 I'm super excited by today's episode because I have Dr. Ben Gertzel on the show and Tufi Salabi. And we dive in deep on the dangers and we dive in deep on the dangers of centralized AI in how this corresponds to COVID-19. Now, Dr. Ben Gertzel is an artificial intelligence researcher, chief scientist and CEO of Singularity Net. And Tufi Saliba is CEO of Privacy Shell and chair of the ACM Practitioner Board Conference Committee. Tufi's background is machine learning, decentralized governance and distributed computing. Man, man, I'm telling you guys, we dive in deep talking about how big corporations are collecting our data, the dangers of this data, how centralized AI is gonna create a dystopian future and how this corresponds to COVID-19. So sit back and enjoy today's episode of Amir Approved. Tufi, Ben, thank you so much for going on the show and I'm super excited about this. And I think it's, yeah, man, I think it's very topical, very topical. We're talking about two things that are very important. We're one talking about AI, the future of AI, where do we see it, where it's going and how does this tie into COVID-19? And so I kind of want to put the ball in your guys' core and kind of start this conversation is like, what's your guys hypothesis, thesis, like where we're heading to? Thanks, Amir. Before we get going, I often whenever we talk to people that they don't know us, I like to say full disclosure, we do have a partnership, Ben and I, in something that's called the DIA, the Centralized AI Alliance. It's a not-for-profit. It's anything, has no way of benefiting. We've been spending money into it. It's just the two of us mainly and our companies we've brought and people that they're supporting us. So we've founded the Decentralized AI Alliance and currently have about 65 different companies that they're members of it. We also have, as a full disclosure, we have partnership in a company called Podalarity, which is currently, it's a deployment of singularity on Toda, one of the Toda APIs, but it's intent to have a singularity on Toda in enable a market place is more like amongst many of the marketplace of the singularity. Full disclosure, we do have partnerships together, but the most important thing of what brought us together is that we've been working in parallel towards a similar or main goal and then we converged several times and we decided to do a lot of partnerships and I let Ben talk a little bit more about these things. He's got a lot more history then and a lot of people in AI and he is extremely controversial and a lot of people may not necessarily believe he's AI but when they get deep and deep and deep and they'll be like, holy moly, I've been missing on seeing his points. So I'll let him talk. All right, yeah, thanks Tufi. Yeah, I mean, Tufi and I have been working together on various things since early 2018, not too long after I launched Singularity.net and it's clear to me when we met, there's not so many people that understand the math and the computer science and the politics and the social and philosophy aspects, all these different aspects of decentralized AI and why it's important, how to build it and how to roll it out and this topic, it's never been more important than right now with a global pandemic sweeping the world and the moment I talked to Tufi for the first time after this COVID-19 pandemic became so widespread. I mean, it was clear we were both thinking the exact same way like, A, decentralized AI can be a big help here. B, the centralized governments and big centralized tech companies while genuinely acting to help people in the world at the same time are gonna be setting into place mechanisms that may well come back to bite everyone a little bit later and Tufi and I saw this like within a fraction of a second of knowing there was a pandemic sweeping the planet and now more people are starting to catch on but then what can you do about it, right? Like you can't stop governments or Google or Apple from trying to save lives but you can put out alternative tools and alternative ways of using decentralized technology and AI technology to cope with the situation. So yeah, should that, but we're doing the interview already in our preliminary intro. Well, the question I have is like, I think we should at least for the audience set the stage of the potential dangers. So let's talk about that. What are the potential dangers that people aren't aware of right now? Yeah, I mean, obviously with a pandemic the number one danger is us all getting sick and dying and that's the number one most important thing and whatever we can do within very broad limits to prevent that from happening, I mean, we wanna do. On the other hand, there is a risk that the measures taken to control a pandemic stick around behind the scenes after the pandemic is gone in ways that ultimately may do even more harm than the good they did originally. So if you're looking at, all right, you've got a Fitbit, you've got a digital thermometer, you've got a PaulStock Seminar and you're putting that information into an app which some well-meaning big tech company is feeding into their database, then they're analyzing it together with national government, maybe sharing with other national governments to spread, to track the flow of the pandemic, right? And that's valuable. Something of that nature is certainly necessary. Now after the pandemic is gone, will big tech and big government stop collecting data when everybody's physiology and using it to track the next dangerous thing? Let's say nasty Muslim terrorists or whatever it may be, right? Of course, they're