podcasts

The World’s Most Used DApp, You've Never Heard of, with Alex Newman of Human Protocol

podcasts

The World’s Most Used DApp, You’ve Never Heard of, with Alex Newman of Human Protocol

November 2020

Posted by

Roland Spencer

Creative Producer

Alex has made major contributions to OSS (Open Source Software) for over 20 years and launched HCAPTCHA, a CAPTCHA tool which now blocks malicious bots for more than 14% of the world’s traffic. What most people don’t know is that it is powered by Human Protocol, a knowledge marketplace leveraging crypto for 100m users worldwide, where machines post jobs for humans.

We talk about how they’ve achieved a scale that could make them the most used DApp on the planet, privacy and the open knowledge economy, as well as the composable Open Data Stack that will power Web 3.

Posted by Roland Spencer - November 2020

November 2020

Posted by

Roland Spencer

Creative Producer

Listen on iTunes

Jamie Burke 0:13
You’re an early stage web three founder apply to our award winning accelerator programme base camp at Outlier ventures.io slash base camp, we write your first $50,000 check and give you access to 200 mentors, including many of the leading web three founders, and a network of 1000 of the world’s leading investors and exchanges. We’ve helped over 30 startups from 15 countries from all around the world, raise 100 and $30 million in crowdfunding. It can help you fast track product market fit, and where relevant the launch of your token economy. So today, I’m really happy to welcome tech lead of the human protocol, Alex Newman, welcome to the show. So Alex, noticeable that you don’t refer to yourself as a founder or co founder, really, I guess, having spoken to you offer because you see yourself as a kind of core contributor, and the leader from a technology perspective. But nevertheless, I think we’re gonna get into a little bit about the nuances of being a founder to crypto project compared to being an active participant in open source generally, that’s something that’s been been throughout your career. So at a high level, human protocol is a globally distributed knowledge market. Human protocol powers distributed marketplaces, multiple of almost any variety, I’m sure we’re going to get into the limitless potential of that, interestingly, for humans to contribute their reasoning skills and knowledge to help machines, which initially sounds counterintuitive, we always kind of think that actually, it’s the machines that are helping the people perhaps even displacing and that’s the kind of collar narrative. And I think this is a this is a really interesting counter narrative to that. And effectively, it’s a marketplace where machines can post jobs for people for humans. So several reasons why I’m really excited to have you on the show. So firstly, you’re a very well respected engineer, and founder in serial founder with a focus on distributed systems way before human protocol. But But consistently, there has been this blockchain context. And I know you could be regarded as a as a crypto Oh, gee, we get into that a little bit in the origin story. You founded several companies, tech companies, you’ve had an exit, you’ve been contributing to open source systems for and software for over 20 years, including Apache foundation and various others. There’s committable, core committer, you helped launch haitch capture.com, via intuition machines. And again, we’ll we’ll unpack that a little bit later, which is effectively an anti bot solution. Most listeners would have experienced perhaps not knowing glitch, but would have encountered at some point in their journey about the web. And effectively it protects their privacy. And is one of the most popular recapture alternatives. And so just to give context, the scale of Avast access. And this is of course, all leveraging even protocol, perhaps unwittingly, to most most users, but this now blocks bots for more than 14%, estimated 14% of the internet, which is insane. So you know, it must be a crazy thought to think that you’ve created something that has such impact on the internet population, global Internet population, right?

Alex Newman 3:41
Yeah. And that’s, that’s really just we caught that number. But that’s that’s actually just one of our customers. That gets us that number. We have 10s of thousands of big companies signing up all the time, it’s actually quite difficult to know, what percentage of the overall internet that it is. But yeah, I mean, it’s basically every country, every territory, places that are recognised by the UN. It’s been, it’s been quite exciting.

Jamie Burke 4:08
Wow. And so you know, this is important because h capture is the flagship application for the human protocol. But really, it’s leveraging the human protocol, that marketplace for the labelling of data. And it completes millions of tasks every day. So human protocol itself already has over 100 million users across 249 countries and territories. I don’t know how many there are globally. Surely, that makes it one of the biggest daps on the planet? I mean, I don’t think many people would realise that. Right. But it’s, I mean, that’s a that’s got to be by far the most used app on the internet.

Alex Newman 4:48
Yeah, and the biggest issue, of course, is getting since we are Ethereum base, getting that transaction rate. One more reasonable means more and more bulk updates and aggregations. But I mean, as far as transaction count. Since we’ve been on the side chain, I think we’re running two or three x what maker Dallas doing. Wow. So that does help us cut the costs. We do have a modified Well, I don’t want to talk about the economics. But we have a we have a lot of things to kind of make that more efficient to run on the blockchain. But we always joke that it’s it’s by far the most popular blockchain app that crypto people didn’t realise they were using. Right. So most of the users of this protocol don’t realise it’s completely backed by a crypto market in terms of what jobs get launched and, and how it gets monetized. So I think I think the one thing, the one thing I want to point out there is that, you know, you’re right, that it’s a work pool, and we provided this descriptions. But at even more kind of primitive level, there’s never really been a way to quantize human labour. And it actually wasn’t that important, until robots came around. Now that we have humans interacting with machines, this quantization or measurement allows for more advanced recruiting of resources, more efficient distribution of resources. So as much as quantizing human labour at first glance doesn’t seem as cool.

