Scaling Laws: Matt Abrams on Trust and the Migration to the Digital Age

So, besides the end of the song and the new "Bundestliga" song, as far as "Türstet" is concerned,

“"Mutigia, I am mental," and "Sport Director".”

Because all the games are played, then it's even more like the kick-based side. And if you say kick-based, what does that mean? The kick-based is the fourth fantasy "Fußball Manager".

But that's the first one.

It's a really simple one. You are surrounded by an opponent. You can easily read your own rules. Or you can read the real "Bundestliga" profile in your card. So, the time you show your "Fußball Wissen" and "Bundestliga" is the first time you know.

The counter-office of the "Bundestliga" is the end of August. The "Bundestliga" is the third time you know. And the second time you know, the kick-based side, the kick-based side, just download and release. Good kick and fun.

It's the law fair podcast. I'm Alan Rosenstein, associate professor of law at the University of Minnesota, the senior editor and research director at Law Fair. Today we're bringing you something a little different. An episode from our new podcast series, "Scalic Laws".

It's a creation of law fair and the University of Texas School of Law,

where we're tackling the most important AI and policy questions.

From new legislation on Capitol Hill to the latest breakthroughs that are happening in the labs. We cut through the hype to get you up to speed on the rules, standards, and ideas shaping the future of this pivotal technology. If you enjoy this episode, you can find and subscribe to "Scalic Laws" wherever you get your podcasts and follow us on "X and Blue Sky".

Thanks for listening.

“When the AI overlords take over, what are you most excited about?”

It's not crazy, it's just smart. And just this year, in the first six months, there have been something like a thousand laws. Who's actually building the scaffolding around how it's going to work, how every day folks are going to use it? AI only works if society lets a work.

There are so many questions have to be figured out, and nobody came to my bonus class. Let's enforce the rules of the road. Welcome back to "Scalic Laws" the podcast brought to you by Laugh Fair and the University of Texas School of Law that explores the intersection of AI, policy, and of course the law.

Kevin Fraser, the AI Innovation and Law Fellow at Texas Law, and a senior editor at Laugh Fair. Today we're joined by Matt Abrams, Matt is a technologist, investor, and co-founder of

“one billion for trust, and he's focused on one of the defining governance questions”

of the AI era, how to build and maintain trust. After decades working across enterprise technology, AI, healthcare, and venture capital, that is now helping lead a new initiative that's aimed at building trust and information institutions and in one another. The timing of this conversation is hard to ignore, as I recount at the outset of this episode

public trust in nearly every major American institution has fallen sharply. So what does this mean for the rest of society and what does this mean for us as stakeholders in AI policy and governance? It's a big question and we try to provide some answers. To get in touch with us, email scaling laws at lawfairmedia.org, or follow us on X or Blue Sky.

And with that, get it up for a great show. Matt, welcome to Scaling Laws. It's great to be here, Kevin. So I want to start off, unfortunately, on a bit of a depressing note, so I've got a few

stats, unfortunately, that we have to address first of adults under 30.

61% say that AI is going to worsen people's ability to think creatively. 58% say it's going to erode their capacity to form meaningful relationships. And just 14% of adults under 30 expect AI's effects to be positive on society. Now if you thought that was bad enough, I'm going to keep going. More generally, only 15% of Americans say they trust AI companies to make decisions about

how AI is developed and used and at an even higher level of abstraction. Just 17% of Americans now say they trust the government, Washington, to do what is right. And the final depressing quota to all of this. Only 34% of Americans say that most people can be trusted.

And you and I hit it off at the ASP workshops sponsored by Fathom.

“And I thought, Matt, I've got to have you on because I love institutions.”

I love a functioning society. I think we all love those things. And yet it seems as though we can't acquire it. So who are you and why are you so focused on trust as a policy inquiry and as a broader inquiry for us all to tackle?

Kevin, I hope that you are as well still a naive optimist as I am. So despite all those stats, I call myself the naive optimist because I don't think I could bear with the opposite and there's not much point.

And I don't want to be a cynic.

So with all those stats you threw out there, Matt Abrams, I'm out here in Central Oregon. And I would say the reason why I'm so focused around these arenas is it goes back to where I grew up in deep southern Oregon. And the lumber town, USA, it was called in the 80s and it was the height of the spotted owl debate.

And we're the epicenter of the spotted owl debate. And I grew up to a father who was an Orthodox Jew, a mom who was a good methodist. So I grew up going to church and to Sunday school of a father who was a doc as well and seen loggers who that's all they knew how to do is cut down a tree, come into the ER or

my dad worked on because a radical environmentalists would have put a spike in the tree.

And the chainsaw would hit that spike, kick back, cut them up. And then a tremendous period of automation during that time where the mills were going through that, but they weren't telling the workers that hey, you're losing your job to a lot of automation, not just, they're pinning things on the environmental aspect. And so I grew up around this perspective of seen multiple sides and perspectives.

