Understand the terms behind the conversation.

there should be a lot more companies going public, not at the SpaceX level, not OpenAI, not Anthropic, but the $100 million IPO. Because if AI does what AI knows, we all know what it will do in terms of disruption, then you want to have some sort of currency that allows you to buy all those companies. Like you guys were talking about buying different legal companies, right? Well, if you don't have that currency, the stock is currency, you're going to have to go out and raise money to do it. If anything happens, and that money is not cheap, that money is really, really expensive. Versus if you go out and you do $100 million IPO or a $50 million IPO, and you can do a secondary or whatever as many times as you need if you're performing.

Seek to 4:16

kids will not have an adaptive learning platform as their primary way of learning.

Seek to 1:06:53

And then in the last few years, what's emerged is what's called a harness, right? And a harness just wraps the model, right? Like Codex or Clodcode or Hermes or OpenClaw, these are harnesses. You know what the interesting thing about open source versus closed source models? When you take a closed source model, it does really well on a benchmark. When you wrap it in its own harness, it decays in capability. When you take an open source model and you use any other open source model, it improves in capability. What is that trying to tell you? What I think we are finding out, and this data is just going to be a tsunami over the next year, is there are all kinds of ways of using open source technologies that are meaningfully more performant and dramatically cheaper than the closed source alternative.

Seek to 27:46

They have to be put in a harness that creates context and memory

Seek to 26:00

That will make this thing go so freaking viral. If you can just add, because then what's cool about the agents is you actually have an agent swarm. You actually have lots of different agents performing different tasks and they get good at different things. Some people ask why don't you just have one agent? It turns out that if you have multiple agents, then the agents develop more context and expertise. So they get more specific. But yeah, you could then use each of those agents as a channel and invite people to it and then they participate in that.

Seek to 8:29

There's a whole, you could say team of agents.

Seek to 25:27

I think AGI has basically been here since the beginning of the year.

Seek to 2:47

And what you now need to do is make that autonomous agent much more informed and give it context and allow it to do a job much more intelligently. So meaning, you take a brain, you give it arms and legs and limbs and eyes, etc. But then you have to train it to be a lawyer or be a customer service rep or be a sales agent or what have you. And to do that next step, you need a ton of contextual information. And so the people that control that today, the large systems of record, what I said before and what I'll double down on now is that those guys hold an incredibly special place in the ecosystem if they do it right.

Seek to 12:09

That will make this thing go so freaking viral. If you can just add, because then what's cool about the agents is you actually have an agent swarm. You actually have lots of different agents performing different tasks and they get good at different things. Some people ask why don't you just have one agent? It turns out that if you have multiple agents, then the agents develop more context and expertise. So they get more specific. But yeah, you could then use each of those agents as a channel and invite people to it and then they participate in that.

Seek to 8:29

So the way I would frame it is Capex is high alpha, low beta in data center infrastructure, that capital. And model development theoretically could be high alpha, but it's very high beta. It's a very risky way to deploy capital. So if I'm the board, I'm the management, I'm deploying more capital in computing infrastructure, less capital into model development. That's what I think is going on.

Seek to 3:42

the guardrails have prevented them from having the latest and greatest American models.

Seek to 35:22

twice yearly AI literacy classes designed to help them think critically about AI before they begin to rely on it.

Seek to 1:12:05

anthropomorphizing, fancy word for making computers into humans or sentient beings

Seek to 19:56

Then at some point in the workflow, a frontier model may or may not come in to check it, add to it. That's what I mean by a composable model, when you have a symphony of models working together with the frontier models being maybe the conductors. But that's what I mean when I say composable.

Seek to 51:16

Isolating dictators, waiting them out, is always the better approach than starting full-scale wars with them because you don't know what could happen in an escalation. So simply not worth the risk, much better to just send the air troops in and clean it up. And we did not need to go collect this material, obviously, because we have these bunker buses and we can just slam it every X number of months and that would have been a much better approach, just like we've contained North Korea for all this time and we contain other dictators with weapons.