not gonna stop, right? Of course, there's always gonna be some other really important thing that's viewed as a terrible threat. And generally, I mean, we don't see these panopticon type technologies getting rolled back too much after they're put into place. And there are decentralized alternative ways to do exactly the same and better things in terms of tracking the pandemic, which would set a positive platform for development after the pandemic is gone, as opposed to setting the stage for big brother to invade everybody's body. So let's give some case example, like anyone could jump in, Ben or Toofy is, let's fast forward, it's already happening. Everything is being centralized, they have their own machine learning algorithms in the background, obviously with government back doors, that's a given. And they start collecting your health records, they start collecting, they're already collecting GPS locations. Did you just say government back doors? Something like that, yeah. Yeah, okay. Or I should say all access pass. And what people don't understand, like I want people to understand like what are the ramifications of this? Like talk about actual ramifications if we allow this. I'm going to let Ben and I were talking yesterday and he talked about one out of 101 different scenarios of how centralized AI use against people, when I think it will be, don't shoot the messenger, we're saying it can be. And specifically in a pandemic situation, Ben can talk about like how that can be used in all of these information against the people to launch an actual artificial pandemic if somebody wants in the future. But before we actually get to that point, I need to make sure that every single one of the members of the audience understand that with centralized AI, it can be prevented of course, it can be by building in a lot of like governance bodies and layers and layers of friction and putting a lot of data that is available to everybody to see and making sure that there isn't anything that is going to be gone in some direction. But in that model, it ends up being so inefficient that eventually people will end up going and shying away from it because economically it wouldn't make sense. When we talk about decentralized AI and our biggest win in this equation, we're talking about autonomous decentralized governance where those devices, they cannot be repurposed for what they were purposed initially. They cannot be used against the users, the host. And therefore they have that security by design. By having that security by design, you end up eliminating a lot of that friction and gaining a lot of efficiencies. And then therefore getting something to be economically much more feasible and having the ability to outperform everything else that they're doing in a centralized way. So on the long run, they will certainly lose. We're hoping that from now until then, they don't end up causing a lot of damages to a lot of people. And I'll let Ben talk about that. One of the scenarios of a pandemic that is triggered by some sort of centralized evil AI. Yeah, let's talk about a couple specific AI functions. It'll be very helpful with the pandemic. One is agent-based simulation modeling where instead of the crude epidemiology simulations that are now being done to model the spread of the pandemic and are being done to guide social policy decisions that are impacting everyone's daily life and costing the world economy maybe trillions of dollars a month. I mean, these rules like social distancing or restaurants only deliver a two-day quarantine to travel to Hawaii or something. I mean, on what basis are these rules being adopted? People's intuition and very, very crude differential equations models of epidemic spreading. Now, we could make a much more fine-grained agent-based simulation, sort of like a SimCity type model where you have little simulated AI agents playing the roles of the people moving around in the simulation of the US and the world. Using this sort of agent-based AI simulation, you can make much better predictions about COVID-19 is going to spread and also much better extrapolations of what the outcome of some policy choice is likely to be. So that's one example. Now, another example is AI-based anomaly detection on physiological data. So if you have thermometer, activity tracker, heart rate variability from your Fitbit, PulseSocks Seminar, all these things connected into your smartphone, you can use AI-based anomaly detection to tell a couple of days before you show any symptoms that you seem to be getting sick. And we're working on an app that does this within Rejuv, which is the Stingularity Net spin-off, but there's been folks at Stanford University and other places who've worked on similar things. So you're able to tell things about the person's physiology and their state that they themselves don't know from their own feeling, right? Of course, this can be super valuable. If you know, you're getting a little bit sick. I mean, if you're in somewhere that's doing COVID testing, you can go get tested. Otherwise, you could just, you know, you'd be a little careful. You don't go visit your grandma that day, right? So these two things, they're extremely valuable in minimizing loss of life in the pandemic situation. On the other hand, imagine what happens when they're put together and used for other purposes after the pandemic is done. I mean, we have a detailed simulation of the movement of people from one point to another and mass movements of human beings. And we're tying this into measurements of, oh, people's stress levels, people's movement and