Loading summary
Transcript54 lines
- [00:00]
A
Out. Our think tank series continues with the new breed, the insurgents, the barbarians at the gate. Caleb Watney, co founder of the Institute for Progress and Max Bodak, executive vice president at the foundation for American innovation, IFP and FAI respectively. Welcome to ChinaTalk. All right, Caleb, you wanted us to start with the prompt applied think tank versus white paper mill. Discuss,
- [00:32]
B
discuss. So when trying to understand what do think tanks even do? It's funny when I'm trying to like explain to my in laws, like what even is the think tank, what do they do? It's a very niche cottage industry that is, you know, very particular to Washington D.C. but it helps as with a lot of different industry analysis to start with what is the goal, what is the institution trying to achieve? And if you go back in time, you know, some of the first think tanks that got off the ground were places like Brookings and Carnegie and a lot of these places like they' inherent conception of themselves, what it is they were trying to do was a university without students. There was an important role for sort of policy relevant research that was like taking one step further in the applied direction than the more traditional pure academia route. And fundamentally they were playing an information role. They were trying to distill and make applicable recommendations across a bright swath of federal agencies. And they continued to do that role. Different voices can think about how well or how poorly, but that is fundamentally what they're trying to do. I think a lot of the newer organization, certainly I would put IFP in this category and Max, I assume you'd put FAI in this category as well. But like the goal at the end of the day is to actually change how policy is operating. We talk about this a lot internally as counterfactual policy impact. Is the world and the laws and regulations that folks are dealing with actually different because of the set of actions that people at the think tank took. And I think if that is your fundamental reason for being, it changes a lot about the way that you're structured, the set of like impact criteria you're using, how you're structured, what incentives you're sort of putting on on folks. And so we at the end of the day don't think the job is done until like the policy or the regulation is actually different. If we've published a cool paper and people said well done, like you know, that's step one of 20. And so I think there's a lot of things downstream of that. But just like what is the point of the organization, what is it trying to achieve is I think the first question you have to ask.
- [02:24]
A
So 1940s Brookings seems like a bit of a, bit of an easy target. We did have over the course of the 90s, 2000s and 2010s this idea of a think and do tank, which was trying to sort of square the circle in a slightly different way. Why don't you go a little deeper on like what you think the sort of like organizational cruft of even the slightly less pedigreed think tanks still have that weigh their staffers and their, you know, that weigh their scholars down from having the impact that they could have otherwise.
- [03:01]
B
I think there's a couple of ways you could look at this and actually maybe the particular thing I want to hone in now is the way in which like funder incentives end up being downstream or upstream of also the kind of work that you can do. So in my like day job at ifp, one of the main areas I work on is science policy and especially meta science. How do the incentives of the way that we fund scientists change the kind of work that they do? And I think there's actually a ton of interesting overlaps between the problems in the way that we fund science and the problem in the way that we fund think tanks. One of the primary culprits is project based funding. And so this is, I think like the primary way the vast majority of think tanks end up being funded is you enter an agreement, usually with a foundation or a philanthropist in some way and you've agreed ahead of time to embark on some project, a well defined, well scoped project, ahead of time, and you articulate a theory of what it is you're going to do ahead of time. You're going to write some paper, you're going to do a number of OP eds, you're going to host some webinars or a conference or something. But once you've agreed that those are the ways of measuring impact, you're now basically just in a box checking exercise and you can basically sell well done once you've written the op ed. And a lot of this is actually like policy change is quite hard to measure. And there's like what economists would say, a principal agent problem between the funder and the fundee. Our funding model, at least at IFP is pretty different. We try to rely primarily on general purpose funding that is just like organization wide. So we try to have a smaller number of high trust relationships with funders and then try to sell people on basically the larger policy vision of ifp what it is we're trying to do. And like then we actually share behind the scenes, kind of like case studies of from beginning to end. How did this policy actually end up being changed? Because of kind of the work we took. And we try to make sure that we're only measuring things that actually happen in the world and not just like the number of specific inputs.
- [04:54]
A
Five minutes in, we're already talking about money. Max. You guys have a slightly different philosophy when it comes to this sort of thing. How do you, how do you approach the money question?
- [05:07]
C
Zach and I, Zach Graves, our executive director, we like to joke that, you know, FAI takes money from everybody so we can't be bought. And that is actually true. I mean, when you look at like the average check size, you know, for us we have around a little over 200 donors, average check size around 150 to 175 depending on the year. And so that does mean we can walk away from anyone when we, you know, happen to, you know, disagree on a particular policy choice or we are, you know, just drifting in different directions and don't have the same mission alignment. You know, most of a large majority of our funding is philanthropic. So we work with right of center foundations, center left foundations, some sort of like heterodox ones that don't fit neatly into an ideological box. And then certain corporations, and particularly for event sponsorships, like if you go look through our gala last year, you can check off all the Frontier labs and also check out lots of philanthropists and some individuals who came and supported us and we're grateful for that. And then we also worked with a lot of individuals and VCs. One thing we're really excited about is that Silicon Valley is generating a lot of new wealth. And I think we think that's going to pretty fundamentally destabilize the philanthropic ecosystem. There's a lot of folks who are going to come online in the next two to five years who are willing to do things a little bit differently from a lot of institutional philanthropy and are skeptical that whether it's these hundred year long foundations that just change and remake themselves over and over again, or it's the sort of more time bound sunset clauses that particularly conservative philanthropists tend to put on their foundations. You know, I suspect that a lot of them are going to be really excited to, you know, as Caleb was talking about in the science funder landscape, right. That you know, try to invest in, you know, more time bound projects, you know, invest in teams and less sort of well scoped white paper mill operations. So yeah, so we're excited to do a lot of that donor education. I think there's a lot of shifts, you know, ground level shifts going on there that will destabilize the sort of traditional New York finance or Texas oil and gas sources of wealth that tend to dissipate proportionally dominate the philanthropic ecosystem in Washington, DC.
- [07:14]
A
Max, you mentioned gallows. And one snarky line I received from the peanut gallery was fai's theory of change is throw great parties and know some guys, while Caleb and IFP's theory of chains is just get the good facts and scream them into the abyss. Let's start with you, Max. How does, how do, like, how does FAI think about sort of interpersonal relationships in Washington and how and why they're useful for making change happen?
