[Auren Hoffman] Welcome to World of DaaS, a show for data enthusiasts. I'm your host, Auren Hoffman, CEO of SafeGraph. For more conversations, videos and transcripts, visit SafeGraph.com/podcast.
Hello fellow data nerds. My guest today is Glen Weyl. Glen is an economist and a principal researcher at Microsoft Research and the author of Radical Markets, which is one of the more thought provoking economics books. Glen, welcome to World of DaaS.
[Glen Weyl] Hey, Auren, it's great to talk to you.
[Auren Hoffman] All right, I'm super excited to dive into quadratic voting. And later we can talk about quadratic funding. So I think this is the better building block. For people who aren't familiar with quadratic voting, how do you describe it in a tweet?
[Glen Weyl] It's a system for voting where rather than getting one vote on everything, people get a pool of credits to allocate to the things they care most about. But it's more expensive for every additional credit.
[Auren Hoffman] Kind of exponentially more expensive.
[Glen Weyl] Quadratically, yes. And that means that you have more incentive to have a little bit of influence on a lot of things, than a lot of influence on a few things.
[Auren Hoffman] Okay, but if you super care about those things, then it makes sense to spend your credits for those particular types of things.
[Glen Weyl] And the particular reason for the quadratic is that it gives you an incentive to vote in proportion to how important things are to you. So you get everyone's votes to be equal to basically how much they care. So rather than doing the majority, you have the greatest good for the greatest number basically.
[Auren Hoffman] Walk me through the credit system. I've got a total of 100 credits that I could vote on different issues. Let's say it's for corporate governance, for a public company or something like that. Instead of one share per vote for everything, I've got maybe one credit for every share that I own, and then I could spend them based on different things. Is that what you're imagining?
[Glen Weyl] Sure. And it's not even just necessarily across elections, it could even be across candidates. So you know, the way that we currently deal with the fact that it's hard to have an election among many people, is that you have some parties, and they nominate someone and then there's two candidates. But an alternative would be that you could vote in favor of or against all the candidates. And you could then let the candidate that people on net most support bubble to the top, and that would sort of eliminate the need for this whole primary system and focus on particular parties and so forth.
[Auren Hoffman] Okay, got it. So let's talk about something a little bit more specific, like a shareholder vote in a company or a public company or somebody that has a bunch of different shareholder votes. Some of these are like binding shareholder votes, some of them are non binding shareholder votes, how could you imagine a system like this working in that more micro environment?
[Glen Weyl] Yeah, so I'll give you a real example where it is working right now, which is in hackathon projects in Taiwan. So there's a bunch of different projects that civil society groups put together to get the support of the public and the public sector. And people have these credits. And rather than just voting for the one thing, approval vote or whatever, they actually get to say “no, this is really important, I really want to see the public support this, this may be a little bit, this more”, etcetera. And then they rank all of the projects based on that quadratic vote. And that's how they give out the prizes that then help people get support from local governments to fund these ideas they have for air pollution, or dealing with disability issues, etcetera.
[Auren Hoffman] Everyone in society is a judge, or are there a certain number of judges who judge the hackathon project?
[Glen Weyl] Anyone who's part of the V Taiwan platform, which is their civic democracy platform, can participate, and half of the population is registered in a quarter or monthly active users.
[Auren Hoffman] Okay, so and then how does it work? So you get like 100 credits each and there's 15 different hackathon projects and then you can allocate those credits accordingly? What do you mean by the rank order?
[Glen Weyl] So the people vote up and down on them, and whoever gets the most net positive votes is like the first place. And then there's the second place winner and so forth in the hackathon.
[Auren Hoffman] It's different from some sort of partial preference voting or something like that, where you're ranking. In San Francisco elections, you have a partial preference voting. In concept, what are the big differences? And what would a partial preference voting lead you that a quadratic voting would lead to something different?
[Glen Weyl] Yeah. So partial preference basically tells you the order, but it doesn't tell you how much you care. So like, it could be that like the first one you really like, and everybody else, you're basically indifferent. Or it could be that the first three you really, but it just doesn't really show you what actually matters to you. It just shows you an order of things.
[Auren Hoffman] Got it. To one person one vote in a partial preference? A simplified version of quadratic voting would be me getting 100 votes across these 15 different teams, I could allocate some teams zero, I could allocate one team all 100, if I really like it. Why the quadratic? Why is that?
