --> 00:46:25,600 very important. We converge into one. 765 00:46:25,600 --> 00:46:30,000 So, what one goal is to make sort of a venue for many 766 00:46:30,000 --> 00:46:34,300 different AI components to join forces to make it Collective. 767 00:46:34,300 --> 00:46:37,300 AI mine with a whole is greater than the sum of the parts. 768 00:46:37,300 --> 00:46:41,000 So you may have it one AI that like uses our opencog algorithm 769 00:46:41,000 --> 00:46:44,800 to generalize an abstract and reason you can have another AI 770 00:46:44,800 --> 00:46:46,700 that recognize patterns in DNA data. 771 00:46:46,700 --> 00:46:49,400 Another AI that uses deep neural Nets, recognize patterns, and 772 00:46:49,400 --> 00:46:53,100 visual data, and you connect them all together into a mind 773 00:46:53,100 --> 00:46:54,800 that self-organizes. Adapt. 774 00:46:54,800 --> 00:46:58,400 So the AG I could be in the whole network, not in any one 775 00:46:58,400 --> 00:47:02,100 particular node in the network. And then the other thing is 776 00:47:02,700 --> 00:47:05,800 okay, but if we're going to have this network of a eyes like who 777 00:47:05,808 --> 00:47:10,600 controls that that network, is it all sitting inside Google or 778 00:47:10,600 --> 00:47:15,300 Amazon, or is it just more like the internet, right? 779 00:47:15,300 --> 00:47:16,800 Which is not controlled by anyone. 780 00:47:16,800 --> 00:47:20,200 It's a network of networks, which is controlled by the 781 00:47:20,200 --> 00:47:25,400 different participants, right? And so, it seemed Use blockchain 782 00:47:25,400 --> 00:47:29,800 to achieve both of these goals to make a network of a eyes. 783 00:47:29,800 --> 00:47:32,800 It's controlled by the participants in the sort of 784 00:47:32,808 --> 00:47:37,200 democratic self-organizing, and open way and also make it so 785 00:47:37,200 --> 00:47:42,100 that the design encourages the a eyes to collaborate with the 786 00:47:42,200 --> 00:47:45,400 with each other and join federations with collective 787 00:47:45,400 --> 00:47:49,600 intelligence and so forth. And so this Yeah. 788 00:47:49,700 --> 00:47:52,700 Of course it's easier said than done but I mean we we did the 789 00:47:52,700 --> 00:47:56,900 initial token sale for this December 2017. 790 00:47:57,300 --> 00:48:02,300 We're launching like the initial beta version of the platform. 791 00:48:03,400 --> 00:48:06,700 The end of this month, the end of February. 792 00:48:06,700 --> 00:48:10,100 After a symbol Alpha was launched September 2017. 793 00:48:10,100 --> 00:48:14,300 And then, during 2019 post the February launch of the beta. 794 00:48:14,800 --> 00:48:17,300 We're going to add more and more and more features to the 795 00:48:17,300 --> 00:48:20,700 network. As well as adding more and more 796 00:48:20,700 --> 00:48:23,800 of our own, our own AI into the network. 797 00:48:23,800 --> 00:48:27,000 And they're, you know, there's a huge, the biggest part of the 798 00:48:27,000 --> 00:48:30,600 struggle remains in the future because, I mean, we have a beta 799 00:48:30,600 --> 00:48:33,200 version of the platform. We have some nice AI we've put 800 00:48:33,200 --> 00:48:37,600 in there, but still, you know, our competitors are trillion 801 00:48:37,600 --> 00:48:41,600 dollar companies with humongous server, server farms, and not 802 00:48:41,800 --> 00:48:45,400 Amazon has 10,000 people working on Alexa, right? 803 00:48:45,400 --> 00:48:48,500 So so to counteract that we need, not only a good Good 804 00:48:48,508 --> 00:48:52,500 design and smart AI. We need to attract a developer 805 00:48:52,500 --> 00:48:57,500 and user Community, which is even bigger and better than than 806 00:48:57,500 --> 00:49:00,100 the armies of highly paid employees at these big tech 807 00:49:00,100 --> 00:49:04,800 companies have and this I'm in. This is one of the reasons why 808 00:49:05,300 --> 00:49:09,100 I'm happy to talk to you guys and your audience because 809 00:49:09,100 --> 00:49:12,900 getting a community crystallized around decentralized. 810 00:49:12,900 --> 00:49:20,400 AI is absolutely critical to really making the Centralized AI 811 00:49:20,800 --> 00:49:25,300 Vision happen. Hiring a stressful. 812 00:49:25,700 --> 00:49:28,000 Let's face it. It's a long process of sifting 813 00:49:28,000 --> 00:49:30,400 through resumes and interviewing candidates, without any 814 00:49:30,400 --> 00:49:32,900 guarantee of quality but it doesn't have to be this way. 815 00:49:33,100 --> 00:49:36,000 Companies all over the place are experiencing a new way of hiring 816 00:49:36,000 --> 00:49:37,600 with top. Tell if you go to their 817 00:49:37,600 --> 00:49:40,200 trustpilot page, you'll see that of the hundreds of people that 818 00:49:40,200 --> 00:49:43,600 have left reviews, over 98% were four or five star ratings, 819 00:49:43,900 --> 00:49:47,600 including one guy who wants to give his developer bear hug that 820 00:49:47,600 --> 00:49:49,700 says a lot. Top dog, gets all this great 821 00:49:49,700 --> 00:49:51,500 feedback because they focus on their clients. 822 00:49:51,700 --> 00:49:53,100 And their top priority is quality. 823 00:49:53,600 --> 00:49:56,800 The only Except the top three percent of applicants including 824 00:49:57,100 --> 00:49:58,600 highly-skilled blockchain engineers. 825 00:49:59,700 --> 00:50:02,500 One of these Engineers is Roddick, ostrowski, Roddick has 826 00:50:02,500 --> 00:50:05,200 experience as a lead software, engineer and data scientists for 827 00:50:05,200 --> 00:50:08,100 Sony and Expedia, then he discovered blockchain, and he 828 00:50:08,100 --> 00:50:10,400 became totally consumed with etherium. 829 00:50:10,900 --> 00:50:13,800 He worked as a consultant. For the firm start on chain and 830 00:50:13,800 --> 00:50:16,900 is time-lock tap when the top quarter consensus, you port an 831 00:50:16,900 --> 00:50:21,100 identity blockchain hackathon, then he expanded his reach 832 00:50:21,100 --> 00:50:22,600 through top towel. He worked with a bunch of 833 00:50:22,600 --> 00:50:26,000 clients on projects, such as smart, contract development and 834 00:50:26,000 --> 00:50:28,800 a POC that leverages blockchain if you want to hire Engineers 835 00:50:28,800 --> 00:50:31,700 like Roddick. Your team go to top tile.com 836 00:50:31,900 --> 00:50:34,800 epicenter for a no risk trial a top tile. 837 00:50:34,800 --> 00:50:37,600 Director of engineering will deliver your next higher in as 838 00:50:37,600 --> 00:50:40,500 fast as 48 hours and you'll get a thousand dollar credit when 839 00:50:40,500 --> 00:50:43,500 you decide to hire, we'd like to thank top towel for their 840 00:50:43,500 --> 00:50:47,100 support of epicenter. Let's go a