- [07:48]
C
Yeah. Well, if you read this profile in Inc. Magazine from last year, we were, you know, dubbed the, the rave throwing think tank Shaping the tech. Right. And you know, I think it's a bit tongue in cheek, but there's an element of truth there too. You know, there's a lot of events in D.C. but there's very few good parties. And so if you become an organization that's known for throwing great parties, people will actually make time to come to your events. Even, you know, they're busy and over scheduled and you're hosting them, you know, sort of in far northeast D.C. rather than downtown next to the White House or next to Capitol Hill or what have you. But our theory of change is, you know, in some ways very similar to the work that Caleb and IFP do. I think, you know, our leadership teams come from very similar, you know, intellectual backgrounds and educational backgrounds. And you know, many of us have spent times at other think tanks where we were just really frustrated by the lack of real work that would happen. Right. Or the, again, the sort of, you know, getting a gold star once you've completed a white paper and run it through a big institutional review. And you know, that's not really the, you know, if that's the point of your job, if that's why you're ultimately drawing a salary, you know, then that's great. You know, you'll do well in your performance review, you'll keep the money train flowing. But, you know, I think, I think we were also very impact oriented and impact obsessed in the way that Caleb was discussing. You know, we also, we're big believers in doing the homework or doing other people's homework rather. We think that to, you know, in some ways this is all downstream of Tom Khalil's lessons in you know, kind of how to be a policy entrepreneur. But, you know, most people don't follow up and most people that you want to influence particularly are too busy to do, do the work that you want them to do. And so if you can make it as easy as possible for them to get a decision over the finish line, you have a much higher probability of, you know, you know, making the policy change that you'd like to see in the world. Right. And so we're big believers in not just, you know, publishing the research or you know, even having some snarky tweet threads or what, or throwing a great party, but also just doing as much work as we can behind the scenes to help the folks that, you know, have some decision making power, get the stakeholders in line and get all, get all their ducks in the row such that you can have that counterfactual policy impact.
- [10:00]
A
One of the sort of like interesting meta critiques of this like, think tank as an extension of people who are too busy in the executive and legislative branches to do their work. It actually comes from something that Kevin wrote about of this like golden age of bureaucracies, when you had like a forestry Bureau in the 1920s, which like actually had the best forestry experts in the world and they were uncorruptible and like very proud of their work. So, you know, is the lane which you guys are filling of kind of like being that sort of like homework helper grease, like uncorruptible, like greaser of the wheels that's working towards kind of like net positive things, not just like one company or one industry's interest, like fundamentally something that is broken about Washington and is like the actual more structural fix to just like get these bureaucracies to be as good as they like somehow were in the 1920s, such that, you know, you don't, you don't. The need for what you guys are our offering isn't as pronounced as it may be today with the way civil service is incentivized.
- [11:10]
B
Yeah, I think there's parts of that that are right and parts that are wrong. I think what is true is that like, obviously I think both FAI and IFE would love to see like a much more empowered civil servant class or just at least a much more informed, like more robust people talk a lot about state capacity, the idea that can the government actually do the things that it sets out to do. And there's a lot of ways in which we have just not set up our institutions across government to actually be able to achie those Ends. I mean we do a lot of work on say like export controls and you know, how can you actually make sure that top end chips are not getting, you know, smuggled to China? And I think like one, one proxy that keeps me up at night is that I think the, the overall like enforcement budget of BIS is about half that as like one, you know, H200 shipment that ends up getting smuggled. And so we are just like vastly under resourcing or under supplying kind of these key institutions that are supposed to be doing work. And regardless of whether you are like a fan of big government or small government, I think we can all agree that we'd like to have the things the government is doing be able to do them well. And so that defin implies like better paid civil servants. And I think one reason you've seen kind of a departure of a lot of high caliber talent from the government over the last several decades is just like pay discrepancies between the public sector and the private sector. And to some extent that's good. Like I'm very glad. I think one of the things that makes America so special is like kind of our just amazing private sector and they're doing cool innovations. And not everyone good should be in the government obviously, but I would like to have more really top caliber folks kind of in the government than currently exist. And so if you had, you know, every senator basically has their own think tank that's akin to IFP or fai, I think it would be great. We don't give them the capacity or the staffing to be able to do that. I think even if you had that, that would absolutely improve things, but that would not solve all the problems that think tanks are trying to solve for. Part of it is just having a little bit of distance from the specific institution, being able to find areas of interest or overlap that are really cross cutting. And you see something that, oh wow, this weird regulatory structure the Department of Transportation is using might have actually a lot of analog to this thing that we should do over in hh. And if you're just kind of housed within one agency, it's hard for you to kind of like see those cross cutting themes. It's also the case that there's a lot of weird restrictions, some of them political, some of them like statutory, in terms of who can talk and how much overlap can you have between the executive branch and the independent agencies and Congress especially? And so there's just huge information gaps actually. And I think one of the things think tanks do is they serve as Kind of a clearinghouse for information, trying to make sure that the relevant folks have the information they need to be able to act. And you try to like solve some of these weird communication barriers or serve as kind of trusted third party intermediaries. And so I think even if you had really well staffed, ambitious state capacity efforts, which again I would love to see, I think they would still absolutely be an important role for think tanks.
- [13:52]
C
When you look at the Gingrich revolution in the 90s and the cuts to things like OTA and a lot of the in House research capacity for Congress in particular, you see these experts eventually filter out into think tanks, into civil society and yes, into lobbying shops. So lobbying gets a really bad rap. It's sort of a, you know, a byword for the, you know, the scamminess and the scumminess of the D.C. swamp. Right. And it creates a lot of bipartisan ire at these, you know, shady lobbyists who only shill for corporate interests or whatnot. But what a lot of people don't realize is that lobby shops also host a lot of these subject matter experts that are, you know, providing a lot of useful services for particularly folks in Congress, folks in the executive branch. And this, this is true across every administration. I don't think it's something that's unique to, you know, Republicans or nor to Democrats, you know, and different, you know, lobby shops could rise up and down in prominence based on the networks and relationships they've developed. But yeah, when you look at the major must pass bills that go through each year or the major legislative reforms that have managed to garner bipartisan majorities, in large part they're written not by in House staff but by outside lobbyists. And while I can certainly sympathize with the desire to build a lot more in House state capacity, we've called for this, Caleb and IFP have called for this. Many other groups on both the right and the left for different purposes have all said that like more in House state capacity and expertise is good. You know, it's also just like worth noting that, you know, that still exists. It just is housed in different places and the compensation structures are different and it's perhaps less of a dire situation than you might sometimes hear about in the news.
- [15:33]
A
I think the sort of when, when would you sunset a, a sort of like techno, you know, like a more technocratic think tank. Another answer to that is, hey, we're not just technocratic. Like we, we have sort of more partisan goals that we're trying to pursue or just More ideological goals. I mean, Caleb, you guys picked the Color Purple presumably for a reason, but let's throw this one to Max. I mean, how do you guys, how do you guys think about partisanship in the type of work that you do?
