Governing democracy, the internet, and boardrooms
01:35:38 Share

Twenty twenty four is an important year for democracy. This year, over sixty countries have, and we will head to the polls representing just about half of the global population. But governance extends past the leaders of nation states.

In fact, as the revenue of some companies to our the economies of some nations, the way these entity es govern themselves is both more interesting and more important than ever. Listen to this episode originally published on our sister podcast, one through a cy, as they explore topics ranging from content moderation to incentify participation to new governance experiments, from those to overside ports. I hope you enjoy.

Hello and welcome to web three twenty sixteen z i'm Robert tacos. And today we have a special episode about governance in many forms, from nation states to corporate boards to internet services and beyond. Our special guests are no a fell in constitutional law professor at harvard who also architected the meta overside board, among many other things.

He is also the author of several books, and our other special guest is Candy hall, professor of political science at stanford, who is an advisor to sixteenth crypt u research and who also co authored several papers and posts about web 3 as a laboratory for designing and testing new political systems， including new work will linked to in the show notes. Our hallway style conversation covers technologies and approaches to governance, from constitutions to crypto, blockchain and those as such. We also discuss content, moderation and community standards, best practices for citizens, assembly courts versus legislators, and much more where governance comes up throughout.

We reference the history and evolution of democracy from ancient greece to the present day, as well as examples of government from big companies like meta to startups like anthropic. As a reminder, none of the following should be taken as tax, business, legal or investment advice. C A sixteen e egypto dot com slashed disclosures for more important information, including a link to a list of our investments.

So I want to start with like a really broad question. Maybe it's too broad, but what is the right model for internet governance? We have all of these companies that host platforms that people participate in, that they build on top of, that they connect and communicate on.

And so a question is, like, what is the right way to govern those platforms right now? The most popular ones, at least, start owned by corporations. And they sort of get to call the shots there.

Is that the way that I should be idealistically? And like a world that we all live in and work in and playing. And we .

got a bit on the regulation. Corporations don't exist in the either, much as we sometimes like to fantis ze you positive negatively that they do. Their products of law usually dellow a law because of the weirdly of how the american federal system works.

And they're governed by a plethora of obligations and regulations that are not only in their place of incorporation, but are also state law everywhere in federal law. And a lot of the duties that they have, they can get rid of even if they try to get rid of them. So if you make a cellar product, you're liable for the best step that happens to people as a result of that product, even if you pretend you're not and even if you don't want to be.

That's statistical example of the duty you can get away from. That means that all corporations are intensely regulated all the time already, even before we get to the specific regulations that are relevant, say, to a user generated contents social media platform. So the first thing is you're gonna be governed by state and national laws, and in a more tanee sway, you going to be god by those international laws that your government bothers to apply to you, in which.

by the way, there are plenty of governments that are democratic, formulated and know some of bunch of these laws are presumably coming from the .

will of the people yeah, and some they were undemocratic and laws, but they still affect you if you want to do business in a country. So, you know, all of these things are are in place. That's the first thing I would say.

So their R D is a lot of governance that comes from governments like the good old fashioned kind of governance. Then there's a further and really important question of whether there should be more governance coming from governments and if so, of what kind. And that's very hot topic right now at the U.

S. Supreme court. This year is deciding three different sets of issues, all related to what are the rights of platforms, what are the rights of users, what is the role of government with respect to social media.

And until now, there has been exactly zero supreme court doctrine on social media companies. And now we're going to have a whole bunch of IT. So this is A A major transformational year.

unfortunate. Don't know the answers to any those things yet definitively. We have sort of tea leaves from the oral arguments. So that's another really huge issue. Like how far should I go? How much regulation should the states of large on taxes be able to weigh in on content moderation?

yes. So you're referring to the the natural case that the supreme court is reviewing. And um what's that issue there? Is these states would like these social media platforms not to sensor or moderate certain types of speech.

I mean, that's how the states would put in the company is would say that the states want to force them to Carry content that they may not want to Carry or want to limit the ways that they can moderate the content on what they call their platforms. And so that it's like any supreme court case. You've got two sides, and those are the two main perspectives.

And then beyond that, there's to me, a superfast sing question I know andy is not great ebay about this two of modes of regulation. The platforms can choose first of all for themselves and use internally. Then there's potential collaborations as in between different platforms on collective uh, self regulation, which is a form of governance.

And then there is the we're kind of hybrid structure that the facebook oversight board represents them in a different way. The anthrops long term benefits from represents. And those are methods whereby a company create a kind of hybrid structure with there some independent actors whom the company agrees to be governed by for some limited set of purposes. And that's another really interesting, into my mind, kind of innovative and cutting edge way of doing this, which, like any innovative cutting anything, has its pressing its cons and is also very much in an experimental phase rate.

I think this we're already getting into something very deep and interesting, which is like where are the boundaries? Where do we think the real world governance ends? And when or why would any type of organization want to go beyond IT? Seems like there's at least a few reasons, but not because so you think I mean, there's obviously just a sense as a companies and organization that you want to do more.

Um the government has told you to do. That's like one possible motivation. Seems like there's also something about the global reach of some of these platforms.

We know that global coordination around regulation is very chAllenging. And so in the absence of that coordination, you might feel the obligation or the strategic need to do something, uh, to coordinate that yourself. And then there's also something this hasn't come up yet, but the I think a third explanation is something to do a competition.

And I think going back to original question, Robert, you know, in a world where we had perfect competition in a very classical sense between all different types of internet services, I think a lot of the governance beyond traditional regulation would be done by users voting with their feet. And the chAllenge we get into is the scale of many internet services in the network. Effects put them in this kind of interesting position, where there are a very, very large scale, which makes IT, in some ways, hard for users to vote with their feet, because they like being where their friends are.

But at the same time, there's still a lot of economic competition across services. It's not obvious that these companies aren't really monopoly in in the traditional entitles sense. And IT put them in this weird zone where there is a sense that users are unhappy being locked in to any particular platform, but the same time, traditional anti re tools don't really seem to apply. And in that weird grey area, that might make sense for large platforms to play with additional modes of giving their users the ability to decide together how the platforms gone to work in a world where they can freely individually move between platforms for the same service.

And you would disagree this, there's also the advertisers who perfect economic picture would just go where the people were, where the users were. So that sense, they seem less important. But as we've seen in the um when you might call the x files, um the advertisers have turned out to be major players because they don't wait to see what the users will do.

I mean, they also do that, but they take preemptive steps based on their perceptions of what might happen in reputational costs and all of those sorts of things. And so there are big players, too. And there are another reason why a company of platform might want to have its own self regulatory mechanisms.

And here I just want to add, in this super polarize world, there are known neutral decisions that you could say, oh, well, everybody leave me alone because I made a neutral decision. Whether you take content down or leave IT up, your not neutral. Whether you amplify IT or don't amplify IT, you are not neutral, depending on how much you applied y IT.

You're not neutral to being on your algorithm, not neutral. Everyone is sort of realized by this point, certainly this ecosystem that there's nothing neutral and where a lot of people are matter you and there is no neutral position, you might have an incentive to upload some of the decisions just so people will be met someone else. And also because people might not users might not trust you as the platform and then they might you might think it's Better off. The people will trust somebody else more than they trust me, even though they may not trust fully either.

I think an trying to rebuild trust. And I think what something knows that really important, which is like trust in tech companies, is falling in a lot of parts of the world, not not all part of the world, but in many parts of the world. But at the same time, faith in government is falling in a lot of those same places.

And so there is no obvious actor to make some of these very hard calls in a neutral procedurally fairway, where the key is that I think you, as a user, you'd like to be in a position where you know a platform besides on a piece of content or what APP is allowed or what transactions are allowed in a way such that even if you disagree with a particular decision, you still buy into the process by which the decision was taken in today. I think there's a lot of people who question both a company's ability to build such a process, but also a government's ability to build into the process. And in the absence of trust and either of those processes that makes sense from a business perspective um to try to improve your trust in society as a company by finding a third option for a fair way to make that decision.

This sounds like a really good segway into the oversight board because that is an attempt to satisfy these conditions that you're talking about to kind of offload some this responsibility onto a third party to enhance transparency and trust, hopefully, and create a process that is reasonable and rule bound for moderation, all sorts of decision making behind the scenes. What let's talk about the formation of that and the needs that LED to its creation.

