Live from Emory Law: Stacey Abrams joins Alex Bores & Kevin Frazier

So, Leader Abrams, welcome to Scaling Laws. Thank you so much for being here. Thank you for having me. So, let's dive into the current stories that we're hearing about our democracy. If you could give us a sense of the doom, the utopia, and where you think we perhaps are sitting right now at this moment. I cannot regale you with utopia. I can do doom and less doom. So, I'll tell you what my posture is. And Kevin, thank you so much for having me. And it's always great to be with you, Assemblymember. The way I think about the current status of democracy in America, and we can widen out and look at it internationally, we are in the midst of a decline in democracy, but it is not a straight shot. We have fewer active democracies today than we had 20 years ago. We are as a nation in the midst of what's referred to as competitive authoritarianism. And Though I don't have to translate it for you all, I'll tell you what I tell lay people when I'm talking about it. For 250 years as a nation, we were all headed in the same direction towards a more perfect union. We just disagreed about how we were going to get there. And in modern parlance, Republicans were using Google Maps and Democrats were using Apple Maps and independent days and crashing into people's houses, all heading in the same direction. For the last, I would say, say 20 years, we've had a hitchhiker. And that hitchhiker has been willing to use democratic institutions, accomplishment of authoritarian ends. And we have seen that with the erosion of human and civil rights. We've seen that with restrictions on access. We've seen that with the rollback of corrective actions that have been taken. And you can insert into that post Watergate decisions that were made to make politics less susceptible to money. You can see that in the expansion of human civil rights, whether you're talking about the Americans with Disabilities Act, the Civil Rights Act, the Respect for Marriage Act. We saw the phalanx of expansion of access to rights that have been very assiduously rolled back over the last 40 years using the mechanisms of democracy. So using the executive power, legislative power, and the judiciary. And then last year we saw. Hitchhiking turned into hijacking. And that became the very aggressive and explicit decision to. We have seen a complicity in the congressional authority. And again, this isn't. And when you have justice of the Supreme Court, he is not willing to. ...or compel the enforcement of certain laws because the president won't follow the directive, you also start to see an impotence in the courts. That's competitive authoritarianism. Now, Lucan Way and Steven Levitsky write about it much more authoritatively than I do, but the reason it's an important frame is to get to your point about doom and gloom. There are 10 steps that every single authoritarian in modern times have followed. And what has happened in the U.S. happened under Bolsonaro. It happened under Orbán. It happened under Putin. I remember days when Russia was a democracy. It happened under Modi. It happened under Erdogan. It always happens in the same way, maybe not in the same order, but it happens. But if there are 10 steps to authoritarianism and autocracy, there are 10 steps to freedom and power, because we've also seen that work. We saw the students overthrow Milosevic in Serbia. We saw Nepal. We saw South Korea or Korea when Korea was briefly. Under martial law, so we know that these things can be reversed. So if there is any, there's no utopia, but there is possibility. And so we have to accept that we are in a darker time than we should be. But there is light and our responsibility is to figure out how to navigate our way back to the light. Leader Abrams, let's stay on the 10 steps to freedom and power. What does that path look like? And for someone who wants to join that fight, what's the best way to tune in? You should go to 10stepscampaign.org, but I'll tell you what you'll find there. So one, we lay out the 10 steps to autocracy and authoritarianism because it's different. In a modern, sophisticated, well-financed nation state, it looks different than it does in a smaller nation, a more homogenous racial country. There are a lot of analogs that people try to draw between Hungary and the U.S. Hungary has 10 million people. That's basically traffic on I-85. So we can take lessons from what Peter Magyar was able to do in Orban, but there is not a direction. Not India. India has been a democracy for a very long time, but India faces an entrenched level of... They were not always a democracy, and so the evolution of their identity is different than America, whose, our core identity is that we were born, had sort of a pre-West existed, this is who we are. And so we have a different salience to how we think about it. And so the 10 Steps layout, and it's based on research by Kim Shepel and others. Erica Chenoweth, it lays out, here are the 10 things that you always see in the fall of democracy. And it's the expansion of executive authority. It's the complicity of the competing powers, the judiciary and congressional members. It's the attack on the truth. It's breaking government. It's the fact that in America, we're afraid of lettuce right now. That's authoritarian because when you break the systems that should protect you, people stop fighting for a system that doesn't seem to serve them. The more aggressive pieces are step seven, where you essentially scapegoat vulnerable communities. And this is true across the board. You cannot find a nation state that falls into authoritarianism that doesn't blame somebody who can't fight back. In the United States, we are really efficient, so we just call it DEI. And then step nine, the normalization of state violence. We sit in a nation right now where our capital city has been under military occupation for more than a year, and we don't talk about it. We have a body