AI and Its Relevance to Philanthropy

Artifical Intelligence is here and everywhere. That includes the nonprofit and philanthropic worlds. In this episode, Chris Koopman of the Abundance Institute offers us a perspective on the problems with the pessimism that is seemingly everywhere, how AI is uprooting old ideas, whether AI is a new Renaissance or a new Industrial Revolution, why everyone hates data centers now, the many ways public policy philanthropy intersects with AI, and how nonprofits should think about protecting themselves in this new era.

Note: This transcript was generated and cleaned with AI

Peter Lipsett: One of the reasons I like being a conservative on the free market side of the ideological spectrum is that we argue that prosperity is available to everyone. In a free society, people have the agency to have a better state for themselves through their labor, through their ideas, through problem solving with creative ideas for other people. You know, one way or another, there’s a pathway for everyone to get to their desired level of prosperity, which could be something different for everybody.

That’s the real essence of the American dream, right? That people get to their desired level of stability and comfort. And here we are in this technological era with artificial intelligence taking over the conversation. And in a lot of quarters, you hear that that’s going to be great for bringing about prosperity. But there’s this other faction, sometimes a louder faction, preaching that there is nothing but doom and gloom and sorrow ahead for all of us as our robot overlords take over.

So what kind of future are we really looking at? And why on a philanthropy podcast are we talking about artificial intelligence? Is that just because everybody else is? Well, no, we’re going to get into that: why this is an important topic for donors, for nonprofits, for everybody. It does all tie together as this AI debate is raging and the broader technological changes take place, because it’s going to affect everything.

And so joining me to talk about that is my friend Chris Koopman. Chris runs the Abundance Institute, which is focused on creating space for emerging technologies to grow and thrive and have a chance to reach their full potential. We’ll talk more about what that really means as we go on. Chris started his career at the Mercatus Center at George Mason University and went on to be executive director of the Center for Growth and Opportunity at Utah State. Chris, great to have you with me.

Christopher Koopman: Thank you, Peter.

Peter Lipsett: So there’s a lot of techno doom and gloom right now. Social media was already ruining our lives. And then data centers came out of nowhere to be this eater of farmland and all of our hopes and dreams. And now AI is gonna turn us all into paperclips. But this doom saying is not new, right?

Christopher Koopman: No, not at all. I think the introduction of every new technology has come with a story of ruin and disaster. It doesn’t matter how big or small the technology is or was; this has always been a part of the conversation. There are always people who refuse to see the upside and focus on the downside. My guess—and I wasn’t there—is that when that first caveman brought fire back to the cave, there was someone who tried to chase him out. And we see this time and again in history.

I’ll give you a couple of examples that I think about a lot and think are really illustrative—it’s okay to be afraid, but what we do about that fear is very, very important. The bicycle. Turns out there was a wild societal panic over the bicycle, for a couple of reasons. One, there were actually studies done on a phenomenon they thought would occur called “bicycle face”: that people traveling at such extreme speeds would be stuck in a constant state of shock on their face. This is a real thing, the bicycle face. The telephone—people were worried about the telephone.

Peter Lipsett: Even Alexander Graham Bell was worried about the telephone, right? Didn’t he famously not actually use it?

Christopher Koopman: Yes. John Philip Sousa, the king of the American march—every kid who’s ever played in marching band has probably played a Sousa march—was against recorded music. There’s always been someone who looks at every new thing and can only help but think about the worst-case scenarios. But it’s our job, as people who work in policy or in shaping culture, to make sure that we don’t let that fear and pessimism drive us to make bad decisions. We were talking about this before, Peter: the doom and gloom had already infected the conversation around technology and the internet in the context of social media. And now it’s just bled over to AI.

But I think about the radio as a great example of this. There were stories when the radio first came out. We have this documented at a project we help support called the Pessimists Archive, a project we’ve been working on with and through the Abundance Institute. It documents all these historical panics. Louis Anslow, who runs it, has done an amazing job. He’s like the archivist of global historical technopanics. And he’s pulled these examples that are funny now, right? I think it’s always interesting to think about the context of the conversations we’re having today. How will we look back on them in 50 or 75 years and laugh at ourselves?

But the radio: there was a story of a headmaster at a boarding school in New England who said, basically, that the vitality of his young men was totally destroyed by the radio. All they want to do in the evening is lay around and listen to stories of cowboys and Indians, fighting these Western fights, and the idea was that they were losing their creativity, their vitality. It was melting their brains. It was destroying their ability to read or to write or to be creative. Almost every technology that you can point to, even ones we don’t think of as anything more than mundane today, sort of all had this pessimist technopanic moment about them. Even artificial ice—even ice machines, Peter. There was a real challenge among some people in accepting the idea that we could just make ice in our homes.

Peter Lipsett: You know, change is tough, right? And any change—it’s not always just daisies and sunshine. There are always downsides. There’s gonna be disruptions. We saw each other last week at the Philanthropy Roundtable conference, and you were speaking about some of these issues, and one of the things you said really struck me: you said we shouldn’t confuse churn with ruin. Can you unpack that for me?

Christopher Koopman: Yeah. So I think, as people who have studied or understand the market process, there’s a concept called creative destruction. This is how things get better, right? The old way works until a new way comes along, and then there’s a transition from the old way to the new way. There are costs to that transition, right? Think of a taxi driver circa 2012, the advent of Uber. It dramatically changed your life, but in the long run I think it has made a lot of other people’s lives better. And I think what happens in the AI conversation is many people have lost the plot, and they’re confusing creative destruction for permanent economic ruin. They’re saying the way things are right now is the way things will always be going forward, and history just doesn’t prove that out.

There will be transition costs in the advent of every new technology, every paradigm shift. There are going to be costs to that transition that are borne by real people, and how we respond to those is really important. But we shouldn’t confuse the churn, or the creative destruction caused by this transition, with anything more than that. And I think to begin to try to project out that humans will just live in this ruinous world as a result of the advent of AI is totally missing the opportunity that’s before us.

Peter Lipsett: Yeah. You know, one of the things I grapple with in the whole AI conversation is this Sam Altman sense of, like, everyone’s gonna be richer, we’re all gonna be wealthier. And I have not been able to mentally make the jump between being able to use ChatGPT really well and being able to lie around all day like I won the lottery. Are you able to connect those dots for me? How are we all gonna get wealthier from this?

Christopher Koopman: Yeah. So—not to pick on Sam—but I wrote a piece in Fortune recently about this. I think his read of economic history is a bit off. He spoke in front of the UN several weeks ago and said something to the effect of: we have two choices before us, we can have a new Renaissance or a new Industrial Revolution. And he paints the Renaissance as this beautiful, serene time when everyone was doing great, and the Industrial Revolution as this horrible time when people were not doing great. My argument against that is that the Industrial Revolution did more to lift people out of poverty than the Renaissance ever did. The Renaissance was a time of patrons and massive accumulation of wealth among a few, and a lot of pain and strife among the many. And I think the Industrial Revolution changed that, right? You look at all of the stats that we like to brag about, the benefits of modern society, and all of that was essentially the direct result of the Industrial Revolution. Infant mortality rates, life spans, poverty rates—all of this was dramatically improved by the Industrial Revolution. I think we’re on the dawn of a new industrial revolution, and trying to say that we must stave that off is a problem in and of itself.