Seek to 1:21:32

And we're now looking at a 30-year 5.2 and a short-term rate over 4 So the federal government has a problem because over the next 12 months, they have to refinance $10 trillion of debt. That debt is coming due. Those bonds are now due. They have to pay the principal back to the bondholders, and they got to go back to the treasury market and sell more treasuries to borrow more money to refinance. So the borrowing cost now is going to climb up. And when that borrowing cost climbs up, the federal government's burn goes up and the fiscal deficit goes up. So my theory and my argument on this is there is no action that Bessent can take that's actually going to have a meaningful effect on the long end of the curve. We have fundamental fiscal spending problem with the federal government right now.

Seek to 35:43

Yeah, so my understanding of HBM stands for high bandwidth memory. Again, this is part of the GPUs that go in the data center to run AI is that you take the DRAM wafer or die and you actually stack them. And so I think HBM 3 is like 8, it's stacked 8 dies high, but now they're increasing to 12 and even 16 And to basically stack them and then package them all together, that's an advanced technology in and of itself. So you're seeing now, like Gavin's saying, there's only 3 companies that can do it. But also this is creating significant price pressure for all the consumer electronics businesses. Apple had huge news today where they announced massive price increases. And again, it's because DRAM now is less available because it's just being hoovered up by all the data centers.

Seek to 1:07:29

Sacks, the long-term issue has been the enrichment of uranium to the point that it can be used as a nuclear weapon. Is it part of the deal that all that enriched uranium that we do know there's a stockpile of gets removed from the country? Because if that goes, I personally feel very good about, let the citizens deal with what they have to deal with.

Seek to 1:19:27

After 110 days, President Trump has announced for the 37th time that the war has ended and that the initial agreement on June 15th will be codified. A lot of details here, a lot of unknowns. Conflict began obviously on February 28th. Formal signing happens Friday, June 19th in Geneva. The deal was mediated by Pakistan. It extends the ceasefire for 60 days. It includes a Lebanon ceasefire. The US gets the Strait of Hormuz reopening, which it was open before the war obviously. Iran commits to not developing nuclear weapons. Iran agrees to destroy its stockpile of enriched uranium under IAEA supervision and that they will freeze its nuclear program at current levels for 60 days. Iran gets all sanctions lifted.

Seek to 1:14:26

And so, to me, that's what's going to happen. To me, a model is like the database market. It's got its own differences, but I sort of somehow think that there are definitely going to be frontier models that are closed source. There are going to be open source models that are going to be frontier class. In fact, if anything, I think in this next year, what will be probably a big part of the discussion is, what's the future of a firm? A firm should be able to take the tacit knowledge it has and embed it inside a weight in a model that they control.

Seek to 23:14

Next topic for sure. Oh my god. Strong first topic. I didn't know you guys were going to go All-In, so to speak. Topic two. Chinese open source models appear to be catching up with the US frontier models. Let's start with a GLM 5.2 released by China ZAI. This is a frontier class, open source free to download Anywhere model, 744 billion parameters, 1 million token context window, and it's under the MIT license. If you don't know what that is, open source licenses have very specific.

Seek to 45:04

Yeah, so my understanding of HBM stands for high bandwidth memory. Again, this is part of the GPUs that go in the data center to run AI is that you take the DRAM wafer or die and you actually stack them. And so I think HBM 3 is like 8, it's stacked 8 dies high, but now they're increasing to 12 and even 16 And to basically stack them and then package them all together, that's an advanced technology in and of itself. So you're seeing now, like Gavin's saying, there's only 3 companies that can do it. But also this is creating significant price pressure for all the consumer electronics businesses. Apple had huge news today where they announced massive price increases. And again, it's because DRAM now is less available because it's just being hoovered up by all the data centers.

Seek to 1:07:29

After 110 days, President Trump has announced for the 37th time that the war has ended and that the initial agreement on June 15th will be codified. A lot of details here, a lot of unknowns. Conflict began obviously on February 28th. Formal signing happens Friday, June 19th in Geneva. The deal was mediated by Pakistan. It extends the ceasefire for 60 days. It includes a Lebanon ceasefire. The US gets the Strait of Hormuz reopening, which it was open before the war obviously. Iran commits to not developing nuclear weapons. Iran agrees to destroy its stockpile of enriched uranium under IAEA supervision and that they will freeze its nuclear program at current levels for 60 days. Iran gets all sanctions lifted.