count of American citizens murdered by masked armed paramilitary, and it is no longer a conversation piece. And so those are the 10 steps to, you know, autocracy and authoritarianism. And step 10 is that you end democracy forever. And that's called the SAVE Act, but we'll talk about that later. On the other side, though, are the 10 steps to freedom and power. And it starts with committing. It's understanding where authoritarianism gets its power, because if you don't understand it, you can't fight it. Step two is sharing. You've got to tell other people. You can't simply hold it to yourself. Three and four is mobilizing and organizing. It's finding the ways that we saw in Minneapolis and in Memphis and in L.A. and in Detroit. And recently in Danbury, Connecticut, it's when communities organized to fight back against the very pointed realities of authoritarianism. Then it's litigation, which I don't know if any of you are familiar with that term, but it means... So litigation is one of the central pillars of fighting authoritarianism, using the courts, using the law. Then it is being disruptive. It is denying the propaganda. It is engaging the existing political edifice. It's electing new people and then it's demanding the democracy you deserve. And this is a very nice way Alex has offered me to tell you that in order to do that. We've created a website, we have created a BuzzFeed quiz that tells you what kind of democracy defender you are, but we also are in the process of developing an AI tool, a civic coach that we currently call Marion. I don't like the name, but I'm arguing with people about it, including AI. Thank But the whole idea is that we've got to meet people where they are, and most people aren't used to being freedom fighters. And so the whole idea of Marion is to create a bespoke. civic map for you, a coach that can grow and evolve with you, help you understand how you enter the steps you fit, maybe your three steps, you. maybe your five steps, maybe your half of a step, and then help you map your place forward. And we're using AI as an undergrounding tool to get that done. I love you talking about arguing with AI. One thing you said there about how the U.S. has a self-conception from our founding of being a democracy, despite some very anti-democratic sins at our starting and clauses in the Constitution, some of which have been eliminated, some we still live with today. Does that self-conception of us being a democracy? Make it easier to slide into authoritarianism? Make it harder to recognize what is happening? Is there a denial aspect? I wonder how you think about that. I think that's accurate. There is a sense of denial because we have always been thus. Therefore, any argument against us being what we think we are is hard to process. It's the normalization bias. You get used to things being as they are. You see yourself in a certain way, and then you see a photo that someone took of you on Instagram. You're like, I don't look like that. Well, that's democracy for us. And each time we've been forced to confront the image of ourself that others have seen. I mean, let's think about the rise of authoritarianism in Germany in the 1930s and 40s was premised on post-reconstruction laws in the U.S. They learned from America's response to the Confederacy and our response to fixing the Confederacy. That's how they built the Third Reich. So we, and when we confront that, we recoil because that confrontation either requires that you take corrective action or that you accept that you are not as noble as you thought. And our self-identity does not allow for a diminution of our sense of nobility too often. When we do well, and so I want to give us, Your question, Kevin, I want us to give ourselves credit for the corrective actions we do take. I just need us to recognize that it's an active decision that we have to make and you can't decide to fix something if you don't acknowledge that it is broken. That's where I think your work, Leader Abrams, on helping more people realize their agency is so important. You, in an interview, somebody asked, if you could go back in time and tell voters one thing while you were running, I believe it was along the lines of, you matter, and your vote really does count. And that claim of agency, I think, is not only relevant to conversations about democracy, but also to conversations with respect to AI, Where in both of those domains, there's a sense of... Whatever I do, democracy is going to do whatever it's going to do. Whatever I do, Sam Altman, Dario Amodei, they're going to do whatever they're going to do. So how do you see a through line of this need for seizing more agency in these domains? And Alex and I, I promise we are not being paid by anyone. But I also want to recognize in your great book, Code of Justice. You identify a need for folks to realize their responsibility to act, even when they may not have the authority to act. And for folks in this room who are wildly smart, who have great futures ahead of them, thanks to professors like Matthew Sag. I don't know if Matthew's in the room, but shout out. There he is. Okay. Thanks to folks like that, inspiring people to get involved and to have relevant skills. How do you see that bridging both the democracy domain and the AI domain? First of all, you have to call me Stacey, please. All right, fine. I appreciate the title, but I'm good. Good. So you mentioned Avery Keene. So the heroine in my three legal thrillers, legal slash political thrillers, her name is Avery Keene. She starts out as a clerk to a Supreme Court justice. And she is given this. responsibility in the first novel where and it's the through line for the stories where she's given these grave responsibilities but no authority to accomplish it and so we follow her as she tries to solve problems that she's not permitted or resourced to solve which is what we call you know Thursday and the challenge that I think for democracy We treat democracy as