little bit in their 841 00:50:47,100 --> 00:50:49,900 van, you know, seeing where you need and what that looks like. 842 00:50:49,900 --> 00:50:52,400 So you can you speak about, you know, what are the different 843 00:50:52,400 --> 00:50:56,100 kind of components of the system and you mentioned, you know, 844 00:50:56,100 --> 00:51:00,700 developers getting involved. Let's say now, there is some AI 845 00:51:00,900 --> 00:51:03,400 algorithms, you are /. Like, how would they? 846 00:51:03,400 --> 00:51:05,800 How would an interaction with Singularity net. 847 00:51:05,800 --> 00:51:11,600 Look like if you have an AI algorithm integrating it with 848 00:51:11,600 --> 00:51:14,200 singularity in it is actually not. 849 00:51:15,200 --> 00:51:17,900 Especially difficult. I mean it's a container based 850 00:51:17,900 --> 00:51:20,100 system like most Cloud systems now. 851 00:51:20,100 --> 00:51:26,200 So you you put your, your AI in a Docker container or lxc 852 00:51:26,200 --> 00:51:31,300 container and, and, and then there's a simple API to 853 00:51:31,300 --> 00:51:35,600 integrate it with, which then lets, that's your a, I accept 854 00:51:35,600 --> 00:51:39,000 payment for services in our AG a cryptographic token. 855 00:51:39,000 --> 00:51:43,000 Let's it and then announce what API it wants to use to get data 856 00:51:43,000 --> 00:51:46,600 and queries. And then it can give responses 857 00:51:46,600 --> 00:51:51,100 and in Json or whatever API it wants. 858 00:51:51,100 --> 00:51:55,100 So it's really just it's a system of containers and then 859 00:51:55,100 --> 00:51:58,900 there's a payment system using a token and I meant for cases 860 00:51:58,900 --> 00:52:02,100 where an AI outsources work to another AI? 861 00:52:02,100 --> 00:52:03,800 Which outsources work to another AI? 862 00:52:03,800 --> 00:52:08,900 There's a there's a multi-party escrow framework on the back end 863 00:52:08,900 --> 00:52:13,100 and there's a system that allows a lot of AI to AI transactions 864 00:52:13,100 --> 00:52:15,900 to occur off. The blockchain for Speed 865 00:52:15,900 --> 00:52:19,300 purposes, but all that's really behind the scenes. 866 00:52:19,300 --> 00:52:23,100 I mean, from the point of view of an AI developer, it's really 867 00:52:23,100 --> 00:52:25,500 pretty simple. To put your AI in the container 868 00:52:25,900 --> 00:52:31,400 and take, I mean, 15 minutes to two or three hours to integrate 869 00:52:31,400 --> 00:52:36,000 integrate with the singularity net wrappers. 870 00:52:36,000 --> 00:52:39,500 So I get that right? So I put my a algorithm into 871 00:52:39,500 --> 00:52:41,000 Docker. Container kind of make it 872 00:52:41,000 --> 00:52:43,200 accessible through Singularity net. 873 00:52:43,400 --> 00:52:47,000 Let's say now I'm On the other hand, somebody I have a bunch of 874 00:52:47,000 --> 00:52:51,200 data, I would love to get a better understanding of maybe 875 00:52:51,800 --> 00:52:53,700 what's actionable, what it means. 876 00:52:54,100 --> 00:52:57,600 So could I didn't go and basically say, you know, kind of 877 00:52:57,600 --> 00:53:02,200 hire the services of different of these AI algorithms to get me 878 00:53:02,200 --> 00:53:03,200 results. Yeah. 879 00:53:03,300 --> 00:53:10,600 So I think the decentralized protocol Could actually be used 880 00:53:10,600 --> 00:53:13,900 by anyone. And I mean we use 881 00:53:13,900 --> 00:53:17,000 behind-the-scenes a component called drizzle which allows like 882 00:53:17,000 --> 00:53:21,800 decentralized search of any network of ethereum node. 883 00:53:21,800 --> 00:53:27,900 So I mean you could, if you are a reasonable scripture, I mean, 884 00:53:27,900 --> 00:53:31,300 you could just put out script your own query to go search, the 885 00:53:31,300 --> 00:53:35,300 whole network, and find any AI that broadcast sets able to do 886 00:53:35,300 --> 00:53:37,500 the kind of thing that that you want. 887 00:53:37,500 --> 00:53:40,900 Now, on the other hand, we're making it easier than that. 888 00:53:40,900 --> 00:53:44,200 So along with the beta we're launching just a Marketplace 889 00:53:44,200 --> 00:53:46,900 user interface which is a website and you can go to that 890 00:53:46,900 --> 00:53:52,200 website and can see what AI services are listed and and what 891 00:53:52,200 --> 00:53:56,900 sorts of things they do and you see their addresses and and so 892 00:53:56,900 --> 00:54:01,700 forth. So I mean that's Right now in 893 00:54:01,700 --> 00:54:04,200 practice that's a bit centralized, right? 894 00:54:04,200 --> 00:54:08,200 Because I mean, we make this web interface which lists a bunch of 895 00:54:09,600 --> 00:54:13,100 AI on there. And I mean, we are legally 896 00:54:13,100 --> 00:54:16,000 liable for what we list there. So we have to do some wedding of 897 00:54:16,000 --> 00:54:18,100 what we allow on there, just like the Google Play Store, 898 00:54:18,100 --> 00:54:20,700 those or something. On the other hand, the 899 00:54:20,700 --> 00:54:24,800 underlying protocol is completely decentralized and 900 00:54:24,800 --> 00:54:29,200 open. So I mean for example if Like 901 00:54:29,200 --> 00:54:33,600 we're Incorporated Singularity net Foundation, which is 902 00:54:33,600 --> 00:54:36,500 building the singularity Network now is Incorporated in the 903 00:54:36,500 --> 00:54:40,200 Netherlands. So suppose that Netherlands law 904 00:54:40,800 --> 00:54:44,200 said, we weren't allowed to list on our user interface, you know, 905 00:54:44,200 --> 00:54:49,000 in AI based in Iran or North Korea or something, then we have 906 00:54:49,000 --> 00:54:52,300 to take that off our interface. On the other hand, the 907 00:54:52,300 --> 00:54:55,500 decentralized network is whatever it is, right? 908 00:54:55,500 --> 00:54:58,300 So someone someone in Iran, could build another interface, 909 00:54:58,300 --> 00:55:02,900 which is Like an interface to all the Iranian and North Korean 910 00:55:02,900 --> 00:55:05,000 AI nodes on the network or something, right? 911 00:55:05,000 --> 00:55:08,400 Because the so this is the beauty of this architecture. 912 00:55:08,400 --> 00:55:12,700 You have this decentralized protocol which is controlled by 913 00:55:12,700 --> 00:55:17,400 no one and anyone can put an AI online and then it just 914 00:55:17,400 --> 00:55:20,100 announces it to their to the other a eyes in that Network. 915 00:55:20,100 --> 00:55:23,700 And then it can be found by decentralized peer to peer peer, 916 00:55:23,700 --> 00:55:25,500 to peer interaction. So that's there. 