[Glen Weyl] So when you do what you said, most people do exactly what you said, they choose their favorite thing, and they put everything on that. And that doesn't give you nearly as much information as when you have an incentive to put something on everything, but not everything on anything. You want to get not what's your favorite, but exactly how much do you like and dislike each of the options. And that's exactly what quadratic voting draws out, because the first vote is really cheap. The second vote is more expensive, the vote after that's more expensive. So you don't want to ever put everything on one thing, because that would be a waste. But you also don't want to just be completely indifferent because you care more about certain things. And this pulls out exactly how much you care about everything.
[Auren Hoffman] Got it. To basically buy one vote, it costs one credit, to buy two votes, it costs four credits to buy three votes, it costs eight credits.
[Glen Weyl] Nine.
[Auren Hoffman] Okay, got it. And then it kind of scales out in 4-16, or something like that. And it kind of scales from there?
[Glen Weyl] Exactly.
[Auren Hoffman] Okay, interesting. Okay, so it's in this hackathon, but where else do you see this? The United States is obviously not going to change its presidential election tomorrow to this type of thing. How do you see this evolving in the economy?
[Glen Weyl] I'll tell you some of the other applications that have been really exciting. So one is the most popular strategy game of all time, Civilization VI. I don't know if you've ever played that. The latest expansion pack for Gathering Storm uses this as its diplomatic voting mechanic. So when you’ve got these countries, and they're deciding on global policies, they use these vote credits and quadratic voting to decide on what the global policy is. It's a game played by like eight million people or something like that.
[Auren Hoffman] You're having all these people in Taiwan that are getting used to it, you have all these people playing like Civ VI that are getting used to it. If more people like it over time, you could see a sense how this could evolve into many other types of things, including maybe even things in our democracy or other ways that we're voting.
[Glen Weyl] But the other thing I found quite remarkable about it is, and I didn't expect this, honestly, we haven't had, as far as I can tell, a single negative experience, where people using it like didn't think “wow, this isn’t an improvement over what we were doing before”? That's pretty hard for anything in politics, like politics is an area where almost always there's dissensus, you know, and so that's been really a fun element of it. The Colorado State Government has been using it for a bunch of purposes, they've been using it to allocate the state budget, and to make a lot of executive branch decisions. There's all kinds of examples.
[Auren Hoffman] Is there a way somehow you can bring this into a prediction market? Prediction markets tend to work pretty well in the middle, but as you get to the tails, they tend to have weird things that go on. Could you imagine some sort of quadratic voting in a prediction market?
[Glen Weyl] Absolutely. That's a great example. When you get in a prediction market, if you think that it's more likely you bet on one side, if you think it's less likely that on the other, but you don't actually get people's probability estimate, you just get whether they're above or below the current number. And so you then get some sort of weird weighted average of what everybody thinks, right? But you might not want a weird weighted average, you might want the median or you might want some other statistic of what people think and you can't get that out of the prediction market. All it gives you is this weird weighted average. You can actually do a quadratic scoring rule to elicit people's full probability estimates. So what you basically do is you say “you can buy from me $1 if this thing happens, but the amount that you pay is a quadratic cost”. It turns out then that the amount that people buy will always be their probability estimate. Now, that requires some subsidies to run that market. But anyway, normal prediction markets need to put some liquidity in the market to make it work. But that's actually a better way. I mean, it elicits more total information to use that quadratic scoring.
[Auren Hoffman] And if you're going opposite, let's say you're having an auction system. Is there some sort of way with quadratic voting where you could do a more fair spectrum auction or some other type of system?
[Glen Weyl] So quadratic voting is really for collective decisions, it's good if we all share information, or public goods, allocating public budgets. Spectrum tends to be a little bit more of a private good, it's something that you want to allocate to one person or the other. Rather than quadratic routing, there's a dual opposite system, which is this thing called the common ownership self assess tax, there are different names for it. It's not gonna sound like quadratic voting, but there's a sense in which it's kind of like the opposite of quadratic voting or the reverse version. And what it is, it's a system where you own assets, you self assess the value of them, and you pay a tax based on that self assess value, but you have to stand ready to sell it to anyone at that price that you assess.
[Auren Hoffman] And why is that? I understand the theory, why is that the opposite of quadratic voting?