I think the core background situation that then facebook was was facing, no matter when IT one mark locker decided to create the oversight board, was the recognition that there were some very, very hard content moderation questions. They were so pretty simple to time, they were sort of to leave up content you take IT down.

They were not as fine grain as they now have become, but in which he might have a strong intuition about what the right answer was. But reasonable people could differ. And I think the insight that he was able to see is that actually there there might be a right, wrong decision.

I'm not i'm not a relativist. I don't think that they are like always two decisions. No, there could might be right, wrong decision on this, but reasonable people could defer about what the right decision is. And IT probably didn't matter that much from the same point of platform governance, which of those two decisions were taken. But if he made IT, IT was going to taking a lot of time, he was going to give him a lot of effort.

He had no special expertise in any of the underlying questions, and people were gonna really angry at him and distrust him, so that no matter when you decided IT would decrease the legitimacy of facebook, as IT then was. And so the idea that I I proposed them and that I think resonated for him, was maybe this decision doesn't have to be made by you in the first time since the company is going to make a decision that gonna have a policy and they gonna try to file out policy. But if it's controversial, when I create a group of independent experts and bring those questions, the hardest questions to them, have them answering the question in light of facebook principles, international principles, common sense, good judgment, have them explain their decision and see if I worked.

And I would say that the secret sauce there is the explanation of their decision. The secret sauce is that if you're answering a really hard question where you admit there are different possible answers, you should explain why you're doing IT. That explanation that practice of reason giving has the capacity to confer a lot of legitimacy when IT comes to hard decisions because I think IT reassures people that got four people gave thought to this, that they had some degrave expertise, that they deploy that expertise and that they tried their best.

It's not a magic solent. Some people will still be really mad about the decision, but everyone will be able to say, I hope this is the closest thing to a fair decision making process that we could have come up with. And my last point on this is obviously this idea of legitimate fairey did not come out of the blue.

It's borrowed from real world institutions, especially courts that often are not elected. Often their decisions aren't interpillar able. If it's like a supreme orter, a high court, they're the last word.

They're not always right. They're often controversial. People get mad about them. But the institutions, on the whole have a fair amount of legitimacy. I mean, it's a funny mom to be saying now in the supreme, as you know, that a lot of self harm on the on the legitimacy front, but the circuit works as much greater legitimacy in the eyes of americans.

And that's even after you range of other things that the supreme court has done that, that they don't care, that it's the majority doesn't care, it's undermine the legion. So the idea that you can get legitimacy from a recent decision making was borrowed from that round and applied experimentally to the opposite poor and I think on the whole is working not perfectly but that IT is working. But a that's a topic for further discussion .

mentioned that this is a funny moment to be uh to be talking about the legion acy of the courts um now of course it's an institution that hundreds of years old but based on the way things have shaken out, is there anything you would have done differently in your conception of this sort of analog body to a court system for will actually I can .

the reverse way there is some yn that did that I am actually kind of proud of which differences us from the supreme court we have the over support is a pretty big body of people who um decide cases and sometimes all of them but sometimes in subset and IT isn't partisan so it's not set up so that when you change the composition by a voter too, everything flips on you, which is what happens is that the U. S.

Are printable and so and then retirements and depth. And so that kind of IT can all change in a second is highly partisan and it's very responsive. Some of the time to political changes are all features that we intentionally did not put in the oversight board to lower the temperature around its decision mainly process.

Is this something that the government should take into a cat, like if you were setting up a new government and a new court system, would you make changes of this sort?

Yes, absolutely. Mean, it's crazy that the supreme court, I mean, you have to go back to seventy. Eighty seven people died, a lot Younger. Being on the supreme group is not that good a job a bunch people who who had that job would quit and go back to private law tracks because they couldn't stand the job in IT.

You know I didn't pan out and you had to get on a horse and ride around in your cases all that which just wasn't that pleasant um now people live longer. They get into that job and they never leave you know they believe of the court when they die. A lot of them, which is very I mean, it's unhealthy.

It's not good. And we also have a much more polarized politics than they certainly at the beginning, although there have been experience of intense polarization in U. S.

history. So yeah, you were designed from scratching. Now you would have staggers terms with time bound.

IT wouldn't be, you know, the lack of you. Justice cla died. Baracouda a was president, but micmacs o was able to block the confirmation hearings. I mean, that kind of crazy town hardwar politics should not affect the composition of the supreme court. And since one of my, one of my jobs is that I watched the court closely, part of my jobs shouldn't be having to have like a not quite expert, but pretty good amateur s knowledge of, like the health of the individual justices. If they are diagnosed with a certain cancer, what are the probabilities that they are live and how long like that? That's detestable and also like absurd that that should be part of our way we think we shouldn't be thinking in those terms.

I want to raise one thing that I I want to make sure we spend more time on how the board is doing because I think there's a lot of interesting different aspects. So that unpack one thing I just want to raise that no one I have talked a lot about the past and lots of other people thought about that thing is important is one one important difference between, say, the supreme court or U. S.

Courts or courts in general. And the oversight board is sort of whose laws they are interpreting. So no, I reference this earlier that they are basically they are considering cases on the basis of facebooks principles or medas principles. And I think that's theological way to start.

But one thing people obviously then point out is, well, you know, the supreme court isn't referencing a single corporations rules, is referencing democratically written rules that the legislature right, nevertheless is an interesting, I think, thing to consider in the long run. Can you get a sufficient amount of legitimacy? Is now to call IT from a court online that interpreting the rules of the company when part of the concern is whether the company is setting the rules in the right place or not. And so one thing people have talk quite a bit about, but I don't think we've correct, is how would you democratize the legislative process as well so that something like the oversight board would be making decisions with respect to democratically written policies.

And this is super. And let me add one more twist to the way things have been playing out because this is something that has not been broadly discussed or broadly covered, except in a very narrow group of specialist, mostly academics. And you know, in ng s who follow the oversight board, the oversight ard has been really worried about the thought that they're following metas rules they don't want.

They are more ambitious than that. They want to follow something more democratic, but they can't use the laws of any particular country, because there isn't one particular country this binding. So what theyve been doing is they're claiming to apply international law, the international law, free speech, which is not anywhere near democratic as the laws of a democratic country.

But IT is sort of second order democratic, because these are treated that were enacted by countries, some of which were, many of which were democratic, and which all countries recognize as legally binding. Now the upside of that is that IT sounds a lot more legitimate than were deciding these important questions based on metas rules, and it's also a little more democratic. The downside of IT is international actually isn't.

That will set up to answer these questions because it's not designed to govern what a platform should do. It's gover. It's designed to govern what countries should do.

That's why it's international. It's the law of stuff between countries. So the oversupply has had to be really creative. They basically playing on some work by by scholars. They're basically created a whole new body of what they call international law.

The they say buying mea and buying shouldn't principal buying other platforms and they're out there making this law and there's no one to tell them no. Um you know so I I just provide a student or dissertation saying this is unjustified um but my response that was always that might be true but you know who's gonna stop them and they are trying to do that. So that's a bid that is out there and it's interesting to watch.

And as I say, IT hasn't very well covered. But I also I do want to get back to the point that and is talking about IT, which is what are the experimental, creative things that a big platform like media could use to bring democratic input to bear on. Some of the hardest decision that IT makes .

well is away lipica service at the history, which is that this is something apparently mark server has thought about for a long time because back in the two thousands, when the platform was so pretty Young, they they had sort of a disagreement among users about what the new community center should be. And mark, uh, held a global referendum on the platform. And I mean super interesting.

You can actually on the internet archive, you can go and a see the video he posted encouraging all the users on facebook. They're about three hundred hundred million at the time to vote uh and and what they learned through the process was actually it's really hard to get people to vote. So out of the three hundred million users, I think something like sixty thousand of them voted.

And the reason for that probably because you know most people are not logging on to any part of the internet to do the hard work of governance. They're logging on to have fun or see their friends are, whatever. And in addition, the the decisions being considered where they're pretty up truth like you to read these like hundred page documents himself.

So I think there was a really innovative idea that suggested early on this idea that there's a way to potentially make decisions about a platform democratically, but IT turned out to be chAllenging to to make work in practice. And so now we fast forward to today you I think lots of people are interested in ways to set at least some kind of broad, simple policies were the values that users all the relevant to the decision through some kind of democratic process. No one has really figured out how to solve this core problem of. Participation as well as informed participation as well as if it's a platform that global in scope, how do you actually get people in different countries who speak different languages to even work together, figure IT out what if they have radical contrasting values or goals? There's a lot of really hard questions, but I think there's there's at least two veins of experiments worth going deep on, one which I think .

no voter voter participation rates are a problem, not just an online voting.

but you know, I are .

well voting to. That's true. That's true. Although thing and a andy, you make you open instagram and you can use instagram unless you vote on these issues, but then you're not going to get highly informed voting right?