this heirloom or this artifact, this thing to be admired and polished, this construct to be investigated, but we forget that democracy is an action. It is the, and for ordinary people, it is the act of being in community and in democracy. In authoritarian parlance, it's about power being taken from the people, it is hoarding that power to a very small group of people, and it's never being held accountable for what you do with that power, which is corruption and voter suppression and all of the things that stop people from using that power. In democracy, it's sharing that power, it's leveraging that power for the benefit of others, and it is being held accountable for what you do with that power. When you bring that conversation back to the active engagement in democracy, we keep waiting for democracy to fix itself. Democracy has no agency. We do. And we decide how much we're willing to concede or give. How much is... Sometimes we don't want the negotiation that comes with it. AI is the same way. You said you argue with AI and sometimes it wins. I will ask, and I've got two different LLMs that I use because one is nice to me on like alternate days. I've programmed to give me sort of somebody who hates me, someone who's like deeply disrespectful of me but pretends, someone who likes me, and then someone who's just about the world as I am. And I'll ask it questions, not because I need it to think for me, but because I need to test myself against it. And then there's Claude. And Claude just doesn't like me. And it's okay. But the point of both of those conversations is that I have to decide what I'm going to do with the information. AI is not designed to solve my problems. It's designed to inform my decisions about how to solve my problems. And democracy is the same. Democracy can't do a thing without us. It is up to us to decide how much are we going to do with the information that we have about the situation that we're in. And for ordinary people, that's a question of How high does my rent go? And do I have access to health care? And those feel like very parochial questions until you look at the systems that decided whether or not you get to even have that debate. And so it's a very long and winding way to come back to the fact that whether it's democracy or AI, agency is something that you have to seize. It is never something that is given. And we often will. cede to either the abstract of democracy or the construct of AI, that because it has information we don't possess, that it has the ability to make decisions for us, or that we are powerless in its face. And ultimately, all of these are power conversations. How much power do we give AI? How much power do we give elected officials? How much power do we give systems? And when we decide That power belongs to us. The debate changes, the navigation changes, and the use of that information changes. Let's double down on that point on our agency, especially around AI. There are some issues, you mentioned the rent, where people feel very comfortable expressing their opinion, getting involved, thinking of that as politics. And then there are some issues where you have this claim to authority that's used to silence people. You see that discussing financial regulations. You see that discussing even the gun debate sometimes. Well, if you can't name a specific magazine, you know. Yeah. We're seeing this more and more in AI. Obviously, you're highly credentialed. You have a law degree from Yale and a public affairs degree from Texas. Kevin made me do that. I did. That was in the show notes. Alex gets to stay on as co-host, which is very exciting. But obviously... People are going to turn to you and think your voice matters in this conversation. How do we invite everyone to see these issues the way they see their rent or other things, as this is a values conversation that everyone needs to be a part of? I think it goes back to storytelling. And one of the reasons I write the novels that I write, I want people to feel a sense of agency about topics that should not be accessible. So Coded Justice is about AI, DEI, and veterans health care. And the premise is that a tech bro creates a very well-intentioned one. He was a combat veteran who creates an AI system, an LLM with a God complex, and a chatbot that is highly overconfident that may or may not be a murderer. And so Avery, who is helping them deal with, navigate these financial processes, has to figure out what's going on. But I. I wrote it because when you finish reading it, you should understand the lingo of AI more. You should be able to watch something. You should be able to read the 3,800-word screed by Dario Amodei and feel like you understand what he said and what he didn't say. You should be able to listen to Sam Altman and demystify. You should be able to navigate. You should be able to listen to Alex Boras when he is explaining why data centers have to be managed and controlled and not feel that you're outmatched because part of the power of politicians is the ability to convince you you're not smart enough to do this on your own. And so I write novels to tell you, yeah, you're smart enough. I write novels, the predecessor novel to Coded Justice was Rogue Justice because the failure of the energy grid in Texas overwhelmed people and they didn't understand why it happened. So I wrote a novel about, that one was cryptocurrency, the FISA court and Farrah's Doctrine. And, and, uh, The energy grid, but when you finish, you now know about the Ferris Doctrine and how we treat our veterans, or how we treat our military. You know that the energy grid is basically held together by duct tape and prayer, and that the FISA court is way too powerful without us having insight, but because when there was a FISA justice who emailed me and said, I knew what that was. I didn't know about it the last time it happened, and so... Storytelling is how people gain a sense of agency. When we are afraid that we don't know enough, the