“And that is something which I think has been a defining piece throughout my life as to how do”

we engage in trust and trusted conversations with trustworthy individuals and have trustworthy information on which to make decisions on. After 9/11, that became sort of the, the, the movie dick for me when I had a friend who is ostensibly a smart, smart guy at Duke Law School grad, who was sending around email change.

I mean, think pre-facebook, maybe the very early days of Googles more Yahoo then. And even back then, you know, him sending around these, these misinformation that were bullshit versus from the Quran, you could spend 10 minutes either researching back then or you're on logic to say, hey, this isn't real. And that led to, to really me driving forward is to how do we have high quality trusted

and trustworthy information? And I differentiate trust versus truth, by the way, you know, what to call that out, trusted trustworthy information on which we can have debates and dialogue to make decisions whether there are personal lives or behind the firewall in their presence and fast forward that led me to, even if we have high quality information on which to have debates and dialogues,

we've lost the ability to do so because we're just demonizing one another. And so what are the solutions that exist to solve for not just the trust and information, but the trust and one another as well? So it's a, it's a quick nickel summary there, but, you know, at the end of the day, this is, this is what we've got, you know, that's our job to make it better.

And I'd say it's a nickel summary that's worth a dime and something that you and I share in common that you may not have known is I'm from Beaverton and I fancy myself. I didn't know. Oregonian scholar, I've, you know, published my own sub-stack before Apple CDI, before any scaling laws, nonsense, before law fair, I was writing a sub-stack called the Oregon

survey that focused extensively on the manifold issues that populated throughout the state as a result of what folks in Oregon call down state issues, right? So there's this idea that there's just Portland and then there's everywhere else.

“And that's how we really did a tremendous amount of damage to Oregon's political ecology”

for a long time, such that folks don't even know about the spotted owl words anymore. You may have heard there was a congressional candidate a few years ago who got asked about the spotted owl words and they didn't know what it was and I think that it's just wild. And for those folks who are thinking, what the heck are these two Oregonians nerdy now about this was the issue and it's emblematic of, you know, folks that I've experienced

the rest of the economy.

And what I appreciate about your narrative, Matt, is that you were experiencing the human

side of that technological change and seen and hearing about different experiences.

“We have a lot to get into but I think a lot of this circles back to just that question”

of, have you been exposed to the actual problems you're trying to solve and to the plight that others are facing and if you haven't, then you're operating in a artificially removed environment and you're not able to build that trust and really connect with other humans. I think a hundred percent absolutely, I would also add to it though that an Indian guilty

of charge even as a kid, you know, I think that we, we tend to view the world through black

and whites right or wrong, yes or no, so these very binary aspects.

And I think that's human nature and some respect because it probably gives us comfort in relation to being able to choose a team, choose a side, you know, choose sources, whatever may be. And I think, you know, whether it's the spotted out debate, whether it is relation to, you know, coal miners in West Virginia, whether in relation to whatever it may be, everything in general are shades of gray and nuanced.

And we've lost, we don't have nuance engines, we don't have context engines which allow us to debate and have the dialogue or interrogate information, you know, with one or another around, oh, like people are not binary. And so all that frames part of our discussion today as well as to, like that, you know,

on our ongoing, going in roots, I tell folks that look, Oregon has always been a state where

you're either pro business or pro social and never the two shall meet and I've never understood that as to why that is because we're handcuffing ourselves and this happens broadly across society.

“So well, and I think to this has particular parallels for the tech discussions we're going”

to get into because we don't have those opportunities to necessarily engage with folks in a lot of these AI questions in particular, you're either pro data centers or you're not. You are either pro API or you're not. And what people are often not accounting for is a lot of this debate isn't necessarily going to be resolved on the facts, a failure to address that fear of being left behind,

that people experienced in Lumbertown, USA in Oregon and throughout the West, that people experience in Colorado, that people experience in West Virginia, it's playing out now, that fear of being left behind. And if you haven't encountered people who are experiencing that, it's easy to gloss over their concerns by just saying, oh, they don't know the facts.

And if only they knew the facts, then we could resolve this debate. So before we get all the way into the tech stuff, let's just zoom out again back to the question of this moment in distrust. How is this moment in distrust distinct from, let's say, watergate or let's go even further back, you know, the printing press, you pick the analogy, why is the distrust we're experiencing

now? Again, I mentioned that only 34% of Americans say that other Americans are trustworthy. Why are you so concerned and animated about this distrust moment as compared to prior waves of the lack of trust?

“I think it's distinct in some ways and again, it's not distinct in other ways at all.”

It's distinct now and the respect that versus watergate and probably our most timely aspect is that one, everything across society is losing trust simultaneously. The existing institutions that we have known, everything is being changed. And we're going through this, you know, it's that massive S curve and flexion point that isn't just, you know, in watergate, it is specific to politics and this respect.

But people still had trusted information sources in which they were, you know, reviewing or, and we had, you know, back then, it was a few television channels, right? Now we are all, it's the flood the zone aspect where we're all flooded, we're overwhelmed with information wherever it may be, up is down, left is right, we don't know where to go to, so on so forth.

