S5E01: Our own Titanic: Overturned tables, a Gashed Hull and the Search for Lifeboats

Foreign. This is Larry Lessig. This is another season. Can't even remember what season this is. Let's call it season five, season five of the podcast another way we launched in 2019 thinking of other ways to get us a democracy, if not back then. For the first time, this season is going to be focused around the development of an argument for a book. Sounds pretty boring, I know, but I don't think it's boring, actually. I think actually, having done most of the interviews already, as I wrap up the framing for the season and introduce it, I think it's quite an amazing set of thinkers and writers and activists who will be in this season. The season is organized around a metaphor. Here's the metaphor. So think about the Titanic and imagine you are the captain on the Titanic. You've just hit the icebergs and you don't quite know what's the status of your ship. So you climb out onto the deck and you look across the deck and you see many overturned tables and ice across the deck and glasses that have spilled and broken and lots of disruption. And you think to yourself, we can fix this. This is fine. We can clean this up. We can write the tables, sweep the deck. Everything will be in order. And then your staff comes to you and informs you that there's a gash in the hull. People used to think it was like 200ft long. Actually, it was just 30ft long. But that allowed five compartments in the Titanic to fill at about 1400 liters per second. Which meant, as the captain would realize, that didn't matter if you righted the overturned tables. The ship was going down. And the challenge then would be to convince the passengers on the ship to climb into the lifeboats, which, in the middle of the Atlantic, in the middle of the winter, in the middle of the night, standing on the Titanic is not an easy task, because, at least initially, when the boat is just sitting listless but not obviously sinking, it's not clear why anyone would leave the unsinkable Titanic to climb into these tiny white lifeboats, to float aimlessly in the Atlantic. But that, as the captain, is your task to convince them to take the steps that might save them. Okay, that's the metaphor. Here's how that metaphor fits to the story I'm going to tell across the episodes of this season. And these episodes, as I've described, are the kind of working material that I'm using to write a book that will follow the very same form. So you'll hear the season. If you listen to the season and you'll understand the argument of the book. And then maybe you'll want to read the book when the book is published, or share the book, or have the book on your shelf to make it seem like you've read the book, even though you know the story of the book by hearing this free podcast. So here's how the metaphor fits the overturned tables in that metaphor are democracy's flaws, or, let's be precise, representative democracy's flaws. Anyone listening to this podcast, or anyone paying any attention to the way democracy in America functions right now, or doesn't function right now, recognizes there are many flaws in our democracy. I've focused for many years on the core flaw of money, the way money corrupts the political process. But gerrymandering, vote suppression, partisan primaries, the filibuster, These are all the dimensions of the flaws in America's democracy. And the argument I'm going to make in the first part of this podcast or this season of this podcast, is that all of these problems are fixable. We understand how they could be fixed. That's different from saying we could build the political will to fix them. We came close when we came close to passing the for the People act or the Freedom to vote act in January 2022. But we could fix them. These are all things that could be addressed through legislation or the like to make this unrepresentative representative democracy actually representative. But I want to argue fixing these flaws is, in a critical sense, useless or irrelevant because of the gashed hull. And the gashed hull, the parallel to the gashed hull in the metaphor that I've described is the emergence of what we should think of as AI writ large. Now, when people talk about AI today, they have in their head ChatGPT or massive machine learning systems, or the machine learning systems that feed social media. Those are all versions of digital AI. What I want to first argue is that we should think about analog AI as the precursor to digital AI. And what I mean by analog AI is any artificial entity or institution that humans have created for some instrumental purpose, for some reason, and that institution functions or operates for that reason. So in this sense, democracy is an analog AI. Democracy is an institution we've built for the purpose of resolving or aggregating or working out our common problems and implementing them in the form of governance. It's an AI, it's artificial, and it's an intelligence in the sense that it takes facts and inputs, and it structures what it does in light of those facts and inputs for the purpose of advancing its instrumental purpose. And Its instrumental purpose, at least in theory, is to advance the interest of its public. In the sense corporations are also. And analog AI originally structured by state legislatures, charters given by state legislatures telling them what they could do and the conditions under which they had to operate, the price they had to pay or the limits they had to accept in order to do the thing they were charted to do. But evolving into these entities that have a general legal freedom to act as an entity with the purpose, the instrumental purpose as it's come to been recognized, even though this doesn't actually make sense of the history of corporations, but this is the way it's thought of today, thanks to Milton Friedman and the Chicago School. Its instrumental purpose