917 00:55:25,500 --> 00:55:27,300 Which gives a lot of robustness to it. 918 00:55:27,500 --> 00:55:32,600 On the other hand for easy, Views were putting up a simple 919 00:55:33,100 --> 00:55:37,800 website, which just lists lists the AIS that are that are on 920 00:55:37,800 --> 00:55:39,900 there, which then can be interacted with from a 921 00:55:39,900 --> 00:55:43,200 customer's view. Just like you're getting AI as a 922 00:55:43,200 --> 00:55:46,600 service from any other directory or somebody's website. 923 00:55:46,600 --> 00:55:52,200 Now that the beta still has some limitations, in the sense that 924 00:55:52,900 --> 00:55:56,100 we accept payment in the beta, only in our API token, which is 925 00:55:56,100 --> 00:55:59,400 near see, 20 token, then one of the things, We're going to do in 926 00:55:59,400 --> 00:56:04,400 the months after the beta, is integrate, third-party, no Fiat 927 00:56:04,400 --> 00:56:09,700 to crypto payment system because, of course, most most 928 00:56:10,200 --> 00:56:13,800 companies who want to use AI inside their, their website of 929 00:56:13,800 --> 00:56:16,800 their product, not from the crypto space. 930 00:56:16,800 --> 00:56:19,800 They don't want to deal with trip to wallets and so forth 931 00:56:19,800 --> 00:56:24,200 until this point. But this This isn't isn't a 932 00:56:24,207 --> 00:56:27,400 really big obstacle. It's just something we hadn't 933 00:56:27,500 --> 00:56:29,800 hadn't done. Use its more regulatory thing 934 00:56:29,800 --> 00:56:31,700 than the than the hard technical problem. 935 00:56:32,800 --> 00:56:37,500 So you mentioned AI, HEI token. So what's the role of the token? 936 00:56:38,800 --> 00:56:45,500 Well, the token is used by a eyes to pay other a eyes for 937 00:56:45,500 --> 00:56:50,700 services that they provide, but having our own token economy, 938 00:56:51,500 --> 00:56:57,300 lets us nudge the incentive mechanisms in an interesting 939 00:56:57,300 --> 00:57:00,300 way. So as well as using it for 940 00:57:00,300 --> 00:57:04,000 payment of one day, I buy another AI. 941 00:57:04,000 --> 00:57:11,900 We We also will issue token bounties as rewards for people 942 00:57:11,900 --> 00:57:17,200 who contribute a eyes that are requested by the community. 943 00:57:17,200 --> 00:57:20,900 And then We will Implement later this year. 944 00:57:20,900 --> 00:57:26,500 A curation market, where if I want to rate your AI is good. 945 00:57:26,800 --> 00:57:30,700 One way I can do that is to stake some token on your AI and 946 00:57:30,700 --> 00:57:34,600 then if you're AI comes out to be rated goodbye a lot. 947 00:57:34,700 --> 00:57:37,200 Diversity of other people, I'll get some reward. 948 00:57:38,000 --> 00:57:41,700 Whereas, if you're AI turns out to be horrible, then the number 949 00:57:41,700 --> 00:57:44,600 lose, I will lose some of what I, some of what I staked. 950 00:57:44,900 --> 00:57:50,200 So, having our own token, it's both an A secure and private way 951 00:57:50,200 --> 00:57:53,200 to do transactions and it lets us do things with you. 952 00:57:53,200 --> 00:57:57,300 No bounties for development and staking and curation markets, 953 00:57:57,600 --> 00:58:03,300 which I think can sculpt and guide the economy of a eyes. 954 00:58:03,300 --> 00:58:07,700 And this is, this is quite important, because there's, 955 00:58:08,100 --> 00:58:11,300 there's something in AI in cognitive science, called the 956 00:58:11,300 --> 00:58:14,400 assignment of credit problem. Which is when, when you have a 957 00:58:14,400 --> 00:58:18,000 complex network of Agents cooperating to do some function, 958 00:58:18,600 --> 00:58:22,000 I mean, how do you ensure that the agents like deep in the 959 00:58:22,008 --> 00:58:24,800 bowels of the network that indirectly helped achieve the 960 00:58:24,800 --> 00:58:29,000 function are actually getting getting rewarded, right? 961 00:58:29,000 --> 00:58:31,500 And the human brain somehow does this, right? 962 00:58:31,500 --> 00:58:35,800 Like if you do something that gets you food or sex or money or 963 00:58:35,800 --> 00:58:39,800 intellectual satisfaction, whatever is good, you know that 964 00:58:39,800 --> 00:58:43,700 the neurons that moved your arms and legs don't get all the 965 00:58:43,700 --> 00:58:45,600 reward, right? There's reward that goes to the 966 00:58:45,607 --> 00:58:48,500 neurons deep in your brain that help you get whatever those 967 00:58:48,800 --> 00:58:54,100 Goodies were the US economy. For example, doesn't do so good, 968 00:58:54,100 --> 00:58:58,300 a job of assigning credit internally, which is why Bankers 969 00:58:58,300 --> 00:59:02,100 make so much more money than programmers or kindergarten 970 00:59:02,100 --> 00:59:03,900 teachers or artists or something, right? 971 00:59:03,900 --> 00:59:10,500 So we and arguably the Bitcoin and etherium economies, although 972 00:59:10,700 --> 00:59:12,400 they're really cool and some ways. 973 00:59:12,700 --> 00:59:15,100 I mean, there's a strong tendency towards like, all the 974 00:59:15,100 --> 00:59:18,900 god, Polly and oligarchy in these economies and the They 975 00:59:18,900 --> 00:59:22,200 don't necessarily do a brilliant job of assigning credit to 976 00:59:23,100 --> 00:59:27,100 genuine value either. So by making our own token on 977 00:59:27,100 --> 00:59:31,000 make economy and sculpting the reward system in it. 978 00:59:31,000 --> 00:59:37,400 We hope to make the economy of a eyes operate better than other 979 00:59:37,400 --> 00:59:40,900 existing economies so that the AI is really contributing most 980 00:59:40,900 --> 00:59:44,000 to the overall Network and its intelligence and the value it 981 00:59:44,008 --> 00:59:46,500 delivers. The AI is contributing most to 982 00:59:46,500 --> 00:59:50,800 the overall Network are Actually getting rewarded significantly. 983 00:59:50,800 --> 00:59:55,500 And that this is a hard problem. We're like economic design 984 00:59:55,600 --> 00:59:58,600 matches cognitive science, right? 985 00:59:58,600 --> 01:00:01,000 It's these are fairly subtle things. 986 01:00:01,500 --> 01:00:03,100 Yeah. They just thought I would ask 987 01:00:03,100 --> 01:00:06,400 you this. So if you take something like 988 01:00:06,900 --> 01:00:10,300 I'm not sure if you're familiar with the system, the blockchain 989 01:00:10,300 --> 01:00:13,000 true bit so it's like a distributed computation 990 01:00:13,200 --> 01:00:16,100 blockchain and just distributed computation systems have been 991 01:00:16,100 --> 01:00:19,300 around for years and more recently, you know, People have 992 01:00:19,900 --> 01:00:23,600 embedded them in with watching system so that you can have this 993 01:00:23,600 --> 01:00:27,000 reward mechanism. And so with virtru, bit you, you 994 01:00:27,000 --> 01:00:33,300 have these actors of the network who watch for, who, who can 995 01:00:33,300 --> 01:00:37,300 potentially validate or verify the computations. 