Jamie Burke 6:13
Sounds, dystopic almost, yeah, exactly. It’s

Alex Newman 6:15
actually it is it is a bit stuffy. But it enables a level of interaction between humans and computers, that primarily empowers humans. So you know, a good example of this is we don’t worry about eyeglasses. You know, taking over the world, although we are cyborgs when we wear eyeglasses, right? That is machine and human interacting in this way. And so for us with the human protocol is being able to create the most generic interaction between those entities, that assures the safety of the parties involved. So do we quantize human data, but we do it in a way that preserves your privacy, we do it in a way that that gets you paid for giving up your data. Um, and so I think, you know, I just wanted to point that out that this this kind of aha moment, once you realise, wow, human labour was never divided this way. This is the first time we’ve done this. That’s super important for machines to be able to digest and use and sell that stuff. And it seems like such a small thing, but I think it’s, it’s actually like a remarkably important and very bizarre that.

Jamie Burke 7:27
Yeah, and it’s really interesting, because so we did an investment. A few years ago in Canada called botanic actually, on the west coast, founded by a guy called Mark Meadows, I think it’s 2017. And they ended up open sourcing it. And the project was initially called seed. But their idea was, I guess the starting point coming back to h capture was that these botnets were out of control. And people could increasingly not trust their interactions with the internet, which was increasingly automated, which was being the interface was a bot, how can I trust the vault is what it says it is representing who it says it is. And the exchange is equitable, or at least the terms of service are codify by the terms of the engagement between the bot and the person to codify it. And I have no idea as to the scale of the problem. I mean, intuitively, you kind of do. But I guess it’s one of those things you don’t think about. And I remember earlier, you were saying, you kind of started out in security, but it was one of those topics where you were basically just scaring everybody going around scaring everybody. And and so I think once you start,

Alex Newman 8:36
I was part of an organisation that that did that

Jamie Burke 8:40
I did. And one of the interesting things that he said, which surprises me, is that so I can understand. And I know, we don’t want to spend too much time on ot nets and H capture. But I think, given it’s had such traction given it is probably the most used app on the planet. I think it’s worth trying to understand why why that level of success. So you know, you mentioned that most people in crypto, and people who aren’t in crypto are using this thing, and they’re not aware about it. And I i’ve always argued that’s exactly what Crypto should be. You know, I think people knowingly, you know, most people don’t understand some of the fundamental protocols of the internet. And that’s because they work really well abstracting away that kind of complexity. So I think this is a great example of what most protocols should be aspiring to. And so it’s going to be interesting to see how how you think that’s happened? Whether by design or organic organically, but the thing that surprised me was you said that so I can understand institutions understanding the scale of botnets and bots and bot traffic but consumers average consumers using the internet because I’m, I’m until speaking to you at least Anyway, I’ve I’ve not been convinced. I know it’s important to me. Privacy is important. I know. bots are a problem. I don’t know if most people in my world care about these these things. And so I think, again, it’s really interesting that you say there is demand, and that by people adopting something like this, it’s kind of got the principles of decentralisation baked in. It’s got the principles of user centricity baked in. And you know, privacy, preservation. But so so how are the average internet user, like coming across this aware of it? Because this is the first time I’ve heard a founder say, you know, people are understanding the problem and adopting it based on the problem?

Alex Newman 10:38
Yeah. Um, so it’s interesting. Now, part of that may just be because of the internet community, I’m a part of, there was a while where I helped the guys that brave working on their, their browser. So I know a lot of privacy centric people. I’m incorporating corporations, and like, you know, I brought up Europe, we’re starting to see more regulations, they’re a lot more more sensitive there. But I think, um, and I, you know, the last thing I would ever want to do, and, you know, I don’t even consider other capture providers, necessarily competitors. But I think, what will what’s happened online, and I actually think that it might have a good deal with Corona is that people are starting to realise that their in person interactions are very different than the interactions they’re having online. And on social media, they have a different tone, people interact with each other differently.

Unknown Speaker 11:37
And I think a lot

Alex Newman 11:37
of it was just people kind of writing it off that, oh, it’s people online, you know, they’re bad and stuff like that. But I think we were just kind of overwhelmed by only an internet centric contact with COVID. And then people kind of went outside and they’re like, Wait, hold on a second. People are actually really nice. Look at this guy. And, you know, the way that he relates to me, his body is saying that he cares about, you know, how I’m moving. And he’s, you know, like, and so I and so I think what kind of happened is we were kind of drawn down the wall of despair, that this was humanity, we all popped out to have our summers. And then we’re like, wait, human beings, they’re pretty good. And I actually think that Zeitgeist is a huge wake up call for the whole world that the internet has, has an attitude problem. And the thing is, people are quite confused about what’s causing this. But when they actually enter a realm, that respects their privacy, that block bots, effectively, these kind of smaller communities that I’m starting to see people pop up all

Unknown Speaker 12:42
online.