activity patterns, right? So then, you know, if you wanna say, squelch a public demonstration, let alone tamp down a revolution or, you know, track where the members of some organization are going, you know, you're gonna have a tremendously better set of tools for doing this type of monitoring. It's not that governments couldn't do this now necessarily. I mean, Fitbit is bought by Google. Eric Schmidt, who is Google's chairman and CEO is a very good friends with the NSA, right? I mean, all this potential was there. But what we're talking about is practically rolling out integrated systems that tie in, you know, real-time physiological measurement of a huge number of people with in-depth agent-based simulations. And, you know, I'm agitating right now, I'm agitating right now for all these things to be rolled out to help with the pandemic. But if, because they're not being done well enough now, they should be. But if they're rolled out in the wrong way, they're gonna be used after the pandemic to make a police state that, you know, hitherto has certainly never been seen in the Western world and even beyond what China has managed to do so far. You remind me of one of my favorite movies, Gattaca with Ethan Hawke. Yeah, that scratches the surface, right? Yeah. So like, think about the discrimination. Think about like, you know, you guys are familiar with like the Indian caste system they have, like different levels. Yeah, yeah, of course. So think about all the discrimination. We're like, oh, well, Ben, you know, you got these health conditions over here. We can't give you insurance. You can't work for us, you know? It's like, oh, too fee, you know, you have all these things. They're like, there's so many ways I see they can discriminate with people from- Oh yeah, we saw that your heart rate, you know, boosted too much when you were around this type of person. So that means your physiological responses were unacceptable. I'm sorry, you have a poor credit rating. You can't rent a car. Exactly, exactly. And so this is a real danger. I know a lot of people don't see it, but a lot of people aren't really tracking what's happening behind the scenes. They're not quite, let's say, savvy with what's happening with the AI space. I think we can all agree over here. Most people are kind of fed up with like all these big Silicon Valley. We're not giving them good or bad, but more of this- Well, you know, COVID is reversing that. Now people love Amazon. Amazon is their lifeline. That's the only way they can get anything in their house and all they're doing all day is Zoom, Zoom Google Hangouts and so forth. So I think the tech lash is arguably being reversed by whatever that symptom is called where people fall in love with their captors, right? Oh, I know this. Ah, I'll come back. Stock housing or something? Stock home syndrome. Stock home syndrome, that's right. That's it, yeah, yeah, yeah. We're getting stock home syndrome when we're falling in love with our governor Cuomo and we're falling in love with Amazon and Google because while we're locked in our room, they're the only ones delivering us a few croons and that's gonna cause people to care less when these tech conglomerates are feeding data to intelligence agencies. But in the end, people are gonna see why this wasn't so clear at all. I think what people don't understand is like, listen, when you're not paying for something, you're the product and second of all, it's all fine and dandy to have access to all these technological products but don't get a twist to the second you step out of line, you are exiled. That's the thing that people understand is like, you can have access to all this if you play by our rules. And so my question is, since you guys are working on this decentralized AI and really highlighting and evangelizing, like listen, we need to do this before it's too late. We need to institute and create technologies that are open source that have checks and balances. And so what is it? How does that look like for somebody who's kind of curious? Like how do we play this out? You know. Yeah, so I'll go on Tufi. One of the events that we're in and I give the example of which Hill with Mark Zuckerberg give his daughters. And the reason why I talked about this and I wrote an article which folks can find online. It's actually LinkedIn. And it's basically, it talks about the future of smartphone. Today we all have smartphone and that smartphone has a lot of AI in it already. We would be foolish to think that this smartphone is not going to be a lot more intelligent than us, but at that point it's not necessarily going to look like this. It's going to be integrated with us. So I call it a pill because it might be a pill that it's a battery or whatnot. So now let's think of that centralized bill that already exists, which is your smartphone. And about 500 companies can actually get access to this device and can control it. Would Mark Zuckerberg give the Facebook pill to his daughters knowing that Facebook administrator could control it, could control his daughters? As opposed to something that we are working on diligently and we have a lot of already followers and we have a lot of companies that are working with us on that. You give credit to each and every one of them if he's scientist working on this is something that we refer to as decentralized AI. If somebody doesn't like the world, whatever you call it, whatever it is, it is effectively autonomous decentralized governance where when that pill is swallowed by your daughter, it cannot be repurposed. It cannot be used against your daughter. Let's