But to answer your question, I think the idea that we’ll all be living in this utopian future in which we live permanent lives like infants or toddlers, where we just play games all day and lay around and everything’s provided for us—I don’t think that will ever occur, because that’s just not how the world works. But what I do think is that you can have a tremendous amount of opportunity and prosperity generated that is completely disconnected from this utopian idea of, well, you can just go be a poet or live the life you want and not worry about the economic implications of your choices.

And this was something I was talking about last week as a part of the conversation: the collapsing price of expertise, right? I think you said you have a hard time understanding how using ChatGPT will bring about your utopian society, as Sam Altman promised you. I don’t think that will occur, but what will occur is that you and tens of millions, if not hundreds of millions, of people, not just in America but across the world, will be given access to information and expertise that they had never been able to grasp at any point in history. Information and expertise that was only reserved for the richest and most elite among us will now become readily available at almost zero cost. Some may say that’s utopian in a different sense. But that, to me, I think, is the impact that a tool like AI and these large language models will have on our ability to live lives of meaning and purpose and find flourishing and prosperity that don’t get us into this world where, don’t worry, you can lay around all day and everything will be provided.

Peter Lipsett: There will be an abundance, which brings us kind of to what you do all day. I mean, we shouldn’t just listen to you because you’re young and hip looking. You actually run an organization that thinks about these problems as a research institution, a nonprofit institution, the Abundance Institute. And I mentioned your mission at the top of the show: create space for emerging technologies to grow, thrive, and have a chance to reach their full potential. It sounds like you’re fully focused on making the optimistic case for technology change. Is that right? Are you only looking at the bright side?

Christopher Koopman: I’m unapologetically optimistic, but I’m not Pollyannaish. So I’ll say, the origin story of the Abundance Institute: we launched just about three years ago. We’ll be three years old in April of ’27, so we’re like two and a half, I guess, and change. But really, the entire origin story of Abundance was that I’ve spent my entire career working in policy and at the intersection of technology and regulation. And going back about three years, I was watching this conversation get away from us really quick. As people who advocate for free markets and are optimistic and believe in the power of human agency, we were losing the narrative very, very quickly. And you can go back in time to those early Senate hearings that Sam Altman was a part of, in which we were told essentially that this technology will destroy us unless Congress gives Sam Altman the power to write the rules, and he’ll keep us all safe forever. That was a moment for me where I was like, that cannot be the only option available. And if we want to see all of the right numbers going up and to the right on all the charts we care about, we have to reclaim the narrative.

The launch of Abundance—as you said, the idea of it at a top level is helping emerging technology grow and thrive and reach its full potential to support human flourishing. This is all about the people who benefit from these technological changes. As a part of that, there’s a tendency, as we already mentioned, for the advent of any technology to be met almost immediately with pessimism, doom and gloom, and worst-case scenarios. What if we helped people understand a different shade of these technologies? The way I view Abundance’s role in the broader conversation is sort of twofold. One, as best we can, we want to shift that conversation away from pessimism and toward optimism. As I said, it’s not utopian, it’s not Pollyannaish. It’s just a belief that what history has taught us is that almost every technological advance we’ve undertaken has led to tremendous benefit to humanity.

The other part of this is the way that the rules are made: the people who have a seat at the table are the people who exist today, right? So when you have a conversation about the regulation of AI, who do you invite to the room? You invite the frontier labs who are building the technology today. The people who are helping shape that conversation are people who exist in the room today. And oftentimes what we see in the history of the regulation of technology is we pass a bunch of well-intended policies in an effort to make things go right, and almost immediately realize we screwed it up. The EU AI Act is a perfect example of this. They wrote all of these AI rules very early on, and before they even went into effect, they realized they were unimplementable. You couldn’t do it. They were totally wrong.

Peter Lipsett: And this is why there are no labs in the European Union, right? Because they just can’t do it. They regulated themselves out.

Christopher Koopman: Yes, this is why AI will not be a European thing. But a part of this, and what we try to do at Abundance, is in some ways present-proof the future. It’s ensuring, as best we can, that we represent a point of view with a bias toward the future in these conversations that are happening in the present. So in some ways the way I would describe us is that we’re representing founders who haven’t founded anything, industries that don’t yet exist, companies that haven’t even started, the next generation, to make sure that we’re writing rules today that allow for subsequent iterations in all of this that wouldn’t be possible if we just wrote rules today with the incentives and the perspectives of the people today making them.

Peter Lipsett: Yeah. So it’d be similar to any public policy shop, except you just happen to be working on the sexy, hot issues of the day that will determine the future issues.

Christopher Koopman: I guess I got lucky. Yeah, when we launched the org focused on artificial intelligence and energy three years ago, did I imagine that they would be this issue today? Sure. Yeah, totally. I knew it. But I think also, a little bit of what makes this different is that we don’t approach these issues from a traditional think tank perspective. If you were to look at what we produce and how we produce it, we overinvest in our video and our storytelling content, and we underinvest in your traditional white papers. Just knowing how the conversation is unfolding, and the way in which politics and policy are downstream of the cultural conversations, the more we can push upstream, the more likely we are to have an outsized impact on those downstream conversations.

Peter Lipsett: Yeah. No, that makes sense. Well, let’s take an example of an issue. Let’s set AI aside for a moment, because AI is a big thing and we could get into what the AI debate really is, because I know you’ve broken that into various different pieces as well. But the data center debate is another thing that you work on. It’s a little bit more traditional public policy, one that people can get their hands around, I think. That just completely went off the rails recently, right? Nobody cared about data centers. Data centers were just kind of a thing. And then data centers went from a thing that’s out there to this absolute boogeyman. And here we are recording a few days before the midterm elections, and it’s become an issue in a way that I don’t think anybody, even back in March, would have said, “That’s going to be a driver in the elections.” What happened there? And let’s take for a second that data centers are neither good nor bad. They just are. They’re a part of this process if we want innovation, right? So it’s tough to really plant your flag on either side if you’re looking at it rationally, in my opinion. But where did this go off the rails? What should people who care about public policy, particularly future-oriented public policy, be thinking about in that debate?

Christopher Koopman: Yeah. So I guess the question is, why has America become obsessed with data centers the moment it has? Why this issue now? And I think—you didn’t want to talk about AI, but you cannot have this conversation without understanding the AI conversation. Here’s how I’ll put it. I would say there are three distinct groups of people who have all converged on this issue.

In the one group, you have, let’s call them, traditional enviro groups that are worried about resource use, worried about increasing energy consumption, worried about the impacts on the environment. They have always been worried about data centers. That’s very, very clear to me. Then you have two groups who enter the chat new. But first, I would actually bucket in that same group the NIMBYs, the anti-development, not-in-my-backyard type folks who just generally don’t want these things built in their communities and their cities and their counties and their states. I would put them all in that same group because the way they’re approaching the issue really is the same, right? These are a problem, we don’t want them. And it doesn’t matter the degree to which you can push back. Kevin O’Leary was trying to build a data center out in the Mountain West, and there was a stat from an environmental sciences professor that said the heat generated from it would be the equivalent of dropping twenty-nine Hiroshimas. Which is sort of like, well, yeah, but look at that in the context of cities themselves; they’re also consuming a lot of energy and generating a lot of heat. It’s a very strange thing.