996 01:00:37,500 --> 01:00:39,900 And so therefore, there's an incentive to those providing the 997 01:00:39,900 --> 01:00:42,800 computations to provide correct computations because they have a 998 01:00:44,200 --> 01:00:47,200 there's the possibility of been getting slashed if they don't 999 01:00:47,200 --> 01:00:50,500 provide accurate computations Now this is for computations 1000 01:00:50,500 --> 01:00:54,200 that are somewhat trivial to to achieve with a given 1001 01:00:54,200 --> 01:00:58,800 general-purpose computers. But with with a I if I send some 1002 01:00:58,800 --> 01:01:02,200 tasks to an AI and it it returns a result. 1003 01:01:02,200 --> 01:01:08,900 It returns some some sort of data set how can I as a user or 1004 01:01:08,900 --> 01:01:13,600 even other users of the network verify that and also I think 1005 01:01:13,600 --> 01:01:17,500 with AI there might even with Jen general intelligence 1006 01:01:17,500 --> 01:01:23,900 becoming Feel closer to reality a ice could have their own kind 1007 01:01:23,900 --> 01:01:26,500 of subject, subjective bias, perhaps. 1008 01:01:26,900 --> 01:01:30,500 And so patient may be different between one and II in another, 1009 01:01:30,600 --> 01:01:33,800 how do you test for that? And how you verify that the 1010 01:01:33,800 --> 01:01:37,100 result of an ally is providing is actually actually accurate. 1011 01:01:38,400 --> 01:01:43,400 Yeah, I mean there's there's clearly no General solution to 1012 01:01:43,400 --> 01:01:46,600 that, that problem just as there isn't among humans, right 1013 01:01:46,600 --> 01:01:51,000 because they eyes are going to be doing so many different, so 1014 01:01:51,000 --> 01:01:53,400 many different types of things. I mean, you could have an AI 1015 01:01:53,400 --> 01:01:57,500 That's proving proving math theorems are coming up with 1016 01:01:57,500 --> 01:02:02,000 science hypotheses, to help with with biomedicine or, or 1017 01:02:02,400 --> 01:02:04,300 predicting the stock market or something. 1018 01:02:04,300 --> 01:02:07,100 Right? So then it's like, is your stock 1019 01:02:07,100 --> 01:02:11,000 for Diction a? I giving subtly biased 1020 01:02:11,000 --> 01:02:15,800 predictions that it's then using to make money by trading itself 1021 01:02:15,800 --> 01:02:18,400 in the background against what you traded or something. 1022 01:02:18,500 --> 01:02:22,000 I mean that there's a lot of subtleties that could come up 1023 01:02:22,500 --> 01:02:24,000 and they're going to be different for the different 1024 01:02:24,000 --> 01:02:27,800 kinds of AI that that you're doing. 1025 01:02:27,800 --> 01:02:33,200 So I think that if you're doing a specific type of computation 1026 01:02:33,600 --> 01:02:37,700 for a specific type of problem, then you could come up I would 1027 01:02:37,700 --> 01:02:40,100 like a formalistic solution for this, right? 1028 01:02:40,100 --> 01:02:43,600 Like if if you have an AI That's generating programs. 1029 01:02:43,600 --> 01:02:48,200 According to Specs, you can do some like formal software 1030 01:02:48,200 --> 01:02:51,100 verification to see that the software actually performs 1031 01:02:51,100 --> 01:02:53,700 according to spec. And if you're if you have an AI 1032 01:02:53,700 --> 01:02:58,400 That's analyzing DNA data. I mean, if you if you have your 1033 01:02:58,400 --> 01:03:02,400 own human DNA data, you can do out of sample, testing on that 1034 01:03:02,400 --> 01:03:05,700 data to see if it's valid, but there's really going to be no 1035 01:03:06,500 --> 01:03:09,200 General. All-purpose solution and 1036 01:03:09,200 --> 01:03:10,300 singular. Jeanette is really a 1037 01:03:10,300 --> 01:03:13,700 general-purpose network. So we I mean, we put a bunch of 1038 01:03:13,700 --> 01:03:18,000 work into designing, a reputation and rating system, 1039 01:03:18,600 --> 01:03:22,300 which is sophisticated and hard to game. 1040 01:03:22,400 --> 01:03:24,300 And this is not part of the beta, but it will be. 1041 01:03:24,300 --> 01:03:29,200 It will be rolled out rolled out later in 2019. 1042 01:03:29,200 --> 01:03:33,100 But I think that that's been, like, a holy grail for every 1043 01:03:33,800 --> 01:03:37,200 Online Marketplace, right? And we really need to get it. 1044 01:03:37,400 --> 01:03:38,600 We need to get that. Right? 1045 01:03:38,600 --> 01:03:44,600 Because in the end, Verification that things are accurate 1046 01:03:45,200 --> 01:03:49,000 unbiased or not too biased or inappropriately biased. 1047 01:03:49,000 --> 01:03:53,100 I mean this is really hard and it's domain-specific and 1048 01:03:53,900 --> 01:03:58,200 ultimately each person isn't doing that on their own, right? 1049 01:03:58,200 --> 01:04:02,800 Like if the point of delegating tasks to in a eyes. 1050 01:04:03,500 --> 01:04:07,900 Yeah but he can't give them on our own so what day I do 1051 01:04:07,900 --> 01:04:14,100 delegated to right so like if I'm if I'm Want to verify that 1052 01:04:14,100 --> 01:04:22,000 someone's AI for analyzing DNA. Data is accurate and then not 1053 01:04:22,000 --> 01:04:26,300 many of us are going to, you know, write the code or even run 1054 01:04:26,300 --> 01:04:29,000 the code for that ourselves. We're going to go to some 1055 01:04:29,000 --> 01:04:34,200 service that that does that and then which service do we trust? 1056 01:04:34,200 --> 01:04:38,600 Like is it the singularity, net Foundation, certified service? 1057 01:04:38,600 --> 01:04:42,100 Well then that's like a centralized delete or or Do you 1058 01:04:42,100 --> 01:04:44,900 have a variety of you know competing Services out there 1059 01:04:44,900 --> 01:04:48,400 then you choose which one so in but then it comes down to 1060 01:04:48,400 --> 01:04:51,600 reputation systems again because then you're choosing the one 1061 01:04:51,600 --> 01:04:55,500 that you think has the highest reputation May because it comes 1062 01:04:55,500 --> 01:04:59,900 from Harvard University, or from the the NSA or like, who do you? 1063 01:04:59,900 --> 01:05:03,500 Who do you trust, right? So I'm in ultimately even when 1064 01:05:03,500 --> 01:05:07,900 there's a formal mathematical solution there you're, you're 1065 01:05:07,900 --> 01:05:11,000 placing trust in someone right now. 1066 01:05:11,100 --> 01:05:14,200 If something is When generic enough you could bake 1067 01:05:14,200 --> 01:05:16,500 verification into the protocol, right? 