best people, the nicest people shut down because we don't want to make mistakes. We don't want to cause harm. And in the absence of causing harm, for many people, the solution set is stillness. It's silence. Or worse, it is assuming that the person who's louder is smarter. And if we've learned nothing else from the last 10 years, that is not true. And that's that's not it's not about political party. It is about. Power. Power often presumes your complacency by overwhelming you with information. And so I try to tell stories that give people a sense of agency. I make sure people who know more than I do read my books and tell me, like, I want someone who's an expert to read it and think, oh, she's smart. And someone who isn't an expert be like, oh, she's smart. But if I've done my job, both things are true. But I know I've done my job when someone else finishes the book and thinks, oh, I'm smarter on this issue. And when we invite people into storytelling, when we invite people into putting themselves in situations, like I wrote a book about AI that was about whether or not a vet got health care, a woman who was homeless because she faced sexual assault, and this AI model knew about her history and was able to help think about the kind of care she would need. That's a good use of AI. But it was complicated by who was in charge of the model and what the model understood. And if the model hadn't been trained on algorithmic bias and DEI, then that woman would have been overlooked and her care wouldn't have been permitted. But if I came up to you and said, let me tell you about that, you'd glaze over. If you're reading it while someone's being murdered, you pay attention. And so storytelling is one of the ways we create that space for people to... their proxy and their role. And it's so important, too, because at this moment, if you go to the Bay Area, which, unfortunately, when you're in the AI space, you find yourself sucked into San Francisco way too frequently. They have a saying there, I'm sure some of you have heard about this, the need to touch grass. These folks are so online all the time that they have to remind each other, hey, maybe get out of your apartment, get out of the Waymo, go touch grass. I think a way better argument, and to his credit, Jack Clark, one of the co-founders of Anthropic, talks about this a lot, the need to actually go talk to people about how they're experiencing AI. And what I love about this emphasis from you, Stacey, is you don't have to become a computer science. I'm a master's student. You don't have to go get that PhD. You don't have to know what back propagation is. But even having some degree of literacy about knowing what questions to ask is so empowering and gives you a chance to experience the positive sides of AI, as well as to recognize what those negatives may be. So for a tool like Marion. How do you see use of an AI civic engagement tool just generally fitting into getting more people involved? Because on the one hand, admittedly, sometimes I get a little skeptical. I voted on Saturday. It was a heck of a lot of fun. Somebody right next to me, I saw them on their phone and a thought popped into my head. Are they using Claude to ask how to vote? Would that be a good thing or a bad thing? I don't know. Would that be cognitive de-skilling or democracy de-skilling or is that something that helps them be a more informed voter? How do you weigh that tension of losing that sort of formation of our civic muscles while also recognizing that AI may be the thing that pulls people into the conversation? And it's navigating exactly that tension. So the way Marion works, It starts with this sort of BuzzFeed quiz. It asks you... It has you answer questions about what things are worrying you. Are you someone who likes to be up front or behind? Are you Moana? We don't do that because it's from Disney. But it kind of tests out where do you see yourself. So it asks you to self-identify your proclivity to just the way the world works. Then the next set of questions are, okay. If you were going to do something, do you want people to know you did it, or do you want to be behind the scenes? Do you want to talk to people, or do you want to email them? Do you want to have patience? And so we ask you those questions, start to bring you in. The next thing is that we then give you a set of options. Like, this is the kind, you could be this kind of person, or this kind of person. It's kind of a version of, you know, your Myers-Briggs. It tells you, like, you're a very rare kind of option you want to do. But then again, Once you select, it then says, okay, so here's something to try. Before you commit, we're going to give you a few options for what your civic engagement could look like, and then come back and ask questions. So if it says you should call your congressman or your state legislator, do you need a script for how to do that? You have to go and talk to the person. Yes, that's your human contact, but there's a helper that can walk you through what you say. Ultimately, I have this Pokemon Go idea. And we're working with a bunch of young people who are younger than you. Y'all are young people, too. And the Pokemon Go for me is it's based on this idea of geocaching, the scavenger hunt that Zoramandami did when he was running, and just fun that I had when I was younger, which is what if we set up civic side quests so that across the country you are out doing these things? You are finding... A VA hospital, you're going to a food bank, you are attending a zoning committee meeting, but you've got to get certain things done while you're there and you have to post them. And that to me is the integration of the technology as a tool and a wayfinder, but the humanity of actually touching grass and touching people and the overarching mission for me, which is there are a lot of people running our lives that we don't know about who are not I'm not smart enough to be making all these choices