[Glen Weyl] Because it elicits, truthfully, that private value that you have on that thing in the same way that quadratic voting elicits a value you have on a collective good. And there's this economic theory around the notion that money that comes out from taxes on these private goods should be used to support those public goods. So there's a sense in which they form like a whole system with each other.
[Auren Hoffman] You've recently proposed a system of quadratic funding. And I was having a little trouble following it. So can you kind of help me walk through what exactly is quadratic funding? And what big problems are you trying to solve with it?
[Glen Weyl] The idea of quadratic funding is basically the spirit behind Kickstarter. Kickstarter is supposed to be a way of democratic funding. But the problem is, it's not really that democratic because only certain people have a ton of money to give to certain things and not to others. And then similarly, you can have a charitable funding thing. But ultimately, it privileges the people who have the money to give away, right? Quadratic funding is a system like that. But that's much more genuinely democratic in character, but not for egalitarian reasons. But to get at the issue that motivates Kickstarter in the first place, which is that there are public goods, or these things we share, someone makes some t shirt and someone can buy it, but main thing is that it was created in the first place, the game was created in the first place, this open source software was created in the first place, the journal was created in the first place. And in those contexts, there's what's called a free rider problem. Nobody wants to contribute, because they think it'll get made anyway. Or if it doesn't get made, they're not really gonna push you over the edge anyway, and so forth. And so you don't contribute even though you have a significant value for something. And the natural way to overcome that is through matching funds. Because if my contribution gets matched by everybody else, then I'm no longer just free riding, I'm deciding whether everybody is contributing, right? And quadratic funding does that in an optimal way. It basically matches every dollar that you give inversely proportional to what share of the community you are. So if you're a small share, then you really have a free rider problem, right? If you're like most of those who support this thing, then you don't have much of a free rider problem, because you're capturing most of the benefits, right? And so what quadratic funding says is that you'll be matched one for n, where n is like the number of people in the community, but not just number. If you're like, very small player, you'll get matched more. So it matches small contributions more than large ones, to things that have many individual contributors more than the ones that have few individual contributors. And it does it according to this particular quadratic formula that for the same reason in quadratic voting is optimal.
[Auren Hoffman] Could you make budgetary decisions with quadratic funding?
[Glen Weyl] Absolutely. I mean, that's basically what they're doing in Colorado.
[Auren Hoffman] Interesting, okay. Could a company even decide “okay, we're going to make budgetary decisions” or “we're going to allocate capital based on this or have the employees make decisions”?
[Glen Weyl] Yeah, so I can even go further in that direction in a moment, but the simple way of doing it is to just say, most companies have some divisions, and they face a problem that the divisions don't always cooperate with each other. And it's in the interest of the company to do that. And one thing that you could do is you could allow the divisions to spend funds out of their budgets to support things that they think are cross cutting infrastructure. But of course, when they do that, they're not never gonna fund it enough. That's the whole point, right? So you could have a matching pool that the company keeps in central headquarters, and uses to match those infrastructure spendings by individual subdivisions. If one subdivision spends on it, it won't get any matching funds, but if multiple divisions are spending, then the headquarters will match.
[Auren Hoffman] Then it becomes internal lobbying, like “Hey, we all need this public good, or something for all of us”.
[Glen Weyl] And this sort of thing happens in companies anyways.I'm in the office of the CTO at Microsoft. Basically, what we do is in an informal version of this, when there are cross cutting pieces of infrastructure that different parts of the company need, we supply the matching funds to match the investments that each of those parts of the company is making that we know that they wouldn't do on their own, or would have trouble cooperating on. But this gives a way of formalizing and decentralizing.
[Auren Hoffman] I've had many conversations with you over the years and one of the more relevant ideas to our listeners is this idea of data dignity which is basically where a person can truly own their own data. How do you envision that working?
[Glen Weyl] One way I envisioned it working is that the term ownership is not quite right. And the reason ownership is not right is that most data's interpersonal. This call that we're having right now is an interesting data stream, but is it yours or is it mine? Maybe we signed some contract and maybe it's yours and who signed up to your podcast or whatever. But the reality is, most data is not created in such a strict contractual relationship. And people are just like doing something socially. And that thing naturally sort of collectively belongs to the people who are participating in that social relationship. All the social graph data is like that. This is actually what caused the whole Cambridge Analytica thing was that my social graph belongs to me, and it also belongs to you and so forth. So I think that you need to have some infrastructure for that sort of collective management of that data in order for there to be any meaningful concept of ownership. And that's this idea of data trust, data cooperatives, data coalitions that has been circulating a lot and displayed a really key role also in Taiwan and some of the things they've been doing.