So, okay, you’ve got the NIMBYs and the environmentalists who are just anti-development of these things. I think two groups enter the chat late, and it explains why this moment now. The one group, I would say, are everyday Americans who have heard a very troubling story from the frontier labs developing the AI technology: that they are building a future in which Americans don’t have a place. They’re telling you there’s like a 90% chance you’re gonna lose your job. There’s like a 50% chance that there are zero-day exploits in all of the banking system and you’re not gonna have any money when the AI is powerful enough, because it’s just gonna hack the system, and there goes all your money and the financial system is ruined. And by the way, there might be a 10% chance we kill you, right? If you’re very concerned about this technology, and you find yourself dismayed by the lack of urgency that policymakers had maybe taken to regulating the development of the technology, but what you find out is there’s physical infrastructure necessary to build out the next frontier of these models and it exists in data centers—well, you’ve got your perfect villain for the AI future that you’ve been sold, in my mind, poorly. And so you fight the data centers. You don’t like AI, so choke the infrastructure. You can’t stop software, but what you can stop are the pipes and tubes and chips that make the software work.

The other group—and some of this depends on what conversation you’re in; people respond to this differently—but there are foreign adversaries who have amplified and fomented the pre-existing distrust that people have for this technology to make this the moment that it has become. So you have people who believe it, and now you have foreign adversaries who are making it seem like everyone believes it, and you’re stupid if you don’t, or you’re naive if you don’t. So I do think this moment is a perfect storm of classic environmentalist, NIMBY, anti-development folks mixed with people who are deeply concerned about AI, all being spun up into a frenzy by living terminally online and believing that everyone also believes that this technology and the physical infrastructure is the end of the world. That’s really the way I see the data center moment playing out over the last six months. It’s only gonna get worse, not better. This is where my optimism starts to fade a little bit. The conversation really took heat, as you said, in May, which is after most state legislative sessions had already gaveled out. So the 2027 legislative sessions are the first opportunity that most state legislators have to do something. And now many of them have just come out of an election in which they believe, rightly or wrongly, that they’ve been given a mandate to do something about data centers. So I think the 2027 state legislative sessions are going to be a frenzy of policy proposals around data centers.

Peter Lipsett: I haven’t thought about that, and I think you’re absolutely right. And that is certainly something for donors out there who invest in public policy to watch. Because, well, let’s get into this. When we say the AI conversation, what I heard you say the other day was really that it breaks up into a bunch of different buckets. And I think that’s actually really relevant for the donors out there listening, because the AI labs are not nonprofits. I think OpenAI is no longer a nonprofit, even though it once was. So AI is moving on its own for-profit schedule, but there are all these other things going on where there is actually a place for philanthropy to engage. Talk about those different pillars of engagement.

Christopher Koopman: Yeah. And I especially think, for your donor community and the people who are advocating from a free market perspective on a lot of these issues, there are a bunch of these layers of this cake where these are home games for us. These are not foreign issues. But I think the best mental model to understand this is one that was popularized by Jensen Huang at NVIDIA: the idea that AI is actually a five-layer cake. It’s not one issue. And oftentimes I think people just bunch them all together as a single issue, and then it gets very muddied and convoluted as you try to sort out the various threads of the conversation. But you can think about these as stacking on top of each other.

At the base, you have energy, right? You don’t get anything if you don’t have energy. That is the base layer for everything in our economy, but in AI in particular, it’s really important. The next level up from this would be chips. How do we actually make silicon, how do we turn sand into thinking machines? That’s the chips that do that. You take the chips, you combine them with energy, and you get the digital infrastructure, the data centers, which is how you actually train and run these models. So that’s the third layer: data centers. Then on top of that, you get the models themselves, the largest versions of Claude and ChatGPT and Grok, the actual models that are repurposed in a bunch of ways. And at the very top, you have applications, which are the direct and discrete uses of these technologies in very specific use cases. So applications would be healthcare AI apps, education AI apps—Google Maps is an AI application.

And I think once you start to think about them in five layers, the questions and the answers and the right venue to have them become a little bit clearer. I don’t think data centers should be a national conversation. It’s a state and local conversation. This is just the building out of infrastructure. You wouldn’t want Congress to pass a bill on the location of Amazon distribution centers or Walmarts or office buildings. That’s a state and local conversation. The energy conversation is a mix of federal and state. The chips is really a federal conversation, though some of it’s state: where do we build them? If we build them in the United States, where are we manufacturing them? That becomes a state conversation. The models: we should have a uniform, harmonious set of rules for models across all 50 states. A model being built in California shouldn’t have to comply with 49 other versions of how that model should operate in each state. Not only is that an affront to interstate commerce, but it’s really hard for the next generation of founders and startups to grow and thrive. You know who can handle that? The frontier labs can handle that. It’s the next gen, the people who are going to unseat them from their perches at the frontier, who are really going to have a tough time.

And then the applications. The applications really are a state and local issue. There’s some stuff that might be federal, but for the most part, how AI is used in the classroom, that’s a state conversation. So once you start to break these down into layers, you start to understand, okay, what are the tools we have, what are the questions we’re asking, and who are the people we’re actually supposed to be bringing these questions to? Because right now a lot of it is all being pushed, rightly or wrongly, to the federal level. And then when the federal government doesn’t do anything, they say, well, the states must. Well, just because the federal government won’t do anything doesn’t mean the states are the right venue to write these rules. We shouldn’t just pull on any lever available. And I think breaking them down into those component parts helps you better understand how to tackle each one of these questions as they emerge in the conversation.

Peter Lipsett: And they’re all a reminder that there are policy fights in all of them. The chips one is probably the weakest of them. The models, as you say, are kind of at the federal level. But energy—there are plenty of people who listen to this podcast who have been in the energy fight for a long time. It’s just taking on a different cadence. And so it’s also remembering the school fight you referenced at the end with the application layer. Some of these fights that people have been in for a long time just now have this different valence, but it doesn’t make the fight different. It’s just that the issue is a little different. And frankly, that’s a mix of legacy organizations working on these things as well as new groups like Abundance.

Christopher Koopman: And it’s completely changed. For a long time, the conversation over nuclear, small modular reactors, wasn’t something people were talking about 20 years ago as a real hot policy issue. And yet now you’re seeing half a dozen, a dozen states really taking an active role in pushing both the federal government and themselves to see this technology deployed. You don’t even get small modular reactors if you don’t have the labs, the hyperscalers who are pushing for demand and creating a market for technology that for a long time has been viewed as not the most economic. So it’s actually bringing up old fights. We’re debating and talking about building nuclear again, which we haven’t done in 50 years. And all of this is a direct result, I think, of the moment being created by these frontier labs.

Peter Lipsett: I know the answer to this, but I’m gonna ask it anyway. Are our legislators, are our elected officials equipped to deal with these conversations?

Christopher Koopman: No, not at all. I mean, when you have legislators that still pride themselves on the fact that they don’t use email, and the ones that do are having their emails printed off by their staff and read to them—this is not ageist; this is not a comment on anything other than the reality that exists. You’re seeing probably the most vocal, active participants at the federal level pushing AI policy—think Bernie Sanders. He introduced his own version of a “pace the frontier” bill that is exactly what you would expect from someone who is not a digital native trying to do what digital natives are telling him is the right thing. And the practical outcome of this is that his bill actually proposes criminalizing even just holding open source AI models on your computer. Being caught with them in Bernie Sanders’ world would be a crime, punished almost to the same degree as having CSAM, child pornography, on your computer.

Peter Lipsett: Having an open model?

Christopher Koopman: An open source model, yeah.