- [16:02]
C
Yeah, I mean, so the, the cop out answer is that, you know, as a 501c3 organization, you are not allowed to engage in partisan political activity. And you know, that's true for us. Right? We are that we do have an affiliated C4 that's getting stood up. More to come soon on that. It's not public yet and we're excited to start getting more involved in that side of the sort of advocacy cycle. But also, it's kind of obvious the FAI is a team of conservatives, folks who come from across different sides of the movement. Some folks will describe themselves as basically state capacity libertarians. Others might describe themselves as social conservatives. Still others might describe themselves as more MAGA New right populists. And then others may describe themselves as more, you know, sort of a, you know, traditional conservative. And they're, you know, not so sure about this whole Trump thing. But I guess this is the, you know, the sort of the hand that we were dealt. Right. So we see ourselves as a big tent for folks on the right who want to come think deeply and critically about technology. You know, many of our scholars and fellows have written about the deeply transformative effect technology has on the social forms that, you know, we care about. And at the end of the day, if you're a conservative, you know, regardless of what flavor you know, might describe yourself as, or if you're right, of center writ large, right. You know, you do ultimately care about preserving the things that make, you know, American society in particular quite good. And technologies can be used in service of that, can also be deeply disruptive and, you know, work as a sort of solvent against the sorts of things that you think are worth preserving. And so we see ourselves as a place for conservatives who might disagree on lots of different policy questions and happy to get more into this later in the show if you'd like, but as a sort of, you know, meeting call tent or something like that to help hash these debates out and try to come to a more fulsome and built out conservative theory of technology.
- [17:56]
B
Yeah, and then for ifp, I think this is one of the main differences, I guess, between FAI and IFP is that we explicitly try to embrace what we might call cross partisanship, which I would sort of contrast with bipartisanship. At least the connotation of that is I think like both sides meeting in the mushy middle on a compromise that neither side is particularly happy with. Whereas cross partisanship is, you know, very earnestly finding ways for both sides to actually end up supporting the same thing, but for different reasons. And so I think one nice example of this is the Chips and Science act, you know, where it got bipartisan cross partisan support, you might say. And I think both sides were really excited about the thing that they were getting. I mean, Republicans were obviously invested in this vision of like American scientific and technological leadership, especially vis a vis China, onshoring more kind of domestic manufacturing capacity. I think Democrats were excited to kind of reinvest in our public institutions of science and of a strong public role in innovation. But at the end of the day, enough members voted for it in this durable way to actually get it over the finish line. And one reason I think we find ourselves liking those kinds of deals is they feel much more durable. So much of policy today, I mean, outside of the rare thing that happens like TIPS and Science act is happening to the executive branch and for a lot of the things that we most care about, whether it's, yeah, steady investments in meta science and new ways of funding science, whether it's new ways to get drugs into the market sooner or, you know, better pandemic prevention investments, like there's a whole host of things you work on, but it's actually incredibly counterproductive if your theory of change implies that those things are going to be swinging back and forth every four years. You should roughly plan on, you know, roughly 50, 50 power split across Congress. And unless you can find a way to like durably connect this to the, each party's kind of inherent identity, I think it's hard to have durable programs that actually kind of like last over time. And so that was, I think, part of the original conceit of why we started IFP and why we wanted to form it in the way we did. I in particular kind of worked on think tanks on both the center right and the center left. And we're trying to specifically find issues that are kind of, I like the term, you can pull the rope sideways. Like, one way to have impact in the world is to see that there's this giant tug of war match and say, I think this side is correct and I'm going to like pull on the rope in that direction. But very likely you are one of, you know, 10,000, a million people kind of pulling on the rope in your direction and your marginal contribution to shifting the tug of War match is like very small. And that's what it felt like working on a lot of more traditional tech policy issues, whether it was like privacy OR antitrust or section 230 content moderation or whatever. And it felt like there was an entirely orthogonal set of issues that did not have any sort of partisan valence in the Mata Glaciasis terminology. You could win through secret Congress or secret executive action placement, trying to specifically avoid having this become a highly salient polarized issue that's on the front page of the New York Times or getting covered by Rachel Maddow or Tucker Carl and instead find genuine organic ways of building this into the narrative that each party tells themselves and then yeah, trying to build a cross cutting coalition and actually move the ball forward. But there's not a strong Republican or Democratic position on should the NIH do more person based funding or project based funding. And yet that question matters a tremendous amount for the future of science. And so we find ourselves really trying to pinpoint what are those issues that the parties don't have a strong sense of. But you could actually really move the ball forward.
- [21:18]
A
Max, what is Caleb missing by being in this like sad little Venn diagram of stuff people don't feel strongly about?
- [21:27]
C
Yeah, at some level I agree with Caleb's theory of change here. And I think, you know, as we both know, like we know FAI also engages in a lot of this sort of cross partisan work. You know, I think particularly on like our state capacity work as well as our energy and infrastructure work. We've assembled these sort of strange bedfellows coalitions on issues that, some issues that are quite high salience and then some issues that, that are flying very much under the radar and as long as you can get a lot of different folks from a lot of different ideological backgrounds or partisan persuasions to have a stake in the outcome. You want to see, as Caleb says, does improve the durability of the issue and it does increase your probability of success. I think where the risk you have there is that sometimes the high salience issues are actually the ones that matter. Sometimes there are just policy fights that you know where like your counterfactual impact might not be on pulling the rope sideways but on adding your marginal heft to one side or the other. And you know Caleb, I don't, I don't want to throw you under the bus here. I'm sure you would agree that there are certain certainly issues like that, that, that where it does matter quite a lot to, to get to pick one side or the other.
- [22:38]
A
Yeah, let me, let me in the, in the ring for you. So you know how the Democrats are defining abundance, Right. Feels kind of ifp, but also like, maybe a little too spicy is the wrong word, but yeah, maybe a little too partisan to like, weigh into that. And that was a very big moment that, you know, that was a very, very big discussion that occurred over the course of 2025. Yeah, I don't know, Caleb.
- [23:07]
B
No, no. Yeah. I mean, I think these are good questions and there's a version of this
- [23:10]
A
answer that we won't air. But I'm curious. Yeah, I'm curious, like what you feel like you're missing.
- [23:16]
B
No, no. So I just want to say is like, IFP has picked a very particular theory of impact and a particular set of issues because we thought it was like, neglected and no one else is doing it. I certainly do want to claim this is the only way to have impact. And absolutely. It's the case that there is clearly a impact case for being like, we are the Democratic organization, we are the Republican think tank, and we're trying to change what our party thinks about these things. Elections have consequences. And if you want your issues to really come to the forefront, absolutely, you should try to have voices in each tent or coalition that are kind of like genuinely trying to their party's views on the issue. That's just, I think, not central to what IFP in particular is trying to do. Not saying that that's like a bad theory of impact on the abundance issue in particular. I mean, this is interesting because I think different factions or different voices are trying to get different things out of the term or out of the discussion. And like a lot of terms, something like abundance is like inherently contestable. Like, there is not a predefined definition like what abundance actually is. And it is in some sense this emergent process that comes out of a bunch of different conversations and shelling points and kind of like who's, you know, framing or whose positioning ends up like, you know, winning particular ground. And so, you know, I really enjoyed reading Derek and Ezra's book. I think they also have a particular conception of abundance as kind of being a message that the Democratic Party can use to like, win. And I'm simultaneously excited for like, their efforts to try to get these issues to become more central in the Democratic Party brand. I also do have real concerns about this ending up polarizing on some dimension part of these issues. I mean, I think one of the like iron laws of politics today is like, counter polarization. And as soon as like a party creates like a strong flag planting like this is what we believe, it creates a ton of natural incentives for the other side to basically just immediately, in kind of a knee jerk way take the opposite side. And so it would be very bad actually I think if like something like meta science or science funding ended up becoming like a polarized issue. The good news is that actually a lot of these issues are just inherently like pretty, I don't know what the right term is. It's hard to turn metascience into ultimately a kind of a partisan issue. I think it's just sufficiently in the weeds of implementation. And one of the things I like about these sorts of issues and state capacity more generally is that both sides earnestly want to achieve the goals that they are setting out. And so you want to have effective institutions and the ability to actually correct in real time and use experimentation to improve the way that you're functioning these institutions. And regardless of what your end goals are, the more effective instrument is useful. And so my hope is that there are some of these issues that regardless of attempts to kind of polarize them, might actually continue to stay out above the fray because they are instrumentally useful for actually both parties agendas.