[Auren Hoffman] Why can't you just say, “my data is owned by me” and why can't we say “we both own the data that's collected here and it's fine, it's all good, unless we stipulate otherwise, that's kind how we're going to go about it, and you could take the data with you and go to another podcasts”. What's problematic with that?
[Glen Weyl] The problem is that you end up in a race to the bottom in that case, because basically, whoever's willing to sell that information for the cheapest, will sell it and undermine the other person's rights to it, because once sold, the other data loses all its value effectively. Effectively, if I think you're going to sell it right, then I’ll undercut you, and you'll undercut me and I’ll undercut you. And in the end, we got nothing, even though we both should share that value together. And, on the other hand, another role that you could make, rather than “everyone can do whatever they want with it” is “no, we can only do it if everyone consents, but then that completely gums up the market, right? So both of the extremes of like, everyone can use it, and no one can use it unless everyone agrees, don't work. You need something that's intermediate between those, which is where institutions like voting, quadratic voting or whatever, come in for making those decisions when you don't want either a race to the bottom or the whole thing to be gummed up.
[Auren Hoffman] In this case, you're talking about this is not like a company stock price or some other type of data. You're talking about data where you really feel like an individual has ownership of it.
[Glen Weyl] Yeah, where it pertains to some individual but many of the cases you're talking about, company stock price or traffic patterns or something like this, actually pertains to the group of people who are participating in those traffic flows, right? So most data is actually neither impersonal or personal. It's actually interpersonal. That, I think, is one big mistake we made. And that's one reason why we see some of the dynamics that aren't working so well in these markets.
[Auren Hoffman] But how does it work? We do want to know collectively what the current stock price of Microsoft stock is trading for right now. If we're all interpersonally managing that data, and not allowing NASDAQ or whoever to adjudicate that and give them the right to set a price or at least publish those prices on it? How does that world look in the future?
[Glen Weyl] At some level, NASDAQ could be one of these organizations that does manage it, there could be all sorts of organizations that manage its data. It doesn't make sense when you have collective management of something to just have everyone in some completely decentralized way doing it. We don't have direct democracy. For most democracies, like you have some responsible fiduciaries who collectively represent a set of people. And those people have some voice in some way over the process, but it's not like they're deciding on everything. I'd like to see those institutions be more democratically accountable to the relevant people involved, in some fashion. But ultimately, you're gonna have to have a fiduciary or manager, administrator, etcetera at some point, or some kind of AI that represents people making decisions at high throughput. In Taiwan, they've been pioneering really interesting, fast and efficient, deliberative democracy procedures that people can opt into.
[Auren Hoffman] There might be certain things where there's these multiple parties involved, but then there's also the public, that may have some sort of right to this data, or at least feel like they have a right to some portion of the data to help them. And it could be a very specific piece of data. Let's say a politician cheated on their spouse or something like that. And the public really feels like they have a right to know that data, even though maybe the politician doesn't want the public to know that. How does that work in these types of markets?
[Glen Weyl] That's a great question. I think that's almost an eminent domain type issue over data property. I don't have an immediate prescriptive answer on that. I think it's probably somewhat analogous to what we do for public purposes for other things, whether it be taxation of certain assets, or seizing private property for public use.
[Auren Hoffman] If you think of the New York Times, it is a data company with some prose around it, right? I mean, presumably, it's news, so it's publishing facts and then it's got a bunch of prose around that data. But it's not that different from Experian. In many ways, it's got core data that it's releasing every day, and maybe there's some sort of editorial decision about what we release, and what we don't release, or this is too hot for national security. So we're gonna not release it or this, we don't feel like we cooperated in some sort of way, so we won't release it. But they're essentially just a data company, right? We could distill down any New York Times story into a collection of facts.