for us. I've met them. And it is sometimes infuriating and sometimes deeply terrifying. Making decisions about AI cannot spell AI. And that does not disturb enough of us right now. And part of my intention with the 10 Steps campaign and all the work that I do. Technology to treatment, but never exempt from the responsibility of actually being connected to each other. I love the idea of using technology to get people to touch grass, right? Of sort of inverting that, of a Pokemon Go style way of reaching out. You've talked about how we sometimes overcomplicate organizing, overthink what it means to organize. Yeah. How do we think about the tension of... Building new tools, using new tools to do a very old thing of organizing. Thank When are we overcomplicating it? When is it enabling us? How should people think of that trade-off? I start with what are you, what's the pain that you're feeling? you What's the problem that you're solving? And often we will complicate the solution very to diminish how much work we actually have to do. much. You may have heard of, you know, analysis paralysis. So we will create complications because it helps forestall moment of action. And so when I think about this, I think about what's the smallest unit of measure for solving the problem. And that's where you begin. I'm hungry. Get a cracker. You know, unless, of course, the cracker is in a locked box and the locked box is on a shelf and the shelf is in a building and I don't have the keys in the building and I need a code and I need a helicopter. Smallest unit of measure and then expand from there where technology comes in is helping us understand these solutions, not thinking about being hungry. No. So, OK, what else is out there? What are the alternatives? So technology is a great generator of alternative. solution sets. It can never be the decision maker, but it can help be an ideator. And that's how I use it. I love Star Trek. And when I use AI, I think about it, if any of you are Trekkies, apparently none of you are, and I don't know what's wrong with any of you. Let the record reflect, there were no Trekkies. Seriously, come on. We've got an objection. Thank you. Thank you. Okay, so as a Trekkie, it's like being on the holodeck, But without the, I'm sure that, you know. Sorry, I was going to do that soon. But you're able to ask the computer questions. Geordi never let the computer solve the problem. And I know I've just taken half. You're like, who in the hell is Geordi? So Geordi La Forge, you don't care. Anyway, so Geordi never let the computer solve the problems that will face the Enterprise. But Geordi never failed to ask the computer to give him ...about what else he needs to know. So in organizing the tools that you can use in any solution set, you always want more information that can be useful. You also want information that draws from a part of your thinking that you haven't imagined. I refer to myself sometimes as the Costco of democracy. I try lots of different things. I pull in from very different spaces. I write romance novels about ethnobotanists because I wanted to explore the issue of how we were deforesting in South America. Now, most people don't pick up romance novels to think, oh, I want to read about ethnobotany and deforestation. But for me, that was a way to connect dots. Technology helps us do that and at a speed and with an intentionality and a creativity that not everyone natively has, but once you have it, it sparks in your mind the ability to do that in other ways and across other issues. And when it comes back to organizing, organizing should always be human in its intention and its output, but technology can speed it up, layers to it, and I think amplify its effect. And to take things all the way back a little bit further to thinking about the fact that we're here at Emory Law, we've got these great law school students who are thinking, this is so inspiring. I am so jazzed. I can't wait to go act on one, if not all of these 10 steps. I fear that for a lot of law students, and now I'm reflecting on your own experience of getting a JD from Yale, that small school in New Haven, I guess we'll say. It's no Texas, but- You don't want me to tell you why I didn't go to Texas, but- Oh, geez. Yeah, yeah. Let's leave that for the after show. So for the folks who are thinking, what is it I can actually do? Because I'm concerned that for a lot of law students, This is a moment where... You know, the economic conditions aren't looking super promising. You're hearing from some of your professors that AI is going to take your job. You're reading headlines that say, who knows what's going to happen with law firms. And the safest decision looks like, huh, you know, I got this sweet gig at this major law firm. I'm just going to go do that. It seems like a good spot to go and ride the wave. But I think we have to dare lawyers to be social engineers. We need new institutions. We need to redevelop and reinforce some of those institutions in light of all of the pressures that they're facing. What's your message to these soon-to-be newly formed lawyers about how to engage and seize this moment and exercise some of that agency we were talking about? The person who spearheads the 10 Steps campaign is a lawyer. He actually started working for me right out of law school when I was Minority leader of the state legislature. He, at the time, was the best lawyer I could afford, but it turns out he was really smart, so I made a good choice. And he had to do that. We had no resources. In the statehouse, when I was minority leader, Officially, my budget was $13,000 and the majority leader's budget was $1.3 million. I was able to claw out of the speaker an additional $75,000. It took six years to get myself to that level. But I had to get a lot done for my caucus with very few resources. And I hired Robert Sills as a lawyer to work with me because I needed someone to help with legislation. So that's one piece. Robert had, you