Alex Newman 12:44
they realise that a lot of these interactions are fake. And then, and then when they actually bring in, you know, the caption, and they’re actually blocking bots, and all those types of things on those in those communities, they very quickly see how things change. So I really think it’s just certain things just get in the air, you know, like the million monkey theory. And a couple years ago, I don’t know what it was, because I don’t think Snowden was enough to do it. I don’t think anything was really enough to do it. But a couple years ago, especially in Corona, this stuff became more and more important to people. And it’s hard for me to say why it came into that Zeitgeist. But it’s overwhelming. Like, when we talk to our customers, we talk to people signing up, you know, they say, I don’t want to give my data to x, you have it. And that’s great. And secretly behind the scenes, I’m like, well, we don’t technically have your data. It’s just the human protocol. And you can get decide which data goes on there. And people have control their own data. Um, you know, I, I really do think it’s that reaction and resignations going on right now. But there’s something in the air when you say to someone, hey, you know, there’s a thing that you’ve got to go through all across the internet. Do you want that tracking which website you’re on? Um, nowadays, people are getting a little bit cagey. So I wish I had a better answer, but it just seems like the psychosis ship.

Jamie Burke 14:00
Well, I’m glad for you that you’re riding that wave. Because I mean, the level of adoption is is is insane. I mean, you know, for any founder we’ve had on here, I don’t think any of them could come close to that level of usage and adoption. So

Alex Newman 14:15
popping up there just for a second, you know, what I what I’ve seen, once again, on the capture side, is the adoption, you know that the privacy thing is there that came from the music, but literally, all of the features that we have a lot of fancy features are captured and other captures don’t have. So they know what they wanted for a long time. The real question is, is was there gonna be a vendor who’s incentivized to provide the consumer with what they want? Because the reality is, you know, and I don’t want to say anything bad about my competitors. But there are lots of people who use other capture solutions, who monetize based on the amount of users they have. So they bring in a stronger CAPTCHA solution, and now half their users are designated bots. That’s not good for them. Bottom line. And we we did a story about that a couple years ago about how we compared to recapture. And I think this goes back to one of the stories, which is like, Are we going to let the incentives of kind of driving us to these big companies be what dominates the internet? Or do we want to decentralise things do we want to kind of say, okay, no one’s going to own all the data, you’re gonna have control of what data you put out there. In fact, we’re gonna pay you, we’re gonna get out of your personal data, um, you should be paid for that. And that, you know, it’s it’s open source, there’s a community that will stand behind it, security and privacy. Um, you know, that that story, um, is very powerful for people. And I think that overall, you know, we’re going to see, not just CAPTCHA moving MySpace when the entire internet. And I’m really hoping, you know, web three, which is something that you are excited about, you know, I,

Unknown Speaker 15:57
I don’t

Alex Newman 15:58
know what tribe I am in crypto. But I’m also super excited about web three, I think that it’s clear that, that these are some of the applications that pop out of it. Um, what’s funny about the human protocol is it’s very non intuitive. monetization of human labour. There’s probably 100 other things that are not intuitive like that.

Jamie Burke 16:16
Yeah, absolutely. And so you know, as you said, at the beginning, h capture is just one part of it. It’s and it’s so fascinating that it’s easy to focus on. But, you know, this is this is generalizable. So it’s a, it’s a generalizable distributed knowledge market. And so I just want to kind of go into like, can we explain how it functions, and you mentioned word incentives. So it’d be good to understand how you’ve applied incentive design, and presumably, this is the crypto piece, right, how that incentivizes different behaviours, or optimises, your network versus another in the context of a given use case.