to be maximizing the shareholder value of the owners of the corporations. That's an instrumental purpose which this artificial entity advances through its own instrumental intelligence. The structure of figuring out what it should do in order to advance the objective of maximizing shareholder value. Corporations, of course, long before digital AI entered the field, had been criticized for their intervention in the democratic process. Democracy is puttering along, trying to figure out what it should do to advance the interests of the public. Corporations, having their own distinctive interests, have intervened in the democracy to bend the rules of democracy to make it so that the corporation's interests can be advanced. There's a competition here between these two analog AIs, and most believe, I certainly believe, that the corporations are better, more effective, more efficient AIs in achieving their objectives than the democracy is. Democracy, in this sense has been overtaken by, by the corporate AI. Okay, but then those analog AIs are tiny as compared to the potential that we're seeing in digital AI. And the first contact, as Tristan Harris puts it, we had with these digital AI in a big social sense was the way digital AI drove social media, especially in the 2016 and 2020 campaigns. Because social media spread throughout our political society, the platforms on which it spread had an instrumental objective. Their objective was not primarily to win for the Democrats or win for the Republicans. Their objective was to maximize engagement. That's what the machines were trying to do, to maximize engagement. And they deployed their techniques of tweaking you with certain stories or dragging you down a rabbit hole of feedback to help you focus on some obsessive or some unhealthy or maybe some healthy ish interest for the purpose of making you addicted to the platform. That was how it got rewarded. The more you spend your time on the platform, the more the platform would earn revenue. And the objective function of the this AI was simply to maximize the revenue on the platform and the conflict between that objective and a safe platform is what we see revealed in Francis Haugen's the Facebook Files, which Haugen, of course, was the whistleblower who turned over the extraordinary cache of documents from inside of Facebook, where again and again we see the engineers, these decent, really brilliant people trying to architect the platform of Facebook to make it a safe platform for democracy, being overridden by management from Zuckerberg to others in the political control of Facebook, overriding safety decisions for the purpose of maximizing engagement and thereby maximizing revenue. So that objective function maximizing revenue had a consequence in politics. And that consequence was to radicalize all of us or those of us who spent their time on Facebook. That consequence wasn't chosen. Mark Zuckerberg doesn't hate America. He wasn't trying to destroy a democracy. He was just trying to make a lot of money. But the consequence, the externality from his perfectly legitimate activity to try to make a lot of money, the consequence, the externality, was further polarization of the public. But this first contact with AI is tiny as compared to the contact we're about to see in the 2024 election and in the elections going forward. Because what we're going to see is an AI, a digital AI, an order orders of magnitude more sophisticated in its ability to engage with us, to get us to do what it wants us to do, taking advantage of relatively weak minds of all of us, tempting us, teasing us, tweaking us, reaching down the brain stack of all of us to produce a reaction that does what the objective function of the AI wants you to do. Now, this won't just be the objective function of maximizing engagement. Indeed it might be. That's not the most important objective function to worry about. As I'm recording this, Taiwan is preparing for a very significant cyber attack from China as Taiwan enters the their election in January 2024. And they expect that cyber attack will be deployed for the purpose of undermining the election or weakening the results of the election. Especially as they might signal the resolve of Taiwan to remain independent from China. But there gearing up to protect themselves from that attack. And that attack, of course, is not aiming to maximize engagement. It's aiming to achieve the political objectives of China inside Taiwan. Now, you know, you should know, if you don't know, you should wonder why you don't know this. But you should know that's going to happen to us too in the election of 2024, certainly by the Chinese, probably by the Russians, who else who knows? But the point is, there will be a large number of political actors who are acting not just for the purpose of maximizing engagement the way TikTok or Facebook might. They will be acting for the purpose of achieving a certain political result. And they will be deploying these super sophisticated AIs to do that, to figure out exactly how to tweak or bend or engage or not engage with a public to turn people out or turn people off or get people to spin up their anger machines or to turn away from the candidates they otherwise would be supporting. All of this will be happening in 2024 and in the elections afterwards, because AI, digital AI, will make it so. And as that happens, our democracy will be threatened. And indeed, I want to argue. It's a hard argument to sustain. I get it. But I want to argue that as it happens, we should recognize the capacity of our democracy to do its work, which is to figure out how we solve common problems in a way that we can all live with and work with and move on with. That objective of democracy will be defeated. This is the gashed hull. We could fix the problems with our unrepresentative representative democracy. We could fix them. But even if we do, AI, in this