know, I think Robert had been out of law school for maybe two, three years, but I gave him this platform and said, please help me solve this problem. You are uniquely qualified to solve problems with limited resources because you know how to find answers. And often we focus on the law as a degree and it is a skill set. You are uniquely positioned to find information. One of the challenges we had at the Capitol was that as a minority leader, I would often come up with ideas for our caucus to offer as legislation, and Alex can speak to this, only he's never been in the minority. We were in the minority, which meant every idea I would give to our shared legislative council, somehow, sometimes those ideas made it to the majority before I got my legislation. So I needed Robert to help me write this legislation before we got it. Basically, it's like you have to give your playbook to the other team, and they would give it to the referee. And so we then hired law students from across Emory, Georgia State, and John Marshall, and we brought them in to be our, we had law students who were helping write legislation because we couldn't afford to pay for other lawyers. But some of the best legislation that came out of the state started with law students who did not know that they weren't supposed to think about things a certain way. You are still... Protean enough in your thinking because you are remembering all the things you had to cram in last month that you connect dots and you imagine ideas in ways that those in those out of the third thing that happened was that once I left and ran for governor. Robert went off. He worked for Amazon. He did some other jobs. We reconnected a couple of years ago, and when I created the 10 Steps campaign, I brought him in because I needed a way to organize all of these ideas, and the whole point of the 10 Steps campaign, we're a wayfinder. We pull in organizations from across the country. We've got more than 120 organizations. More than 250,000 people have signed up with us, and I needed someone who could help me think about How to keep it on the rails, but also what else was out there. You were incredibly creative. You have to solve new problems. Every single creative capacity of lawyers gets trained out of you, but that if you hold to it, you become this information source for groups that don't know they need a lawyer. And then the last piece, as someone who's a writer, I'm the writer of interviewing. John Grisham for his new book, and we got to talk about how he uses his law degree to inform issues like exoneration and the death penalty. And I pushed him on it because he said, well, I don't want people to take my books seriously. And I said, things come behind and everyone thinks that Atticus Finch, like that's a good thing. And I think to your point, Kevin, in this moment of crisis, You are smart enough to know that there are problems. You're smart enough to find answers. You can go work for those law firms because a lot of those law firms need the courage of their conviction. You can go and work for organizations that believe that we've already lost and you can remind them that there are still things to be won. You can work for politicians or become one yourself because part of the solution set is being able to, it's the crack method, having a conclusion. Creating the rationale, doing the analysis, and then coming back to that conclusion. And you are trained to do that. I'm excited about being here because I need to know you exist because I need a nap. And your ability to go into the world, especially in a moment where we don't know what AI means. We don't know what it can be, but we do know what humanity is. We do know what people have faced, and I think, Alex, you referenced this. This is not new. Nothing is new. So what do we know from what we have done? What do we know from the Magna Carta that should be imported into any AI-controlling legislation? How do we help school boards think about what's happening to the data of seven-year-olds? You don't have to wait for Congress to take action, and you all are uniquely positioned as native technologists to think about laws in ways that those who've been thinking about this forever can't imagine, but also there's an urgency for you because you've got to live with whatever you come up with. And so I think law students in particular are positioned to get on it. Wow. And I'm also fairly convinced, but I think we have some excellent questions from this esteemed student body. And we have Matt and Nadia who can help us out with asking some of these questions. So we'll now move to the Q&A and take turns asking questions students submitted in advance. So our first question is, are there decisions that should remain exclusively human, even if AI could produce equally accurate outcomes, especially in sentencing and judicial decision-making. Yes. So in Code at Justice, I think about this, and it's one of the tensions between the characters. They are really good friends, but they fight about this. There's an objectivity that AI can mimic. AI is not actually objective, but it can mimic objectivity sufficiently to reveal our biases. And so I do think there's a role for technology to play in, but humanity, understanding of context and the nuance of effect. It's why Les Mis is still such a visceral novel for so many people. The story matters. And so AI can never replace the human understanding, but it should be used to point out where human bias has actually created more harm. And that's the other part of the code of justice narrative, which is even the best intentioned technology is going to be infected by humanity's tendency to be selfish and vain and mean and cruel and vindictive. And so sometimes technology can interfere with that. Thank you. So our next question is, could AI meaningfully help close the access to justice gap by making legal assistance more affordable, or might it create new inequalities between those who can afford them