- [25:55]
C
Of course the cynical take that I have is the abundance agenda is great because it's just taking a bunch of longtime Republican ideas and trying to repackage them as Democratic ideas.
- [26:05]
B
Supply side.
- [26:06]
C
Exactly. Now the thing that I'm interested on again is a bit of an outsider to the abundance debates. Well, on the one hand, FAI tries to be the right of center partner to a lot of these conversations. You know, we helped produce and co sponsor the Abundance conference last year. You know, we're obviously like, you know, very close with lots of folks in the community here and you know, enjoy like you know, both spirited argumentation and you know, surprising amounts of agreement with lots of the folks in that community. But one thing I'm really interested in is if to what extent is abundance going to be, you know, basically accepted as something that like the Democrats can really get excited, you know, get motivated, have that be sort of like a flag planted that rallies the, the base around it. Right. Like it's a, as Caleb says, like correctly, like a lot of abundance issues are, you know, sort of wonky supply side reforms or wonky meta science and science funding debates. Right. Like it doesn't motivate people in the same way that much more high salience like cultural issues or you know, getting really pissed off about whatever the Trump administration is doing. You Know, this day or this hour. Right. You know, there's an open question, I think, like, to what extent these more like wonky technocratic ideas can motivate people and push people in a sort of, you know, in the increasingly sort of like hyper partisan and also like highly fractured media environment that we live in. And you know, if that's just the thing that gets people out of bed and to vote in the morning. So, you know, I don't worry myself as much about the, you know, the sort of like the left coding of abundance just because I think that like the ideas themselves, while I think they're quite defensible in almost all cases, I also just don't think they get people out of bed in the morning.
- [27:45]
B
Yeah, I think it's true actually. And I mean, I'm sure this is probably not how Derek and Ezra would answer this, but I would say like, I think fundamentally abundance and this agenda is like not necessarily a mass politics book. It's really kind of a book for staffers in many ways. I mean, I think it's actually been kind of commonly remarked that people found it surprising that, you know, Joe Biden during the 2020 campaign ran as a moderate and yet then in many ways had like a quite progressive kind of like agenda in terms of the things that they were running on. I think a lot of people attribute that to staff level differ, that actually like the kind of Elizabeth Warren crowd had won a lot of mindshare amongst relevant staffers that were going to be naturally kind of the ones going into these positions. And so then you could really punch above your weight from kind of an agenda perspective if you've captured mindshare amongst a certain kind of like person who's going to actually be staffing this administration. There's so many ways in which political leaders inevitably have to rely on kind of the implementation capacity of the people in like the rungs, you know, below them and then five rows below them. And that just like this is actually the way the political machine operates. And so I think maybe like one of the stronger cases for abundance from an impact perspective is that you sort of create a new durable set of issues to prioritize as you're pursuing kind of whatever your pre existing ideological commitments are. And that if just on the margin you care more about the implementation and making sure the thing actually happens, rather than just passing the bill saying that you're going to give money to the thing, that would be really good. And I think, Max, to your point, regardless of whether you call it abundance or Call it something else. I think the more important thing is actually that this focus on implementation, this focus on really taking innovation and the supply side seriously as kind of like the underlying root cause of a lot of this political dysfunction. I think that genuinely can have its roots and kind of become a part of the governing ideology across both parties. This is actually kind of like a Niskanen point. But when the Niskanin center launched, they wrote this conspectus kind of about what was their theory of impact. And they focused specifically on what are issues that are kind of in this like durable, bipartisan kind of elite consensus, regardless whether or not we talk about it or run elections on it. And you see issues that, you know, for, for decades and decades de facto have a kind of like monopoly. And so, you know, something like free trade prior to the Trump era was, I think, an example of this where, you know, across many different election cycles and Republicans or Democrats winning, like people who had a particular set of views on free trade were just kind of de facto winning. And obviously like all, all consensuses eventually begin to erode. But like there was, there was a period where like the reformers in that sense, like won a durable. And if you could achieve the same thing on a set of issues around science funding and supply side reforms and whatnot, like, I think that'd be great. And I would love to see like the Republican and the Democratic version of that each have their chance in the sun.
- [30:29]
A
Max, throwing it over to you. As our resident conservative, where are the youth staffers thinking nowadays?
- [30:37]
C
To my earlier point about the ideological heterodoxy or diversity within fai's team, I also forgot to note that, you know, there's at least a few registered Democrats on our team as well. You know, we're also very bullish on center left Democrats who get increasingly radicalized by their own party and become effective center right apparat chicks. So we're certainly interested in that pipeline. But in terms of the staffer class in the GOP or the sort of the Trump Vance staffer class, I think there's a somewhat similar dynamic to what Caleb was discussing during the Biden years, where I think one of the big differences you can see between Trump 1 and Trump too is that there's a lot less ideological diversity or you sort of. And there's a lot more alignment. Right. This is like a term that gets thrown around a lot in right of center circles, right with the President's agenda that's shared among the junior staff or the junior to mid career staff. You know, I think some of this is due to the. Just a much stronger vetting process. I was, you know, I think, I think that, you know, the folks at the presidential personnel office and folks within the transition team more generally had a very like, you know, sort of smooth running transition operation that was able to quickly identify and vet and, and kick out a lot of folks who would have otherwise gotten a job in a more permissive, Trump one style, more chaotic administration. This administration, we've seen, for better and for worse, a much more streamlined and efficient operation that has a lot of folks kind of rowing in the same direction. I think that's also, there's certainly the case that Twitter and Signal group chats and whatever have been like, sort of like great coalescing forces for young people who care about politics. And what can you see? Right. You can see this in sort of like, if you want to go on the left of center, right, you know, the sort of dogpiles from folks in the sort of like, Liz Warren, Bernie Sanders camps or the, you know, the antitrust crusader camps. Right. And I think you also see this in the Republican Party, whether it's debates about anti Semitism or debates around, you know, to what extent is the free trade orthodoxy going to be allowed to continue to exist or, you know, what have you.
- [32:47]
B
Right.