I mean, this describing I just to make them concrete because I think there are two kind of good examples. One is the problem that I don't know people who are listening. All i've heard of this or not, because IT may be little dated, but they called the body mic boat face problem.

Oh, this was red. I didn't read IT IT was the british. The british .

government decided they were going to use um you know this new amazing thing, the internet to gather votes on what they should name a new battleship and you know the the winning vote gather was both .

mic boat face, which I think we can agree is a fantastic .

name for the problem yeah her .

majesty ship I guess was her majesty ship body, big body and the royal navy. Um and then the other example that I like to use and andy mention this, is let's imagine you're setting a new udi policy. Well, there's one possible that works on the south of france on a beach which is more permissive than what we have in the U.

S. And then there's saudi abia, which is a lot less promising than we have in the U. S. And it's not obvious that each of those should control the entirety of the platform.

It's not obvious that ARM models to fall the the platform there, IT, is mostly a default that you know a lot of these platforms of U. S. Space companies.

And so U. S. Space standards end up controlling americans. Think of those are the reasonable standards. But of course, there's no one inherent reasonable is answered to the question of like how much skin you should show.

That's a classic example of sending the different cultures are equally confident that their way of doing that is the only right way to do IT. So that's that's a genuine chAllenge. And then that leads to thinking, well, maybe we should have regions, but different rules in different regions.

And IT turns out it's harder to do that on a global platform than you d think. And most of the global platforms have tried as much as they can to avoid that because they don't want this kind of extremely organized platform where IT all depends on where you logged in from. And you know bit you have A B, B N, you can get different standards and you can imagine what what a mastic IT can quickly become.

So those are part of the chAllenges um and then the last part of the chAllenges, all of our examples of democracy and you can you can correct me if this is wrong because you're the political scientists, but at least the ones that come to my mind, all of our examples of democratic c governance assume a domaines or that is the relevant group of people is the people who live in your town. You're going to vote on town policy or your state, your country. We don't really have true democracy at the global level.

And so that raises the question of who's the who's the dominate, who's the relevant community to vote on these issues. And maybe it's the users, but a lot of issues affect non users. And so you know is that everybody, including non users, is IT a subset of users. That problem is, I think, the very first problem to grapple with. And in a lot of ways, it's the most chAllenging.

Let's unpack the voting with both faced thing for second because I think it's really important. It's a great example of two or really three deeply related problems in the use of voting to make collective decisions. The first is just the lack of participation.

So very few people actually voted on the boat with both face decision, because why would most people pay attention, even know about IT? Second, which is closely related to the lack of participation, is that who then are the weirdos who choose to vote? Were not very many other people are voting people who have very extreme, weird preferences, like the kind of people who think you will be hilarious if there is a vote named voting with both base.

And so those two things go together quite tightly, low participation and then selection for unusual, unrepresented, tive extreme views or preferences or trolling. So I think that's like an important thing that understand. And then the third, which is also the deeply related, is the just the general problem that the you are asking people to vote on decisions and they have no skin the game, and that again lowers the incentive to participate and lowers the incentive for people of representative views to participate. yes.

Would you have to serve on that trip .

like little bit more well? That kind of gets for a noas dominated point. I think there's a very broad chAllenge is much broader than just democracy, which is how do we govern things that are not economic in nature, in ways they are smart when the people are asking to do the governance, don't have skin in the game.

And that that was the problem, you know, that was the problem with open the eyes. Governance has been a chAllenge with university governance. The trustees, and some sense, don't have skin in the game.

You measure the skin by how much money they put in. They would say, I mean, i'm not not that i'm always sympathetic to them, but the only ones skin in the game, you know, we're going on eating the money you guys are doing is taking the salary.

What's a good point is a good point.

They say, I know i'm not saying i'm anyway.

go on yeah yeah. And when we look at at all these voting problems, if you're asking people to vote on things that don't clearly affect them, that's a reason you might not get very for participation.

Furthermore, when you ask large groups of people to vote, you have this problem we call the paradox of voting, which is that even if you do care a lot, even if you do feel like you have skin in the game, if lots of other people are voting, then you know your vote doesn't matter. It's not going to swing the outcome. And so no, you can care infinitely much about the decision and yet you're incentive to actually pay attention of that is very, very small.

So we put those things together and we ask, okay, how are you gonna something democratic to make a you know, big decision and a big online platform? I think you have to start with this fundamental set of problems that are, you want to get people to participate, you want them to pay attention, you want them to feel like IT matters. And how on earth do you do that? And I think there's been several directions of experiments that are pretty fascinating.

One which no elude a to earlier is what anthropic as well as meta and others have been playing with, where you you don't use voting, you use these things called citizens assemblies. And then the other is what's going on script, where you're experimenting with different types of voting to try to draw in only the people who have some sense of skin in the game through the tokens that they hold. So those, I think, two two sets of experiments we can learn a lot from.

And I would just add one more of framing thing here. Most forms of democratic government in the history of the world. I tried to address these questions by saying, we're not gonna have direct governance where every single person has to vote on every issue, because that's also like a time consuming thing. And and so one way to solve both the how much time you put in and the how much you care and the skin in the game, is to have representative decision making.

This is the move from like ancient athens, where everyone can show up and vote every, to be clear, every free man can show up as a citizen and vote in the assembly to a system where you elect people and your representatives are professionals, or quiz I professionals, and they do the decision making in the voting. They have a skin in the game. They have the job of caring about getting reelected based on whatever incentence they have.

And that gives them the incentive to try to guess or figure out what you're thinking. And one of the really interesting things about the internet, and I think it's important background to you when we dive into both cypher and citizen assemblies, is that prompts early days. The internet has been, broadly speaking, the ideology of the internet, the ideologies of the internet have been really skep portal of the idea of elected representatives.

There's this impulse, and it's really interesting to explore why is that because of pcrp culture? Is that because of color culture is at the personality type of the people who first got really good using the internet before everyone use IT. I mean, is a lot of possible explanations, but there has been a deep impulse to get around the idea of elected representative.

And both of these these experiments, the citizen assembles and the crypto do or organizations are attempts to like, say, we are going to reinvent the way, like we're not gonna the thing that all democracies have done, namely rely on elected representative, we're going to do something different and Better. And that kind of appealing because especially if you have some scepticism in our elected representatives, it's good time to try to reinvent what they are doing. But there's also a reason to be modest when all of the countries in the world that uses a certain technology named democracy have converged to a first approximation on one solution to this problem named the elected representatives. And you're like now they don't know what they're talking about, what we can do Better, maybe you can it's worth definitely worth trying, but you want to have like some degree of modesty about the probabilities coming up, something that in twenty five hundred years of thinking about this, no one has yet really managed to solve in any other way.

I wants to build on that quickly because I think it's really important, and I completely grew no one thousand percent. There's something about technologists and people online that makes them really crave direct democracy. And one of the key claims they make is that technology is going to allow for direct democracy. That I didn't work in the physical world because .

IT was too bird. This was going to be my question like that the internet is IT a step change like we entered a new era where. Based on the tack that's available now, this is conceivable.

There are no certainties in the social sciences, but this is as close to a certainty as I really to go is the technology is going to make no difference to this problem.

And in the year is just the way is IT .

goes back to the paradox of voting I was talking about. It's very burdensome to be asked to understand and become informed on every possible decision that a group is going to make. And you might think, and this is, I think, the fallacy that a lot of people in tech of committed, you might think that having more coin quote, democracy, by putting more vote to the people, he's gna get you, you know, a more empowered user base.

But the opposite is, in some sense, true. The more things you ask them to do, the more burden symm is for them to do IT, the less that they'll choose to do IT. And that then creates a really important and dangerous vacuum, because now you have very few people voting.

And then interest groups can come in and capture the decision making process at a relatively low cost because there's not that many other votes out there for them to compete with. And this is, I think, fundamentally why no successful society on earth is stuck with duct democracy for very long. And it's why technology is not going to solve the problem because it's actually not a problem about voting in person being too difficult to anything like that.