So there's this, you know, broader unease and the pace and speed at which everything is happening. We haven't developed from, I would say, a human species perspective to even get to a point where we can interrogate that information at the speed that things are happening, so on so

So that's, that's one aspect.

“You know, it isn't so different in some respects.”

This goes way back to Plato's time and the Republic and the allegory of the cave, like he was contending with this, to, if you look at the Renaissance and the printing press and the distribution of material, you know, so-and-so forth is that, look, people talk about the Renaissance, but the early days of the Renaissance as well. It was, it was bloody and ruthless, and it took probably 200-plus years until we developed

as a human species, a set of epistemic sources and methods and processes and systems, and what Jonathan Rauch calls the Constitution of Knowledge, that we developed that over, you know, literally of centuries, so that we could have trust across our societies and which led to human flourishing. And the digital world is just taking a wrecking ball to all of those institutions, if

you will, that if we were to design those institutions for the digital world today, they would not look as they do now. And so, optimistically, I look at, look, we've been here before. Even the human species has been through this. This is just our time, whether we like it or not, for how do we develop the epistemic

systems and methods and practices and culture, because I think at the end of the day, people are, look, whether it's conspiracy theories that result from things or, you know, people are trying to, they're unmoored. So when they're unmoored and they don't have a bearing in an navigational environment, they will grab on to anything.

And in the information environment, I tell people, people will crawl through the desert, looking for water and they'll drink the sand because they don't have any other alternative.

“And that's what we're in in the digital and information world.”

But I'm incredibly optimistic, and I think this is probably a, both frustrating

but incredibly optimistic aspect is there are so many incredible entrepreneurs and

people that we're doing amazing things right now, that a building trust. And oftentimes it's hyper-local. Oftentimes it's at that community level where I would argue trust and relationship building and community is flourishing in many areas that people don't hear about. We hear about all the bad and the negative so it's self-worth.

And so I'm not even quite sure that in some respects, even the data that you presented at the outset, that that is to be trusted and based upon the nuance of geographic domain. And that isn't in relation to country aspect. It's from hyper-local on-up. So this is what's older than it is again.

Yeah, and what I love about your historical deep dive is to keep in mind that when the Constitution was written, for example, information traveled at the speed of a horse. One of my favorite statistics is that-- Or the speed of a ship. Or the speed of a ship.

And in that domain, right, we had the battle of New Orleans was fought in the war of 1812. Despite the war being over. But that news just didn't get to anyone fighting that record. And so boom, we have this whole battle against fought because information doesn't move fast enough. And now, information moves so quickly that we don't know how to respond in kind and how

to just sit and digest all of that information. And one of the things that Alan, co-host of the show and I have talks about, for example, is think about an executive officer, whether it's the governor, whether it's the president. It used to be the case that if they wanted to promulgate an executive order, they would need to physically go walk to their general counsel or to their team.

They would need to sit and deliberate for a while about how best to implement or effectuate whatever the goal of that policy was. They'd need to bring out a pen and a quill, you know, an ink, just to write that executive order so and so forth. There was so much friction.

And now, all of our technologies are about removing that friction.

And in some ways, that should be celebrated, and I want to get to that in a second.

But in a lot of ways, as you and I have discussed, this sort of constitution by way of a physical world.

“Isn't preparing us or isn't easily adjustable to a digital one?”

And so where else are you seeing this transition from a physical world to a digital world manifesting and causing some of these issues about connection and trust? I think it's such an important framing because if we were, and I will frame this as I do,

this, that to your point, the Constitution was made in a physical world for a physical age. When that inflow of information and decision making, so on and so forth, was it a different speed and symmetry, if you will? And that is effectively in my mind, been replaced with terms of service in the digital

world.

“And where I see it playing out is that I was just really a piece, I think it was out”

of wired on a study that don't quote me specifically on this, I think it was one in eight applications for military service members, evidently if been found to have foreign code from China or otherwise in these mobile applications. And this is, the framing I would say is that this is the quartering of soldiers in our digital homes.

When Amazon or Google or Microsoft or Oracle or picked, you have a cloud provider or in Thropic or OpenAI, there is no requirement for any form of digital border control.

And so, and part of that is an economic piece because look, the reality is they've spent

a shitload of dollars on CapEx to build out pre-AI, these data centers and cloud capabilities so on and so forth, and with AI that's even gone higher, and they don't have any requirement or economic incentive to block the users who sign up for their cloud services for their AI services so on and so forth. And so, from a constitutional perspective, our physical home and our privacy so and so forth

was enshrined for us and our life liberties, you know, etc. Very simply, and I'm not a lawyer, so this is just a simplistic, common sense framing for me is we have shredded the constitution writ large in relation to digital terms of service which have taken away our digital home privacy, which forced us to quarter enemy soldiers within our digital environments and companies that are reaping massive amounts of dollars

from that with no consequences. And so, we have to change how we think about, you know, I view this as to, it's fascinating to see in terms of EU versus US regulatory aspects in our humans, and we talk extensively about border control writ wide right now, nobody talks about digital border control. And that, like, what does that mean?