1068 01:05:16,500 --> 01:05:19,700 I mean, as it's done, as it's done with with cryptographic 1069 01:05:19,700 --> 01:05:23,500 checking and but I think checking if an AI is correct or 1070 01:05:23,508 --> 01:05:27,300 not is just it's not going to be that simple, it's going to be 1071 01:05:27,300 --> 01:05:30,700 variety of different algorithms for different, checking, 1072 01:05:30,700 --> 01:05:33,200 different types of problems in different domains. 1073 01:05:33,500 --> 01:05:38,000 And then you need reputation system to be able to place To 1074 01:05:38,000 --> 01:05:43,900 know which verification Checker to to trust and then people will 1075 01:05:43,900 --> 01:05:49,600 try to game game that reputation system by giving a hydrating to 1076 01:05:49,600 --> 01:05:53,200 bogus like very truth verification Checkers. 1077 01:05:53,200 --> 01:05:55,400 Right? So you need the machine learning 1078 01:05:55,400 --> 01:06:01,700 based Reputation, police this to try to stomp out people gaming, 1079 01:06:01,700 --> 01:06:05,200 the reputation system, and then you have to believe the machine 1080 01:06:05,200 --> 01:06:08,300 learning based reputation police itself, isn't corrupt, right? 1081 01:06:08,400 --> 01:06:12,500 So I mean, this, this, this is the world that were that, were 1082 01:06:13,100 --> 01:06:16,500 that were in, but on the other hand, like the real world 1083 01:06:16,500 --> 01:06:23,000 economy, Isn't all that clean and and safe either, right like 1084 01:06:23,000 --> 01:06:28,400 which major government is not is not corrupt in some in some 1085 01:06:28,900 --> 01:06:32,200 serious way. So I don't think like the AI and 1086 01:06:32,200 --> 01:06:34,500 blockchain economy is not creating this problem. 1087 01:06:34,500 --> 01:06:37,400 This is a problem of human beings being assholes, right? 1088 01:06:37,400 --> 01:06:40,900 And this is this is just it's manifest itself in everything 1089 01:06:40,900 --> 01:06:44,300 that human beings do. Okay, so this is great. 1090 01:06:44,300 --> 01:06:49,100 So I a you very much ties into another thing that you know, I 1091 01:06:49,100 --> 01:06:51,500 really look forward to addressing here a little bit. 1092 01:06:51,800 --> 01:06:57,600 So when you spoke about kind of the division mission of 1093 01:06:57,600 --> 01:07:01,400 Singularity, net in a different interview, I heard, you know, 1094 01:07:01,400 --> 01:07:06,600 you mentioned that Tingler narratives to objectives first, 1095 01:07:06,600 --> 01:07:08,900 its objective of maximizing intelligence. 1096 01:07:08,900 --> 01:07:12,900 The other one, this objective of kind of pursuit The maximum 1097 01:07:12,900 --> 01:07:16,400 benefit for all beings, I'm curious. 1098 01:07:16,400 --> 01:07:19,100 So, you know, we spoke a little bit now about okay. 1099 01:07:19,100 --> 01:07:22,900 How do you evaluate in a way? I and, you know, how can you 1100 01:07:22,900 --> 01:07:25,000 check with it? What they're doing is correct, 1101 01:07:25,000 --> 01:07:29,600 this new interest. So now I understand also the 1102 01:07:29,600 --> 01:07:33,800 concept of creating this efficient market place for AI. 1103 01:07:33,800 --> 01:07:36,900 And so now I is a normal small business owner. 1104 01:07:36,900 --> 01:07:41,000 I can kind of use Ai and maybe have something almost as good as 1105 01:07:41,000 --> 01:07:43,800 Google and I don't trust Tim Foley, or maybe something better 1106 01:07:43,800 --> 01:07:48,400 than Google Right Down the Line. But like, how can you? 1107 01:07:49,700 --> 01:07:55,300 How can you make sure that this system is going to end up being 1108 01:07:55,300 --> 01:07:58,700 a system that pursues this you know, benefit for all beings? 1109 01:07:58,700 --> 01:08:05,000 And that kind of embodies is this value what we can't make 1110 01:08:05,000 --> 01:08:08,800 sure of anything and I would say if if we don't create 1111 01:08:08,800 --> 01:08:13,400 singularity in that if I decide to go do something more relaxing 1112 01:08:13,400 --> 01:08:18,600 with my life instead, then how do you know for sure that you 1113 01:08:18,600 --> 01:08:24,300 know, she's Vladimir Putin, Donald Trump, Google IBM, 1114 01:08:24,300 --> 01:08:27,300 tencent, all the companies out there. 1115 01:08:27,300 --> 01:08:29,300 How do you, how are you sure that those guys are going to 1116 01:08:29,300 --> 01:08:32,700 create an AI, which is for broad human benefit. 1117 01:08:32,800 --> 01:08:36,200 And if everyone stops making AI, how do you guarantee that? 1118 01:08:36,200 --> 01:08:39,899 No one's going to like, send synthetic viruses out there to 1119 01:08:39,899 --> 01:08:44,300 poison everyone to the right. So I or that the proliferation 1120 01:08:44,300 --> 01:08:48,600 of nuclear material in the, you know, Eastern Europe isn't going 1121 01:08:48,600 --> 01:08:50,200 to be used to blow. Everybody up. 1122 01:08:50,200 --> 01:08:55,500 So I mean I think we're not in a point in human history where 1123 01:08:55,500 --> 01:08:58,800 there's a great amount of certainty, there's probably even 1124 01:08:58,800 --> 01:09:02,399 more uncertainty than in the past and there's there's always 1125 01:09:02,399 --> 01:09:04,899 been a lot. So but really that the question 1126 01:09:04,899 --> 01:09:11,800 to ask is On average, are we better off creating a 1127 01:09:11,800 --> 01:09:16,000 decentralized, you know benefit oriented AI platform? 1128 01:09:16,000 --> 01:09:19,000 Like, Singularity, net, or are we better off? 1129 01:09:19,000 --> 01:09:21,899 I mean, not nice. Not having that there and and 1130 01:09:21,899 --> 01:09:23,899 having all the other shit going on in the world. 1131 01:09:23,899 --> 01:09:26,200 Right? I mean that that's that, that's 1132 01:09:26,200 --> 01:09:28,000 the question. That's the question to ask. 1133 01:09:28,000 --> 01:09:29,700 So, so I mean, that's a fair point. 1134 01:09:29,700 --> 01:09:33,000 But then I mean that's just sort of rephrasing the question a 1135 01:09:33,008 --> 01:09:34,899 little bit. So then I guess my question is, 1136 01:09:35,399 --> 01:09:36,899 what are you doing? You know what? 1137 01:09:37,100 --> 01:09:41,200 Order such that, you know, this objective is in this value is 1138 01:09:41,200 --> 01:09:46,000 embodied in the platform. Yeah, I mean, there's two parts 1139 01:09:46,000 --> 01:09:49,800 of that. So one part is in sort of the 1140 01:09:49,800 --> 01:09:53,399 token omics of the singularity in that ecosystem. 1141 01:09:53,399 --> 01:09:57,500 The other part is in the a, is that the singular unit 1142 01:09:57,500 --> 01:10:01,900 Foundation itself? Is our our building and putting 1143 01:10:01,900 --> 01:10:05,600 into the network. So I mean in terms of the token 1144 01:10:05,600 --> 01:10:12,300 omics, I mean there's curation markets and An intelligent 1145 01:10:12,300 --> 01:10:14,600 reputation system, which is designed. 