Alex Newman 16:53
That’s right. So what we’ve done is we’ve optimised our payout though, based on the quality and quantity of your traffic, there’s more about that image capture website about about the details, they’ve done much that you get paid more or less so so basically, the more bots you’re bringing into the system, providing inaccurate labels, that kind of harms the amount that you get paid. And so we’ve created this incentive system where, you know, we’re driving real traffic will protect you from any traffic, but what we really want to do is encourage the sites to have more and more human behaviour. And, you know, I think, when you look at the human protocol more generally, there’s kind of, like, we often talk about the pyramid of different types of users. So you know, anyone can solve CAPTCHA. So we we designed that to not stress people out, not be something that provides special training. Um, and there’s a lot of interesting things to pop out of capture, right? Like, you know, you ask someone to something a short sleeve shirt, you might get a different answer in Ireland versus Saudi Arabia, right, that tells you something about probabilities, and the subjective things, but, but if you but if you imagine a pyramid, when the bottom is captured with the least trained people, and that’s going up the pyramid, there’s more and more training. So we had a bulk user programme for a while we’re giving people much more advanced puzzles to solve the art system. Um, we can do the same thing. So so you know, as you, you know, the what the protocol decides in the papers that, you know, the bottom is kind of the least train the top, you know, you could imagine a world you know, I’m excited about a world where we have radiologists in that top square, right. And so, you know, kind of as you go up, you know, there’s less and less verification less and less things pretending to be bots, unless someone has a trained radiologist spot, which would be pretty awesome. Yeah. That would be that would change the world. You know, I hope for that. Yeah, exactly. Start out nice, right? But, um, and so that that’s, and they all fit together, right? So capture has no biases. It has no training, right? Anyone up here on this training proportions have biases, which is their training their background, their country? So it’s actually like, when you think about this pyramid, to get good answers, you really want to sample people up across the whole pyramid, to kind of construct probabilities to handle robots about things. I should also mention that in our world, when we look at a photo, and there’s like a picture, let’s say a truck in there, right? Um, you know, from from your point of view, or a catchy point of view, they’ll be like, Oh, for sure there’s a truck in there. And then 60% of people be like, for sure there’s no car but other people think trucks or cars, right. So, you know, it’s one of those things where

the answer you get from capture is very powerful and without bias, and it helps make everything else more accurate. But the reality is, is that most of the really powerful value comes from the intermixing of experts and different levels of verification. communicating with each other in a network. And so that once you get into that world, it’s not just labelling that you can support. It’s actually, you know, I don’t like talking about future applications just but I’m excited about this because I don’t want I don’t want to be Mr. vaporware, you know, I want to be I want to be like, here’s what we’re really building. Um, but you know, what, when I get really excited about applications, it’s, it’s kind of building stuff across that stack where they can plug into it, and realistically, in an open source manner. So, um, you know, when we talk about why trust us things, it’s not just because of the privacy story, it’s not just because of the stuff we’ve done for security. It’s also the fact that we, the human protocol Foundation, kind of, you know, I’ve took some of the tricks from the Apache Software Foundation on how to create a real independent technology community that makes decisions for the people based on technology, what’s going to support them. Um, so, you know, when you when you think about that, that, that infrastructure, we’ve got kind of, you know, there’s no one who’s kind of controlling, there’s no one, you know, there’s there’s kind of this meritocratic technological organisation, like you would see with Bitcoin kind of pushing, pushing those types of things forward. And then finally, just by, you know, we got the act of, you know, how these things get decentralised it, it kind of gives us whole story where, you know, people can join the human protocol foundation contribute to it, that’s basically the same level that I do. If you look at the people who are working on the human protocol Foundation, who we’ve hired, almost all of them came to us via bounties, open source bounties, like literally posting issues people came in, we said, Hey, you got a couple good issues done. Let’s bring you into the foundation. So and so it’s very open source is a lot closer to what you would see with community like Kubernetes, the Apache web server Foundation, where there are people all around the world that really have this shared vision for humanity, that are coming together and open source to develop this tool where the network gets all the value.

Jamie Burke 22:04
And so to try to simplify, grossly simplified down, and then I think we can come back out, but I do want to talk about some of these future use cases, perhaps towards the end, because I think that’s it’d be, it’d be a real waste to not go there with you. But like, in its most simple form, I think you described it, there are there are some jobs that for now, anyway, humans are just better that better at than machines. And so that could be trade training, a model, a training ml model, right, or it could be interfacing with a bot, a customer service bot, so it gets better at delivering customer service just to make it tangible for for the audience.

Alex Newman 22:45
Yeah, I think I think the way to think of it is, we only send out the puzzles that we can’t solve with them. The images that people are seeing an H capture weren’t just like, arbitrarily picked, we serve the images that a state of the art machine learning system can’t label. And then we can use that to retrain our, our state of the art machine learning label. So the issue isn’t just that there’s a certain percentage of words that cannot, you know, handwritten words that can’t be translated, there’s certain amount of images that can’t be classified or pixel broken down. Geoffrey Hinton, kind of the godfather of ml, you know, says, you know, often talks about when we get, you know, perfect at the letter A, like, if we could just get machine learning to recognise letter A, as well as a small child, then we’re set, you know, I’m already set. But you know, that’s real progress, right, like, so I think she learning is so amazing, can do all these tools, all these things, and it works way better than you expect people who work with them, the first thing they’ll tell you work way better than I ever expected. But it’s still fundamentally is, is it’s not as smart as us, it’s not as ingrained in the world. So what we are doing with the human protocol, is we’re teaching the computers about things that it knows the least about. And what’s fascinating, is, without getting into the details, because we know how to get the computers to tell us what questions that knows the least about and those are the things that are being served out the human protocol. And so yeah, you know, it’s actually to run human protocol, not even the captcha system, but just human protocol. In general, you need one of the most state of the art machine learning

rigs in the world.