What does that look like?

“And I think that these are important topics to what are the economic incentives, what are”

the structures, what are the requirements, and the digital world we have in hand requirements, this includes for, for software developers, as Bob Lord has said, Bob was that the former CSO at Twitter and Yahoo, he says, there are more requirements to be a petition in the state of California than there are to be a software developer, and yet our software and our

digital ecosystem is critical infrastructure and it is our new home, but we don't treat

it as such. Yeah, you know, I will leave it to the next set of scholars to dive into third amendment jurisprudence and scholarship. I think it's a fascinating inquiry, and I hope some of our more nerdy legal scholars listening to this treat that as an invitation to host to symposium and invite you as a speaker.

But I call me, you know, I'd love to have a one-on-one call, like tell me where I'm wrong. There we go. There we go.

The third amendment, I will say, is one of the least studied amendments, so there's plenty

of threats to go down.

“But to the broader point about this question of digital borders, right?”

We're talking in the middle of July, there's a lot of conversation right now about open source. There's a lot of conversation around the export control reform act, and the extent to which we can prohibit access to things hosted in the cloud for folks who are non, non-US citizens or otherwise station abroad.

And so that is an emerging issue, but what I think you're drawing on too is the fact that we didn't anticipate the digital world becoming such a rival to the physical world, and did not expect that private parties operating in that digital domain would begin to have so much power and so much control, such that they were impacting the physical domain, and to just nerd out for a second, and I try not to do this, but I wrote a paper called

can go and you can see that the founders were very concerned about other entities, exercising

authority over individuals, and this included mobs, this included militias, but even included

“things like corporations and churches, and that's why we have a whole chartering process”

of corporations was to make sure, hey, you are within the domains of the state. Yeah, I'll push back, it used to be a requirement for to be a corporation, you had to show what you were doing for the public good. No, and that's legitimate. So not only were you subject to the powers of the state, and therefore subject to the powers

of the people, but also having a more explicit tie to what are the ends that you are going

to advance, that align with the public interest, and I think now we've gotten so much

away from, what is it you're trying to actually build for? To your point earlier, we were discussing, there's a hell of a lot of slopp, and folks don't like slopp, especially when building that slopp requires a data center in their backyard, or perhaps it's perceived that building that slopp requires more of their personal data.

“The question, understandably, is what the hell is this all for?”

And this is my kind of chief concern with a lot of the AI companies right now is there a nunciation of what they're actually building for is incredibly in precise, very vague, and then they'll just throw out something like, "Oh, well, we're going to cure cancer." Yeah, you know, I see how you're trying, and you're making some investments, and so on and so forth.

But the American people deserve a much more compelling story about what it is you're actually doing to advance the well-being of the general welfare, than just wheelchair cancer at some point. Do you agree with me where am I going wrong, and then we'll get to the positive stuff, because I know you have some good examples there.

Yeah, I mean, I generally agree across the board. I think this is built on that tech, while in some ways, has been tremendous for us all. It opened up creativity and new opportunities and economic prosperity for some at the end of the day. There has not been trust that has been created, so pre-AI, there was not trust created

across the tech industry to the general public writ large. And that includes the fragmentation of how workers were viewed. If you weren't in software and you were doing the trades, for instance, you were viewed and I'm firing for effect year, you were viewed as lesser than. And that was, hey, you got to be in the in-club and in relation to tech and be a software

developer and remember, all of these folks were preaching that if software software software

“and STEM, STEM, STEM, and that's what you have to do if you're going to have a livelihood.”

Well, now that's pivoted and like, oh, now we're going to have the software developer apocalypse. So as these tech oligarchs, if you will, and whether, you know, people view oligarchs and prejorative or positive, that's, I will let folks translate on that. It's just a reference to the monopolistic aspect here where, guess what? This is where now you told us one thing and now something else is happening.

And then we're fighting for our privacy, we're fighting for our data rights, we're fighting for our data ownership, we're fighting for our algorithmic choice and so and so forth.

And they never built any trust out of that from a broad societal standpoint.

And so the narrative as well with AI is, to your point, it's been, you know, horribly, you know, put out there. I think a lot of it was, I am an optimist where I think that there will be a tremendous amount of jobs that are created and we're not focusing on what those jobs can and should be. And I'll fire for a fact here, perhaps, in this respect, engineers do a really shitty

job of telling stories. Engineers do a really shitty job. Like, you know, we have a lot of Steve Wozniacs running around, but not many Steve Jobs. And without Steve Jobs, there would be no Apple. Woz would have built the best home brew club kit computer and no one would have heard of Apple. And we have in the Zuckerbergs of the world, or in the most of the world, it is this perspective of telling these narratives and stories that

does the, the lumber worker? How does the coal miner? How does the farmer? How does the doctor?