- [32:47]
C
Like, there is much more sort of like public totem policing and sort of like ideological policing going on. And so I think people are just a lot more sort of subject to this, like, panopticon of surveillance both on their, like, you know, whether it's their past statements or past jobs, but also, you know, they're sort of the extent of their alignment to the president's agenda as it's articulated. The last thing I'll note about this is just that I think this, you know, this creates a lot of, you know, if you're in the sort of like, center right, center left reformist community, right. This creates a lot of problems because, you know, you're, you're the, a lot, a lot of the people in the current administration just don't want to talk to you, don't want to listen to you. Right. But I also think it creates a lot of opportunities because these people are by and large disconnected from a lot of the elite institutions that traditionally help produce the elite consensus that Caleb was talking about. Right. So, like, particularly on foreign policy. Right. These people are, you know, people across the State Department or across the Department of War are, you know, by and large, like, disconnected from most of the legacy institutions that do a lot of that Opinion shaping. And so that does create the sort of the avenue or the route for new institutions certainly to take their place, but also to have more informal or private channels of influence. Just because there's not as many resources for these folks to draw on and they don't have as many relationships with the blob or the establishment that would often have to fill that gap. And so I think this certainly creates problems for people, but I think there's also opportunities if you squint and look carefully and, and try to think carefully about your issue selection to some extent,
- [34:24]
B
like not staffing is itself a policy choice and that your inability to kind of like drive a long run agenda is in many ways connected to like not having staffing capacity. I think like there was a really interesting podcast between Ross Douthat and Chris Rufo, you know, about like sort of the Department of Energy. And yeah, the extent to which if you are not able to capture some part of like the preexisting elite, you're just going to be incredibly short staffed and then that limits your reform options to basically like destroy the institution or kind of ignore it entirely. You don't have the ability to have a strong guiding hand, but still fundamentally agree that the institution should continue. And by taking those options off the table, you're actually restricting your own policy avenues.
- [35:04]
A
So on staffing, I want to talk about internal think tank staffing. I think you two are both famous for having rather junior staffers who have potentially outsized impacts. Caleb talked earlier about having staffers how like the Brookings model of your staffer as rainmaker and, and event host and writer who like almost like owns their own book, has some positive positives and negatives. But when you are trying to sort of hire and incentivize like actual like impact and impact and legislative change and the government doing a thing it wasn't doing beforehand, you know, what do you look for? Why do you think 20 somethings can have that impact? And you know, maybe how has your view, how have your views on this change changed over time?
- [35:52]
B
Yeah, for sure. One of the key theses I think we had or hypotheses we had coming into IFP was the fact that there might actually be too much specialization within think tanks. I think in a lot of again, traditional think tanks you have a very senior fellow kind of like, I'm just a researcher, I sit in my office all day and I turn out white papers for 40 hours and then I go home and that's the end of the day. And occasionally maybe I'll be brought in to testify in front of Congress. But that's about where my engagement with the political process begins and ends. And then you have a whole other set of special specialized actors who work on communications. And all day they're just trying to like send stuff out to journalists or put out press releases or write blog posts synthesizing what the big white paper said. And then you have a set of folks in sort of outreach and like they're pounding the pavement on the Hill and trying to like bring the white papers to staffers. And again, I think that model has some virtues. I mean, obviously we know specialization has value, but I think in many ways actually there's too much and you get a lot of economies of scale of being able to like, be vertically integrated across that whole stack of like policy entrepreneurship, to use the Tom Kalil kind of phrase. I think staffers are way more interested to talk to you if they know you're the person who wrote the 40 page white paper and can actually really go blow for blow with them about every single part of kind of the process or getting into the nitty gritty details. And likewise with journalists, they actually want to talk to the person who has been deeply engaging in the world. They don't want to talk with a comm staffer who then hands them off to somebody else. And as soon as you actually have like a third party intermediary who is sort of like trying to hand off the relationship without the requisite like relationship capital, you're, you're basically killing a lot of the relationship context that makes them willing to spend political capital on your behalf. And so I think there are enormous economies of scale of basically having researchers that can to some extent do their own comms and do their own outreach. This not only makes you better at doing the comms, but I think it also makes you better at doing the research when you have a much more granular understanding of, you know, the actual evidentiary needs that kind of need to be filled. What are the holes that if I could actually show convincingly saying X, Y and Z, that would move the thing forward. You're just like way more looped into those decision making loops and which parts of like the research landscape are actually important to cover and what are the particular nuances or things you have to watch out for in the wording of how you're thinking about the proposals or oh man, I gotta make sure we don't like accidentally cross into somebody, some other committee's jurisdiction because that will dramatically increase how complicated it is to get like this piece of legislation across. And so I think this vertical integration has like big advantages from both being a better researcher and being a better communicator. It does inherently mean you have to be somewhat more limited in terms of the number of topics each person is covering. So I think we think of ourselves as, yeah, being like much more narrow in terms of how much we're asking each person to do, but then much deeper across the entire policy engagement pipeline from early stage ideation all the way to like agency implementation, actually making things happening. Like we're looking for people that can kind of be these like multi tool athletes. They have the ability, at least the potential to be really cool, good at obviously being the deep technical expert and doing their own research, but also can like really socialize well in a happy hour and kind of like make friends. They can sort of get, get a sense. They have like a good spidey sense for like what's actually going to really like pop off in the larger discourse. Know how to operate in group chats. And so it's, it's hard and you're sometimes looking for unicorns or people that can like do all those things at the same time. But at the very least we're trying to create like a conducive environment that like encourages people and sets themself up well to kind of begin developing all those skills. Um, let's talk about, let's talk about
- [39:21]
A
that for a second. Like I've been trying to talk to potential donors about the idea that the time I spend in group chats is actually perhaps the most impactful time I spend over the course of my day. And like doing the actual counterfactual policy impact thing.
- [39:41]
C
Don't let that get out. Come on. That's.
- [39:42]
B
You're stealing all of our alcohol trade secret. No, but I feel like it's not
- [39:46]
A
a trade secret because there are so few donors who get it. I don't know what are the special words you say that like what appears to just be socializing is in fact the way things happen in Washington in 2026. I mean you would think that signal gate would like be enough, but I guess not.
- [40:06]
B
I don't know.
- [40:07]
A
Thoughts.
- [40:08]
B
I mean it's interesting. The way we've tried to tackle this is by rather than showing aggregate statistics in terms of like here's the number of op EDS we wrote, here's the number of meetings we took, we inst really try to focus in on very specific case studies and walk folks through how like every single part of the Process ended up like linking together to achieve some outcome. So we actually kind of like take like a quasi VC approach to policy, I guess you could say, like we want to maximize shots on goal for high expected value policy outcomes. And you assume that because policy is actually really hard, you're only going to succeed, I don't know, 5 or 10% of the time. But if you're taking 50 shots on goal over the course of like a two or three year period, you should actually have some very concrete wins at the end of the day to be able to point to, you know, just probabilistically so long as you are actually well calibrated in terms of your, your odds of success along that margin. And so then you basically hope to have a couple of successes that more than pay for the overall portfolio of failures. And because like policy actually matters so much to get right, I think that like it's very easy for one or two successes to pay for an overall portfolio of approaches. And so then really what we're trying to just show is like in great detail, like here's the specific set of actions we took and like we texted X person at Y time about Y thing, they responded in Z way. We then introduced like person to this other person who wouldn't have met otherwise. And actually here's an email from the relevant person saying that without our effort this thing wouldn't have happened. And no single part of that process is definitive. But you link them all together and I think it creates a pretty compelling picture that actually, yeah, this reform at the end of the day wouldn't have happened without our efforts. And so I think once you can show funders that kind of process, I think it makes it much easier to pitch this broader, more, more kind of open ended pathways to impact.