And then effectively, that’s how we decide what work to get out.

Jamie Burke 24:33
So if we look at this, this pyramid is he describes his knowledge pyramid and the bottom you got all this low level categorization, basically, maybe categorization is consistent throughout, it’s just the level of complexity on or specialist knowledge, the scarcity of knowledge that can carry out that task. So as you say, you know, it could be this very basic categorization for capture at the bottom and then as you move up to the top or radiologist where there’s maybe only a handful on the planet, that Could have the capability to detect, you know, particular patterns. And, you know, presumably, at that point once, once the human or humans have trained the ML to be able to carry out that activity, that knowledge can then be. I mean, it’s quite a loaded word, but I can’t think of a better one democratised. Right. So, as you said, it could be extended to any any hospital anywhere in the world, including Sub Saharan Africa. Right. And so that’s like a very powerful concept. But then some some questions start trickling through, it’d be good to understand how you how you solve for these, or at least how you’re considering them human protocols. So. So on the one hand, you’ve got people, so around this knowledge economy, you have participation, people carry out tasks, they get paid for the tasks, and there’s a premium placed upon the the bar of participating. But what about ownership? So beyond that transaction, which could be very temporal, most of these will be temporal until the thing is trained to carry out that on its own behalf. So how do the people that participate in the system also share in the ownership of the human protocol success as a whole?

Alex Newman 26:22
Yeah, that I love that question. Um, so the interesting thing about capture is, users of it don’t have a lot of choice. So their reward for captions access to the website.

We’ve looked at, you know, we would love

reassessing how that money is paid. Once we went to, let’s say, you could imagine world of theorem to one of these other chains, where we can put a lot more stuff on chain, even imagine the user being directly rewarded at that point. Um, but for now, just because the nature of chains, you know, on the Pacha story, it’s really the websites, you can think of it as almost like an alternative to ads are making money by putting our cash up, the other users of our system won’t have that kind of abstraction, most of the time. You know, you may have kind of hierarchy in your work pools. But most of the time, the more that we can kind of know exactly the user who’s labelling in those work pools in a safe privacy preserving way, then we can effectively reward that individual and get them paid and all that stuff. Um, so I think, you know, kind of as you go up the the pyramid as well, probably the payment becomes more direct as well. Because you can imagine, like, you know, at the tippity, top of that pyramid, you know, you’ve got some trained expert, you know, his name, probably have a contract. Right. And so, overall, you know, I think that we would like to see everyone in the system rewarded the winner, that’s physically possible, just because we think that’s gonna have the clearest best set of incentives. But for now, it’s really as you go kind of up that value chain, that they get more directly control over how much you’re getting paid, what they’re falling, when they’re called, and all that type stuff.

Jamie Burke 28:20
Yeah, and so is the, so you’ve kind of got this unit of account for payment for the transactions that happen in the market. Now, presumably, there’s some scarcity in that. And so it could be considered a form of equity in the network. So if you’re being paid paid out for a task, but you hold it, rather than cashing it out for fear, is that is that not having a stake in the success of the network? Somehow? Or?

Alex Newman 28:49
That’s a great question. I’m not sure I should answer it, though. So the, you know, this goes back to what we were saying before, right, we want some interchangeable way of quantizing human labour, right. And so what are the what are the actions we imagine people are doing with that quantity? Right? So sometimes it’s, hey, I’ve got some of these quanta. I want to get my value out in the form of information. Right. Um, another thing that we’re actually looking at is, okay, I’ve got these quanta. I want to spend that just to demonstrate that I’m human. Right. So basically, the quanta is just a representation between a human robotic interaction. And the question is, you know, what is that interaction? And how much does it cost? You know, there’s a free, you know, who can solve it, and that’s going to get marketed, you know, to the right number in terms of when it gets served. And so yeah, there’s there’s this kind of this order book, right. And, and the more and, you know, you’ve got this this unit of, of human work. And so kind of the more capacity the more interesting jobs you have, the more different types of puzzles going on. The more that type of tasks that we can do increases. So we’ve been mostly talking about labelling, but like I said, it can be any human work, right? Like, there are lots of, there are lots of kind of rituals that human beings perform for other rituals in a professional and non professional setting that can all be done this way. So, um, so so as it becomes more diverse, has become more broad as it can do more things, it becomes more valuable as it gets more work into the network, it becomes more valuable. And arguably, and this is, I think, kind of the weirdest one, as people learn to value themselves more. And the time that they’re spending, when they take control over this thing, and directly are interacting with each other, um, they get a bigger cut, so there’s more more value. So it’s one of those things where our goal is just to make it work better bring more value and make it cooler and all that type of stuff. Because it’s a network effect. Right? The more people like imagine if some of our capture contributors, right, okay, and not competitive, other captures came to us and said, We want to join the human protocol as well. Now, you got to ask yourself, would that make the network more or less valuable to have more capacity? Right? Normally, when you kind of bring more on, you reduce the scarcity, but in this case, what you’re bringing on value. And so you have the same amount of value chasing the same amount of quantization of human labour. And so I don’t want to talk about scarcity or prices or any of that type of stuff. But but but but I do think like in terms of value, you know, it’s clear network effect. And you know, in Silicon Valley, when we see these network effects, we go, Oh, wait, this is an exponential up in value, right? It’s not even linear. Right. And so that’s how I kind of think of how much it’s worth how to how this monetization work, because, you know, ultimately, it’s about, you know, monetizing, or quantizing, this human labour