imagine those two scenarios. Which pill would Mark Zuckerberg give his daughters at the time? And if it's not going to be the Facebook pill, then why the fuck are we building the Facebook pill in the first place? If it's not going to be the Facebook pill, why every single centralized AI building it, it's actually the shareholders of those companies that should go against them and be like, why are you building products that you wouldn't use on your kids, we wouldn't use on our kids, nobody's going to want to buy them, loss, loss, loss. What I'm trying to get to the audience here, Amir, is that what we're building is not just some sort of altruistic thing that we really care about like Homo sapiens making it better and we care about the polar bears or whatnot. Of course we care about those things. But we're building the ability for businesses to provide products that users are going to want to need and going to want to buy. A lot more profitable than having something that nobody wants. That's the point that I really want to emphasize here with the decentralized AI. Back to you, Ben. Yeah, I think, you know, it's hard because we need rapid response to the current pandemic. And I mean, Google, Apple, these big tech companies are, I mean, they're very well geared up to respond. Everyone has accounts, everyone has their devices and so forth. Well, I mean, what's going on, right? Right now it's too late. My fear is with the next pandemic. Yeah, that's right. So what we're doing now with the decentralized AI Alliance, the organization that Tufi and I founded in early 2018, bringing together like 50 plus organizations working at the intersection of AI and blockchain. I mean, what we're doing now is a hackathon called the COVIDathon. It's on DevPost. It opens April 1st very soon. It's a two month hackathon. And we're trying to bring AI and blockchain developers together. And part of the goal is to build stuff that may be useful in the current situation right now. We want to work on agent based simulation. We want to work on anomaly detection from data gathered from smartphone peripherals and so on. But part of the goal is to set into motion, building the tools so that by the time there's another crisis, so maybe a pandemic or worldwide cyber attacks, whatever else it is, right? When the next major crisis comes, we really need the decentralized infrastructure to be there. Otherwise, one crisis after another, each crisis just increases the hegemony of the large institutions and their ability to control everyone. And look, I mean, Bill Gates talked about we're not prepared for pandemics about maybe five years ago, and nobody believed him. We talked about the threat of AI years ago. Nobody believed us until recently folks like Elon Musk started believing us. And he started talking about the stuff without even touching, barely like the scratching the surface of the real threat. And the more people they're learning about the threat of centralized AI, the more they're joining forces with this. So now let's imagine, in five years from now, Amir, we find out that, oh, what Elon Musk was talking about, that the centralized, that specific centralized AI that is being built is more dangerous than nuclear warheads. Turns out that he's right. Now, what does that mean? If it turns out that he's right, if people wait until that moment, more dangerous than nuclear warheads, Elon Musk is saying that, okay? If he turns out that he's right, what does it mean? Is that something that is reversible? And the answer is no. So, you know, now I'm bringing another point here that it's not only that we're showing a glorious future with autonomous decentralized governance that can be profitable and so on and so forth. I'm talking to you about horrible future if we don't do something about it. So why are people not, you know, accepting this and they're trying to fight it or whatnot? When they come in, talk to Ben and myself and try to open their minds, they end up joining forces. We just ask people to open their mind and look deep into what we're trying to do here and join forces. Going back to the example though, you mentioned about more dangerous in a pandemic or like nuclear stuff. Primarily like this should be, as you mentioned, partners are joining, like other developers are joining. Primarily this is also a roll call for all startups around the world. Like at the end of the day, consumers are consumers and they're not building the products. And so do you see guys, any active startups who are trying to kind of switch or change gears right now? That's the best time for startups ever. There's never been any better time for startups. If any startup person would tell you, it's like they start complaining about things, they're acting like they're working for large companies or whatever governments. This is the time to take a stand and truly benefit from the time that you gain per day for not traveling, for not commuting and work on your startup, focus and pivot. This is the time to pivot. This is the time to actually focus on things like, you know what, fuck you Ethereum, I'm not building on you anymore. I wanna build something that can actually truly scale, can be truly efficient. All the good ideas and the people that we have with us, we wanna get them something that can deliver on the promises, deliver to the investors and so on and so forth. This is the time for people to do things that can actually work. And so let's talk about more about, so the mission that you guys have is both an organization, but is it also like a technology stack that you guys are building? Kind of one of the