How does the insert here? Whatever profession? It's much more of a, hey, get out of our way and we're going to tell you what to do and this is the world we want to create and we don't

“need your input and we're not automated, everything. And everyone should, you know, so I think”

it's a terrible narrative that's been put out there. When in fact, there could be so much positive aspects of in many respects, a JFK type aspect is to, hey, we're at this, this S curve and function point, this evolutionary tipping point, look what we've developed. Now from a JFK

like perspective, we choose to go to the moon. What's the world we want to build? Not have

shoved on us and not have thrown at us. And I think that's the frustration where people don't feel like they have agency and they started with social media. And it's just amplified, you know, now and people are hungry for being able to problem solve and have agency and feel empowered and back to the constitutional freedom and all those are implied there. And instead, they're feeling plastic freedom. And we can told, hey, look at everything you have and this kind of thing

and it's plastic because they recognize that we don't have the freedom or agency and it's digital

world that that we thought that we should have. So there's more there and maybe rambling a bit,

“but I think it's again, it's complex and complicated as to the aspects around us.”

I think that this also begs a bigger question, which is I was nerdy now recently on the difference between a trusty model of elected representatives versus a delegate model of representation. And the understanding was at the founding, you would want to elect a trusty, someone who you thought could represent your community and advance the best things for your community. Now, there has been a shift over time, not only to a delegate model of representation, where

people want them to specifically do what a community asks them to do, regardless of whether they think that's in the best interest of the broader community. So not asking them to defer to their on judgment. We've since moved to a model of what some people call a personal delegate model of representation, where people will only support what their elected official does if it's what they want them to do. Regardless of whether a majority of the voters or the rest of the community

make in test that view, they're instead just advocating for adherence to their own personal values. And that's where I see that trust component coming in because people don't trust their officials to be able to have that awareness of what's going on and representing their views faithfully.

“But instead, it's just this atomic vision of democracy that I think is really concerning and can be perpetuated”

when you spend so much of your time exclusively online or only in these plastic environments that you're referred to. So that was my own rant and ramble. And now I want to talk about how we get out of this. But do you have an initial response there? Yeah. I mean, look, I'll go right to the next part of how to get out of this. And then as you know, I'm deeply involved in citizen or civic assemblies and deliberate

deliberation technology. And that includes where AI fits in as part of that. And for those who are not familiar with what citizens or civic assemblies, it's a very old idea, you know, but it's you know, I'm simplifying. Think jury duty for for issues and it's civic issues, but it's also can be used in business settings. It can be used in organizational settings writ large. But think of it in this way that, you know, a random set of citizens is called up on an issue.

We did a citizen assembly on youth homelessness here in Central Oregon. There's one being done in

those delegates, those assembly members. But it has to be statistically representative of the

“demographics of the community, whatever the size of that community is. That is not how jury duty”

works, it's not statistically representative. And so in this case, I have seen personally, you know, as to the trust that this builds, you know, in Central Oregon here, we're at the the dividing line of what's called the greater Idaho movement. Where the Eastern portion of the state wants us to see the Idaho in part because they don't trust Portland, they don't trust Salem. Like they don't feel representative. And they're not necessarily wrong in some respects. In fact,

they're right in the aspects that they need to be heard and understood. But it's the solution

aspects around this is we haven't developed the mechanisms of the systems to say, how do we sit down

and as citizens together have the agency and the power to make the decisions for our true representation here. And this is where the trust building that happens in these citizens and the civic assemblies is tremendous and incredibly proven. And you know, I'm happy to go into detail on this respect, but those are the types of solutions. And on the deliberate deliberation tech side or we talk about bridging algorithms and right now we have algorithms which are putting

us in our bubbles and our silos and reinforcing our belief systems in our and ensuring that, hey, my belief system is not going to be impacted. So back to your point on, hey, I'm electing

this representative or here's my choice and they don't specifically follow my belief system.

This is something where how do we change and how do we require from a technology perspective or standardizations as to how do we encourage bridging and healthy dynamics and healthy conversation

“instead of the worst of us and incentivizing the village idiots?”

Yeah, well, so to go down the con law route again at the foundation of our Constitution is a Republican form of government and three things if you look at federalist paper 10, three things really define a Republican form of government. Number one is deliberation. Number two is cutting against sort of a impassioned response. We want to make sure we have cool deliberation. And number three is making sure that we're countering interests. We're countering things like

factions, right? And we want to make sure we're bringing in enough groups. Well, in the 18th century, we didn't have AI to do things like make sure we could actually scale deliberation to entire communities to be able to engage in these conversations all at once. We didn't have AI tools to assist with countering our own impulses and our own biases to point out how we could think through issues in a way that's more reflective of their impact on broader society. And so

“what I think is a really exciting moment right now is to ask, how can AI actually advance the”

core aspects of Republican governance rather than advance Republican governance as we know it today? Because I think that's a big differentiating factor that people just struggle to have the imagination to think about what a future Republican governance could look like. Look, I'll give it, you know, perhaps a simplistic example that I have this, this dream on is that I want to see from an AI perspective, us focusing on what is, I want to assume, but in some

part I will, that the audience is familiar with SimCity, the video again. SimCity allowed us to simulate what is this city going to look like when you build it out so and so forth. We are not in have not focused on, and this is where AI I think has, and I want to separate out AI, AI is not just LLM, it's not just transformer models, there's a lot more beyond it, but but so I don't want to quate AI when we talk about it to just generate AI so on and so forth.