Peter Lipsett: Open source is essentially—I mean, this is a man who is a self-proclaimed communist who believes in sharing everything. How is an open source model not closer to that than something that’s patentable and closed off?

Christopher Koopman: Yeah. Because it cannot be controlled and it cannot be contained. It’s all about the accretion of power. So the idea being, well, if we just have—and this is not a new idea; the Biden administration talked about this—a few labs developing these frontier models, they’ll work very closely with the federal government, we can control it and pace it and make sure that it operates the way we want. Open source is an affront to that, right? Because it says anyone can build it, anyone can adapt it, anyone can use it without asking permission. That’s sort of the whole point of open source. So some of it is less about distributing the means of production and more about making sure that those who operate in this space are aligned and well controlled.

Peter Lipsett: Because at the end of the day, socialism and communism ultimately rely on an oligarchy that thinks they know best for everybody else. And that’s what they’re looking at here.

Christopher Koopman: Yeah. But back to your original question: there was a Senate debate in Massachusetts where one of the candidates was asked, “What AI do you use?” And he said, “I don’t know, whatever one comes with my phone.” And now he’s asking not just his constituents but the rest of America to trust him to write rules, and he doesn’t even know what model he’s using. Just to re-answer your question, I think our policymakers at every level of government are deeply ill-equipped to handle this, which is why I think you actually see the increased interest in pausing and slowing it down. Because if we’ve conditioned ourselves as a society to trust that the people we elect know everything they need to know and are operating with full and complete knowledge, then when a technology like this is running so fast and they don’t have that knowledge, it’s really easy for them to say, well, wait, we need to pause this until I understand what’s happening. And only once I understand it and can say what needs to be done will anything be done at all. So I do think that’s fueling the “pace the frontier,” the pause—those types of arguments in policy circles.

Peter Lipsett: Well, and it’s why groups like yours, and the traditional think tanks, so many of the organizations that are out there that can be supported, are so important. Because elected officials don’t know a lot about a lot of things, but they still make decisions on it. But there is a place for the influence and the conversation with groups like Abundance, American Enterprise Institute, AIER, think tanks out there, AFPI, that are actually looking at it and taking the time and informing them to get them up to speed. There just has to be a certain level of trust. I think part of the problem is the public is so confused and uninformed that they don’t know who to trust either. And I think we have room in our movement to try to build some trust there and be a fair arbiter on some of these things. Do you think that’s right?

Christopher Koopman: I hope so. I think so. And I think even to your point—the scariest stories benefit the frontier labs. This is all of them rushing to tell everyone about the latest hack. It’s in some ways cheap marketing. They’re just saying, look at how powerful my model is, it can destroy everything. And so I think you always have to keep the messenger in mind when you’re trying to discern the validity of the message. And I think groups like ours, who are not unbiased—I have a bias toward the future and a bias toward innovation, and I believe that history proves that allowing these things to develop freely leads to better outcomes than not—but at least in that way we can bring more voices to the conversation and ensure that people are at least being told enough, and then trust people that they can decide for themselves what is true and right.

Peter Lipsett: Yeah. All right. I wanna close with one more practical thing. Folks who are listening who are maybe board members of nonprofits or who run nonprofits, work at nonprofits, what are the questions they should be asking or things they should be doing to make sure that they are both leveraging AI internally in a good way but also creating the right safety vectors and guardrails to make sure they don’t accidentally step in it somehow, legally, compliance-wise, or just training a model on things that it shouldn’t know?

Christopher Koopman: Interesting. Yeah. So I think there’s a couple of things. And I’m sure a lot of the CEOs out there won’t like the questions I’m going to tell their board members to ask them. But right now, especially, we are up against a tidal wave of philanthropic capital that is going to be entering this space and overwhelmingly falling into groups that do not view the world the same way we do. And we will not outspend them. So we have to become more strategic and more effective at what we do. And I would say for board members—and some of this might be uncomfortable, especially if you’re not up to speed on how AI works—challenging orgs to leverage AI as much as possible in increasing their effectiveness is really important, because again, we’re going to be outspent and outstaffed, but it doesn’t mean we have to be outmaneuvered. So I think making sure that groups can expand their impact through the smart and strategic use of AI internally is really important. Really understanding how, not just using chatbots to edit your documents, but actually using it in your operations and in your day-to-day business, to ensure that not every solution becomes “we must hire a new person.” I think that’s one aspect. As we started with, AI is collapsing the price of knowledge and expertise and offering it at near-zero price to people, and if we aren’t taking advantage of that, we’re totally missing out.

I think the other aspect becomes a really important question for organizations to ensure that they aren’t stepping in it: what AI are you using, and what do you own? Because when you invite these companies into your organizations and into maybe the most confidential aspects of your org, what happens when you begin to rely on this? Imagine this on the other side. And I don’t mean to go too long on this, Peter, so I’m sorry, but think about the reliance that a lot of orgs have on Salesforce now. You can’t get out of Salesforce, right? You could imagine a world in which, through a series of well-intentioned but miscalculated steps internally, you become almost as beholden to an AI company as your development team is to Salesforce. And you can’t get out of it. The switching costs are too high. And now all of a sudden the question is, well, who are you working for? Who’s making the decisions here? This is where I think Alex Karp, Dr. Karp at Palantir, has talked a lot about sovereign AI: own your data, own your AI. This is the importance of open source, custom-built AI solutions, ensuring that the tools you have actually work for you in the long run and you don’t end up drinking seawater. Because it may feel like you’re quenching your thirst. It may feel like you’re being productive, but in the long run, you may actually do your organization a tremendous disservice in over-reliance on the wrong AI in the wrong areas.

I think the other thing for board members is, look, play with it yourself. Get up to speed with it. See how it really works in your own lives, in your own work, and I think you’ll begin to understand the tremendous impact this can potentially have. And then the last thing, and then I’ll leave you with this, on this question at least. The last thing I would say is that almost every issue, unfortunately, is becoming an AI issue. And those orgs that say, well, we don’t work on AI and we’re not covering that issue right now in our lane because we’re focused on X or Y or Z—if they’re not prepared to address the way in which AI is going to tremendously change their focus area, they’re going to end up playing catch-up rather than owning the space. Think education, fiscal policy. AI is going to dramatically change the way we think about taxes and fiscal policy. If there are policy orgs out there that say, well, no, we’re a fiscal org, we don’t cover AI—well, you may not care about AI, but AI certainly cares about you. It’s coming for your industry and it’s coming for your focus area. And if you’re not ready for it, you’re going to find yourself on your heels playing catch-up and being responsive rather than being proactive in owning the space before anybody else moves into it.

Peter Lipsett: I think that’s a really important point. I mean, just look at this conversation: we talked about policy issues of energy, of regulation, of trade with the chips, of education, of zoning, so many issues, not because AI is the hip thing and we just want to talk about it everywhere—although it is—but because it is part of everything now. It is part of the conversation and is going to be critical and change the way we do stuff.

Christopher Koopman: Yeah. It would be like thinking electricity is a fad.

Peter Lipsett: Yes, and I’m keen on electricity. It allowed us to have this great conversation, which we’ll draw to a close here because we could otherwise just keep talking. There are so many rabbit holes on all this. And it’s a conversation that’s gonna continue in all the different issues that we cover. So, Chris Koopman at the Abundance Institute, keep doing the good work, keep us optimistic and rational, and help those poor elected officials understand what the heck is going on out there.

Christopher Koopman: Thank you. I’ll do my best.