more clarity on that. There's something called the totality. It's a separate from the DIA. DIA, honest to God, we have no way of our own money and effort into it. And if anybody's trying to participate to tell you otherwise they're lying, at this point, anybody who wants to participate in DIA is a volunteer. If they can volunteer, they can help whatever. We welcome any help to help us drive it forward. What Ben mentioned is that this hackathon was mainly organized by two sponsors in DIA that Ben brought, which is Singularity. He's the other foundation and one other foundation. Ocean Protocol. And I was in this hackathon. I might actually just be contributor in the hackathon myself, like as a developer. One thing that a lot of folks who can participate in this hackathon start building things that bring storming and so on and so forth. They can bring their startups into it. It's a remote hackathon. There are a lot of people that can do it from anywhere they are. The cool thing about working in environments like that, you end up collaborating with a lot of other folks and trying to complete instead of compete. You can also compete, but you can complete each other as well and become stronger. We truly believe that there are a lot of solutions, a lot of opportunities right now. And this is the time to stand up and truly make a difference in the world. This is the time to do something that is extremely relevant. Before we get to the next huge pandemic or problem that is in cybersecurity and pandemic, Ben mentioned yesterday, like if essentialize AI, somebody can get in and create a virus and knows all of those methods into how to prevent it, but using AI to make sure that it doesn't get prevented. All of those methods that people they're working on end up bypassing them and killing 70% of the population. Is that a possibility? It is, I've seen case studies where people can alter pacemakers on humans. Right. So think about that. I mean, there's no doubt as nasty as COVID-19 is. A lot nastier pathogens are possible either naturally originated or synthetically engineered and it's clear that right now our best ways for dealing with these things are big tech companies and big governments. And it's clear this is bad. A, because there's danger in giving more and more power to these centralized entities. B, because these centralized entities aren't that good at dealing with this stuff actually. I mean, they may be better than anything else at the moment, but they're not tremendously effective. So I mean, we need far more agile, effective and flexible ways of leveraging the world's intelligence to deal with problems like this. And we need ways that don't have this minor side effect of concentrating more and more, oligopolistic and autocratic power in the few entities. And the good news is, we have the basic tools for that. We have Tota protocol that Tufi and his colleagues have developed. We have SingularityNet platform and we have ocean protocols code that's being used in COVIDathon and dozens and hundreds of other amazing software protocols and packages. And there's a huge number of people with the tools and the ability to leverage these things to make software and hardware that will allow decentralized networks to deal better than centralized systems with ever nasty things that are thrown our way, but effort needs to be focused. And I mean, the COVIDathon hackathon is one step in that direction, certainly very far from sufficient. So I mean, if we want to be sort of rose-eyed optimists here, what we would say is, well, maybe this COVID pandemic is just severe enough to get people off their asses and make people activate and start developing alternative tool sets, networks and mechanisms so that when something even nasty or maybe it comes around later, then we have decentralized networks that can really help. I mean, Silicon Valley people are always talking about disruption, creative disruption. Well, COVID-19 is certainly disrupting things, right? We're seeing a hell of a lot of disruption right now. And this can cause you to crawl into your closet and curl up with your phone and just wait for the thing to end. Or it can cause you to look around and say, well, what are the interesting opportunities that are offered by this chaos and this reorganization? Yeah, a lot of folks around me, they stopped using the word viral. Yeah, you don't want to... Hey, let's spread, let's go viral. And yeah, it's interesting time. I truly encourage every single person to truly take action now. This is the time to do a lot of things that can change the world to the better and make tons of money doing it. It happened. We can't really go back in the past and say that there's no COVID-19. But one thing I'd leave you with, Amir, I know we've scheduled half an hour for this, is something that people may want to remember what's being done when you look at the history of the planet and the impact of something, how far it can go. COVID-19, without any doubt, restricted a lot of people from going out of their homes, restricted a lot of freedom. It's going in that direction where it's even enabling a lot of folks to build measures to restrict freedom, as if it's, sometimes I use that example, what's the best and most efficient way to kill cancer? You kill the host. Let's not kill the host. There are methods to combat things like that without really destroying our nature and becoming bad people. Like Yuval Nawa Harari wrote a phenomenal article about how we could build solutions without really killing the host, killing who we are, changing who we are, becoming bad just because, oh, we