That said, there's so much of an opportunity for us to be able to use AI, not just in relation to bridging algorithms and deliberation and how to we encourage the healthy communities and folks like whether it's Demnext, whether it's folks like Google Jigsaw, pull us others, doing some remarkable stuff here, but this is where what is a SimCity 2.0 look like? Imagine if we as citizens, imagine if if whether it's existing representatives on

so forth could actually, I want to understand, hey, we are proposing putting this new process or this new legislation in place, well, what are the incentives? What are we trying to solve for?

consequences? How do we have these simulated environments in this predictive just as we're doing this

with drug discovery? Why aren't we doing this with our form of Republican governmental aspects on

“so forth in liberal democracy writ large? And I think this is some of the power for us to”

really lean into is to building out that type of capability, that is type of a civic engineering solutions that they don't exist today. And I love that because I've had this debate actually with a couple AI skeptics in particular, which AI has many flaws, but one of the things it's continually improving on is simulation. And I think that future generations are going to look back at this era and ask why the hell were you not mandating that your legislature simulates the likely outcomes of

that piece of legislation. And if they don't achieve stated outcomes X, Y, Z, then they can't pass it or maybe they need a super majority to pass it. The fact that we haven't expanded are regulatory technology to include the advances in technology we're seeing in other domains is just bonkers to me. But I want to transition to how we can make sure that we realize this JFK moment. I am a big fan of a piece by Harry Law of the Cosmos Institute who highlighted the fact that being a part of

a Republican form of government involves having formative experiences. Even if we had an AI model that could perfectly express our values and beliefs and so on and so forth, if you and I don't go out and engage with those people. If we don't go meet those other Oregonians and learn what it was like to go through the spotted owl wars and so on and so forth, then we still don't have that glue that binds us all together as a political community. And that's something I'm increasingly

concerned about because I go and I talk to folks in SF and they live in one universe where spinning up agents is a normal activity that that makes sense and then I go and I travel the Morgan town West Virginia and if I said spinning up agents, they look like I just had 15 mountain dudes or something and I worry that we're going to see a further separation of these political communities such that it's going to be really hard to say these United States of America as opposed to these

very distinct communities based off of very different technological paradigms. So do we need national service? Do we need a CCC? What is it that can help restore this glue? Because yes,

we can have AI improved deliberation, but ultimately I still think we need that human to human

connection and I would really love your thoughts on how we can manage that. There are many of others who have spoken on this, besides myself so and some things I'll repeat from others.

“Two minutes now I respect for the national service question. Yes, I think that there should be”

required national service. I think and national service doesn't mean military service. That just means service and that could be back on the citizen assembly front. That is service. The friend of mine who I brought on is Executive Director to lead it up here. He was 25 years of Air Force Special

Operations and he opened it up with saying that he always felt it was disingenuous for being

thankful for his service when folks from NGOs, from State Department, from the Intelligence Community, for the press, were never thankful for their service and he opened it up saying, I want to think each and every one of you assembly members for your service here. And so I think people are hungry for that as well. They're hungry for that. Again, we're all every human being wants purpose. They want to be able to contribute. They want to be able to do not just be alone for the

“ride, if you will. I inherently believe that. Some don't. I inherently believe in, I think what”

Scott Calaway refers to unremarkable people. And that I think is the power of this country. And I think that we're at this moment where, and you and I have discussed this, I use the phrase civic engineering. I want to see an army of civic engineers that is helping build the new country. And we discussed that with my son goes to West Point. And West Point was created as a, as a, by an engineer to go train the next generation of engineers to go build the new country.

And we're at that pivotal point of what is the country look like that we all want to see. I believe that people, the common things that bind us generally is around the, the broad aspect of freedom.

those are you with his father who were, they were first generation immigrants from the civil

union. He was outlining that the, to his father he said, look, freedom and what this country stands for is for you to believe in something that I might not agree with. And that's a beautiful thing in some irrespects. And we tend to be afraid or run from that. And this is where the engineering the country we are, we want to create. What is the, what are the solutions? Trust is at the foundation of it. Like if we don't trust one another, if we don't have trust information, like we don't

have freedom, if we don't have, you know, epistemic, you know, systems and structures, if we're

“not giving people purpose and agency. And so there is this, this moment where I think that we need”

to lead into. And again, I call it the civic engineering pieces. What are the solutions to soliform bill and trust and information, trust and one another, trust our institutions? And reimagine and not expect that we're going to have a billionaire who's going to solve our problems because they were able to create a company that landed a rocket with with chopsticks for nominal.