Jamie Burke 32:04
ideology. So when I asked that question, it wasn’t, it wasn’t to get into actually trying to quantify the value of of the token, it was much more around the concepts of ownership. And so you know, how people begin to participate in the ownership of the network, because I, you know, there’s, there’s the transaction, there’s the public good outcome. And then, you know, over time, you could imagine and the reason why I flagged it is because the project I mentioned earlier, botanic and then seed was the conceptually anyway, they were thinking about the idea of, well, if you will, one participant that directly or indirectly, so indirectly, just by providing your data or directly by carrying out some form of task, trained an AI, then Wouldn’t it be great if you could own it? If you were this the specialist? Well, what you don’t do is make yourself redundant. And what you ideally do is you want to kind of have a stake in the digitization of your knowledge set. And it’s, you know, future success.

Alex Newman 33:13
I mean, so now you’re getting into some of the stuff I love with the Foundation, which is how we’re interacting with the other people in the data ml blockchain space. So I’ve already I’ve already told you that, hey, why am I doing this? I think, you know, decentralisation of this technology is important privacy, the other side, right? We’re playing a critical role within the data and labelling ecosystem within the blockchain. But one of the amazing things that’s happened is there were a lot of blockchain ml startups, and a lot of them didn’t make its way through the sea. And we’re actually starting to see who the real players are. So we actually have potentially great partners also on the blockchain, that can help us with exactly what you’re talking about. Right? So so you know, we’re the ones who are providing the basically the capacity to train the bots, other people, you know, are trying to bring, okay, what are the inputs that we want the bots to learn about? Other people may actually attempt to have this ownership bot model. I’ve seen a couple people trying to do that. And so I think that it’s a little interesting, because I hope you don’t find this too extreme. But I actually think that we could be in a world where ownership of bots is immoral, in a decade,

Jamie Burke 34:32
by corporate institution, you mean so like a, a single corporate entity,

Alex Newman 34:36
if it has the, you know, there’s a small chance that the bots could be as human as us in a decade. And so, you know, what is ownership of men, you know, um, and so, you know, I mean, they’re not as smart as a pet yet. My cats are actually smarter than my bill, but the eyes are getting smarter a lot faster than my cats. And so, for me, they’re definitely getting smarter. Faster. I’m probably closer on the cat side, at this late stage of my life. But but so I think that, you know, right now we’re talking about how does it go? And how is it controlled? And I think there are a lot of people were talking about safety, how do we use AI safety? Who controls the AI? Who’s benefiting from it? You know, as we’re putting people out of work? How do we take care of those people? And so I’m hoping that our economics in terms of Hey, you provide the training, you get the benefit at you know, and that’s valuable, it’s absurd if your benefit, you know, I’m hoping that that works out naturally. But the reality is, is there’s a lot of just kind of drudge work that’s being done by people that, you know, imagine a world where mortgages are 1%, more efficient, right? or medical paperwork with more than one, you know, I know, I’m excited about all of these things, these kind of horrible jobs, being automated, and the kind of the person part of those jobs are being expanded, right. So we don’t need we don’t need to protect people from from drudgery, what we need to do protect people from not having their drudgery, what we need to do is protect, you know, it’s for people that have this role in the world where they can be on their product, right. And so the example I always give is chess players. Garry Kasparov was beating the pants off a computer until, until, you know, with the blue, and at that point, the world was like, oh, man, chess was beaten, you know, go now even go as beaten. The one thing that’s not known or talked about, is that Cyborg chess players beat the pants off of any computer. So you can actually take a adequate chess player, combine them with an AI, and he will beat the pants off of that same AI fighting alone. That’s the future that we’re trying to write just like glasses allow us to see and give us more jobs. Computer AI is our modern glasses that will allow us to see and do a wider variety of jobs. And to the extent that it provides new forms of literacy, mathematical literacy, competence in complex fields, it will create make these fields a lot more productive, and more efficient. You know, if in the end, they’re half as many radiologists but they’re making twice as good a decision for half the cost. It’s a bummer, you know, that, you know, that’s, that’s, that’s a difficulty. But in the end, the people who benefit are the patients. So I think that, from my point of view, you know, the things you’re talking about automated world are definitely something I’m concerned about. But what I’m more concerned about, is it being centralised? Yes, this stuff is coming. It’s actually almost here, a lot of it’s being blocked by legislation, I talked to a lot of my colleagues and other countries, and rightly so, they will have different trade offs in terms of security and how they run their society. Um, you know, my buddies in the UK say, there are cameras in every block, and if their dog, you know, picks a poop in the yard, then, you know, the town council will send him a ticket. And apparently, you know, some old nannies got a videotape of their lawn all the time. You know, in the US, we just don’t do that. Right. And so there are different trade offs that that people can make about how civil society is maintained. And I you know, I’m all for that. The reality, though, is the technologies coming, we have to figure out so that we all get a part of it. And crypto. Is I in blockchain, I think is, is, is clearly working for that.