It's a problem of information acquisition and analysis. And i'll just know on that point. People have also claimed that will be able to provide voter all the information they need to make every decision to some kind of APP.

I've had a lot of underground in my office pitching me on the next APP that's going to inform us all about politics so we can vote on everything ourselves. The problem is that information that did is not enough. You actually need to analyze IT and then decide what to do, and that's actually quite difficult.

And so that being buried in data on what you know your community is voting on is just not enough to help you do this. And so I think it's a completely fundamental problem because wee yond technology that has to be solved to other models besides direct democracy. And I just note almost all of the big voting downs and crypto to now have some for of representative democracy as a result of this this .

thing to this and and you think no, that you know no one and you're wrong. You know.

like to god, just ilus come on.

What I would say, if you're inspiring technologies, ask yourself, what's your vision for the company you're going to found? And in ninety nine point nine percent of cases, people are like, I want to found the company and then I want to have founder shares and I want to control the decision making. And then you say, well, why? And they say, well, because otherwise, like my, the vcs will first involve.

They'll be shareholders. They're gonna delude my votes. And they going to take the company, they are going to do all kinds of stuff I want to do with that.

And i'm the one who knows because it's my company. I spend the most time. I don't I care about IT. That's exactly the proms describing.

So I always find the amazing when you know the students come in and say that because I said that and you imagine a company where you will have no special input into how the company runs, know every decision will be made collectively. They like, are you crazy? Of course not.

Not like well, then there's a contradiction in what you're proposing. So it's the you know the intuition. The technology will interfere.

You can easily fix that intuition by just asking yourself, when you get rich with your own company, do you want to be in charge of IT? And in my experience almost everybody thinks the answer that is yes um although there are some twigs, right I mean so another thing that anthropic did and I I know about this because I was involved in advising them on IT as they've created a long term benefit trust which is a trust that appoints at first just a few, but eventually they will point a majority of members of the board of directors of anthropic and the people on the trust we are going to appoint. Those board members are not themselves travolta's.

So when the topic is doing there is it's not the full open eye thing where the the parent board was completely made up with people who had no stake in the company, which blow up. As we know, this is a more modest in between version where there will still be shareholders on the board directors. But the idea is to have some external check. So it's a little bit Better, but that's very rare.

So let's talk about the example of anthropic. You saying that in this case, there are actual shareholders that are able to serve on the board and that this is different from medas overside board and it's also different from open a eye where people did not have a stake in the company. And so maybe there will be a more representative collection of views in there for keeping the concern going and also uh mitigating risks that could yeah it's .

designed to avoid problems on both sides. So take the open eye problem. You want to have some break class measure. And so the big glass measure was they created this nonprofit entity um and that was its the overarching body that the board of directors belong to.

And then underneath that board directors was the four profil tty that was that we think of his OpenAI, the actual company that does stuff. If the people on the world of directors of the nonprofit who had no financial stake at all believe the company he was going the wrong way, they could break the glass and fire the management of the company. And they did that. They believe that things were going terribly arrived, that this was bad for the world, and they broke the glass, and they fired the directors of the company.

the management of the company, until the back fired in the glass, go back in their face exactly.

They didn't realize that what would then happen would be that sam t. men. We then say, well, out, i'll just go to a microsoft alt. Every single employee in the company with me and the employees all went online and signed up that they would go with him, which is maybe not so shocking because, of course, you're gonna agree to jump ship. And so then IT backfired, and then the members of out where directors had to resign.

So what you need in the real world is something that doesn't go that far where there are people who are not financially incentivized, who serve on the board, but there are also people on the board who understand the real world and who do have financial incentives. And they develop a relationship with each other so that IT is possible for the people who are worried about a problem and want to break the glass to do so in a thoughts and responsible way where IT won't blow up in their faces. So that's what the long term benefit trust is trying to do um and IT does give some real reasonable responsibility to people who don't work for the company.

I'm in a nd sense it's a kind of cousin of the overboard, but it's still from the overside work because those folks don't have day to day decision making responsibility. They're there at the level of oversight and break last an emergency where the oversight is meant to be in every day. You know there are hard problems that met is dealing with and the overside board wising on them so I will call them cousins. Um they have some conception similarities, but they're not they are they're not they're siblings, they're definitely not twins.

IT seems a little bit interesting that you have a certain set of people who have this opportunity working side by side with people who are, I assume, just getting a regular salary and then how you get those people that really care about, you know, the decisions that are being made.

They have to care based on their reputation. And this was raven at the oversight board, too. So how do you get the oversight board members actually to care that they do a good job? It's not so simple, but the short answer is it's nobody's full time job.

And the the jobs that they have are as activists or as scholars or as people who care about free speech. And so the idea was that they will care a lot about their own reputations, and that will give me incentive to do a good job. It's not perfect, but I was the best we could come up with.

If IT was their full time job, their incentive might be too closely allied, you know, with the power of their institution. And they still probably wants to do a good job. But IT wouldn't IT wouldn't be that they had a reputation to preserve elsewhere.

So that was a kind of complicated compromise decision because is not IT transit's a hard problem. You know, how do you get people to care about their jobs? Usually it's they get paid more if they do a well, and they get fired if they do IT badly.

And on an independent body, you can really do inter of those things. You know, you can't like, reward them for making good decisions. And if you can punish them by, fire them for doing for making the wrong decisions, they're not independently. So then you need something somewhere in between. And reputation, I think, is the most powerful motivator.

And there I would just say, you know, you might think o no one's motivates by reputation, but there is a turns out there's a whole bunches of people in the world who take the jobs that there a lot number academics to take the jobs that they have rather than some job that would pay a lot Better because they care about the thing they're doing. They like doing IT. And then so you ask the OK, now you have tenure when I remember, you know, getting tenured and thing out, this is so great.

And then someone said to me, this is you never have to do a days work again. I thought myself like, I never thought of that that way. Like i'm so much of neurotic, i'm onna.

Keep on working hard. But rationally, if I were real rational actor and maybe I would stop. And the main reason not to is that my reputation would be entered.

And people would say, as they do say about some academics, oh, there goes film is one of those people who the day got tenure never did, did anything ever again. And you know, that hurts. So and I care about is not urting.

it's impossible to picture you know I stuff in working after that there 这就 来吧。 So anthropic is also one of the examples of these uh, citizens .

assembly. So might be 起来。

I think the general idea here is you have cases where you're either worried that too many people will participate if you hold a vote. Or that they won't actually know enough about the issue prior to having some kind of deliberation together and or briefings on the issue to make an informed decision.

And so instead of holding a vote, you try to randomly sample a representative group of your citizens in the case of a citizens assembly or users in the case of one of these online, i'll call them user assembly. These other people call them different things, uh, and then you bring those people together, the ideas because you've randomly sapped, they'll be representative of the user base of your platform. You get them together, you have them debate the issue, you have them hear briefings from experts so that they're infor about the issue and then you have them make a decision through voting or some other collective process um on the whatever the difficult issue is. And this is something IT had some history in the real world going way back, and he is closer related to this idea for a mage agrees of sorting tion.

What is sorted sortin .

was this idea that you would randomly choose people to be in charge of various issues.

It's election by lottery. yeah.

So we other autographs.

also the ideas, basically, we need someone to have a full time job of running some part of the government. We don't want to vote for them because I create all kinds of weird incentives. So every year we we have a lottery or every or twice a year and we pick wonder two people who are onna have this job.

And so and they only do IT for one term. They don't have worry about getting elected um but for reputational reasons, they try to do a good job because you know they look bad if they don't do a good job and people will think well of them if they do a good job. So you pick them randomly um and then you do again the next year.

It's jury, the height to the absolute level while the authorities .

had very aggressive jury duty, they had much larger juries and they paid them quite well, which is interesting to but I think the main justification for sortin was this idea. I'm going to paraphrase this, but there's this great close from Douglas atoms, the often of pitch hikers guide in several other great books. That's basically like any person who wants to be in charge is, if so facto, unqualified to be in charge.

And and I think that was the idea of sortin, is we don't want these, you know, untrustworthy, overly ambitious, potentially corrupt or sociopathic people who desire power to be the ones in charge. We want, you know the average well meeting patriotic s citizen to make decisions instead. And so the way we're going to make sure we don't allow that social path to insert himself into power is we're onna Randy, choose us in charge.