broad sense of both analog and digital AI will defeat democracy's purpose. And so the question then is, what are the lifeboats? What are the ways in which we might relocate ourselves so that we could do democracy protected from these AIs? And that's the third part of this season, the Lifeboats. And my objective here is to get you to think about democracy differently. Not imagining that we're going to immediately recreate democracy like tomorrow, but that we start down a path so that eventually we can build a different kind of democratic structure that is protected from the corrupting or corrosive influence of these AIs. If you haven't seen the movie Independence Day, you probably don't have to waste your time, but it's a very dramatic movie about Earth being invaded by aliens. And the aliens are a superior intelligence. They have the capacity to organize, to defeat us, and they have the capacity to listen to all our communications, so we can't even communicate around them. But at the end, they realize that the aliens don't know Morse code. And so all of the armies around the world attempting to defend against the invading aliens begin to communicate in Morse code. And because they can communicate in Morse code in a technology the aliens can't understand, they are protected from the interference of the aliens. They're able to rally and defend the earth. That's a powerful metaphor for what I'm suggesting with these lifeboats, because the objective of this way of thinking about the lifeboats is to think about how we can move what we think is the essence of democracy into a place where AI can't pollute it. And so what would that move look like? Well, this story is going to begin with a pretty big conceptual move. I interview David Van Raybrook, who wrote a book against elections. And that book tells an extraordinary history of democracy, which you could summarize by saying democracy is thousands of years old, but only the last 200 have been dominated by elections. All of the history before the last 200 years is a history of democracy being done through a technique of sortition, meaning random selection of citizens to participate in the acts of governing, whether that's legislatures or administrations, regardless, or overseeing for the purpose of protecting against corruption, regardless. These institutions of sortition brought citizens into the process of governing regularly. And that was the essence of democracy. And as von Rayboeck quotes both Aristotle and Montesquieu. Montesquieu put it in this way. He said, selection through lot tends toward democracy. Lot means through sortition, random selection, selection through elections tends towards aristocracy. And indeed, the Americans certainly, and the French a bit too, but the Americans certainly thought they were trying to establish a certain kind of aristocracy. Not the aristocracy of wealth, not British aristocracy, but the aristocracy of excellence, that they would structure a system of elections so that the very best would rise to the top and they would be the people governing our society. Okay? Of all the things the framers were wrong about, the little eye towards the clowns that now dominate Washington, this was one of the things they were most clearly wrong about. And so the suggestion of von Raybruck is that we think of ways to begin not to eliminate elections, not to get rid of them as part of democracy. And certainly I'm not arguing we should get rid of them, but to begin to complement the democratic process with sortition based systems that help us better determine and represent the will of the people. Sortition systems that are insulated, protected from the corrupting influence of the AI. And so, in this section of this season and this section of the book, I review a wide range of examples across the world and increasingly in the United States, of sortition based systems beginning to identify and do democracy differently. These include citizen assemblies from Ireland to Iceland to Europe and in the United States, as well as techniques for facilitating an understanding of common ground and fleshing out common ground without requiring something like a citizen assembly. Polis here deployed Most dramatically in Taiwan, but now being deployed around the world is an open source platform for identifying common ground. In this sense. We talked to Kim palace, who's developing a platform that will be available atcommongoodai.org for facilitating the identification of common ground among a wide range of people, but in an asynchronous manner. So a much easier way to determine that then it would be determined in the standard way of gathering people together and having them deliberate. But the point here is I want you to see the wide range of techniques that are now being developed and deployed for facilitating an understanding of the democratic will without relying on elections. And once this alternative is clear, I'm going to suggest a strategy. Places where we could begin in America to embed sortition like structures, citizen assembly like structures to begin to identify a democratic will that we should increasingly recognize as healthier representations of us than the representations that get produced through Fox News and Facebook. That will be the lifeboat section. Now, as I've told this story of this book to people, most people my age at least, are kind of impatient, maybe bored, because our focus, of course, is the next election. It's always the next election. It's literally every election that's been the most important election in our life over the last 20 years. And so do I imagine these lifeboats being deployed by 2024? No. By 26? No. By 2028? No. By 2030? No. But I do think by 2030, we're going to see a significant chunk of democratic life in America and across relocated into this protected place. And as we do, we're going to become more excited about it, more confident about it, more