and those who cannot? Yes. So. So let me give an example. One of the organizations I created in 2023 is called American Pride Rises. It looks at DEI because I believe DEI is the central pillar of democracy. You cannot have a pluralistic democracy without diversity, equity, and inclusion. But in the U.S., DEI is actually our shorthand for all of the corrective actions we've taken since 1776. So it's the 13th, 14th, and 15th Amendments, but it's also the Americans with Disabilities Act. It's Title I education. It is the Fair Labor Standards Act. It's the Immigration and Nationalization Act. All of those things are considered DEI. We don't usually define it that way, but it is. And when this administration started outlawing DEI, prohibiting the use of DEI, people often either happily or unhappily thought, oh, they're talking about people of color, they're talking about women. No, they were talking about any corrective action, and we know that because one of the hardest hit populations with the DEI prohibitions were disabled veterans. Thousands lost their jobs, lost their loans, cannot get access to resources that used to be available because they were DEI. We also saw organizations that serve them become afraid to do the work. And what we created through American Pride Rises, if you go to aprnetwork.org, we created a chatbot called Adiva. And Adiva actually answers legal questions. We are not lawyers. It's not a lawyer. But it helps demystify and explain what the law actually says based on we created our own LLM and we trained it only on trusted sources, but it serves as an intercessor for organizations to hire a lawyer because a lot of lawyers were afraid to answer the question about what was the tool Adiva helped. Thousands of smaller organizations find answers because the speed of DEI was crippling their ability to do their work. That said, Adiva is not the answer. And what we need are lawyers who are better trained and willing to actually do the moral calculus of understanding what the law says and do risk assessment that doesn't include, will I lose my client? So we do need, I think technology can be a leveler. But it always needs an intercessor. And yes, anytime you have that capacity, the amount of data you can afford, the amount of time you can afford, the questions you can afford to ask will always create inequality and inequity. And so one of our responsibilities is not to not do it because it could lead to inequity. It is how do you solve for those inequities? How do you anticipate them? Because they always, always, always happen. Just to add one brief point, not to bring up any PTSD from the folks who just took the MPRE or are studying for it. We as lawyers have an obligation to improve the quality of justice and to defend the rule of law. And that doesn't mean defending status quo, and that doesn't mean defending our guild. Right. That means actually looking at to what extent can people access the court system and meaningfully leverage new tools rather than saying, oh, that's the unauthorized practice of law. Sorry, you can't use that tool. Alex, you look like you were going to say something. No, I just was reminded how outnumbered I am as the non-lawyer in the room. I was like... We've got time. Deep, deep. DEI applies to you, too. Thank you. Thank you. We'll welcome you. So our next question is, looking ahead five years, what would successful AI governance in the United States look like and what must happen now to get us there? One, we actually have laws. That would be a start. Thank We are woefully behind every nation state that has considered the advance of AI. you so Even China has created. Now, I don't agree necessarily with all of their decisions, but we are one of the few. much Developed nations hasn't grappled with what it should mean. for Instead, we've spent all of our time fighting over your who doesn't get to have a say. I wrote a piece on Substack. time. About the fact that when everyone was lauding Dario for his 3800 word screed about how AI was going to kill us all and how it should slow down, he said slow down the development of the AI, not the development of data centers. He didn't say speed up the oversight that could come from state and local governments, because those are the only entities that are currently regulating AI at scale. No pun intended. And so the... The reality is even those who can point out the harms are unwilling to entertain the solution sets. So five years from now, we will have legislation at the federal, state, and local level. It would be bespoke based on the characterization, the character of the affected parties. Children should have protections at the, we have given them. Chromebooks, and we have no rules and regulations over the scraping of their data and the tracking of their futures. That should be dealt with. Number two, we should have a universal international set of standards about humanity. I would like for us to agree what a human is. That would be nice. We've been grappling with that since the 1940s. We still haven't gotten there, but hopefully in five years we'll get there because our overlords will make us do so. So, um, That was a joke and none of you laughed. Oh dear. But three, we would have public AI, because one of the challenges is that when a handful of oligarchs control the means of production, they will control how those products are used. And AI is supplanting other forms of information, and we have seen what happens when the aggregation of information into the hands of those who don't like some people, when that happens, it is a harm that affects everyone. And then lastly, it would be money, it would compensate humans for how we built it. I have eight books that were scraped. I'm in the anthropic settlement, so I have eight books that they've got to pay me back for. That's just the stuff I know about. Your data builds this. You should benefit from it. There's every single person who has produced information. And that is a