And I just wanted know it's it's an interesting solution to that problem. IT also raises a bit of its own chAllenges. If you think that a big part of what makes democracy work is what we call accountability, the idea that the person whose job IT is to pay attention, learn the issues, figure out what maps to your preferences and make decisions on that basis, if it's the fact that they're worried about whether they are onna, get to keep their job is what makes them do all those things.

Well, the sedition is a terrible, terrible method because I have no incentive to do anything. I've been randomly chosen to do this job, and now I just have to do IT no matter what. I don't have those strong incentives that are coupled with my desire winter election. So there's no accountability in some sense in less like no is saying considerations like reputation patristic m genuinely just caring about the issue are dominant. So those kind of that's one of the key tradeoffs.

When you do a bad job, maybe you'll get maybe i'll get also sized the next they .

need to do that they would actually do. But there's another key trade off, which is the background assumption of soria an is that you don't need very much that all the expertise you need to do the job you already have or can pick up really fast.

So you know, imagine you're running your start up and you know you're gona choose someone to be your co and you I go to do a by sortin like bad luck for you if your lots fall on me because I know code well enough to be able to do a good job of reviewing the other people's code that's necessary in the startup for the C. T. O.

And so and there's no chance that in the six months or one year period, I could get up to speed fast enough. Like maybe if you gave me a year, i'd make some progress, maybe, but you know I don't have the relent expertise. So it's also probably based on a very amateur, in a positive sense, amateur idea of government.

Imagine one hundred and century british rest craters who think that, like anyone in their club, if called upon, could sit in government and figure IT all out, you know. But life is very complicated today, and the things that government does today are infinitely more complex than they were in a, you know, and and tech society, you know, when a in authors. And so even if you assume that people could have done a good job, and then it's not so clear they could do IT now.

And even the authorities didn't do IT for certain jobs like they didn't pick their generals by sortin. They weren't done enough to think that they could go out and win wars by picking a random person to be the general that that requires some expertise. So that's that's a second problem.

And then the third problem is, is this job the kind of job you get Better at over time? Because if you get Better at IT over time, you're not going to get a chance to because you're in for six months and then you're out. And so you not if it's the kind of thing where there's upsides of doing IT for longer, you're onna lose the upsides.

upsides. There are also downsides, setting someone in the job for too long. But you know this sort of like a sort of like turn limits on on steroid .

ah so I think where I completely do with all that, and so where this is god, so so I was one of who a number of people who help to design this user assembly for for meta, it's called the community forum.

And based on these concerns, but also knowing that having straight up user voting would be very chAllenging, uh, we really wanted to focus on really value late decisions where it's not really a matter of so much of expertise, but really it's about trying to capture in this community what are your core values that, that affects this decision. And that might be a place where this is a more workable model. But I do think it's quite limited and and forward.

And I mean my personality. No, no, thanks. ks. My personality is I think these things have to be expanded in some way to bring in more expertise to allow for delegation as we call IT crypto.

Because of because of the things that is just saying, I think it's particularly hard that someone serves on one of these things, learns a bunch, then disappears. And so developing ongoing expertise in the types of decisions that are socially fraught. Going back to the beginning of this whole conversation, the kinds of decisions of platform might want to give over to its users. We're going to want people to develop ongoing expertise and a sense of accountants for anything.

And A N D says there's also there clearly a lot of upside. So I mean, take A A hard kind of values based, you know ethical and moral question. You know let's go back to the what what is the neutrality policy be on a social media platform, right? It's a hard question.

And you could imagine that if you take a thousand people and they are genuinely representative of the users, which is a tRicky thing, as we talked about, but imagine that they are and you know maybe they all have an impulse one where another way when they start. But then you give them enough information for them to have some portal conversations and then see where they land after, say, you know three longest conversations over three days, they may all land in a place that's a lot more thoughts. They will land in a place it's a lot more thoughtful ful than where they started.

And they may have a different perspective then, say, the people who work inside the company or the outsiders who start off very committed to one point of view or another, either because they're like, I don't know there, there are projects and they want to have as few clothes possible or you know, they are deeply religious, they want to have as many clothes possible. And so you know, you could imagine that you get and I think it's plausible. And I think you know I think in some experiments that andes run with with matter, you get a more thoughtful nuances bounce.

And I think for those purposes, it's great. It's it's Better than a focus group because there's maybe a little less control, you know in focus group in in the problem is if you're good at running a focus roup, you can make a focus group say almost anything. And here there are more people.

There's more space, they're more protocols to stop the people who are presenting the question from driving the answer in a particular way. And IT just makes sense that we would get a in general, we get a Better result, not every time. I mean, because there some scenarios were deliberation produces perverse results. You know where people get into a cast ade where you know this is like how people burn witches.

you think can heard mentality yeah .

you get into that. And then if everyone goes towards some crazy extreme thing that can happen. But there are a lot of techniques that you design um builds in that the design to stop that from happening.

Now part of this community forum is you're trying to provide information so people can make an educated decision about something and maybe um maybe change their mind about something. They arrive at a you know workable solution um but my question there is how do you how do you ensure that the information and education people are receiving is nonbiased objective and uh no because whoever is presenting you the menu of options, they have a great amount of control on steering. You told a certain outcome.

Good question for andy and then the other tRicky promise, how do you make decisions about what information decisions about what's the right information when I imagine what what we're discussing is precisely the misinformation problem. What's our baseline? You know, where do we even start on what what is reliable?

It's super chAllenging problem. I think there's a couple different IT is not of which are perfect and this is A A broad problem in all democracy, not just the online world and not just these assemblies um no IT you to perhaps sharpest in this disabled since the enter the organizing the assembly makes decisions about what information to present. I would take couple things.

What is you try not to have a monopoly on information. And so you know big part of these assemblies, the people get to discuss and delivery, and they're completely free to bring in any information they want gathered from whatever source they want. So that, I mean, that's the first most import thing probably.

Then of course, you try to do as good a job as you can of bringing in experts that are credible to represent the different sides of the debate. I think it's become invoke a little bit a at a certain part of the twitter community x community and of these days to say that there's lot of fraught debates in society where it's obvious that one side is factually correct. And so we shouldn't show both sides that both siders ism is is itself a problem.

Of course, there are instances in places where that might well be true, but I don't think you can get people to make an informed decision about any issue if you don't let them hear both sides or more than two sides of the issue, even if you think the preponderance of evidence rest on one side than the other. You really, I think kid, get to the for decision as you heard all the arguments. And so I think that's A A big effort and this is get uh A D A diverse set of experts um to present all sides. Uh but you know long run I think you want to be in a position if you really can build out a democratic system that goes beyond you'd want to Foster a competitive ecosystem in which different information providers crop up and provide their different analyses and views. And that's something I have talked quite a bit about with double because I think that's an issue in print to to on that point.

There's also a body of literature about um you know people in brainstorming settings and when there are bad ideas proposed to actually enables the group to arrive at Better decisions .

over time yeah I mean that idea which I mean the the the classic modern formation about is john Stuart. No you know his argument for free speech was we actually need the bad arguments because what working through why they're bad and wrong and false will help us get the right argument and that IT sounds kind of like a tried argument.

But if you really think about IT, it's it's pretty deep because it's not obvious um and it's and it's interesting in that in that way, I think another really interesting kind of aspect of this whole chAllenge to me is jumping out to the bigger political world of democracy you so you are asking before Robert about so any time to talk about the us. Supreme model. I the same in a way is kind of true about these ideas of democratizing.

But one of the aspirations of the community assembly reforms, that citizens assembly of community forms that and he's talking about, is to get real disagreement, that is, thoughts that doesn't turn into policed, yelling at each other. And one of the ways to do that is by narrowing down the issue. So point of one weird feature of our polar ization, one web feature of polarization, is that people turn out to have really strong views about things they don't know anything about, and they have a instantaneously.

So you're like, how like how do you you know, this happens to me something like talking to something they're like, oh, i'm sure I think act and I like, I knew you for thirty years, you know, aiming about that. Why do you have the strong position? And the answer is, as a time saving device in a world where there are so many issues and they're so complicated, once we've picked the team, you know, we're blue, we're red or were libertarian or were anarchist, whatever we happen to be.

Once we've picked a team, there is now a list of positions that associated with that team. And so we just default as a defensible time saving hurst ti C2Citing wit h our tea m on tho se iss ues bec ause we kin d of thi nk tha t som eone els e has tho ught abo ut, a whole bunch of people have thought about. And probably this is where i'm gonna end up because I generally agree with this group of people on various things.