- [41:55]
A
But then it's kind of hard to do the staff analysis and promotion and firing stuff if you're doing that VC model right and kind of like disaggregating the final impact from the input of the work that the staff put in is much less straightforward.
- [42:15]
B
Yeah, it's a good question. There is this sort of like in the aggregate you can judge by results. But yeah, in the intermediate you absolutely have to kind of like analyze the process. There's a metaphor I like from actually the world of clinical trials. So when you're trying to design a drug to say reduce cancer mortality 30 years down the road, in some idealized world, if time were not a factor, you'd want to just run a clinical trial and see does this very directly decrease cancer mortality 30 years down the road. But, like, that takes way too long. It's completely impractical as a thing to actually link your cancer trial to. So instead, people have designed what are called surg endpoints, basically, really highly correlated outputs that you can measure much earlier that you think have a strong linkage to the thing, like, later down the road. And so in certain kinds of, say, like bone cancer, we see, like, bone marrow density ends up being really well correlated with kind of like cancer and mortality rates, you know, like 10 or 20 years down the road. And once you have those really well defined, well validated surrogate endpoints, you can kind of begin optimizing for them or at least like using that as one way of measuring your impact. And so I think we're constantly looking for what are the policy equivalence of surrogate endpoints and trying to then measure those. And sometimes that looks like just persuading very particular key moments. You can almost work backwards from victory and say, okay, in worlds where we won, presumably that means that we've been able to persuade parts of the pipeline like A, B and C. If this year we were able to persuade part A, but we are still missing B and C, like, we are closer to our goal than we were before. And what are the necessary but not sufficient conditions in terms of actually making progress and then beginning to measure some of those intermediate steps? If you are keeping your eye on the ball at what the end goal is and kind of being ruthless about your internal prioritization, I think that's like a totally coherent way of trying to measure impact. But there's also some part of this that is inevitably a kind of like, art more than a science. And you have to have a good sense of like, ah, yeah, this seems like the kind of thing that's actually plausible to work. And there's absolutely a part of just like human judgment and being able to assess that.
- [44:13]
C
One thing I think I disagree with Caleb slightly on is that, you know, surrogate endpoints, like, finding those surrogate endpoints for policy impact are important, but I think you can sometimes over index on signals that end up being less predictive than you thought they were. Right. And so we have experimented with a bunch of different ways of trying to you sort of measure people's intermediate progress towards their moonshot swings or their big policy swings. Right. And we've eventually kind of reverted to a much simpler sort of performance model, which is basically like, you know, at the beginning of the year, we ask everyone to write down in their jd, like, what are the big Swings you want to take this year, right? On policy impact. And then what are the big swings you want to take in terms of your thought leadership and your, you know, growth of, you know, both your personal brand or like, you know, placements that like help redound onto the fai, you know, prestige or whatever. Right. You know, so like getting, you know, an op ed in the New York Times or the Wall Street Journal or you know, FT or you know, doing a podcast, you know, with China Talk or with Tokesh or whatever. Right. Or with, you know, but like we, we just try to pick, you know, like, you know, don't go crazy. You don't have to be like hyper specific. But like, what are the like sort of like highly prestigious sort of placements you could land that would again, like help you get your ideas in front of the decision makers that matter. And then if you're responsible for fundraising and attracting support, durable support, so we can build a program around some of these ideas, some of these bets that we're making, then at the end of the year, did you do that? And we've ultimately found that if you're good at sort of traversing as what Caleb was talking about, the full policy life cycle or the implementation life cycle, from idea to research to outreach, to pounding the pavement on Capitol Hill, to all the media work that you want to do and, and all that jazz, these pretty simple metrics of basically, did you do the thing you said you wanted to do? And if not, did you do something else that was of equivalent value? And then were you able to get other people to care about those ideas? And then were you able to get people to pay for those ideas? Those have been the indicators that we've ended up landing on as just pretty good predictors of people's long term success and our overall ability to achieve the foundation for American innovation view of the world. The other thing I'd like to just throw in there into the mix and address one of your questions at the beginning about empowering young people is like one of the great things about Silicon Valley and FAI was founded in Silicon Valley back in 2014. So we have deep roots in the Silicon Valley culture and trying to understand what makes that place special. But one of the things that makes it really special is that as far as I can tell, it's the only place in the world a young person, an 18 year old or a 22 year old with very little credentials or very little credibility or very little to show what, you know, to show for themselves, can raise, you know, millions of Dollars in funding to go attack a problem, to try to, you know, fix something that they see that's wrong in the world and try to make some money along the way. And as, as Caleb said, like there's a VC model, right? You know, in some ways we're both trying to like, you know, take these big swings and you know, we assume most of them aren't going to work out, but some of them are going to hit and the expected value on those is gigantic. But part of that is also being able to take bets on young talent and convincing young talent that policy entrepreneurship is a valid and interesting way to have a huge impact on the world, a counterfactual impact on the world at a very early age. And so that is better than doing management consulting or spending years in the foreign service or doing other high prestige careers, going to big law or what have you, or even going into tech. In some ways tech is over. Index on young strivers who've founded their first startup in high school and are just really trying to get into YC batch 26 or whatever. You need to find new ways to maximize the value on your own time. And we think policy entrepreneurship, you can often short circuit if you do what we're talking about, following the full policy life cycle, you can be a very young person and yet have a disproportionately large impact on final policy outcome that you could shape. You know, trillions of dollars in investment, could save, you know, millions of lives, could be, you know, critical for, you know, the development of, you know, transformative technologies like artificial intelligence. Like, you know, pick your, pick your poison, right? Pick the thing that you're passionate about, can have a, you know, huge impact very early on in career. And I think most people don't realize that yet.
- [48:46]
B
Well, yeah, totally. And actually just like a quick shout out to I guess Zach Graves, who I think Zach was my, my first boss in D.C. when I was at the Archery Institute. And like a big part of going there was that I think Zach very clearly had this philosop of like betting on and empowering young people. And I think there's a lot of other organizations that like, yeah, unless you got your PhD, you can't dare dream of like solo authoring a piece. And I think that's like really counterproductive, especially in the world today when actually most of the people like writing pieces of legislation are themselves like 25 year old. And so then they have like a very different standard for people trying to like help engage in this like world of thought of like trying to, trying to build out better, better ideas around this space. I think you can tell a lot about like the institutional health of an organization based on like, how quickly do they empower and elevate young people that are kicking ass. And I think that's definitely something Zach has brought into a lot of different organizations.