1146 01:10:14,600 --> 01:10:20,600 So that, at least the agents that are contributing value to, 1147 01:10:20,600 --> 01:10:24,700 the network are getting rewarded proportionally to that instead 1148 01:10:24,700 --> 01:10:28,300 of having sort of game game-theoretic Dynamics, where a 1149 01:10:28,308 --> 01:10:31,700 few agents will accumulate all well, which is what seems to be 1150 01:10:31,700 --> 01:10:34,200 happening in Bitcoin and ethereum and this is what 1151 01:10:34,200 --> 01:10:36,300 happens in most conventional economies. 1152 01:10:37,100 --> 01:10:42,900 Also and then on top of that, a certain percentage of the tokens 1153 01:10:42,900 --> 01:10:47,400 that were initially minted are earmarked to be spent on benefit 1154 01:10:47,400 --> 01:10:50,000 tasks as decided by the community, which can be things 1155 01:10:50,000 --> 01:10:54,500 like healthcare, education medicine, and so forth. 1156 01:10:54,500 --> 01:10:59,300 So there there's at least that nudge and put in there to have a 1157 01:10:59,300 --> 01:11:03,000 certain percentage of the token spin on things that are 1158 01:11:03,000 --> 01:11:06,900 considered of broad benefit. I mean and this is this Is much 1159 01:11:06,900 --> 01:11:10,000 like a government does when it spends some percentage of its 1160 01:11:10,000 --> 01:11:12,300 well from on social welfare, right? 1161 01:11:12,300 --> 01:11:16,500 It's just most most projects don't like wire that into their 1162 01:11:17,100 --> 01:11:21,000 economic operation. But then the the AI is that we 1163 01:11:21,000 --> 01:11:26,300 are putting into the network ourselves are largely benefit 1164 01:11:26,300 --> 01:11:29,200 oriented. So Sophia robot which we talked 1165 01:11:29,200 --> 01:11:33,500 about one thing we've been doing is using Sophia as a like a 1166 01:11:33,500 --> 01:11:35,600 therapist and meditation assistant. 1167 01:11:35,600 --> 01:11:39,200 So that's You know, that's not solving all the problems of the 1168 01:11:39,200 --> 01:11:41,400 world but it's different than the Terminator, right? 1169 01:11:41,400 --> 01:11:44,200 I mean it's using a robot to just kind of help people expand 1170 01:11:44,200 --> 01:11:50,200 their Consciousness and we we're working on applying AI. 1171 01:11:50,200 --> 01:11:54,600 That's using opencog framework and wrapped in Singularity, that 1172 01:11:55,100 --> 01:11:59,300 to analyze DNA data of people living 105 years or over to 1173 01:11:59,300 --> 01:12:01,900 figure out what makes them live so long to try to. 1174 01:12:02,700 --> 01:12:05,800 Figure out how to extend other people's lives. 1175 01:12:05,800 --> 01:12:10,400 We're analyzing images of plants from China and Africa to try to 1176 01:12:10,400 --> 01:12:14,100 diagnose spread of crop disease and early stages using using 1177 01:12:14,100 --> 01:12:15,600 deep neural nets for image processing. 1178 01:12:15,600 --> 01:12:21,000 So of course each of these things is a drop in the bucket 1179 01:12:21,008 --> 01:12:25,000 regarding what we need to do to massively improve the state of 1180 01:12:25,000 --> 01:12:29,200 humanity but the hope is that by injecting these things in the 1181 01:12:29,200 --> 01:12:33,000 network is an early stage. You're impacting the You're of 1182 01:12:33,000 --> 01:12:36,300 the community because ultimately this is about the community that 1183 01:12:36,500 --> 01:12:38,200 that you build around Singularity net. 1184 01:12:38,200 --> 01:12:44,300 So if if to curation rewards and benefit tokens and having a 1185 01:12:44,308 --> 01:12:48,300 bunch of, you know, positive beneficial stuff happening in 1186 01:12:48,300 --> 01:12:52,900 the network and then be our largest, our largest development 1187 01:12:52,900 --> 01:12:56,500 offices, in Ethiopia, in Addis Ababa, where we have 1188 01:12:56,700 --> 01:12:59,900 20-something developers working on singularity in it. 1189 01:12:59,900 --> 01:13:04,000 So we're, we're trying to actively We pull people from the 1190 01:13:04,000 --> 01:13:07,200 developing world into development and use the network. 1191 01:13:07,200 --> 01:13:12,000 So hopefully my all these things will be nudging, the community 1192 01:13:12,000 --> 01:13:15,900 in a in a positive direction which is really going to be the 1193 01:13:15,900 --> 01:13:18,200 most important thing. Because I mean if if we're 1194 01:13:18,200 --> 01:13:21,700 successful with this then you know five years from now, the 1195 01:13:21,700 --> 01:13:26,300 work done by Singularity that foundation will be a relatively 1196 01:13:26,300 --> 01:13:28,800 modest. Percentage of all the work being 1197 01:13:28,800 --> 01:13:33,300 done to build the protocol on the network over time and The AI 1198 01:13:33,300 --> 01:13:36,800 is in, the network will be, mostly contributed by other 1199 01:13:36,800 --> 01:13:40,100 random people not not by people paid by Singularity, net 1200 01:13:40,100 --> 01:13:44,600 Foundation, but then we are ceding this community. 1201 01:13:44,600 --> 01:13:47,000 And we're seeing this culture, and let you can see that in the 1202 01:13:47,000 --> 01:13:49,800 Linux right? Like, Linus, Torvalds, and 1203 01:13:49,800 --> 01:13:53,300 Richard stallman and their friends from the old days, 1204 01:13:53,300 --> 01:13:54,700 right? The very small percentage of 1205 01:13:54,700 --> 01:13:59,700 code in Linux right now but the culture of Linux is what it is 1206 01:14:00,200 --> 01:14:03,100 because of how they started. Stood it, right? 1207 01:14:03,100 --> 01:14:08,000 So we want to get, you know, beneficial motives and love and 1208 01:14:08,000 --> 01:14:12,300 compassion and inclusiveness in the cultural DNA of singularity 1209 01:14:12,300 --> 01:14:15,400 in that community. And then it will continue to be 1210 01:14:15,400 --> 01:14:19,100 there in the code also. And this is a bit, you know, 1211 01:14:19,100 --> 01:14:22,400 soft and fuzzy. It's not like a mathematical 1212 01:14:22,400 --> 01:14:25,800 guarantee of beneficial activity, but I think that's 1213 01:14:26,600 --> 01:14:29,800 That's how things actually actually have to work. 1214 01:14:29,800 --> 01:14:33,200 Because in the end, it's about the community of human beings 1215 01:14:33,200 --> 01:14:37,100 who are going to be developing developing this ongoing llege. 1216 01:14:38,300 --> 01:14:39,800 Well, that's, that's really fascinating. 1217 01:14:39,800 --> 01:14:44,600 And, and I think also the fact that you guys have actual people 1218 01:14:44,600 --> 01:14:49,200 in Ethiopia, working on problems in Ethiopia, is it really is 1219 01:14:49,200 --> 01:14:53,800 great. And I are far removed from what 1220 01:14:53,800 --> 01:14:57,000 a lot of people in the blockchain space are doing this 1221 01:14:58,800 --> 01:15:02,400 code down. O cardano is running a like a 1222 01:15:02,407 --> 01:15:06,600 year-long education program where they're teaching 100 young 1223 01:15:06,600 --> 01:15:08,600 Ethiopian programmer. Haskell. 