And one of the things about doing a citizens assembly or a community discussion about a narrow or topic is you can sometimes avoid this habit that we have of defaulting to a polar ized position as a time saving device because you're going to be told about IT like you're gna get the information in front of you. And so it's not only that we're closed minded and that we don't want to listen to the other side, although that does happen, we also partly close minded because were raged with so much information, but we don't have time to consider every perspective on every issue. We don't have the cognitive bandwidth.

Do IT, I actually think, gets one of the problems with our current polarization. It's not the only cause of IT. There are been polarization in many societies that denied any much anywhere near as much information as we're getting. So it's not if i'm not making some like you know, this is the main cause of factor, but IT is one reason that we do Better in a well designed citizens assembly often than we do like out in the wild of politics.

I think that's been one of the most interesting to me, surprising learnings them both in the physical world end online, is there generally organized around the most fraught kind of culture wars type issues. And when theyve been run, one of the big learning S, I think, has been that, you know, if you get a relatives, small group of people together and you structure the conversation well, uh, people generally are pretty reasonable. They know that turns out they don't want to be so partisan, are so hostile when they're put into the right environment. And that connects to other things about our polarization rate, which is that IT is much a larger than IT seems because we consume so much of the through these online platforms where we we hear the loudest voices, but we don't see the in percent interactions that turned out to be usually less polarized.

I had a funny example of that for my own life. I was I was moderating a panel at harvard, and some students decided they're going to leaflet against me. And they were handing out left people as they came in and the leaflet was like IT was like a four page leaflet um but its tax was entirely a download from a twitter likes, a twitter thread that someone had created, attacking something that I that i'd written and I was like this is so IT was just so weird because IT IT didn't translate well to the format of the know to the form of the you know of the leaflet but also I just not to myself the extremely of the formulation that this person was using, you know against me online.

You know they didn't really translate to the form where, like we were having a thoughtful conversation and I I sort of feel bad for the people who were relieved ending because it's one thing to you know call what I said. I think that bid horse IT, which is what they were describing me as saying, like they seemed in context weird and rude to be saying, because everyone in the room was having a rational, reasonable conversation. Where is on twitter? I'm sure they seem great.

And then the punch line of IT was, there was like a substantive issue where I had in the thing i'd written that they made, they made them angry. I'd made an argument, one possibly wrong, but even I gave evidence for IT, and I supported IT. And their response to this argument was, it's obvious that this is wrong.

So we're not even going to entertain IT and I like that just sounded so absurd in the context of the university on twitter. IT sounds grade right, like feldman is is full of ap or so. Why should even bothered review them? All you need to know is that we think he's stupid.

And then that's enough to reach the conclusion. But in personal conversation, very hard to Carry that off without looking silly. I think .

since you brought up harvard, are you feeling this intense tension between the administration? The student body IT just seems like such a powder keg over there.

IT was a very, very, very intensive and I mean, very nice as intense I have experienced on the university campus. And i've been on off of this campus since one thousand nine hundred and eight eight. So IT IT was extreme.

Now things are sort of lowly. Getting back to maybe Normal is too strong word, but towards a calmer way of being. And I will say during all of this, you know, our classes were Normal.

You know, like school went on, you know, people learn stuff, they studied for exams, and we did have a Normal campus life by all of the intensity that was being felt. IT was felt. And so he was IT was an experience and not for the most part. I think it's hard to say for most people, they don't think IT .

was a good experience. Have things .

subsided a little bit? Yeah, I I think so. You know I mean, there is now.

So what are university is bad at? And what are they good at? Universities are really bad at making fast decisions about anything they don't like to do IT. And if they do IT, they make bad decisions.

Because if your personality were that you could really react in real time, you know you'd be a great trade, some job where you could take advantage of that. Um and of course, there are some people in university were really intellectually quick, but the best of them use that intellectually quickness to fill in. They are debt and then you make the decision over time.

So universities are bad at the fast stuff, and the fall was all fast stuff. Responding to a news cycle, making, issuing statements and decorations. Now we're entering a phase where certainly here you know the university has two task forces already set up to say these issues.

There's two more, at least coming, and this is going to be the part universities do pretty well, and take a deep breath, figure what we've done wrong, fired what we can do Better, give reasons, explanations of how we should do Better in the future. And so our kind of the nervous system of the university will return closer to what is best that you know. And again, I wanna be clear, there are some people whose, you know nervous that is always like, go, go, go.

And that's really valuable in a whole bunch dimensions of life. The university is not really one of them. And so we do our best when we read the little bit.

And that's that's now what's happening um and that that's Better. So in that sense, it's we're heading directionally in the right way. By no means are we there yet.

I think it's it's a really good important example of just the general topic we been discussing s because a lot of IT has to do with governance and the governance of these super important. Very in some cases, very old long running institutions that are not straight forward businesses and therefore face quite complex governance issues.

And certainly some of the ones you know i've observed over the last let's eight, ten years that make IT so chAllenging is you have the exact same uneven participation problem. So not every faculty member leads in and participates in the governance in the university. The one who choose to do that may have different views than the one who don't. And then on top of that, you have necessarily mean this in a negative way, but like the mission creep of the university, which is a lot less focused on protecting academic random and developing the very best research in the world and a lot more focused on a lot of other issues that extend way beyond research. And I think it's hard it's hard to figure out how to fix those governance problems because the institutions are so big um entranced so many different areas and trying to spread their attention across so many different things.

And I want to step back from, at least in my view, to step back from thinking that by making declarations, the institutional part of the university contributes to our understanding of the truth, that of knowledge. I I don't think that's true.

So if the deans get together and hold a meeting and announced that the second law, thermal dynamics, is true, I don't think that gives us much special knowledge into whether the second of the dynamics is in fact true. However, you know, when the people are at the cutting edge of some area of inquiry, published their research in a peer review journal, there's a reason to take that seriously. It's not always correct.

It's always subject to revision. But I would stop and listen closely to what they had to say and give you some you know some of physiological you know benefit of the doubt because they're expert and they're speaking on their area of expertise in a thoughtful way. And I think one of things that happened is that are universities, and not only are universities, but it's happened in universities, have have fAllen into this part of the mission creep that is describing thinking that they have to express the public statement on every matter of public importance.

And I understand the moral impulse to say the right thing, but that has to be baLanced against. Are you good at that? And are you contributing to the universities mission, which is to pursue the truth in the broadest sense, you know, through reading, writing and teaching, but that's different from you know being in the announcement and decoration game.

And I don't think universities are very good at that. And I think stepping back from that is not this by no means in overall solution, but it's the first step. This is happening .

all across corporate amErica beyond um you know these calls for activism and for institutions to take a position on various matters.

I have clients coming to meet corporate clients every day saying we're in so much trouble about this, you know and I do try to tell them there's no such thing as being neutral. That's the first point. Genuine neutrality is not possible in the world.

But within the frame of realizing that you can be neutral, you can sometimes step back and say, look, you know, we're not gonna take of you on this. Is that important thing? And there the companies find themselves in his positions because they're lobed and they're lobby by people who are trying to affect consumer behavior often and sometimes they're able to pull that off.

I mean, that has been subject to a boycott. You know. I mean, other company becomes are also a subject to to various boycotting. So you, you you are living in a real world.

If you have customers, you have to worry about your customers, but you also have to be aware that your customers could be all over the place. And that's one of the reasons to have step back policies, including policies of referring something to some other group of people and saying this is a really hard one. We're not qualify to decide this and we're handing IT off to somebody who is qualified to weigh on IT.

I agree, and I think the key value is the time your heads you really you have to make what we would call on the such sites is an incredible commitment. You have to get to the point where the people who would try to force you to take a position on this issue that's irrelevant to your core mission, yet important to them, believe that there's nothing they can do to make you take the position.

And I think the critical mistake that a lot of the most important universities made, as well as a lot of corporations made, was to give in to those demands and then create common knowledge that they can be forced to make those statements and that they aren't committed to not making those statements. And you know, people often point to her, but the one of the few universities that had less of a problem with this stuff has been university chicago because they had made a preexisting written commitment to not take such positions. And that has turned out to be a huge luxury for the the other universities have been afforded themselves because they hadn't made that commitment or the commitment wasn't credible.