- [49:38]
A
Let's talk about pay. So Mike Froman now famously on Chinatalk, when I called out him paying his Ra55 grand, told me to endow them to get that number up. Yeah, when you're trying to talk people out of going into finance consulting or tech, the sort of like long term earnings potential of this world is obviously on a different curve. But you know, how do you guys think about compensation and how do you try to sell comp, you know, sell paying people appropriately to, to the donors who ultimately are going to be funding it?
- [50:17]
B
I would say, I mean, I think for IFP and I imagine FAI is pretty similar. Like we are very concretely trying to pay on the upper range of basically what you'd expect to make from like other nonprofits. I mean, to be clear, like I, we want folks who are doing this for the mission. I think almost everybody at IFP could be making more money if they were someplace else who were working in private equity or in Silicon Valley. And I sort of like the ethos of if you're like one of the top government ministers working in the Singaporean government, your pay is oftentimes indexed to like an index of the top 1000, like private sector employees in Singapore, but then with a 40% discount for kind of the ethos of public service. And so they're still paid like absurd amounts of money by you know, US civil service comparisons. But I kind of think that's like the right vibe basically. Like, you should absolutely not be expecting people to work on poverty wages. And this is again downstream of like what foundations and other funders are willing to tolerate. But like, on the whole, I would love to raise like think tank salaries in kind of the nonprofit space here for excellence and like, not just make work jobs, but if you're actually driving value and actually changing policy, like, yeah, you should pay very competitively. Recognize that probably you're not going to make quite as much as you would in the private sector. But like, you know, we pay, I think very competitively or I mean, above market for basically all of the they think tanks and nonprofits.
- [51:33]
C
Yeah, totally cosigned. I think there's a one macro point here is that FAI and I think to some extent ifp, you guys would agree with this take is that a lot of the traditional modes of influencing the policymaking process in D.C. are broken. Multi client lobbying. Oftentimes these folks just move really slowly. The trade groups are paralyzed and oftentimes don't get involved in the issues that literally only exist to do too many large sort of sclerotic think tanks just move too slowly or don't empower the brightest people at their organization who then leave for government or leave for private sector because they just get frustrated. And so there is an excite, I think there's a small but growing community of young and bright talented folks who understand that a lot of these sort of traditional pathways towards having influencer shaping policy is broken and that there's a few organizations, you know, like across D.C. like where, like FAI and IFP I think are the leaders where you can be empowered to, you know, go change the world. Right. And you know, there, there, there's a, you know, like you will take some discount on that relative to like you know, the kind of money you could make like in a government affairs shop or you know, the kind of money that you could make, you know, going to Silicon Valley and you know, joining your, the next hot AI startup and hoping for your lottery ticket to pay out. Right. Like there, there's certainly like a discount that you're taking on your lifetime earnings here. But the other hand, right, like if you think that a lot of these like traditional models are broken and that a lot of these places are going to collapse within the next, you know, I don't know, 10, 20 years, I think it's actually a good bet to start like developing a new way of, you know, shaping, shaping policy in Washington D.C. and around the country. And I think people are willing to, you know, come take that bet because they think it's going to pay off handsomely for them regardless of where they go in the future. Right. Yeah. And then like to what Caleb spoke about, like you know, paying above market rate, like I think that's, you know, a lot of it is just downstream of what your funders are willing to, you know, like let you get away with. Right. You know, I've worked with some conservative philanthropies that have, you know, they don't, they don't pay their own, their leadership or staff very much at all. Right. Because like they, they really do emphasize, you know, like thrift and you know, sort of a frugalness in all their grantees. Right. And so there, there are, there are real limitations to this just because of, you Know, whether it's like ideological priors or you know, sort of like cultural preferences of individual program officers or individ individual, you know, philanthropic givers. Right. But at the other hand, right. I think people do, God forbid those
- [54:09]
A
program officers aren't paid appropriately though, you know.
- [54:12]
C
Well, I, I mean, I've been shot. I mean we've all heard stories about like, you know, like well endowed foundations that are, you know, lighting 500 grand a year on like, you know, just on their program officers, you know, like, like let, you know, don't even get me started on the program directors or whatever. But there's also like, you know, you would be shocked at how little money there is in conservative philanthropy, right. In particular, right. Like if you want to go work in kind of like Leonard Le. If you want to go work in like, you know, I don't know, like Sarah, like Sarah Scape World or you know, like, you know, the Mercer world or whatever. Right. Like most of these people do not make a lot of money. Like you do it because you, you genuinely love it. Right. Or like stand together world. Right. Like there, there is a, you know, like there's, there's certainly like, you know, well paid leadership jobs that do exist here. But most people in, even in like conservative philanthropy primarily do it because they're mission oriented, not because it's the you know, sort of like revenue maximizing career potential or career path that they could pick. Yeah. So I think there's like more persuasion to do. But I think that the argument that makes a lot of sense to people is just that you get what you pay for and you should pay top talent, like top talent, Right. Many people who've been successful enough to grow a great business and then eventually have enough wealth to start a philanthropy, Right. Certainly follow that in their professional lives and they would instruct their teams to follow that in their professional lives in the private sector. And it stands to reason that the public sector would to operate in similar ways and that if you can add more of those incentives, you can expect better outcomes from the civil society organizations that you decide to fund. Right. So I think that argument needs to be made a little bit more widely and it'd be great if our members of Congress could also raise their own pay, raise their staff for pay. You start thinking about broader civil service reforms that also allow for higher pay, but also make it easier to hire and fire. I think that all of these kind of reforms, they're part and parcel of having a thriving civil society and thriving public sector. Right. But we have to be willing to tie pay to outcomes and performance a lot more explicitly.
- [56:15]
A
Let's close on AI. How do you guys feel it changing the way policy work is done?