1224 01:15:09,600 --> 01:15:11,200 That's cool. The programming language. 1225 01:15:11,200 --> 01:15:15,500 So I think well I've been I've had this office in Ethiopia 1226 01:15:15,500 --> 01:15:21,300 since what 2014 I guess doing AI Outsourcing before we shifted 1227 01:15:21,300 --> 01:15:26,700 them to singularity in it but now cardano's moved in there and 1228 01:15:26,700 --> 01:15:30,500 yeah there's a lot of tech projects throughout various 1229 01:15:30,500 --> 01:15:34,400 African Tech hubs. Now so those powerful forces of 1230 01:15:34,800 --> 01:15:38,400 centralization And wealth concentration. 1231 01:15:38,400 --> 01:15:42,500 But yet there's there's also the, the opposite and that 1232 01:15:42,600 --> 01:15:46,700 peer-to-peer and then positive globalization happening. 1233 01:15:46,700 --> 01:15:50,600 So it's a very interesting time with these two different forces 1234 01:15:50,600 --> 01:15:55,000 are both both surging forward in powerful ways. 1235 01:15:55,500 --> 01:15:58,500 Well, so before I wrap up, I didn't want to ask you one. 1236 01:15:58,500 --> 01:16:01,200 One, last question. And this is, we kind of touched 1237 01:16:01,200 --> 01:16:03,700 on this earlier when you were talking about a ice making 1238 01:16:03,700 --> 01:16:06,400 predictions. So, let's imagine a future now. 1239 01:16:06,500 --> 01:16:11,300 We're a lot of a lot of the economy is is Ronald watching 1240 01:16:11,300 --> 01:16:13,800 systems. You have, you know, powerful 1241 01:16:13,800 --> 01:16:18,000 markets that exist exclusively on blockchains and organizations 1242 01:16:18,000 --> 01:16:21,000 and companies are interacting with these markets on 1243 01:16:21,000 --> 01:16:23,200 blockchains doing business on these markets. 1244 01:16:23,600 --> 01:16:27,200 And these Mark, these markets are run by doubts. 1245 01:16:27,200 --> 01:16:31,200 So there are there are governance mechanisms in place 1246 01:16:31,200 --> 01:16:36,000 which allow the companies that themselves that are operating on 1247 01:16:36,000 --> 01:16:39,300 these Gets to participate also in the governance through 1248 01:16:39,300 --> 01:16:40,400 through staking. Okay. 1249 01:16:40,400 --> 01:16:43,300 So, you know, the companies that use the markets also have stayed 1250 01:16:43,300 --> 01:16:45,600 on the markets and they can participate in governance 1251 01:16:45,600 --> 01:16:48,100 decisions for like protocol updates or things like this. 1252 01:16:49,000 --> 01:16:52,000 Now, it seems like there would be an incentive at this point 1253 01:16:52,900 --> 01:16:56,000 and even for something like prediction markets for these 1254 01:16:56,000 --> 01:17:00,300 companies, to have steak do essentially delegate their steak 1255 01:17:00,300 --> 01:17:04,800 to an AI because the AI is going to make much better decisions on 1256 01:17:05,100 --> 01:17:09,500 what types of Governments are what types of proposals that 1257 01:17:09,500 --> 01:17:14,300 they should be making in order to maximize the network itself 1258 01:17:14,300 --> 01:17:18,000 and also sort of like maximize their profits long-term. 1259 01:17:18,200 --> 01:17:22,600 So it seems like there would be kind of a like a Nash 1260 01:17:22,600 --> 01:17:26,900 equilibrium here. We're at some point if one 1261 01:17:26,900 --> 01:17:31,000 company starts using an AI to manage their governance or make 1262 01:17:31,000 --> 01:17:34,000 predictions and other companies start using a eyes to make 1263 01:17:34,000 --> 01:17:36,300 predictions and then when everybody's making predictions, 1264 01:17:36,400 --> 01:17:41,500 Us or making governance decisions with an AI as we move 1265 01:17:41,500 --> 01:17:46,600 closer to their only, I then it's like they just had a eyes 1266 01:17:46,600 --> 01:17:49,700 competing with a eyes and I like, I guess this also extends 1267 01:17:50,000 --> 01:18:07,500 more, probably, I'm sorry. That's in charge rather than 1268 01:18:07,500 --> 01:18:12,500 any, any one person and who's in charge of Bitcoin, in aetherium. 1269 01:18:12,500 --> 01:18:17,000 Yeah, we don't We don't we don't actually know that but it's 1270 01:18:17,000 --> 01:18:19,400 clear. It's concentrated in the small a 1271 01:18:19,400 --> 01:18:24,400 small number of individuals and investment groups who are 1272 01:18:24,400 --> 01:18:28,200 controlling these things. So yeah, I think You know, in 1273 01:18:28,200 --> 01:18:33,500 the long term it's inevitable that, you know, of a eyes are a 1274 01:18:33,500 --> 01:18:36,100 thousand times more intelligent than human beings and of 1275 01:18:36,100 --> 01:18:38,300 molecular nanotechnology, and so forth. 1276 01:18:38,300 --> 01:18:42,900 It's inevitable, they're going to have more physical power than 1277 01:18:42,900 --> 01:18:45,400 humans, right? I mean it doesn't mean they're 1278 01:18:45,400 --> 01:18:48,700 going to control every little aspect of what humans do in 1279 01:18:48,700 --> 01:18:51,200 their lives. But they're going to have more 1280 01:18:51,400 --> 01:18:54,600 more oomph than we do, right? So I'm in it in the long term 1281 01:18:54,800 --> 01:18:57,300 which may just be like decades from now. 1282 01:18:57,700 --> 01:19:00,700 We're going to have two choices. One is like you wire into the 1283 01:19:00,700 --> 01:19:04,900 network and become become one with a super intelligent Global 1284 01:19:04,900 --> 01:19:07,000 brain. Even if that means giving up 1285 01:19:07,000 --> 01:19:11,000 many aspects of your legacy Humanity or else you know you 1286 01:19:11,000 --> 01:19:13,600 live in the in the people preserve like the squirrels in 1287 01:19:13,600 --> 01:19:16,600 the National in the national park and you know that the 1288 01:19:16,600 --> 01:19:20,800 squirrels the squirrels in the park they can fight over 1289 01:19:20,800 --> 01:19:23,600 girlfriends and hunt for food and play and have fun. 1290 01:19:23,600 --> 01:19:26,700 And people are not trying to regulate every aspect of their 1291 01:19:26,700 --> 01:19:28,500 little squirrel exhibit. Distance right? 1292 01:19:28,500 --> 01:19:31,600 On the other hand, if they run out of the park, they might get 1293 01:19:31,600 --> 01:19:33,200 rolled over by a truck. Right? 