AI and Its Relevance to Philanthropy

Artifical Intelligence is here and everywhere. That includes the nonprofit and philanthropic worlds. In this episode, Chris Koopman of the Abundance Institute offers us a perspective on the problems with the pessimism that is seemingly everywhere, how AI is uprooting old ideas, whether AI is a new Renaissance or a new Industrial Revolution, why everyone hates data centers now, the many ways public policy philanthropy intersects with AI, and how nonprofits should think about protecting themselves in this new era.

Note: This transcript was generated and cleaned with AI

Peter Lipsett: One of the reasons I like being a conservative on the free market side of the ideological spectrum is that we argue that prosperity is available to everyone. In a free society, people have the agency to have a better state for themselves through their labor, through their ideas, through problem solving with creative ideas for other people. You know, one way or another, there’s a pathway for everyone to get to their desired level of prosperity, which could be something different for everybody.

That’s the real essence of the American dream, right? That people get to their desired level of stability and comfort. And here we are in this technological era with artificial intelligence taking over the conversation. And in a lot of quarters, you hear that that’s going to be great for bringing about prosperity. But there’s this other faction, sometimes a louder faction, preaching that there is nothing but doom and gloom and sorrow ahead for all of us as our robot overlords take over.

So what kind of future are we really looking at? And why on a philanthropy podcast are we talking about artificial intelligence? Is that just because everybody else is? Well, no, we’re going to get into that: why this is an important topic for donors, for nonprofits, for everybody. It does all tie together as this AI debate is raging and the broader technological changes take place, because it’s going to affect everything.

And so joining me to talk about that is my friend Chris Koopman. Chris runs the Abundance Institute, which is focused on creating space for emerging technologies to grow and thrive and have a chance to reach their full potential. We’ll talk more about what that really means as we go on. Chris started his career at the Mercatus Center at George Mason University and went on to be executive director of the Center for Growth and Opportunity at Utah State. Chris, great to have you with me.

Christopher Koopman: Thank you, Peter.

Peter Lipsett: So there’s a lot of techno doom and gloom right now. Social media was already ruining our lives. And then data centers came out of nowhere to be this eater of farmland and all of our hopes and dreams. And now AI is gonna turn us all into paperclips. But this doom saying is not new, right?

Christopher Koopman: No, not at all. I think the introduction of every new technology has come with a story of ruin and disaster. It doesn’t matter how big or small the technology is or was; this has always been a part of the conversation. There are always people who refuse to see the upside and focus on the downside. My guess—and I wasn’t there—is that when that first caveman brought fire back to the cave, there was someone who tried to chase him out. And we see this time and again in history.

I’ll give you a couple of examples that I think about a lot and think are really illustrative—it’s okay to be afraid, but what we do about that fear is very, very important. The bicycle. Turns out there was a wild societal panic over the bicycle, for a couple of reasons. One, there were actually studies done on a phenomenon they thought would occur called “bicycle face”: that people traveling at such extreme speeds would be stuck in a constant state of shock on their face. This is a real thing, the bicycle face. The telephone—people were worried about the telephone.

Peter Lipsett: Even Alexander Graham Bell was worried about the telephone, right? Didn’t he famously not actually use it?

Christopher Koopman: Yes. John Philip Sousa, the king of the American march—every kid who’s ever played in marching band has probably played a Sousa march—was against recorded music. There’s always been someone who looks at every new thing and can only help but think about the worst-case scenarios. But it’s our job, as people who work in policy or in shaping culture, to make sure that we don’t let that fear and pessimism drive us to make bad decisions. We were talking about this before, Peter: the doom and gloom had already infected the conversation around technology and the internet in the context of social media. And now it’s just bled over to AI.

But I think about the radio as a great example of this. There were stories when the radio first came out. We have this documented at a project we help support called the Pessimists Archive, a project we’ve been working on with and through the Abundance Institute. It documents all these historical panics. Louis Anslow, who runs it, has done an amazing job. He’s like the archivist of global historical technopanics. And he’s pulled these examples that are funny now, right? I think it’s always interesting to think about the context of the conversations we’re having today. How will we look back on them in 50 or 75 years and laugh at ourselves?

But the radio: there was a story of a headmaster at a boarding school in New England who said, basically, that the vitality of his young men was totally destroyed by the radio. All they want to do in the evening is lay around and listen to stories of cowboys and Indians, fighting these Western fights, and the idea was that they were losing their creativity, their vitality. It was melting their brains. It was destroying their ability to read or to write or to be creative. Almost every technology that you can point to, even ones we don’t think of as anything more than mundane today, sort of all had this pessimist technopanic moment about them. Even artificial ice—even ice machines, Peter. There was a real challenge among some people in accepting the idea that we could just make ice in our homes.

Peter Lipsett: You know, change is tough, right? And any change—it’s not always just daisies and sunshine. There are always downsides. There’s gonna be disruptions. We saw each other last week at the Philanthropy Roundtable conference, and you were speaking about some of these issues, and one of the things you said really struck me: you said we shouldn’t confuse churn with ruin. Can you unpack that for me?

Christopher Koopman: Yeah. So I think, as people who have studied or understand the market process, there’s a concept called creative destruction. This is how things get better, right? The old way works until a new way comes along, and then there’s a transition from the old way to the new way. There are costs to that transition, right? Think of a taxi driver circa 2012, the advent of Uber. It dramatically changed your life, but in the long run I think it has made a lot of other people’s lives better. And I think what happens in the AI conversation is many people have lost the plot, and they’re confusing creative destruction for permanent economic ruin. They’re saying the way things are right now is the way things will always be going forward, and history just doesn’t prove that out.

There will be transition costs in the advent of every new technology, every paradigm shift. There are going to be costs to that transition that are borne by real people, and how we respond to those is really important. But we shouldn’t confuse the churn, or the creative destruction caused by this transition, with anything more than that. And I think to begin to try to project out that humans will just live in this ruinous world as a result of the advent of AI is totally missing the opportunity that’s before us.

Peter Lipsett: Yeah. You know, one of the things I grapple with in the whole AI conversation is this Sam Altman sense of, like, everyone’s gonna be richer, we’re all gonna be wealthier. And I have not been able to mentally make the jump between being able to use ChatGPT really well and being able to lie around all day like I won the lottery. Are you able to connect those dots for me? How are we all gonna get wealthier from this?

Christopher Koopman: Yeah. So—not to pick on Sam—but I wrote a piece in Fortune recently about this. I think his read of economic history is a bit off. He spoke in front of the UN several weeks ago and said something to the effect of: we have two choices before us, we can have a new Renaissance or a new Industrial Revolution. And he paints the Renaissance as this beautiful, serene time when everyone was doing great, and the Industrial Revolution as this horrible time when people were not doing great. My argument against that is that the Industrial Revolution did more to lift people out of poverty than the Renaissance ever did. The Renaissance was a time of patrons and massive accumulation of wealth among a few, and a lot of pain and strife among the many. And I think the Industrial Revolution changed that, right? You look at all of the stats that we like to brag about, the benefits of modern society, and all of that was essentially the direct result of the Industrial Revolution. Infant mortality rates, life spans, poverty rates—all of this was dramatically improved by the Industrial Revolution. I think we’re on the dawn of a new industrial revolution, and trying to say that we must stave that off is a problem in and of itself.