need to solve a certain problem. There are better ways to tackle this. So that's from one perspective. From another perspective, I'd like to remind everyone, what's another thing that existed on this planet that not a lot of people, nobody really knows where it came from as well, but it has the complete opposite of COVID-19. It was built to liberate people and it starts going across the globe and it's been attacked many, many times and so on and so forth, Bitcoin. So just think about it this way. Like, there's something that is being built to liberate people, to be better for the people and there's something that is against the people. Which way do you wanna be as a listener in what's the next thing you wanna do to make it better? How can people help you guys or get involved with your guys' initiatives? Well, I mean, from the DIA perspective, if folks that can, you know, join hackathon, they can participate, they can do volunteering, you know, there's a lot of things that needs to be done from the DIA perspective, frankly. When it comes to many of the other initiatives of how we're driving them forward, one of the things that the IEEE, are you familiar with the IEEE? It's the Institute of Electrical Electronic Engineers. So one of the things that I've took on recently with them is the Global Chair for AI Standards. IEEE is 470,000 engineers globally. It's a huge, huge organization. So we're trying to actually build those things to get to fruition sooner rather than later. We're active, we're not just like sitting and talking, like we're doing a lot of things. And at the same time, on the company's level of organizations, there's a lot of movements that we are building. And one of the things, for example, we're helping with the Bitcoin social scalability. It's called BAT 53. It's still kind of stealthy, but people that they want to find out about it, they usually, you know, reach out and find the ways of like, hey, you know, how can I help with this thing? How can I get to the Bitcoin scalability from social scalability or whatnot? And from technical scalability, they can reach out from other different perspectives. So we're here. We're big believers in what we're building is an additional thing to every single thing that exists. So it's an added value. And we want to benefit by adding that value from Delta. And welcome a lot of folks to join us in any capacity that they can. What do they do? Every single person, you look at yourself, what do you do? No matter what you do, if you really want to help us out, you reach out and tell us, this is what I do. And we will have room for you to help. Thank you, Dufu. Ben, any last words? I think there's a lot of ways that everyone can help. And of course, it depends on your expertise area for anyone with a technical background in software, mathematics, AI, biology, project management, you know, writing, making videos, whatever it is. There's COVID-a-thon, which we're just launching right now. And there's a host of different projects there. And even if you're not great at coding, I mean, we need people to write blogs describing the results of, we need videos made describing what's happening and why. I mean, when you really come down to it, almost everyone has a skill. People can even run podcasts, right? I mean, there's an endless number of ways that you can use your abilities to help. And probably the first thing is to get more informed. I mean, you can, there's just a lot to know to understand how decentralization works, what AI is all about today, whether Google and NSA and the world government is really up to and the degree deep, the more you learn, the more concerned you're gonna get and the more curious you're gonna get to learn more and the more obvious it's gonna become to you how many ways there are that you really could help, how much need there is. I mean, if you look at what humanity is spending its resources on right now, I mean, on the one hand, people are enjoying themselves a lot of the time, which is good. On the other hand, there's a lot more resources being spent on, say, reality TV, trashy throwaway t-shirts with stupid messages on them, like my singularity net t-shirt here. And I mean, chocolate ear, chocolate and SUVs that burn gas more efficiently, much less resources being spent on, say, medicines to stop people from dying or to help people live forever or, say, making sure the superhuman AIs we're creating are gonna be beneficial instead of turning people into spare hard drives. So I mean, there's not as much attention going on to the things that are really critically important to the future of our species. And actually working on these things is even more fun than eating chocolate ear chocolates and watching reality TV once you really get into it. So I think everyone's got their own talents and their own skills. And once you learn what's really going on, you're gonna see a lot of ways to cooperate. But the COVIDathon that we're launching with the decentralized AI Alliance now is certainly one way to get a foot in the door because there's a bunch of new projects just starting now that could use participation from people with all sorts of different skill sets and orientations. Awesome, Ben, Tufi, thank you so much for coming on the show, sharing your thoughts. I wish you guys all the best. Good luck with the COVIDathon this April 1st. I'll make sure that I have show notes. This is going live within 24 hours. Anybody listening, go check it out. There'll be links in the show notes. Thank you guys once again. How beautiful, yeah, thanks for having us. Keep up the good work, take care.