But this country has always been about bottoms up and people working to solve our problems and

create something amazing. And so, look, I am back on the naive optimist. I'm an naive optimist because we are always on this journey of can we live up to the potential of what the founding fathers created. They knew that we weren't, you know, all-menter created equal. They knew that

“that we were not. That was an aspirational. What is our potential? And I think it's really important”

to for all of us and this starts at fire cells. It starts at a community level, ground up. What can we do? These little things to leave things better, to work on improving, to chip away at the wall. So as Clint Smith says, we'll be that much closer to the other side. And we've got to change the minds that there. Instead of this, it is tiresome to be more and to be arguing, to be hating, to be fearful. Right? So what can we do to flip that? Yeah. And I love that idea of building.

People want to contribute. People want to go and help construct new communities, new institutions, new ideas. We're seeing that in so-and-so, so-called "prototounds" across the U.S. Right? Because people are just hungry to go create new things and to be parts of communities that don't feel so bound by the status quo. And now we have the technology and the opportunities to ask even larger questions. And the challenge for us is to just really take that seriously. Because I think

that as much as I'm supportive of, for example, President Biden's moonshot for cancer. Great, obviously, fuck cancer, everyone hates cancer. But that moonshot didn't involve a whole of society approach of STEM education, of new engineers, of new schools, so on and so forth. We need moonshots that are tangible for people, that result in that meaning and purpose.

“And that's how the way. Imagine, imagine, and granted, this was when the frontier malls were just”

emerging as well. But imagine that cancer moonshot, if the frontier labs had partnered on that, and how society would feel about what would we be having the same sort of AI data center debates?

First is, hey, we just created this thing that can do these cool, you know, deep fakes or cool,

you know, later videos or whatever it may be. Who the fuck cares? Like, like, let's solve real problems that affect everyday people, that people can feel and touch and understand and make their lives better. And your narrative is then solved for you. The trust is solved for you. Exactly. Exactly. Well, and I think having new institutions like data trusts and data co-ops that say give us your data and then we'll build tools for you would be so transformative. But Matt,

I'm getting the sense that you and I could talk for many, many hours, and it would be maybe several days if I were hanging out with you at crux in Bend drinking a stop that is way too strong. Even in the summer, I will drink an organ stop because I miss them so much. But Matt, thank you so much for joining scaling laws. This was a who it was great to be on here. So thanks, Kevin. Scaling laws is a joint production of law fair and the University of Texas School of Law.

at our website. Lawfairmedia.org/support. You'll also get access to special events and other content available only to our supporters. Please rate and review us wherever you get your podcasts. Check out our written work at lawfairmedia.org. You can also follow us on x and blue sky.

This podcast was edited by Nome Osband of Goat Rodeo. Our music is from Alibay. As always,

thanks for listening. Already, for example, changed your predictions about which of these plans plan A, B, C, D, or S. You think is most likely following the OpenAI hugging face incident. You now suspect that plan A may be 18 percent likely and improvement on 15 percent likely. You have now diminished the likelihood of plan D, the default do nothing from 30 percent likelihood now to 20 percent likelihood.

You've already changed some of these things. What's been the main additional pushback or additional piece of feedback that you said, huh? That was a really good take. I wish we had addressed more

“that you know, has really landed for you in the team. I think the main one is actually something”

that you may have just covered, which is this domestic regulation thing. I think that Richard know has this critique where he basically says that like we are inadvertently reinforcing this harmful narratives about the race with China by, I mean, if you read any of our work, we're very much talking about the race with China. And that's not because we think that it's good that there's a race with China. It's because we think that's where DC is at and that's where policy

makers are thinking about. And so we want to meet them more that are and be like, yeah, race with

China. It's a serious thing. Here's the way out. Here's what we think should be done about it.

But Richard is thinking that maybe it's better to like deny the premise more and say that like it's not really a race because you're not going to win. Like you're going to lose control of the race. So like, what is different from from most races? For example. And I do feel like maybe he's right about that. And I don't know. Well, we'll see. But we said what we said. And we do think that like even from within this sort of race with China framing for the reasons that we've stated,

“you should do the things that we recommend. And for the rest of the AI policy community, we've”

already talked about the value of scenario scrutiny and really pressure testing your ideas.

Would you like to see more people publishing their own version of AI 2040? And what would that look

like to see a sort of field of scenario scrutiny developing? I hope so. I think yes, we would love to see more of those things. In fact, one of the nice things, I think there is this scenario called EU or Europe 2031 that was clearly inspired by 2077. And we don't agree. We don't agree with the authors about everything. But like, we are very pleased to see people sort of like put things down. And then I think that like once we have a bunch of different scenarios on the table,

then we can have like the argument about which is more realistic. And which is less realistic. And like what are the different aspects of them and stuff? And so I think that's good to happen. I think there's a set beyond that, which I'm hoping to happen, which is like war games. And I want to mention this just as a brief part of the motivation for scenario scrutiny is that when I look at the history of military history, I'm sort of inspired by how seriously they take

their jobs intellectually. Like when even back in World War II, it was common for commanders to have war games gaming out the plans that they had for the war and for the battles and so forth. And so for example, with the Battle of Midway, the Japanese had their plan and they made the plan in part on the basis of various war games they had done. And then even after they had made the plan, they kept war gaming it out. Like as they were like getting ready to attack. And in fact,

if they had taken their own war games more seriously, they might have realized that they were about to lose the battle of Midway. Because in one of their war games, the person playing the Americans had the American fleet waiting in the north to attack them and then utterly wrecked the Japanese fleet. And then they were like, "Well, but the Americans don't know we're coming, so they're not going to be doing that." And in fact, they did know they were coming and they did

do that and they got wrecked. So that's an example of people in the war, people in the military