doable thing. It is a challenge that we pretend is too big. We know how to solve it. Our problem is never that we, we don't have a money problem in America. We don't have a money problem in this world. We have a morals problem. And an AI future should solve for that dynamic and should create economic opportunity for the least of these because not a person on this planet is immune to the effect or was not participatory in the creation. You mentioned having good regulations at the federal, state, local, and international levels, right, that these all work together, and I think that's an important point to underscore in this debate, where I love how you said there's been more debate about who's not allowed to have a voice than there is on what we should actually do. There's been this... The combination of two different concepts, one being, well, the federal government will preempt the states when it conflicts with states, which is natural and should be. And as the author of the RAISE Act, I would happily trade it for a federal rule versus what I call an anti-democracy clause, which is the states cannot pass any new or cannot enforce any new laws in this space, which is really pushing back on who gets a say. And I think keeping those separate. And over the next five years, making sure that all levels are playing their role is the only chance we have to get this right. And I would just add the one very small thing, too. Let's see the labs practice what they preach, right? So there's a lot of good language in that 38,000 words creed, a lot of great blog posts by OpenAI about what they're going to do, who they're going to invite in. Go for it, right? Hey, there are professors who are ready. Civil society is ready. Let's see you all actually hold yourselves accountable. They say they want to. They are free to do so. Although I will point out one little nuance, and you may have noticed this. At the same time that came out, they also quietly went to Congress to ask to be exempted from the Sherman Antitrust Act. So I think this goes to both the need for international, but also a question of why do you need to be exempted from a law that the one law we do have that's putting any type of governor on your behavior, you didn't put that in your screed. That one you just did quietly. You sent Sam to go and talk to folks while Dario wrote this. Although it's nice to see cooperation amongst competitors, I guess. Yes. We have one final question. How can universities, law schools, and students better prepare themselves for an increasingly AI-driven world, particularly as emerging technologies are shaping the legal profession and society as a whole? One, it's learning about it. A project that I lead down in South Georgia, we are training an entire town that only has about 150 people in it. Just training all of them on AI. Their county is likely going to be a target of a data center. And I want them to know about the technology that is likely going to bankrupt their community. And for me, it's important because this is a low-income community. It is 60% African-American, 30% white, 10% Latino. It is largely agricultural or service workers. They are never, their jobs are the ones that will be at risk. Their community will be at risk. Their water will be at risk. Their power bills will be at risk. And I want them not to be afraid of this technology. And so our first job as students of all kinds, and I use student not in the tuition paying sense, but in the learning sense, we should understand what we face. You don't have to get a degree. I read a lot of books. I did a short course on MIT's website to teach myself about AI so I could be conversant in it. Do not do that unless you really want to. I mean, I learned a lot and I'm a little afraid of things, But we should... One, we should know, we should have a passing familiarity with the technology and with how the technology works. We should, two, practice our knowledge. We don't have to like what we get from it, but hiding from it is not going to help us, especially as lawyers. If we do not know the enemy, if we do not know the challenge, if we do not know the opposition, we cannot build a sufficient defense. We cannot negotiate a settlement in an agreement. When I was minority leader, my job, I used to say, they call you leader to make you feel good. They call you minority so you never forget your place. My job was cooperation. I could not get a thing done legislatively without the other side's agreement. It was mathematically impossible. My predecessors sometimes used that as a reason to just not talk to anybody. I saw it as a reason to understand the other side. I got to know them. I went on, I spent a week on an agricultural tour in north and south Georgia. I know how to set a cow now, and I know how to do center pivot irrigation. Even though that's not a necessary indicator, I learned how to do it because I need to understand it if I'm going to ever make decisions about it. We need students to do that. But we also then need to integrate it where it makes sense to make us better at winning the future. And then we need to be so conversant that we are then the teachers of those who don't know they deserve to be students. And more than anything, and this comes back to democracy and civics and social justice, You are privileged. You are in this space because you enjoy an extraordinary privilege of intellect and resource. There are millions who will never imagine these spaces, but will be the victims of these choices. And I need us to be prepared to be their knights, to be their intercessors. And that's what you can do uniquely. If you know it, you understand it, and you can then help defend, they can use it too. Well, that is one heck of a story to end on. And thank you so much, Stacey, for being here. Thank you again to Matt and Nadia, to Dean. Thank you guys for helping. And of course, to my co-host, Alex Borges. And thank you to all of you. This was fantastic. Thank you, guys.