And I think that I think you know, the central problem is that over a ten plus year period, a lot of the top universities demonstrated that they did not have a critical commitment to not taking those statements. And now they're trying to walk that back. But it's hard to create that. Nothing I D never .

missed not to bring the conversation back to some of the subjects we were discussing at the very outset around internet governance. And um you you're mentioning this conceptive credible commitments uh in the context of the intercept. Chicago when IT comes to making commitments with internet services and binding people the rules, I wonder what the path forward is there.

When you look at something like the matter overside board, they have made some decisions that met a doesn't necessarily have to adhere to. Now IT adds a lot of transparency of the process. But I wonder if there are other systems that would you know bind services and corporations and applications to the way that day uh to some some set of you know agreed upon circumstances. Of course the thing that comes to mind is dios distributed autonomists organisations. Are you concerted create these uh commitments in blockchain code that bind all participants?

How to say I mean, my view list, I think there's two things that are interesting about blocking and deals with respect to this topic. The first is the trust problem and being able to write and commit to a process. So one of the first things, you know projects do when they start in crypto to is essentially right, a constitution I E write down in code.

How they're gona make decisions over different issues, and you know exactly how binding those are is up for debate is a little bit complicated. But in the long run with blockchain, I think IT really is true that the constitution you commit to do that process in code is a lot harder to change without resort to some kind of democratic process that if it's not committed that way, especially that goes way beyond what can be protected through Normal real world legal processes. And so I think there is something really interesting about being able to build online community where you can make a sort of promise into the future that any time this type of decision comes up, this is the process by which we're going to make that decision.

I think there's something quite interesting about that. The second is more economic and nature and has to do with, you know, one of the most important types of trust for online platforms is the trust between the different sides of the platforms market. And so, you know, what's all online platforms have this characteristic that they try to bring together two sides of a market, drivers and writers or developers and users for the APP store and so forth.

And you really need the producer side of that two sided platform to believe that into the future, they're going to have an economically beneficial relationship with the platform. And I think one of the biggest chAllenges were seeing the space right now look like the huge dispute between the epic games and apple, for example, is that developers are starting to feel like it's still a tremendous opportunity to develop on top of these extraordinary good global platforms. But at the same time, the taxes they pay to the platforms are going up.

The decisions made around their services are changing and unpredictable ways. They would like to have a much longer term promise about how the platforms going to treat them. And that's another place where I think being able to make a long term, in some sense, immutable able promise over the economic relationship between the platform at its producers, I think could be could be aboard .

the broad sense. That's what a constitution is. I mean, a constitution. The best metaphor for IT is, you know, dishes st in the autism when he's on his boat and he's going to the island of the sirens and he knows they are going to be so appealing to him, so beautiful, their song, so beautiful that he's gonna jump.

And so he, you know, he ties himself to the mast and tells his crew members like, don't let me go that's in one of the leading metaphor, is what a constitution is. I mean, you know, the idea is sort of like we know that when the chips are down and we're in a panic, we're going to take away people's rights, are going to silence them, are going to take away there. You know, there are right against arbitrary rest.

And so we we try to bind ourselves from doing that, and we try to create a institutional mechanisms to bind ourselves, notably in government. There are two parts to that. There is the written rules, and then there is the human ability to interpret and maybe even override those rules. And there's a productive tension between those things.

And IT goes all the way back to a debate, believe or not, and he could be so ancient greek, but have done a lot of ancient greeks today, a debate between play article about which is Better, whether it's Better to have the rules in the end in charge, because no one's perfect, which is the outside of you, or whether it's Better to have the wisest person you can get your hands on in charge because, yeah, because the rules won't give you the best outcome all the time, which is broadly the plate of you. And they are both right. You know you need some you need some back and for the production attention between those things and that's true.

And dose where you know you don't want the thing to lead to a total spiral down, you need to have some break glass measure ah and it's also gonna be true in um you know in something like the oversight board where the company sometimes has committed itself absolutely defy in their rulings. Other cases it's asked for an advisory opinion. But in the end, you know, if ma chooses to stop, listen to the advise to the other airports all together IT, could they would just have to pay a reputational cost for doing IT?

So I mean, I think you know this really is in the realm of art rather than science. You always need to have some of each. You need to have some meaningful constraint and then you have to have some capacity for flexibility.

Um and so IT be nice to say there is like a magic solution to this, but there there isn't and there's no purely technological solution, but there is no solution that has no technology because rules and the buying of our technology. So you you need both and that's um may not be the IT may not be the most um drilling conclusion. Rules always win or people always win, actually sometimes some wins and some the other wins and that's that's real life. Um so maybe that's not a terrible place to to end. What I have to say on the topic.

I want to build that one particular way, which is like one of the things we study a lot in the history of democracy, is this idea of of constitutions and what we call political sides and power economy, self enforcing constitutions. There's no extern authority that combine the country to its constitution. And so the constitution only has power in the long run to force people to follow the rules.

If there some track record of everyone having agreed to IT or long enough has a special power that IT gets that IT creigh over time, because there's nothing to stop someone from tearing and up and ignoring IT. And so that's why we call themselves enforcing. They only really bad. The long run with everyone has a long enough track record of agreeing to be built by IT.

This goes back to the magnet to where king john was forced to sign IT, right? And then he was sort of like, well, actually i'm not going to pay attention to IT that much and then know there was a battle over the legitimacy of the crown and .

that's a tons of history of this and and the U. S. Sitution is somewhat unusual in the link of time, for which is proven to be somewhat self forcing.

That same problem online. And I think nit is right. He was hitting at this. I think that block chain or ways of writing immutable able agreements in the code do not, uh, in a comprehensive way, solve this problem of self enforcement because you still always, people have the option to fork and to leave and to do whatever they want or to change the code.

And you know, as long as they agree to IT and there is still something deep in, like he was saying, arts really than science about IT. However, and do think for the online world, blockchain provide something prety fascinating that I was really blown away by as a political scientist when I discovered IT. And I think it's best highlighted through an example.

So there is a door called lighter. Lighter has been discussing publicly for a while it's desire to build in a vito for certain types of decisions. And what they want to do is essentially update their constitution to say when there's this one set of decisions, maybe the deal just makes the decision.

But once this other type of decision IT goes to an external video. And when I was talking to them into other people about how that would work in practice, one of my biggest concerns was that in the real world, if you try to set some procedure, ural rule, that only applies to one set of legislative votes, not another. That is just obvious, that strategic actors will change IT into whichever category is favorable to them.

So if that they just claim, like if I don't want to veto, apply this claim. This is one of the votes that the V S. Count on.

And because legislators get to make all their own rules, you really can't stop that. And when you seen that in the senate, occasionally the senate parliamentarian tries to stop the senate for doing stuff. There's a weird equilibrium.

Sometimes the senate differs to the parliamentarian, but in the long way that the senate has no real deep obligation to do that and can always change its rules if IT wants to, if enough people want to. And so you kind of thought this video not going to work, because you can just redefine light. Those votes into one category, the other.

But I think the thing that turns out that I learned that really interesting is, no, because these votes, the the topic of the vote, is defined by the smart contracts that the vote actually touches or doesn't touch. You can define in a very deep and immutable able way, whether this is one of the votes that the video s going to apply to or not. And that allows for a form of commitment that I ve never seen before in the legislature.

I don't think it's a panacea. Uh, I think there are still these broader issues of self enforcement that matter for all the reasons to do is saying, but I do think that offers something pretty fascinating. A in the vate of binding commitments to legislative procedures.

Isn't this the miracle of the U. S. constitution? No, I mean, you have written a biography, James medicine. The fact that you can set these up front and yet make them flexible enough that they can endure well.

if we had medicine here, he would agree that IT was his his perfect design. Ah you have to ignore the silver war and some other uh, blips but yeah I mean, the aspiration was to create something that would be able to be both rule based and also responsive to change over time. Some of the design elements didn't work well. I mean the amendment provision we we use IT very, very rarely. Um he probably thought that was generally okay, but there been some circumstances where we really need reforms and we can't really get them more fundamentally, IT was is a system that's designed to enable compromise and to enable compromise in the middle and to push politicians back towards the center.