But to answer your question, I think the idea that we’ll all be living in this utopian future in which we live permanent lives like infants or toddlers, where we just play games all day and lay around and everything’s provided for us—I don’t think that will ever occur, because that’s just not how the world works. But what I do think is that you can have a tremendous amount of opportunity and prosperity generated that is completely disconnected from this utopian idea of, well, you can just go be a poet or live the life you want and not worry about the economic implications of your choices.

And this was something I was talking about last week as a part of the conversation: the collapsing price of expertise, right? I think you said you have a hard time understanding how using ChatGPT will bring about your utopian society, as Sam Altman promised you. I don’t think that will occur, but what will occur is that you and tens of millions, if not hundreds of millions, of people, not just in America but across the world, will be given access to information and expertise that they had never been able to grasp at any point in history. Information and expertise that was only reserved for the richest and most elite among us will now become readily available at almost zero cost. Some may say that’s utopian in a different sense. But that, to me, I think, is the impact that a tool like AI and these large language models will have on our ability to live lives of meaning and purpose and find flourishing and prosperity that don’t get us into this world where, don’t worry, you can lay around all day and everything will be provided.

Peter Lipsett: There will be an abundance, which brings us kind of to what you do all day. I mean, we shouldn’t just listen to you because you’re young and hip looking. You actually run an organization that thinks about these problems as a research institution, a nonprofit institution, the Abundance Institute. And I mentioned your mission at the top of the show: create space for emerging technologies to grow, thrive, and have a chance to reach their full potential. It sounds like you’re fully focused on making the optimistic case for technology change. Is that right? Are you only looking at the bright side?

Christopher Koopman: I’m unapologetically optimistic, but I’m not Pollyannaish. So I’ll say, the origin story of the Abundance Institute: we launched just about three years ago. We’ll be three years old in April of ’27, so we’re like two and a half, I guess, and change. But really, the entire origin story of Abundance was that I’ve spent my entire career working in policy and at the intersection of technology and regulation. And going back about three years, I was watching this conversation get away from us really quick. As people who advocate for free markets and are optimistic and believe in the power of human agency, we were losing the narrative very, very quickly. And you can go back in time to those early Senate hearings that Sam Altman was a part of, in which we were told essentially that this technology will destroy us unless Congress gives Sam Altman the power to write the rules, and he’ll keep us all safe forever. That was a moment for me where I was like, that cannot be the only option available. And if we want to see all of the right numbers going up and to the right on all the charts we care about, we have to reclaim the narrative.

The launch of Abundance—as you said, the idea of it at a top level is helping emerging technology grow and thrive and reach its full potential to support human flourishing. This is all about the people who benefit from these technological changes. As a part of that, there’s a tendency, as we already mentioned, for the advent of any technology to be met almost immediately with pessimism, doom and gloom, and worst-case scenarios. What if we helped people understand a different shade of these technologies? The way I view Abundance’s role in the broader conversation is sort of twofold. One, as best we can, we want to shift that conversation away from pessimism and toward optimism. As I said, it’s not utopian, it’s not Pollyannaish. It’s just a belief that what history has taught us is that almost every technological advance we’ve undertaken has led to tremendous benefit to humanity.

The other part of this is the way that the rules are made: the people who have a seat at the table are the people who exist today, right? So when you have a conversation about the regulation of AI, who do you invite to the room? You invite the frontier labs who are building the technology today. The people who are helping shape that conversation are people who exist in the room today. And oftentimes what we see in the history of the regulation of technology is we pass a bunch of well-intended policies in an effort to make things go right, and almost immediately realize we screwed it up. The EU AI Act is a perfect example of this. They wrote all of these AI rules very early on, and before they even went into effect, they realized they were unimplementable. You couldn’t do it. They were totally wrong.

Peter Lipsett: And this is why there are no labs in the European Union, right? Because they just can’t do it. They regulated themselves out.

Christopher Koopman: Yes, this is why AI will not be a European thing. But a part of this, and what we try to do at Abundance, is in some ways present-proof the future. It’s ensuring, as best we can, that we represent a point of view with a bias toward the future in these conversations that are happening in the present. So in some ways the way I would describe us is that we’re representing founders who haven’t founded anything, industries that don’t yet exist, companies that haven’t even started, the next generation, to make sure that we’re writing rules today that allow for subsequent iterations in all of this that wouldn’t be possible if we just wrote rules today with the incentives and the perspectives of the people today making them.

Peter Lipsett: Yeah. So it’d be similar to any public policy shop, except you just happen to be working on the sexy, hot issues of the day that will determine the future issues.

Christopher Koopman: I guess I got lucky. Yeah, when we launched the org focused on artificial intelligence and energy three years ago, did I imagine that they would be this issue today? Sure. Yeah, totally. I knew it. But I think also, a little bit of what makes this different is that we don’t approach these issues from a traditional think tank perspective. If you were to look at what we produce and how we produce it, we overinvest in our video and our storytelling content, and we underinvest in your traditional white papers. Just knowing how the conversation is unfolding, and the way in which politics and policy are downstream of the cultural conversations, the more we can push upstream, the more likely we are to have an outsized impact on those downstream conversations.

Peter Lipsett: Yeah. No, that makes sense. Well, let’s take an example of an issue. Let’s set AI aside for a moment, because AI is a big thing and we could get into what the AI debate really is, because I know you’ve broken that into various different pieces as well. But the data center debate is another thing that you work on. It’s a little bit more traditional public policy, one that people can get their hands around, I think. That just completely went off the rails recently, right? Nobody cared about data centers. Data centers were just kind of a thing. And then data centers went from a thing that’s out there to this absolute boogeyman. And here we are recording a few days before the midterm elections, and it’s become an issue in a way that I don’t think anybody, even back in March, would have said, “That’s going to be a driver in the elections.” What happened there? And let’s take for a second that data centers are neither good nor bad. They just are. They’re a part of this process if we want innovation, right? So it’s tough to really plant your flag on either side if you’re looking at it rationally, in my opinion. But where did this go off the rails? What should people who care about public policy, particularly future-oriented public policy, be thinking about in that debate?

Christopher Koopman: Yeah. So I guess the question is, why has America become obsessed with data centers the moment it has? Why this issue now? And I think—you didn’t want to talk about AI, but you cannot have this conversation without understanding the AI conversation. Here’s how I’ll put it. I would say there are three distinct groups of people who have all converged on this issue.

In the one group, you have, let’s call them, traditional enviro groups that are worried about resource use, worried about increasing energy consumption, worried about the impacts on the environment. They have always been worried about data centers. That’s very, very clear to me. Then you have two groups who enter the chat new. But first, I would actually bucket in that same group the NIMBYs, the anti-development, not-in-my-backyard type folks who just generally don’t want these things built in their communities and their cities and their counties and their states. I would put them all in that same group because the way they’re approaching the issue really is the same, right? These are a problem, we don’t want them. And it doesn’t matter the degree to which you can push back. Kevin O’Leary was trying to build a data center out in the Mountain West, and there was a stat from an environmental sciences professor that said the heat generated from it would be the equivalent of dropping twenty-nine Hiroshimas. Which is sort of like, well, yeah, but look at that in the context of cities themselves; they’re also consuming a lot of energy and generating a lot of heat. It’s a very strange thing.