Seek to 1:14:26

So then let me make this point. So Sacks, there was this argument that DEM has put out for a FINRA style regulatory system, where industry participants that understand technology, move quickly and are aligned with commercialization, actively review each other's systems and each other's models via an industry consortium that hires and assigns scientists to work on this stuff, and they're just doing it versus having it be owned and managed by the government and create some regulatory capture scheme. But then this week, you were pretty against the FINRA system in your tweets. You seem to be supportive of it the first time around when DEM has originally published it. It seems to me that the proposal from DEMIS that Dario signed up to, Sam signed up to, Elon all signed up to, you're now against, but that could be a solution if there is no genuine concern here, given Chamath's points. The industry can police each other.

Seek to 12:01

Now, how should the industry coordinate with each other? There are tried and true ways of doing this. Companies in the industry publish papers, they publish technical papers, they publish blogs. These are good ways of sharing best practices. They attend symposia and conferences and they present at these things. They do requests for comments as part of open standard setting exercises. There are ways of the industry collaborating on safety in very productive ways. But what is the common denominator of everything I just said? They are all open. They are not in secret. And that makes those judgments contestable to the point that Chamath made earlier about we need to see the thinking behind this. And what I'm against is these companies meeting in secret, whether it's part of some sort of like FINRA or otherwise, to set standards that are not transparent, that are uncontestable. And you know what's going to come out of that? Regulatory capture and anti-competitive behavior.

Seek to 17:33

Auctions are the way. Auctions are the way. And so it's not hard to price if you just do it the way it's supposed to be done. It's hard to price when you want it to be a certain number.

Seek to 1:36:07

He's not selling. He's under a one-year lockup and I predict that he'll hold on for much longer than that because creating SpaceX is his life's work.

Seek to 18:19

Yeah. We haven't seen the S1. I do think if you're bringing a company out responsibly, you want to price it in such a way that you can absorb the lockup, because you don't want to price it really high, create a lot of volatility. That's like very disconcerting for employees. You do not want anyone distracted at this moment. So I think that they will price it smartly. Maybe that is two trillion. Maybe it is two and it trades to three. But you just got to remember these leaks are always coming from the bankers who probably lost lead left, and they want to make lead left look bad.

Seek to 8:01

And they're sitting there and they're pressure testing models and they figure out, hey, there's ways you can jailbreak models to get them to do things that they're not supposed to do. And then you can use them to do things like design bioweapons, make viruses, make infectious diseases that don't exist today. You could use them to design cyber weapons to get them to go and do things online, like break into bank accounts and take people's money and all these sorts of things. And you can prompt them to do things that could broadly influence society by creating fake media, fake videos, fake images, propagate them through social media, hack the algorithms.

Seek to 8:24

our oil is this light sweet crude. That's good for refining it down to gasoline.

Seek to 1:22:37

When you say slush funds, the way the funding process works in the modern era is grants. You apply for a grant with a very specific set of objectives, and that's what the capital is going to be used for. To your point earlier, grants are typically approved if they're lower beta, which means more predictable in terms of the outcome.

Seek to 35:55

They still have to define it, again, this is at the MOU stages, at the level of high-level principles. They have to define exactly how that program is going to work. But what they get in exchange is relief from the sanctions. And they get a more normalized relationship with the US.

Seek to 1:20:29

I think the second is around quantum computing. So the president signed an executive order just last month, launching a new national quantum initiative. And one of the key goals there that we're trying to do is create a scientifically relevant quantum computer by 2028 This is a directive to the Department of Energy, and we're going to be working very hard to make sure we can hit that pretty crazy timeline. I mean, your camp, we really want to get the fusion by 2035, and we're trying very hard to get the resources allocated at DOE and combine with all the private sector investment to get there. And I think the last one, which is kind of our big flagship project for the whole administration, is called the Genesis Mission. And that is where we want to essentially double the scientific output of the United States by applying AI to our hardest scientific challenges and endeavors.