[Glen Weyl] Well, yes. And one issue with data always is that what data we choose to pay attention to, is as important as the values that those data take on. The classic example is, you can look at these image net challenges. And there will often be 100 categories of things. Like there's a dog and a baby, and then there's a baby dog, and what is that? And so, the ontology that the data has, is at least as important, if not more important than the values within that ontology. And so what narrative does is help shape that ontology. Data is kind of a funny thing, because you and I, we're having a very high throughput conversation, we're thinking through all these issues, etcetera. But then we're gonna go off and do stuff and design systems, where people, in quadratic voting, report a number, or in your systems, a position on the earth, and there's a real asymmetry, almost a power there, right? Like you and I have a respect for each other that we understand, I'm not going to just say “Auren 17”. And you're not gonna say back to me “25 degrees north by 30...”. You can say that to me, but it wouldn't be the greatest conversation, right? But as designers, we have that high throughput. But then as observers, other people interact with us through this very low throughput math. I think that what data really means, as opposed to overall communication, is that sort of thin representation of reality.
[Auren Hoffman] You've said many times that we should treat data as labor. Where does that analogy hold and where does it break down?
[Glen Weyl] Well, I think it holds best as a prescription. So, the notion is that I think we'll get better quality data if we get people actively aware of how their data is being used. I don't know if you know about the Toyota manufacturing system and the whole Kaizen thing before that, and Deming. They have this philosophy that if you understand the process of production, you'll find problems in it and correct them.
[Auren Hoffman] ….as you go through the different bottlenecks.
[Glen Weyl] Exactly. And so I think that in the current data world, we don't have people aware of what's going on with their data. And therefore, they're not able to provide the additional value and adjustments. There's this video that leaked out, unfortunately for Google, a few years ago, which showed this notion that they thought of your data as their customer, rather than you as their customer. Basically, if they want to know your weight, rather than asking you your weight, they would just design a scale that, based on what they knew about your preferences, would maximally appeal to you, they would pay for the Amazon ad for that, and then you'd buy the scale. Then, they'd have your weight, more in the terms and conditions. At some level, it sounds really dishonest, creepy and whatever. But I think more important than being dishonest and creepy is that it's just incredibly wasteful. If you knew your weight, they could have just asked you, right? And that and the thing is, there's all sorts of things like that. We have a bunch of pictures of birds, and then we go and have someone label them. But the person who took the picture of your bird was probably a birder in the first place. So why not get that person to label the data, you know what I mean? And it's amazing how low quality data we put up with, because people aren't engaged with the process of designing. Again, for all these image net type things, you draw in all this information that comes from the context, often of the application of the method. And then it all becomes one giant dataset, and then you train against the data set, but then, the actual error on the application that it's interested in is way higher, because the overall distribution of the data was not the same as the context it came from. So why not actually involve the people in the context that it came from, so you get the right distribution of data so they’re aware of that? It's this really weird thing where we've sort of artificially backed into a problem of surveillance, when we could have had a problem with production, you know what I mean?
[Auren Hoffman] So one thing I'm super interested in is just data about the weather. And for whatever reason, I've always been really interested in data about the weather. And one of the things is that data about the weather is always extremely flawed, because the collection mechanisms are not calibrated. We could have a thermometer, your house and my house is right next to your house and they could be off by many degrees, because mine might be in the shade. You might be more in the sun, your thermometer might not work that well. And then of course you have all these other things like humidity and wind and all together. And it becomes really hard collectively to actually even know the truth. How do we actually get to some sort of place where we can at least have a better sense of what the truth is?
[Glen Weyl] Well, I think getting people actively involved in participating can play a huge role in that and they've really shown that in Taiwan. In China, there's a lot of pollution monitoring for these reasons. And most of it is done by companies or by the government. But in Taiwan, what they did is that a bunch of people who were worried about pollution had IoT devices in their house, just measuring air pollution. And they got these people together into a civil society coalition of people who had these boxes. And they basically said, “look, we're going to invest in making sure these are working, improving the quality, etcetera, and in exchange, we want the government to place boxes like these in certain places and monitor them”. So there's a collective bargain there between the people who own the data and participating, the civil society coalition and the government, and it really pioneered this model of data coalition where they have people being active participants in data creation, and using it as leverage to get what they want from the state.
[Auren Hoffman] Interesting. All right. Well, this has been really great. I've got a couple personal questions for you. I've known you for some time and I've always wanted to ask these, so it's a good time to ask them. So I know you wear a star David around your neck, maybe we can kind of see it here in the video, for those who are watching the video. And I remember you telling me that you refound Judaism at some point in your life from being very secular. You've had this interesting religious evolution. What advice would you give to a smart young person who also might be struggling with these spirituality decisions?