willing to look at what those processes demonstrate and think to ourselves. Here's us represented in a way all of us should respect. And that's going to put pressure on the clowns. That's going to make it harder for the clowns to behave as they are behaving now. I don't imagine in my lifetime we're going to see a citizen assembly sitting next to the United States House of Representatives or maybe sitting next to the Senate. I don't think that's going to happen. I wouldn't be against it, but I'm not suggesting that. But I am suggesting we'll begin to find solid ground for understanding what we the people think and believe and want. And with that solid grounds, we can begin to fight back against one of the most depressing trends in modern political science. This is a trend described by this extraordinary graph that Pew has published. They've been doing a study since 1998 where they've asked the question, whether we Americans have confidence in our own political judgment. Not our own individually, but our own collectively. Do we Americans have confidence in America's political judgment? And in 1998, two thirds of America said, yes, we have confidence in our own confidence in our own political judgment. Today those numbers are inverted. Today we don't. Two thirds of us don't have confidence in our own political judgment. And why would we? All we see, all that's fed to us, all that gets the attention in both cable TV and the Internet is the ravings of crazies. We talk about issues that have nothing to do with the most important issues that we as a people face, because those are the issues that get people to focus or to turn on or to get angry and share or retweet. Yeah, no, I'm not going to call it reacting. Why would you believe. Why would you have confidence in what we see? And if you don't have confidence in what we see, then why would you continue to have confidence in democracy? Well, my view is if all democracy is, is the pathetic plaything of these emerging and increasingly powerful AIs, then we're toast. If that's all democracy is, is the conclusion of these elections driven by whatever crazies happen to be driving them, then self government is gone. But I am a deep partisan in favor of self government. And I think there's ways to rebuild self government that are protected from these poisonous alien influences. So whether your metaphor is a shield or moving underground, I don't probably babble on about the metaphor throughout the season, but whatever the metaphor is, we need to find a way to relocate democracy to a place where we all could respect it again, not because it gives us everything we want. We're not everybody. But because it gives us reasons to have confidence in the process that finally determines what we together will do. Sometimes that will be what we want, sometimes it won't. But at least it will be a process with integrity, guided for the right reasons by people who are committed to a public good, not AIs committed to maximizing shareholder value. We're maximizing engagement or screwing up American elections because we're Chinese or we are Russians. Okay, that's the arc of this season. And part one will be the hopeful part. Part one will be the part where we identify the overturned tables and show how they can be righted. We'll begin with reform with a conversation with Congressman John Sarbanes about his for the People Act. What is in my mind the most important reform legislation passed by the House of Representatives since the Civil War. We'll then turn to the root of the problem. It's still my view that this is the root money. Looking both at super pacs, speaking with Ron Fine about whether super pacs are really part of our Constitution, and then going deep on one version of public funding vouchers. We'll speak with a hero in the voucher movement, Alan Durning, who was instrumental in getting vouchers passed in Seattle. Then we'll talk to an academic, a favorite academic of mine, studying these practices, Jen Herwig, who will tell us about vouchers and the data about their impact. Then we'll step back a bit and think more carefully about other flaws rendering our representative democracy unrepresentative. We're going to talk to a colleague of mine, Nick Stephanopoulos, about his conception of representativeness. And we'll talk to reformers aiming to fix the primary system. I'm hedging a bit here because I haven't nailed down that interview, but I'm really hopeful today I'm going to get that nailed. But it's really important to understand how primaries could be reformed, both in eliminating party primaries and introducing ranked choice voting into this to see how we could produce through elections, an even more representative system. Then we're going to talk to one of the most knowledgeable souls about the rules of the Senate, Marty Payon, to understand the filibuster. But after I filibuster a bit about how just how absurd this bit of our so called democracy is. Okay, that's part one of the season, the hopeful part. Then we'll turn to the gashed hull and then we'll turn to the lifeboats. Stay tuned. This has been the first episode of season five of the podcast Another way these podcasts are produced by Equal Citizens in the abstract institutional sense, but they are produced quite literally by Josh Elstrow of Elstrow Productions. You can find more about Equal Citizens at Equal Citizens Us and you can give us your thoughts and feedback at that site. We or at least I love the feedback, especially the ideas. I steal ideas or all the time and love you for it. And of course we're also grateful for the support. Though everything I do for Equal Citizens and in the democracy movement is pro bono. We do have a team that needs to earn money for a living, to eat and flourish and be happy. You can donate to help us keep going at =citizens us by clicking the ever present red donate button. Okay, thanks again. Stay tuned. For episode two.