And that's why you know right now, if you ask about of the big questions, the grand questions, you know, lagging politics and political science, I think it's fair to say that if you have some confidence that the laws of political science are still true, then you think that we will migrate back towards the middle, because our system is designed to push us back towards the middle. And if you, on the other hand, are really panicked about where things are going in the possibility for breakdown, you think that I might take to, you might still believe in the rules, but you think they might take too long to Operate. And then in the short term, you can get a breakdown in overall faith in the system capacity. Any other part of IT is looking at the world around us and seeing that for all of our polar ization lots departments of our society are still functioning and functioning really reasonably well.

amazing. This been a fantastic conversation. Thank you all so much for all your time.

Thank you so much.

Really fun. Thank you. Yes, great.

We sort of took IT by the that direct democracy has problems. We mentioned bodiam c boat face, which is a pretty humorous example. But I like, are there more serious examples from history, from physical world of direct democracy in action and going right?

I mean, the most classic example people point to is, uh, in which the america's founders pointed to was athens. So authors had a pretty robust direct democracy. The reality of heart function was quite complicated. IT wasn't as simple as just like everyone shown up and made every decision. They're actually like a pretty complex layers of decision makers and stuff, but there was a significant component of direct democracy among lendin then. And a one way that people say they went right is that during the war with Spark um you know this is just one telling but like the the people got swept away with passion that were uninformed about the strategic decisions that had to be made and essentially forced athens to open a second front in the war by invading sissy and that's what they turned into athons vietor he was like a disaster um and IT massively sapped athletes wer was super, super costly and that is often held up as the example of like mob rule.

This is because the decision to open a second front and invaded sisily was now you said there are complex layers decision making but yeah in this respect, IT was kind of a bottom up. So like, no.

And the claim is that the mob was manipulated by ego driven generals who wanted to british their reputations by opening a new invasion, and that they manipulated the mob into supporting them. And if you look in the classical era, and you know, you can question the motives of all these authors, that sort of the the stereotypes that arises is.

And so post athens glory days, a lot of Robin writing refers to mob rule in the ability of demagogues to manipulate the mob as the reasons to be very skeptical of direct democracy. Uh there's like a really famous passage and virgils the need where he goes on IT like um about how you know cynical leaders can whip up the bob and what we really need to seasoned statement who will you know make decision to carefully on behalf of people you know more modern examples. Hard to say it's so and functional that it's really not even try to yeah it's really I mean, you can point to like a in switzerland and there is a pretty aggressive local referendum system.

I think there is general agreement that is kind of crazy. But like IT doesn't work as badly as the other thing. You literally at the case on level and switch and ju a vote on who receives passports is really quite well. Yeah, there's a great research on this.

You know in the U S, people point to california is pretty extreme on the end towards direct democracy and that can go both ways um in some ways, I think it's important and potentially valuable institution because IT allows voters to surface issues that the legislature might want to ignore. And so if you think your legislators are due to good job or are captured by special interests, or for whatever reason, or not sufficiently accountable to voters, that giving voters this alternative mechanism to force issues could be quite valuable. The downside is that you end up a tremendous voter fatigue because you have lots of these votes in the car ford baLance.

Ridiculous as many, many pages long. Most people, including myself, can understand most of the issues being voted on. And then the second problem, and this is often a problem, direct democracy, is interest group capture of the agenda.

And so how ballot issues get on to the ballot is complicated but it's actually not that hard for a well resource committed interest group to force car bouts for themselves onto the ballot. It's one thing that happened to california cycle. We have to vote on this extremely abstruse, uh, ballot is that has to do with whether doctors should be required to sit in on all dialysis. And you might think it's doctors who are pushing this because you might think like all they make money from IT. But no, the doctors do not want to do this and they think it's crazy because there's no medical reason why adopted its me present for the administration.

just a ton of time wasted and .

yeah yeah, it's crazy. And it's being pushed by, uh, certain interest groups that are kind of just more generally in a conflict with doctors. And they have stayed publicly that the reason they put on the ballot every two years is to force the doctors and scripts to spend money convince to everyone not to vote for IT. So it's a it's a complete waste of everyone's time and resources. And so that I think, is a great example of how these process is to get messed up.

By the way, the irish referendum that just happened, my brother in law was just visiting from irelands. I was interested in this on um I IT was interesting to see people reject so strongly these proposed changes to the constitution tracking IT no.

I haven't father very carefully. But I do think that's a good example where referendums, if they are embedded into a broader process in a healthy wake, he would be a really good way to get more signal from for voters. And this happened called for people I think was twenty eighteen we had a bunch of pretty like culture or related referendums or about ships yeah and quite consistently, the voters really signals through them that they were in a more centers position than california officials um and I think that had a big impact on how the elected officials that had proceeded for that on after they lost uh these these big ones so I do think I can be valuable. But it's complicated how you enacted practice and I think that kind of were the governance heading there will still be some direct token holder voting, but there will also be a lot more delegation to professional experts of these issues.

S you need to protect against mobile but also protocols. Y and exactly find some sort of um you know middle way .

yeah and I think that the delegation stuff, you know, it's very valuable, but it's not going to solve the participation problem. fundamentally. You're still going to do the token holders or the other voters to pay attention and make sure their delegates don't go rogue.

And that's git a require some pretty careful planning because there's definitely attempt to set IT and forget IT in terms of delegating your tokens. Um but we have lots of reasons to to suspect that if that behavior manifests regularly, the delegates tes won't have those at city. We want them to have to do a good job. So I think a lot of the most interesting work right now, a doll, government, a road. How do you build delegation programs that, first of all, recruit and give the incentives to delegates well at the same time, second of all, still encourage token holders or other voters to pay attention and to think about redelegate their vote on a regular basis so that delegates feel like they're being washed?

What are people putting in place? What sorts of new rules or experiments are happening?

First thing that happens, what I think is really interesting, is a bunch of the largest house of instituted, they need to have a different name, I would call the delegate programs, which is a combination of offered IT, includes paying the delegates sometimes as a function of how many votes they are. Crew and creates like a online web interface that makes IT easy for for token holders to delegate to different delegates.

And so you're putting together your you're creating incentives for there to be delegates in your helping token holders find delegates. Um and we've actually done some research. There's a pretty interesting evidence that will we get out those programs actually does increase token holder voting participation probably because you're asking the token only to do a much easier task instead of voting on like literally changes to the underlying protocols code. You're just asking them like find one of these delegates who you like and give them your voting power. And part of these programs is also having the delegates right platform statements in door post videos about what they want to accomplish as a delicate that again helps the token holders find like, oh, know that to delegate the shares, my views are going to delegate .

to the for faster. So this is a paper that you're working on currently.

Yeah yeah. I'm hoping to have you done the next month there.

See, very cool. And so basically getting people to delegate their votes more, uh, making IT low friction, making the information available for people to make reasonable decisions about who they want to delegate to and the directions that it'll take A A W exactly.

yeah. No, I think it's going to be a really important model because, uh, this came up in the conversation. No, I think a lot of people in other parts of online governance are coming to the same conclusion that we need representatives because we need this kind of expertise in the accountability and we can't rely on direct democracy. But web three is years ahead, literally years ahead, in with how you actually set up representative democracy. Uh, so I think it's going to be yeah really important development.

I mean, we've got this intense like Darwinian combat going on. You know all these experiments just let loose and seeing which one's will flourish. How far away are we from you know actually determining what works, what doesn't work and seeing you know the kind of best practices shake out from all this?

I don't know. I think there's probably two aspects to IT that we need to have happened before will know for sure what is to get dance to a place where they have, you know, more broader killer use cases for society, which will in turn make the governance decisions higher stakes for society and that will see which government structures can stand up to that pressure. Uh in some ways there already involved in high stick decisions.

Uh, certainly these like d five protocols and stuff, but they don't have the same kind of uh, global public pressure on them that other online platforms have faced because other online platforms are much more mature, have many more you know regular users and so forth. So I think that's going to be a really big interesting shift as IT occurs. That's going to bring the down governments more together with what we talk to with no answers of like web two point o governance.

And then the second is we we need over time to have more double with broader distribution of voting power, which is happening over time. You know, big critics of dw historically has been, you know, they are closed to the rhetorical c of democracy, but the voting powers very undefined, distributed. And that may make some of this, like delegation experiment, sort of unrepresented tive of what would happen in a broader democratic system, which is more conflict among users with with roughly equal amounts of power uh, and so that is the trend, I think, in token holding over time.

And again, as this become larger, as they touch up more killer use cases for society. I think we'll see that broader distribution, and that will be another kind of pressure test for these systems. So those are the two things. I keeping my iron.