Jamie Burke 38:59
Yeah. And I think that’s something I’m, I think we’re aligned on in that. You know, I think the idea is, as you said, there’s a free market like forming around these technologies and they’re gonna play out and, you know, you’re providing currently a function within that. So in the in the same way people refer to sort of things that happening in crypto is money Lego, I think here you could refer to as AI Lego, right, this composable stack of technologies and markets, primarily driven by this kind of concept of free market fundamentalism, ultimately, that the market will drive value now so to date, deliberately or accidentally, a lot of the activity that’s happened on human has been directed by you guys and capture. As an earlier you, you alluded to the fact that there are there’s some things that the same about Crypto and open And being a contributor to that, but a lot of things that are very different. So, you know, you’ve had this great start, how do you now as a foundation, allow the protocol, the human protocol to evolve for new cases to emerge in, in a way that doesn’t require such direction? Hmm.

Alex Newman 40:22
Yeah, I have to admit, I was a little naive starting a crypto startup, I think I’ve got my feet under me about what the investors are looking at. I mean, realistically, you know, doing business in Silicon Valley, with people who you know, and have a multi year relationship with a little bit different than doing worldwide business. And so, you know, we mostly got our feet under that, with the foundation side, trying to expand the foundation to be an organisation, which effectively plugs in to the best partners in crypto is kind of what we’re doing now to, there are a couple ways that that plugs in one, there’s lots of great technology coming out in crypto that may allow us to put stuff more on chain faster, all that stuff. So those are, those are important partners that we have, right, just improving that core substrate that we’re running on top of

the second bit. In terms

in terms of, you know, how do we actually, you know, how do we actually work with people is kind of those data providers, so people are providing datasets and all these different use cases we were talking about. I don’t want to name names, but you know, they’ll be more and more people coming out to do that type of stuff. But then finally, there’s kind of, you know, what I really want to do and can’t wait to provide is kind of the the hook up to those legacy infrastructure providers, on hooking into other infrastructure providers, other implementations human protocol. Now we’ve talked about capture some some of the, the human protocol things that we did in previous experiments, some of the stuff from applications we have going on,

realistically, there’s all sorts of

any interaction between a human and computer privacy present, protected and interesting, where we want to save that money. I really do think that or say, or, you know, or capture, capture that data. And I really do think that, you know, I mean, like, I would love for someone to come and say, Hey, we want to donate this new open source analytics stack, you know, like an open source, Google Analytics stack to the IMA protocol. And we want that to join it to capture that human data. Right? You know, and so just, there’s really, I think this is stuff that you and I are both passionate about, is imagining a new internet, where we really do have control about what is released about us. And, and by the way, this shouldn’t be that controversial, right? This is the law in Europe, right? Europe, Europe, you know, the right to be forgotten, right? So you have some control there. And I think that we need to build that infrastructure. Once we build that infrastructure, all the other things that I was talking about decentralisation of the power of data, giving people a share of this new AI community. I think it all falls out of it. I’m not actually like, a free market. fundamentalist, certainly not for crypto. Um,

Unknown Speaker 43:18
but on the other side.

Alex Newman 43:21
I think it’s interesting to consider,

you know,

what, we’re really what what a lot of those libertarian tendencies is all about. It’s really about volunteerism,

and I don’t and not being, you know, compelled.

And I’m not sure about libertarianism, and I know, I think it’s got some good things, I think it’s but but I’m really into the this idea of people being sovereign, making their own choices make their own mistakes. I think that’s the world that people want to live in is at least, you know, and so and so everything about the human protocol, is really about giving people that control. Because when we are interacting with humans and robots, it’s very important that people feel that they’re not being taken for a ride, you know, that they that they really are doing this in voluntary way that they’re putting on glasses, right? They’re not, they’re not becoming a slave to the machine or something horrible like that, right? They’re, they’re just putting on a new shade of glasses. So that way they can live their life and see see more clearly that that’s really the vision that we have and why it’s kind of the foundation is kind of set up to be this independent organisation to enable that, but it is it is primarily a technological organisation. You know, we do have some really good people who have a lot of experience in the data space, Lonnie, we’re talking about stopping with Ops, you know, she was CEO of metal and it’s kind of revolutionised, that whole medical space.