So, okay, you’ve got the NIMBYs and the environmentalists who are just anti-development of these things. I think two groups enter the chat late, and it explains why this moment now. The one group, I would say, are everyday Americans who have heard a very troubling story from the frontier labs developing the AI technology: that they are building a future in which Americans don’t have a place. They’re telling you there’s like a 90% chance you’re gonna lose your job. There’s like a 50% chance that there are zero-day exploits in all of the banking system and you’re not gonna have any money when the AI is powerful enough, because it’s just gonna hack the system, and there goes all your money and the financial system is ruined. And by the way, there might be a 10% chance we kill you, right? If you’re very concerned about this technology, and you find yourself dismayed by the lack of urgency that policymakers had maybe taken to regulating the development of the technology, but what you find out is there’s physical infrastructure necessary to build out the next frontier of these models and it exists in data centers—well, you’ve got your perfect villain for the AI future that you’ve been sold, in my mind, poorly. And so you fight the data centers. You don’t like AI, so choke the infrastructure. You can’t stop software, but what you can stop are the pipes and tubes and chips that make the software work.

The other group—and some of this depends on what conversation you’re in; people respond to this differently—but there are foreign adversaries who have amplified and fomented the pre-existing distrust that people have for this technology to make this the moment that it has become. So you have people who believe it, and now you have foreign adversaries who are making it seem like everyone believes it, and you’re stupid if you don’t, or you’re naive if you don’t. So I do think this moment is a perfect storm of classic environmentalist, NIMBY, anti-development folks mixed with people who are deeply concerned about AI, all being spun up into a frenzy by living terminally online and believing that everyone also believes that this technology and the physical infrastructure is the end of the world. That’s really the way I see the data center moment playing out over the last six months. It’s only gonna get worse, not better. This is where my optimism starts to fade a little bit. The conversation really took heat, as you said, in May, which is after most state legislative sessions had already gaveled out. So the 2027 legislative sessions are the first opportunity that most state legislators have to do something. And now many of them have just come out of an election in which they believe, rightly or wrongly, that they’ve been given a mandate to do something about data centers. So I think the 2027 state legislative sessions are going to be a frenzy of policy proposals around data centers.

Peter Lipsett: I haven’t thought about that, and I think you’re absolutely right. And that is certainly something for donors out there who invest in public policy to watch. Because, well, let’s get into this. When we say the AI conversation, what I heard you say the other day was really that it breaks up into a bunch of different buckets. And I think that’s actually really relevant for the donors out there listening, because the AI labs are not nonprofits. I think OpenAI is no longer a nonprofit, even though it once was. So AI is moving on its own for-profit schedule, but there are all these other things going on where there is actually a place for philanthropy to engage. Talk about those different pillars of engagement.

Christopher Koopman: Yeah. And I especially think, for your donor community and the people who are advocating from a free market perspective on a lot of these issues, there are a bunch of these layers of this cake where these are home games for us. These are not foreign issues. But I think the best mental model to understand this is one that was popularized by Jensen Huang at NVIDIA: the idea that AI is actually a five-layer cake. It’s not one issue. And oftentimes I think people just bunch them all together as a single issue, and then it gets very muddied and convoluted as you try to sort out the various threads of the conversation. But you can think about these as stacking on top of each other.

At the base, you have energy, right? You don’t get anything if you don’t have energy. That is the base layer for everything in our economy, but in AI in particular, it’s really important. The next level up from this would be chips. How do we actually make silicon, how do we turn sand into thinking machines? That’s the chips that do that. You take the chips, you combine them with energy, and you get the digital infrastructure, the data centers, which is how you actually train and run these models. So that’s the third layer: data centers. Then on top of that, you get the models themselves, the largest versions of Claude and ChatGPT and Grok, the actual models that are repurposed in a bunch of ways. And at the very top, you have applications, which are the direct and discrete uses of these technologies in very specific use cases. So applications would be healthcare AI apps, education AI apps—Google Maps is an AI application.

And I think once you start to think about them in five layers, the questions and the answers and the right venue to have them become a little bit clearer. I don’t think data centers should be a national conversation. It’s a state and local conversation. This is just the building out of infrastructure. You wouldn’t want Congress to pass a bill on the location of Amazon distribution centers or Walmarts or office buildings. That’s a state and local conversation. The energy conversation is a mix of federal and state. The chips is really a federal conversation, though some of it’s state: where do we build them? If we build them in the United States, where are we manufacturing them? That becomes a state conversation. The models: we should have a uniform, harmonious set of rules for models across all 50 states. A model being built in California shouldn’t have to comply with 49 other versions of how that model should operate in each state. Not only is that an affront to interstate commerce, but it’s really hard for the next generation of founders and startups to grow and thrive. You know who can handle that? The frontier labs can handle that. It’s the next gen, the people who are going to unseat them from their perches at the frontier, who are really going to have a tough time.

And then the applications. The applications really are a state and local issue. There’s some stuff that might be federal, but for the most part, how AI is used in the classroom, that’s a state conversation. So once you start to break these down into layers, you start to understand, okay, what are the tools we have, what are the questions we’re asking, and who are the people we’re actually supposed to be bringing these questions to? Because right now a lot of it is all being pushed, rightly or wrongly, to the federal level. And then when the federal government doesn’t do anything, they say, well, the states must. Well, just because the federal government won’t do anything doesn’t mean the states are the right venue to write these rules. We shouldn’t just pull on any lever available. And I think breaking them down into those component parts helps you better understand how to tackle each one of these questions as they emerge in the conversation.

Peter Lipsett: And they’re all a reminder that there are policy fights in all of them. The chips one is probably the weakest of them. The models, as you say, are kind of at the federal level. But energy—there are plenty of people who listen to this podcast who have been in the energy fight for a long time. It’s just taking on a different cadence. And so it’s also remembering the school fight you referenced at the end with the application layer. Some of these fights that people have been in for a long time just now have this different valence, but it doesn’t make the fight different. It’s just that the issue is a little different. And frankly, that’s a mix of legacy organizations working on these things as well as new groups like Abundance.

Christopher Koopman: And it’s completely changed. For a long time, the conversation over nuclear, small modular reactors, wasn’t something people were talking about 20 years ago as a real hot policy issue. And yet now you’re seeing half a dozen, a dozen states really taking an active role in pushing both the federal government and themselves to see this technology deployed. You don’t even get small modular reactors if you don’t have the labs, the hyperscalers who are pushing for demand and creating a market for technology that for a long time has been viewed as not the most economic. So it’s actually bringing up old fights. We’re debating and talking about building nuclear again, which we haven’t done in 50 years. And all of this is a direct result, I think, of the moment being created by these frontier labs.

Peter Lipsett: I know the answer to this, but I’m gonna ask it anyway. Are our legislators, are our elected officials equipped to deal with these conversations?

Christopher Koopman: No, not at all. I mean, when you have legislators that still pride themselves on the fact that they don’t use email, and the ones that do are having their emails printed off by their staff and read to them—this is not ageist; this is not a comment on anything other than the reality that exists. You’re seeing probably the most vocal, active participants at the federal level pushing AI policy—think Bernie Sanders. He introduced his own version of a “pace the frontier” bill that is exactly what you would expect from someone who is not a digital native trying to do what digital natives are telling him is the right thing. And the practical outcome of this is that his bill actually proposes criminalizing even just holding open source AI models on your computer. Being caught with them in Bernie Sanders’ world would be a crime, punished almost to the same degree as having CSAM, child pornography, on your computer.

Peter Lipsett: Having an open model?

Christopher Koopman: An open source model, yeah.