[Glen Weyl] I don't think I have a great piece of advice to give, but what I would say is what affected me. I grew up in an atheist Jewish family. And I had every indoctrination and exposure to universal values, humanism, blah, blah, blah. And you know, Unitarians stuff, and so forth. And the thing that I came to realize is, like most of the other people I met who had that same intersection of influences were also secular Jews from atheistic families. And so I realized I was just kind of bullshitting myself by being “No, I don't come from anywhere”, do you know what I mean? Because I did come from somewhere. And in fact, the very thing I was doing was a reflection of where I came from. And it was at that point that I realized that you may not care about your culture and religion, but it cares about you. You ultimately are a product of where you're from. And if you want to move beyond that, if you want to see beyond that, you can't just leave it behind and dismiss it, you have to embrace and make something out of it. That's not just true of religion for me, I was an economist, but I don't really consider myself an economist anymore. I think there's a lot of things wrong with economics. I'm very worried about what Israel is doing, but I love Israel. I think the way that you actually transcend your attachments, is by connecting them and making something of them and not ignoring them.
[Auren Hoffman] Does that kind of philosophy allow you to almost be like more...I wouldn't say the word self critical, but more real about trying to understand yourself?
[Glen Weyl] Yes, I think what yourself is, is ultimately a bunch of different social groups that you're a part of, and liberating yourself from them, is not actually a way of understanding or improving on them. It's just the way of forgetting about it, and therefore falling into the same traps, whereas if you actually take seriously where you're from, and understand its strengths, and its weaknesses, and work through those, and make a commitment to take the good and the bad, I think you get to a healthier, more growth mindset oriented place personally.
[Auren Hoffman] Got it. Interesting. All right. Another interesting thing I know about you is that you met your wife when you were freshmen in college, which is nowadays not common, especially for a secular Jewish person. What advice would you give people about love and partnership and is there any way to apply data to any of these decisions?
[Glen Weyl] You know, I think people mature in different ways at different times. I don't know if you were the same as me, but I matured intellectually much faster than I matured physically and socially. And that's meant that a lot of my growth along those dimensions has happened, you know, as it was in partnership with someone, and there are pluses and minuses of that. But ultimately, I think that a huge part of relationships are what you invest in them. And a huge part of beauty is what you pay attention to. I think the data that shows that match quality may not be as important as approaching things with the spirit of looking into the other person.
[Auren Hoffman] There's got to be some sort of bar that you've got to get above for match quality. But then, once you get above that bar, maybe it doesn't matter as much.
[Glen Weyl] Approaching things with too much of an attitude of match quality, I think, undermines your willingness and capacity to make the investment that I think really makes things work.
[Auren Hoffman] If you're doing some sort of quadratic voting to get your meat, maybe that's not the best strategy.
[Glen Weyl] I think relationships have a lot of complexity that eludes formalism. And I think that we should, as much as possible, aim to make our formalisms capture more and more of that richness. That's what quadratic voting is taking a small step towards, and other things I'm interested in is taking small steps towards, in communications technology, we moved from like writing, which is the thinnest representation of what it is to talk to someone, to this video conference, and maybe we'll move to virtual reality and whatever. And I think more and more areas of life, we need to find a way to make our technologies and ways of interacting with each other at scale mimic more of that richness of investment and commitment and interpersonal linkage, that is possible to conversations we're having in a long term relationship even more.
[Auren Hoffman] Oh, this is great. This last question, we ask all of our guests, which is what you've told yourself, either like in college or in high school, that would have saved yourself either just a ton of time or money, or some sort of emotional well being. If you could go back in time and tell Glen something, what would you have told him?
[Glen Weyl] When I was writing my thesis and so forth, there were two things that I was focused on. One was trying to apply machine learning in economics. This was around 2006, before machine learning was the thing, and on the possibility that financial market arbitrage would actually make things less stable, rather than more stable. And at the time, both of those were totally dismissed. People were just laughing at machine learning, they thought “we've got econometrics, it's way better”, etcetera. The people were like, “you know, the markets are doing great, blah, blah, blah”. And so I worked on other things. I went down the path that got me praise, because I had always been a really good student, and I was always getting good grades. I got addicted to that praise. I lost six years of time that I could have been doing the stuff that was really meaningful for the world, to doing what was getting praise, and then when quadratic voting and all this stuff came along, and I came up with those things, those things got rejected too. At that point, I was like, “screw this, if the importance of my ideas is inverse to their acceptability in my field, then I'm in the wrong field, or I'm seeking praise from the wrong people”. So I think having the courage to see that vision and to turn away from praise from a particular area, and not just in some individualistic way, but to look outside of the narrow community that you're in, and to look to the other communities and see how they would have they make sense of things. Quadratic voting found its community, it just was a different community.