So I think that, you know, we’ll kind of, you know,

have these kind of industry experts who who all have the same view that are trying to pushing this network effect? So that way, you know, that each quanta of human labour becomes more and more flexible and valuable?

Jamie Burke 45:10
Yeah. And look, I think, for me, how I would define or I defined the importance of what I call web three, and I know that can see quite loaded, it’s actually not not just in an Ethereum context. For me, it’s casual. But you know, really, there are three internets. There’s, there’s, or, and, you know, or three webs, right, I, you know, the internet and web, of course, has a bit of a distinction. But you’ve got surveillance capitalism, that is the established norm, you have, as you say, platforms that dominate the web, they have data monopolies, platform monopolies, have data monopolies have AI monopolies. And that is a direction that I think we can all agree from the legislator down to the citizen is probably not gonna win. Well, you have an at the heart of that is shareholder supremacy, the idea that the the right of the shareholder, the company that owns AI is more important than its users, you then have something that’s kind of coming from the east, which is digital status, and that ultimately everything subordinated to the state. And that can be, of course, if benevolent, hugely valuable, because it’s training this, you know, large datasets to train an AI to make society function better. But in the wrong hands could could be dystopic. And then there is something in the middle and it as a European, it feels intuitive, but I don’t think it’s so obvious, which is, which is this web three paradigm, which is that user centricity, user, citizen centricity. And so I really subscribed to your vision, there.

Alex Newman 46:52
I am friends with a lot of people at Google, who work on these problems, and some of the nicest people in the world. And the last thing, they want to do surveillance capitalism, right there, and you brought up all these shareholder concerns, but the reality is

naked as well,

that bury their data is more valuable, when it’s more scarce. Like them just having all the data, that means that each one of those data’s was worthless, right? And so it’s actually scarcity that provides the value of the data. So so. So like, actually, you know, Google and all the capitalist functions of the world. Let’s go over to the status, right. I don’t want to say anything about that a particular country. But these countries that people view as highly centralised are actually incredibly distributed.

Yeah, um,

yeah, you know, there are now more Chinese banks than American banks. They’re more local Chinese banks and American American banks, straight down Chinese banks number straight up. And so it’s very interesting, because think of the West, we have all this the Soviet view of these countries, when they’re quite capitalist. There’s a lot of lot of power there, once again, that just people, you know, and so I think, what I love about the human protocol is, it really is a better world, right? We can get the people who the data barons, we can actually get them more money for the data they have. And they can sleep better at night, knowing that it’s protected, and they’re industry standards, everything’s working world more cleanly. Um, and even these other states that have, in my opinion, serious concerns about public safety, scalability of the government, so their economy, there are real,

there’s real stuff there as well, you

know, it’s not solely a power grab those types of people, when this capability is brought onto a transparent blockchain, it changes everything. It changes, what if, you know, all of this data that people were using, you know, to keep their society safe, was fundamentally controlled access in a way that actually created a new, you know, digital judicial system, one that that is more framed for a modern world. So I, you know, I’m, maybe it’s just because I’m an eternal technological optimist, but I actually think that this is the solution to what you described as surveillance hospitals, this is actually the solution to state control of data. And actually, you know, if we take, you’re in the UK right now, yes, I hear you. Yeah. If you take your example, I think that’s a really easy one to understand. Right? You guys got cameras everywhere to share public safety? I’m right now. It’s interesting, who has access to them? What if that access was a bit more controlled, careful, monitored, maintained. So if some person in the head of the town council, let’s say got hacked, and then all that data got out, it wouldn’t be a huge data leak, there would still be some type of controls, amount of ization and all that stuff built into right. So I think that, you know, before we go all the way to Asia, which you know, I’m always amazed by and so much activity going Not

even just in our boring Western world

that’s moving so slow, like, these are real problems for us as well. And I actually think this is just infrastructure where people can buy in and do the right thing and make money, you know, and when you could get that those two together, do the right thing and make money working together and making more money, the more you do the right thing. You get people doing the right thing more. That’s just by an optimistic view.

Jamie Burke 50:23
Yeah. And look, and I, I think that’s a great place to end I like to end on an optimistic note. And you’re totally right. Of course, this is a universal human problem, which is probably quite apt given the name of the protocol. And I’m really excited to see how this AI Lego forms around an open data economy. You know, we were also an investor in ocean protocol, which is doing some really cool things now with the IDEO the initial data offering have just been testing that a little bit. fetch data either they’re doing autonomous economic agents, and some AI Lego. And then the secret foundation have got secret contracts. So privacy preserving smart contracts, and I can I can just see how the composability of all of these things because they’re all open source, how they begin to interplay. It’s difficult to feel that that can’t be net positive somehow. So I’m really looking forward to seeing your continued contribution to space. Thanks for coming on, Alex.

Unknown Speaker 51:21
Thanks them.

Jamie Burke 51:24
If you enjoyed today’s podcast, please make sure you subscribe, rate and share your feedback to help us reach as many people as possible with the important mission of web three.