“take very seriously this idea that you need to apply a scenario scrutiny to your plans. Not just”

scenario scrutiny, like war games journey, which is like a higher level of scrutiny. It's like not only are you gaming out in detail, like what it would look like to implement your plan, you're then subjecting it to adversarial pressure and like gaming out different possible ways to

they ideally would like to get to that place where the policy makers in Washington are treating superintelligence with the level of seriousness that is routine for military operations.

“Well, and I think too, as I've reflected this summer on July 4th, as we all have in some”

way, shape or form, it's worth noting that the founders themselves were engaged in a degree of war gaming when they were drafting the constitution thinking about what are the ways in which this could break. Let's use our imagination. Let's have a high degree of creativity and yet that's often lacking in a lot of these policy discussions. So again, I will applaud you all for taking that on and embracing that sort of thoughtful creative approach. But Daniel, I know you have many more

reports to author, many more timelines to sketch out any final thoughts you want to leave for our audience. Yeah, thanks for asking. A couple things. I'll try to briefly go over them. So one plan A is different from just like permanently pausing AI until 2040 and then going as fast as possible. It's more of like a graduated controlled scale up over the course of the 2030s and it does

“involve some pauses at various key points, but so that's one thing. The world transforms dramatically”

in the 2030s if we do plan A. One intuition pump for that or one reason why that's happening is that in plan A, you are sort of slowly scaling through the human range and you're starting, you're starting in 2029, they're already at the point where the AI is able to automate some jobs and are having a big effect on the economy. And then they reset to, you know, get to super intelligence in 2030, but instead they go more slowly. But that means that like

you're still having this transformative effect on society and you have these AI's that by 2035

are as good as top human professionals in basically every field while also being much faster and cheaper.

And so the economy goes crazy. Like we we project that countries are going to want to limit economic growth rather than encourage it in plan A and that they're going to want to have limits that are going to look things like only one doubling per year. You know, and things like that. And for contrast, for context, like right now the economy grows at like 3% or 4% per year on average. And so doubling would be like 100% growth, you know. And so I'm we think that that is actually

what you get. And it's actually pretty straightforward, the argument for it. If you have machines that can substitute for human labor at practically everything, but the machines are much cheaper than humans and you can produce more of them much more easily because there are, after all, just just more GPUs. And we know how easy it is to produce GPUs. And we know how

it easy is to produce robots. Then you're like population is basically doubling several times a year

or maybe doubling once a year. It depends on, you know, maybe starts off at once a year and then it gets faster, you know. And so then your whole economy, once that population is the bulk of the economy and the humans are just sort of like a small sliver on top of this giant army of robots and robot factories and robot trucks and and everything, then the whole economy is growing at machine speeds instead of growing at, you know, human reproduction speeds, you know. So we say more

“about this and the supplements and we explain why we think this. But it's I think an important sort”

of high level takeaway is that you still get this insane rapid transformation of the entire world,

abundance for everyone, all that sort of stuff, even if you basically pause at human level,

API. Yeah. And worth flagging the often quoted remarks from Ethan Molakir, which is to say, even if we pause today, not that I'm endorsing a pause of any any kind necessarily, but saying that we have so much room for just integrated the advances from today's AI that our systems and our institutions aren't ready for. And so it's a huge societal task ahead. Thanks to you, Daniel, and the rest of the team for pushing the rest of us to engage in a thoughtful policy exercise and showing a potential

way forward for how AI may unfold. I'll let you get back to it, but Daniel, thank you again for joining scaling laws. Thank you very much, Kevin. Happy to, happy to come on more, maybe. But I really appreciate your covering these topics and, you know, good luck to us all over the next two years. Uh, we'll see how it goes. There we have it. Thanks, Daniel. Scaling law is a joint production of law fair in the University of Texas School of Law. You can get an ad free version of this and

other law fair podcasts by becoming a material subscriber at our website, law fairmedia.org/support. You'll also get access to special events and other content available only to our supporters. Please rate and review us wherever you get your podcasts. Check out our

edited by Nome Osband of Goat Rodeo. Our music is from Alibay. As always, thanks for listening.

“Schicksa's Wall in Saxon-Anheit. For the Landfax Wall, let's ask a lot how the land is going to work.”

In the podcast, there's a lot of different ways. No cliches, but then there's a lot in a

Bundesland Imvanthel. We're from Podcast Radio Detector FM, Schauenhien and Ordnen ein.

“Hintergründe, Stat Vorurteile, Pertrain, in the podcast that's between in Euler Lieblings Podcast”

or on detector FM/Stat Zwischen.