Seek to 22:51

nationalizing our American oil companies over there

Seek to 1:28:42

But just this week, if you pull up this image, China is installing this 582-ton magnet, superconducting magnet at their nuclear fusion center, which at this point is going to be the most advanced fusion system in the world. Incredible. 582-ton magnet, 60 foot by 40 foot for one D-shaped magnet. They put a series of these together and that creates the conditions for them to drive a sustained plasma, which is 100 million degrees celsius, protons spinning around, smashing into each other, creating energy from water, and then they can capture that energy. Unlike Europe, which runs ETER, and the US projects, none of which have actually fired up. This is the Chinese Academy of Science and the Institute of Plasma Physics. They ran this a 30-minute trial last year, and as this magnet gets installed and they start to bring this thing online

Seek to 28:20

I mean, the person that decides is the president of the United States. So a foreign leader isn't getting to call the shots in the United States. I think very practically speaking, the markets are effectively pricing in that this was a small blip. For whatever people think, that's just what the best prediction market that we have is telling us. I think that's important to acknowledge that we're probably in the endgame here. And the second thing to acknowledge is, if I was Israel, I would really be concerned that, unless I help find an off-ramp quickly, the risk that Israel loses America as a predictably steadfast ally could go down. And I think that that's problematic for Israel. Far more than is problematic for the United States. So all of that kind of tells me that we will find an off-ramp. A, because I think economically it makes sense. And then B, geopolitically, I think Israel will want to make sure that this doesn't burn a long-standing relationship.

Seek to 1:17:22

Well, there's two different parts to the Taiwan, one China policy, that's a political statement. And then nobody changes the status quo through the use of force. Those are two things.

Seek to 21:18

Next topic for sure. Oh my god. Strong first topic. I didn't know you guys were going to go All-In, so to speak. Topic two. Chinese open source models appear to be catching up with the US frontier models. Let's start with a GLM 5.2 released by China ZAI. This is a frontier class, open source free to download Anywhere model, 744 billion parameters, 1 million token context window, and it's under the MIT license. If you don't know what that is, open source licenses have very specific.

Seek to 45:04

Chinese open source models appear to be catching up with the US frontier models. Let's start with a GLM 5.2 released by China ZAI. This is a frontier class, open source free to download Anywhere model, 744 billion parameters, 1 million token context window, and it's under the MIT license.

Seek to 45:04

So we now have a Chinese open weight model that is as good as the currently available models from OpenAI and Anthropic. And look, this is a point I've been making really since I joined the administration, is that we are in a very competitive situation with China. I've been saying this from the beginning, our whole AI strategy from the get-go was about winning this AI race, defining it as a race, as being globally competitive. And we cannot afford to do things unnecessarily that slow our companies down.

Seek to 53:12

But there is a cop on the beat there. And I do want to caution that insider trading is not something that's necessarily allowed in our markets. But we do believe that markets are truth machines, that they do create a really powerful source of information. We've seen the hoaxes, the fake news and the manipulation of the polls. The prior administration tried to ban these markets ahead of the 2024 election. And they really increased turnout. It showed that they were correct when a bunch of the fake polls were put out right ahead of the election. So we really have to foster these markets here in the United States and make sure that they don't flourish in Russia or somewhere elsewhere. They really will turn out to be a source of disinformation. So we do believe it's valuable to have that trading and information flowing through the markets. But insider trading is still illegal here in the US.

Seek to 20:57

So we'll have a model out within a few weeks, been a little bit delayed by the World Cup and the redistricting stuff. Prediction markets say Democrats are around, I think, 80 or 85 percent to take the House and 40 or 45 percent to take the Senate. Those both seem pretty reasonable to me. I think if anything, it's a little bit low on the House. I think it might be more like 85 or 90 My models surprise me sometimes.

Seek to 49:56

Well, what if they could do that earlier? And I've beaten this drum on this podcast and my other podcast for a decade on this, but people knew SpaceX, they were watching these rockets launch for a long time and they would have placed that bet 10 years ago when the company was worth 100 billion or 10 billion or 500 billion, but they weren't allowed to because we have a corrupt system in our government that says rich people, the top 4 or 5% of this country are smart enough to buy private company stocks and the 95% are too stupid to. And it's antiquated and it has to change and you should be able to buy whatever you want.

Seek to 28:19

So what they're going to do is they're going to have the standard setting body say, well, we have to set the standards for AI safety. By the way, Dario and OpenAI, they're going to fund the whole thing. They're going to contribute all the compute. They're going to be behind it. They're going to be the ones coordinating with the government officials because, frankly, people in government have no idea how to monitor and control and set standards for AI safety. Technologically, this is way beyond them. So they're going to go to these companies and say, tell us how to do it. And so what will happen is the standards will get set, and then it'll be a very simple matter of fairness to say that the standards need to apply to open as well as closed models. The open models cannot comply in the same way, and gradually they will be shut out of the market.