[Auren Hoffman] That's obviously hard to do, because you may have had to quit your Phd program, because you may have not been able to find anyone at a quote-unquote “establishment” university to have to have worked with or you may have to be okay with getting a C in class, which does have some ramifications, as you're trying to get into grad school or other types of things that happen or get a job. This is not a costless thing, to basically go against the grain.
[Glen Weyl] You have to stay on the march, you have to stand on the edge of the grain, I think is what you have to do. I think it's at those edges that things happen, the notion of escaping, that doesn't work, just like I said, I couldn’t escape my Jewish heritage. But the real place that creativity flowers is at the edges of the continental plates, it's where things intersect and where you see that something makes a certain amount of sense from a certain framework, but it makes a whole lot of sense from another framework, that's usually opposite. And if you find something in between, you can just barely survive in both. But that's where things really take off.
[Auren Hoffman] Is there some sort of heuristic to know, “okay, in this particular situation, I'm going to subvert my own feelings and just kind of go along. And then there's this other situation where I, where it's maybe more okay to be hermetic or or more ok for me to go forward and, and be different from those around me”?
[Glen Weyl] I think you get it by finding the other community and not by just going off on your own.
[Auren Hoffman] Ah, so looking for another community because it's too scary, you can't be the unabomber, right? You've got to find some others to be your allies, essentially.
[Glen Weyl] We talk about children becoming independent as they grow up, but they don't, they become dependent on different people, that's how you form a new identity is, that's the reason why teenagers, as they become “independent”, are actually the ones most focused on peer pressure, right? By finding that other community, it's not by breaking out on their own. And so I think that's the thing you always have to ask yourself. If this feels right here for some other reason, but doesn't quite fit in this world, is there another world that it fits into? And you know, for me, a lot of that's in the blockchain world, I like went from econ thing, the blockchain world did that. I'm super critical of the blockchain world too, because I always like living on the edges, but it was that tension between the econ world and the blockchain world that gave me the ability to do a lot of what I've done.
[Auren Hoffman] Awesome. This has been really wonderful. Thank you very much. Please tell the audience where they can find out more about you Twitter, you know, etc.
[Glen Weyl] Great. Check out [email protected] Radical exchange is a global social movement. We have 100 local chapters around the world, and we use these types of things to try to remake politics and the economy.
[Auren Hoffman] Awesome. Thanks. Thanks again. It's been great.
[Auren Hoffman] Thanks for listening. If you enjoyed this show, consider rating this podcast and leaving a review. For more World of DaaS (DaaS is D-A-A-S), you can subscribe on Spotify or Apple Podcasts. Also check out YouTube for the videos. You can find me on Twitter at @auren (A-U-R-E-N). I’d love to hear from you.
Glen Weyl is Microsoft's Office of the CTO Political Economist & Social Technologist, founder of RadicalXChange, and author of “Radical Markets: Uprooting Capitalism and Democracy for a Just Society”. Glen is reimagining democracy with two revolutionary concepts -- quadratic voting and quadratic funding. Auren and Glen cover how quadratic voting and quadratic funding can revitalize collective decision-making, current applications of both concepts, and how they can be used by corporations. They also explore why accessing high-quality data is so hard and what businesses can do to significantly improve their data accuracy.
Susan Athey, Professor of Economics at Stanford University, talks with World of DaaS host Auren Hoffman. She is also a research associate at the National Bureau of Economic Research and was formerly the first Chief Economist at Microsoft. Susan is also on the Board of Directors of Lending Club, Expedia, Ripple, Turo, and more. Susan and Auren dive into the role of tech economists, machine learning, and causal inference.
Will Marshall, co-founder and CEO of Planet, talks with World of DaaS host Auren Hoffman. Prior to Planet, Will served as a Scientist at NASA/USRA. Planet is a data business that operates hundreds of orbiting satellites that capture images of the entire world daily. Will and Auren cover Planet’s proprietary methods of building a satellite imagery dataset, how its data is used by numerous industries, Planet’s plan to go public later this year, and more.