1294 01:19:33,200 --> 01:19:37,100 So, I mean, I think if you have a superhuman AI That's 1295 01:19:37,300 --> 01:19:41,300 tremendously more intelligent than us, either you join it or 1296 01:19:41,300 --> 01:19:43,800 you're going to remain living a happy, human life. 1297 01:19:44,100 --> 01:19:47,800 Hopefully with a lot of abundance provided for you, but 1298 01:19:47,800 --> 01:19:50,400 I mean in the end, there's something much more powerful 1299 01:19:50,400 --> 01:19:54,100 than you. That does have some regulatory 1300 01:19:54,100 --> 01:19:56,600 control when it when it needs to, which could be good. 1301 01:19:56,600 --> 01:19:59,700 Also, write like a If the squirrels die of some plague 1302 01:19:59,700 --> 01:20:01,700 will come in to give them antibiotics, right? 1303 01:20:01,700 --> 01:20:04,400 And the same way of human society went too far. 1304 01:20:04,400 --> 01:20:08,800 Awry a super AI that loved us. It would let us go about our 1305 01:20:08,800 --> 01:20:11,900 business, but if things went too far awry, it might come in and, 1306 01:20:12,100 --> 01:20:14,700 and fix things. I mean, that's, that's a long 1307 01:20:14,700 --> 01:20:17,500 term. It's upload to the global brain 1308 01:20:17,500 --> 01:20:19,100 or live in the people freezer of right. 1309 01:20:19,100 --> 01:20:22,900 But I mean in the in the medium term, it's going to be really, 1310 01:20:22,900 --> 01:20:25,700 really complicated and the and as you say, there's going to be 1311 01:20:25,700 --> 01:20:28,500 a gradual. Transition from Human 1312 01:20:28,500 --> 01:20:32,400 decision-making to AI decision-making. 1313 01:20:32,400 --> 01:20:36,900 But given given how profound we fucked so much of our political 1314 01:20:36,900 --> 01:20:41,200 and corporate ecosystem is now I see that as a great opportunity 1315 01:20:41,200 --> 01:20:43,100 to improve things. Right? 1316 01:20:43,100 --> 01:20:46,300 I mean there's a lot of if the AI is written, right? 1317 01:20:46,300 --> 01:20:49,100 It's going to do a lot better than the individual humans 1318 01:20:49,100 --> 01:20:51,200 institutions and they're controlling things now. 1319 01:20:51,200 --> 01:20:55,500 So then it really comes down to, you know, creating the AI is 1320 01:20:55,500 --> 01:20:59,800 that are going to be the Decision support systems for the 1321 01:20:59,800 --> 01:21:03,100 people controlling, most of the world's most of the world's 1322 01:21:03,100 --> 01:21:06,600 resources. So look, I'm glad I can always 1323 01:21:06,600 --> 01:21:11,100 go back to the, to the human Reserve Nature, Reserve wherever 1324 01:21:11,100 --> 01:21:15,800 that is like, and I could Chase, whatever whatever things you 1325 01:21:15,808 --> 01:21:17,700 would Chase. Easily said any encumbrance? 1326 01:21:17,700 --> 01:21:19,800 Yeah. We're we're setting aside and 1327 01:21:19,800 --> 01:21:22,700 reach a region in the southern Ethiopia, for this purpose. 1328 01:21:22,700 --> 01:21:24,400 So yeah, I'll show it to you sometime. 1329 01:21:24,600 --> 01:21:27,900 Maybe I got account when the world when the Well, yeah, after 1330 01:21:27,900 --> 01:21:32,800 some but after some global warming, so let's wrap up. 1331 01:21:32,800 --> 01:21:35,100 Let's just want to ask you, you know, how can people get 1332 01:21:35,100 --> 01:21:37,500 involved in single airing at and where they can learn more? 1333 01:21:38,100 --> 01:21:42,100 Yeah, absolutely. So I'm in the center of it all. 1334 01:21:42,100 --> 01:21:46,100 Go to the website, Singularity, net dot IO. 1335 01:21:46,100 --> 01:21:49,000 And I mean there you can find information on how to download 1336 01:21:49,000 --> 01:21:53,100 and play with the with the beta. If you're if you're a developer 1337 01:21:53,100 --> 01:21:57,100 or we have a Blog which has updates in the research, You're 1338 01:21:57,100 --> 01:22:00,900 a pretty frequently, we have a telegram Discussion Group, which 1339 01:22:00,900 --> 01:22:04,200 has a, some percentage of interesting things on it and 1340 01:22:04,200 --> 01:22:08,100 some other things. And so I think lots and lots of 1341 01:22:08,100 --> 01:22:12,000 ways to get involved with the with the community and that, you 1342 01:22:12,000 --> 01:22:15,900 know, I'm still as well as doing some actual work going around 1343 01:22:15,900 --> 01:22:20,400 and speaking at various conferences, so I can meet meet 1344 01:22:20,400 --> 01:22:23,400 some of some of you guys listening there. 1345 01:22:23,900 --> 01:22:27,600 I think in middle of March, we're having tokens. 20:49 1346 01:22:27,600 --> 01:22:31,300 conference here in Hong Kong. So if anyone's there, we can we 1347 01:22:31,300 --> 01:22:34,300 can we can hang out. I'll tell ya. 1348 01:22:34,300 --> 01:22:37,300 This is all in the end while we're talking about building 1349 01:22:37,300 --> 01:22:39,700 superhuman. A, I getting there is all about 1350 01:22:39,700 --> 01:22:42,500 the human Community, right? So we need people to be involved 1351 01:22:42,500 --> 01:22:47,000 in many, many different ways. Yeah, join our communities 1352 01:22:47,000 --> 01:22:50,800 online than the that we're happy to talk to you about about what 1353 01:22:50,800 --> 01:22:52,800 you can do to help out. Right? 1354 01:22:52,800 --> 01:22:55,300 Well then, thank you so much for joining us is a real pleasure, 1355 01:22:55,300 --> 01:22:59,100 talking and great and a diving deep in this is really 1356 01:22:59,100 --> 01:23:01,300 fascinating topic that we don't really always give it. 1357 01:23:01,900 --> 01:23:03,800 Get much chance. To discuss here on the podcast. 1358 01:23:03,800 --> 01:23:06,000 So happy to have you all good for ya. 1359 01:23:08,400 --> 01:23:10,300 Thank you for joining us on this week's episode. 1360 01:23:10,600 --> 01:23:12,200 We release new episodes every week. 1361 01:23:12,700 --> 01:23:15,500 You can find And subscribe to the show on iTunes Spotify, 1362 01:23:15,600 --> 01:23:18,600 YouTube SoundCloud or wherever you listen to podcasts. 1363 01:23:19,000 --> 01:23:21,800 And if you have a Google home or Alexa device, you can tell it to 1364 01:23:21,800 --> 01:23:24,800 listen to the latest episode of the epicenter podcast, go to 1365 01:23:24,800 --> 01:23:27,900 epicenter, .t V /, subscribe for a full list of places where you 1366 01:23:27,900 --> 01:23:30,500 can watch and listen, while you're there, be sure to sign up 1367 01:23:30,500 --> 01:23:33,300 for the newsletter so you get new episodes in your inbox as 1368 01:23:33,300 --> 01:23:36,500 the Reynolds if you want to interact with us guests or 1369 01:23:36,500 --> 01:23:38,200 other. Last listeners, you can follow 1370 01:23:38,200 --> 01:23:41,200 us on Twitter and please leave us a review on iTunes helps 1371 01:23:41,200 --> 01:23:43,300 we've on the show and we're always happy to meet them. 1372 01:23:44,100 --> 01:23:46,700 So thanks so much and we look forward to being back next week.