Peter Lipsett: Open source is essentially—I mean, this is a man who is a self-proclaimed communist who believes in sharing everything. How is an open source model not closer to that than something that’s patentable and closed off?

Christopher Koopman: Yeah. Because it cannot be controlled and it cannot be contained. It’s all about the accretion of power. So the idea being, well, if we just have—and this is not a new idea; the Biden administration talked about this—a few labs developing these frontier models, they’ll work very closely with the federal government, we can control it and pace it and make sure that it operates the way we want. Open source is an affront to that, right? Because it says anyone can build it, anyone can adapt it, anyone can use it without asking permission. That’s sort of the whole point of open source. So some of it is less about distributing the means of production and more about making sure that those who operate in this space are aligned and well controlled.

Peter Lipsett: Because at the end of the day, socialism and communism ultimately rely on an oligarchy that thinks they know best for everybody else. And that’s what they’re looking at here.

Christopher Koopman: Yeah. But back to your original question: there was a Senate debate in Massachusetts where one of the candidates was asked, “What AI do you use?” And he said, “I don’t know, whatever one comes with my phone.” And now he’s asking not just his constituents but the rest of America to trust him to write rules, and he doesn’t even know what model he’s using. Just to re-answer your question, I think our policymakers at every level of government are deeply ill-equipped to handle this, which is why I think you actually see the increased interest in pausing and slowing it down. Because if we’ve conditioned ourselves as a society to trust that the people we elect know everything they need to know and are operating with full and complete knowledge, then when a technology like this is running so fast and they don’t have that knowledge, it’s really easy for them to say, well, wait, we need to pause this until I understand what’s happening. And only once I understand it and can say what needs to be done will anything be done at all. So I do think that’s fueling the “pace the frontier,” the pause—those types of arguments in policy circles.

Peter Lipsett: Well, and it’s why groups like yours, and the traditional think tanks, so many of the organizations that are out there that can be supported, are so important. Because elected officials don’t know a lot about a lot of things, but they still make decisions on it. But there is a place for the influence and the conversation with groups like Abundance, American Enterprise Institute, AIER, think tanks out there, AFPI, that are actually looking at it and taking the time and informing them to get them up to speed. There just has to be a certain level of trust. I think part of the problem is the public is so confused and uninformed that they don’t know who to trust either. And I think we have room in our movement to try to build some trust there and be a fair arbiter on some of these things. Do you think that’s right?

Christopher Koopman: I hope so. I think so. And I think even to your point—the scariest stories benefit the frontier labs. This is all of them rushing to tell everyone about the latest hack. It’s in some ways cheap marketing. They’re just saying, look at how powerful my model is, it can destroy everything. And so I think you always have to keep the messenger in mind when you’re trying to discern the validity of the message. And I think groups like ours, who are not unbiased—I have a bias toward the future and a bias toward innovation, and I believe that history proves that allowing these things to develop freely leads to better outcomes than not—but at least in that way we can bring more voices to the conversation and ensure that people are at least being told enough, and then trust people that they can decide for themselves what is true and right.

Peter Lipsett: Yeah. All right. I wanna close with one more practical thing. Folks who are listening who are maybe board members of nonprofits or who run nonprofits, work at nonprofits, what are the questions they should be asking or things they should be doing to make sure that they are both leveraging AI internally in a good way but also creating the right safety vectors and guardrails to make sure they don’t accidentally step in it somehow, legally, compliance-wise, or just training a model on things that it shouldn’t know?

Christopher Koopman: Interesting. Yeah. So I think there’s a couple of things. And I’m sure a lot of the CEOs out there won’t like the questions I’m going to tell their board members to ask them. But right now, especially, we are up against a tidal wave of philanthropic capital that is going to be entering this space and overwhelmingly falling into groups that do not view the world the same way we do. And we will not outspend them. So we have to become more strategic and more effective at what we do. And I would say for board members—and some of this might be uncomfortable, especially if you’re not up to speed on how AI works—challenging orgs to leverage AI as much as possible in increasing their effectiveness is really important, because again, we’re going to be outspent and outstaffed, but it doesn’t mean we have to be outmaneuvered. So I think making sure that groups can expand their impact through the smart and strategic use of AI internally is really important. Really understanding how, not just using chatbots to edit your documents, but actually using it in your operations and in your day-to-day business, to ensure that not every solution becomes “we must hire a new person.” I think that’s one aspect. As we started with, AI is collapsing the price of knowledge and expertise and offering it at near-zero price to people, and if we aren’t taking advantage of that, we’re totally missing out.

I think the other aspect becomes a really important question for organizations to ensure that they aren’t stepping in it: what AI are you using, and what do you own? Because when you invite these companies into your organizations and into maybe the most confidential aspects of your org, what happens when you begin to rely on this? Imagine this on the other side. And I don’t mean to go too long on this, Peter, so I’m sorry, but think about the reliance that a lot of orgs have on Salesforce now. You can’t get out of Salesforce, right? You could imagine a world in which, through a series of well-intentioned but miscalculated steps internally, you become almost as beholden to an AI company as your development team is to Salesforce. And you can’t get out of it. The switching costs are too high. And now all of a sudden the question is, well, who are you working for? Who’s making the decisions here? This is where I think Alex Karp, Dr. Karp at Palantir, has talked a lot about sovereign AI: own your data, own your AI. This is the importance of open source, custom-built AI solutions, ensuring that the tools you have actually work for you in the long run and you don’t end up drinking seawater. Because it may feel like you’re quenching your thirst. It may feel like you’re being productive, but in the long run, you may actually do your organization a tremendous disservice in over-reliance on the wrong AI in the wrong areas.

I think the other thing for board members is, look, play with it yourself. Get up to speed with it. See how it really works in your own lives, in your own work, and I think you’ll begin to understand the tremendous impact this can potentially have. And then the last thing, and then I’ll leave you with this, on this question at least. The last thing I would say is that almost every issue, unfortunately, is becoming an AI issue. And those orgs that say, well, we don’t work on AI and we’re not covering that issue right now in our lane because we’re focused on X or Y or Z—if they’re not prepared to address the way in which AI is going to tremendously change their focus area, they’re going to end up playing catch-up rather than owning the space. Think education, fiscal policy. AI is going to dramatically change the way we think about taxes and fiscal policy. If there are policy orgs out there that say, well, no, we’re a fiscal org, we don’t cover AI—well, you may not care about AI, but AI certainly cares about you. It’s coming for your industry and it’s coming for your focus area. And if you’re not ready for it, you’re going to find yourself on your heels playing catch-up and being responsive rather than being proactive in owning the space before anybody else moves into it.

Peter Lipsett: I think that’s a really important point. I mean, just look at this conversation: we talked about policy issues of energy, of regulation, of trade with the chips, of education, of zoning, so many issues, not because AI is the hip thing and we just want to talk about it everywhere—although it is—but because it is part of everything now. It is part of the conversation and is going to be critical and change the way we do stuff.

Christopher Koopman: Yeah. It would be like thinking electricity is a fad.

Peter Lipsett: Yes, and I’m keen on electricity. It allowed us to have this great conversation, which we’ll draw to a close here because we could otherwise just keep talking. There are so many rabbit holes on all this. And it’s a conversation that’s gonna continue in all the different issues that we cover. So, Chris Koopman at the Abundance Institute, keep doing the good work, keep us optimistic and rational, and help those poor elected officials understand what the heck is going on out there.

Christopher Koopman: Thank you. I’ll do my best.

More Podcasts

View All