- [56:24]
B
Good question. We recently did a staff retreat and I think I gave a whole talk on this. And basically, if you think about a very academic understanding of what think tanks do is they try to act as this translation layer between a set of, like, academic findings and empirical evidence and like, the actual policy demands of. Of policymakers. But it is just obviously true that synthesizing, translating, and applying a existing body of knowledge into a new format is like the single thing LLMs are best at. And so I think from that, like, very simplistic view, actually think tanks are sort of like going to get eaten away. But the good thing is, actually there's lots of other things think tanks, especially the best ones, do that is not just like, application and synthesis and translation of existing work. And so I've sort of been saying, like, do the things try to specialize in areas of work that are not legible information, at least certainly not to the Internet. So our senior editor, Santi, has this really great newsletter and podcast he runs called Statecraft. Highly recommend listeners who's trying to talk subscribe. It's great, and I think there'd be lots of overlap, but one of the things he does is he'll just call up civil servants who have been working in the belly of the beast for the last 30 years, and they've recently retired, and they're kind of willing to tell all and really get into the wheat weeds of how does policy actually work? Why did this thing that seemed so obvious 10 years ago never happen? Tell us about the interagency process that ended up killing this process. And so much of that information is just not legible. And so LLMs or AI cannot do anything with that. And so there's a certain element of become almost a little bit like an investigative journalist and call up people on the phone, do stuff in person, grab the parts of the information landscape that are not legible to AI, and then basically compliment that and then modify your workflows to be able to really take advantage of these really amazing research assistants. I basically don't do anything without at least running it through an LLM for critique. You can just run it through 5.2 Pro and say, point out literally every single flaw and hole in this, and it'll point out a bunch of stuff that might otherwise have gotten missed even by diligent human editors. And so if you're not using these Tools to some extent, I think you're really missing out not to name names, but have heard of organizations that are unwilling to say, expense the costs of simple Clott ChatGPT subscriptions for their employees. And that just seems like the most pennywise pound foolish thing I can imagine in this environment. And so you have to be taking advantage of these tools. You cannot outsource your thinking to them. And you have to find ways to gather information and make parts of the world legible that are not otherwise legible to these models. And that's where you're going to provide a lot of value.
- [58:54]
C
One of our fun projects that we did last year, this was just sort of a side amusement, I think, for lots of our employees, was the think tank hackathon, right? And so it was just a Tim won know who has worn a hat at IFP for many years and now hangs his hat over at fai, you know, organized, you know, brought together a bunch of think tankers both, you know, at FAI and IFP and as well as, you know, some of our friends from other, you know, other organizations. And, you know, we just spent a day hanging out on the rooftop, you know, hacking away and trying to build a, you know, like little tools that could accelerate think tank productivity.
- [59:29]
A
Right.
- [59:29]
C
And, you know, I like the products that came out of that. You know, there's some really funky and interesting little tools. But what I'm really excited about is to rerun this this year with the advent of coding agents that actually work. So whether it's Claude Code or GPT 5.2 Pro, high on Codex, right? You can get an amazing amount of work done in these kind of vibe coding hackathons that you probably wouldn't have been able to do even six months ago. And I'm really, really excited to see what folks come up with. I will also shamelessly plug some of our team's substacks. You can check them all out on thefai.org, but senior fellow Dean Ball, who was most recently working in the White House as working on President Trump's action plan, he's also a diligent LLM user and I think he's got a really good take about how this does really augment and scaffold a lot of the work that he does. But at the end of the day, the writing and research and analysis he's doing is still mostly powered by him. It's just that the coding agents that he's spinning up and the, the research agents that he's spinning up are producing way more useful and high salience and high signal information that makes his work workflow a lot more productive. And so I recommend checking out his post on kind of how he uses coding agents to augment his work. And I think a lot of think tank researchers can learn from Dean and kind of use in the way that uses it. Last thing is just that I think I disagree that LLMs herald the end of the think tank. I think that's a bit tongue in cheek Caleb, but do ultimately believe that most of the work that makes a think tank worth paying for is the relational work. And as long as human beings trust other human beings to govern them and to staff the folks that govern them, I do think think tanks particularly like more focused applied think tankery in the way that both of our organizations like to engage in is going to be a valuable service that's worth donating to if you care about seeing good policy outcomes in the world. And you know, an LLM can make a lot of the background work that goes into doing that sort of relational work. Well, it can make a lot of that go faster. It can maybe surface new and interesting ideas or routes of exploration that wouldn't have otherwise come to your attention. Right. But I'm skeptical that if anything, I just suspect that the relationships are going to become more and more valuable over time and that again, as long as we have humans running other humans, this is going to be the window the world works for quite a long time. I'll also add this. I don't know if this is useful, Jordan, but to your earlier point about hiring, one thing I think people appreciate about think tank life, as opposed to corporate affairs life or even startup life or working in government, is that they're free, right? They're much more free to think and say exactly what they want to think and talk about. I think this has certainly created some costs for us in terms of FAI has an open voice policy as opposed to the sort of one voice policies that you see at other large organizations. So this means that we do allow staffers to disagree with each other publicly on including on issues of within their policy vertical, which can sometimes create genuine confusion on Capitol Hill or genuine confusion among our donors because they're like, well hey, I thought you guys believed this. And it's like, well, yes, some of our folks argue this thing, some of our other folks argue this thing. But I do think that's really valuable because it has created again a sort of intellectual community of friends where even though we disagree with each other sometimes extremely strongly, we do always want to sort of like model like virtues of good debate and like healthy, you know, contestation. Because we do think that ultimately like elicits a lot better, you know, a better, better outcomes. Right. It lists stronger arguments and stronger research and stronger outreach and stronger impact ultimately. Right. You know, so it's a cost that we're willing to pay. I think a lot of other organizations choose to avoid that. But I think it's something that's also like a selling point to people is that like, you can come here and think and you know, within like a sort of like very broad lane of FAI thought. Right. Like you can read our about page to see our sort of like general values, you know, around like, you know, human flourishing and human dignity and technological innovation. Right. But, you know, as long as people are operating within this very, very broad frame, we want to like preserve a lot of internal dissension and a lot of internal discussion. And I think that's been a really large selling point for people over and against other opportunities that they might explore.
- [63:54]
A
So I had a conversation last night that really bummed me out with a person who was like the perfect Statecraft interviewee who spent a long time in government and just left. And I was like, dude, you gotta be writing stuff. And he says, no, I'm worried maybe the next administration might not want to hire me if we're not 100% ideologically aligned and someone sees my stub stack post saying X but that principle believes in Y. I would just say to everyone who thinks that you're wrong, A, you're not going to be a Supreme Court justice and B, like, like the upside of, you know, making having like more quote unquote renown. The, the nice feeling that you will have to like live not as if you live in a police state or like, you know, are literally subject to a party, be that the Democratic party or the Republican party, but as a sort of like thoughtful individual who has like things to say. Like, yes, if you are working in government and go in and out, there will be times in your life where you can't write and speak freely and like say your mind and you will be part of a team where you're working towards things beyond your kind of like self expression when it comes to ideological or policy opinions, but when you're not in that stage, not. I am deeply convinced that the upsides, especially in, you know, the 2000 and twenties, outweigh the downsides when it comes to writing thoughtful things on the Internet. Caleb and Max, thanks so much for being a part of Chinatown.
- [65:48]
D
Stab a light you think that's where it ends we wrote the draft we line of votes we texted all the friends showed up at the markup language in our hands you were on panel explaining nobody for follows up nobody follows up everybody's got a white paper everybody's got a take nobody does the homework nobody stays up late nobody follows that's what we did did we move. If you need a PhD to offer at your shop we got a B baby face kid who text the counseling won't stop comms team calls the reporter she says put the expert on we wrote it so they close it you were on the panel John project based funding is a beautiful scam check the box and the PDF say damn what a year of impact but what move you delivered the deliverable we delivered the law 40 tags deep in a signal thread your institutional review never saw.