Seek to 29:51

When you say slush funds, the way the funding process works in the modern era is grants. You apply for a grant with a very specific set of objectives, and that's what the capital is going to be used for. To your point earlier, grants are typically approved if they're lower beta, which means more predictable in terms of the outcome.

Seek to 35:55

Well, I think the revenues are a combination of revenues. So the ultimate goal of tariffs, the revenue collection, I think, in a way is a payback for the imbalances that have gone on over the years. But over time, the real idea is to balance trade and reshore manufacturing and bring our economy into balance with our trading partners. So what should happen is over time, tariff income will come down, and US tax receipts will come up, whether it's from factory jobs or more manufacturing and through higher payroll taxes. So we will start off at this very high level, then we will rebalance and come up. So I think it's difficult to know the timing.

Seek to 9:38

But what Nvidia is doing that I think is really smart is they're saying, hey, we are going to help create this market, grow this market, but they're going to provide two very important functions. They're going to say, hey, Blackstone, Goldman Sachs, KKR, Apollo, BlackRock. I'm leaving someone out and I'm so sorry. There are six of them. Come to us. If you have a deal, you can bring it to us, and we will give a residual value guarantee. And I think the way that that residual, which will further lower the cost of financing these, and that residual value guarantee can then be incorporated into those companies underwriting. And what that basically means is they're saying that after three years, four years, we guarantee that these GPUs can be rented at a certain rate.

Seek to 59:27

They still have to define it, again, this is at the MOU stages, at the level of high-level principles. They have to define exactly how that program is going to work. But what they get in exchange is relief from the sanctions. And they get a more normalized relationship with the US.

Seek to 1:20:29

Yeah, I mean, at the basis of what you call the scientific method, for many people here who might be watching that aren't familiar, it's like you ask questions. The process of science is a process of asking questions and inquiry, and that inquiry leads to experiments which collect data, and that data is what informs your view, and then you continue to ask questions. The idea that science is authority is almost antithetical to the basis of the scientific method, which is constant inquiry. Is that kind of fair?

Seek to 4:09

that's just based on the secondaries, you know, the small cashouts

Seek to 16:23

One point that really stuck with me was when the private equity guys said that this was a little bit like what they do for plain financing, where when an airline buys airplanes, they're able to finance it not just based on the creditworthiness of the airline, but rather because the airplanes themselves have value. And that makes it much easier to finance because you know that even if the airline gets in trouble, you're protected, right? You have asset-back financing. And that's what they're basically doing here with GPUs. Now, in order for that to happen, Nvidia has to standardize. They have to create essentially reference designs and so forth and so on, so that it can all be standardized enough for these Wall Street guys to now package it up and securitize it, turn it into securities. And that's basically what they're doing, like they did with mortgages.

Seek to 1:07:11

He gets paid 100 rupees and he's like, you know, the rupee is constantly getting devalued. I'm constantly losing purchasing power. I want to swap that into a US dollar. So he would create a crypto wallet. And what Tether will say is great. Give me your INR, your 100 rupees. They immediately swap it to a US dollar. So now there's a US dollar and there's a token for that dollar. I give Jason the token for that dollar. Now I have this dollar. What do I do with it? If when I accumulate enough of these dollars, 50 billion, 100 billion, I can take that and I would invest it in treasuries so that it's completely safe.

Seek to 15:40

So like you earn rewards and you earn reward points, but it's not really what it should be. It should be that if you earn that interest margin, you should be able to share that. And by the way, you should be able to have different rules in different markets because again, if you're in Kenya, the last thing you're probably thinking is, do I get the 4%? What you're more worried about is the Kenyan currency, whatever it's called, is about to depreciate another 60% this year. Let me just hedge that. And that's more than enough value. Anyways, I thought Paolo was incredibly impressive.

Seek to 18:52

Let's step back. The way I see it is, we got Xi comes to power, and we buy into this win-win stuff that he's selling, and it was never a win-win. China wants everybody dependent on China, and them independent of you. That is their theory of the case. Our markets are going to be so big and important, you can't do anything about it. You're going to be dependent on your economy, on our markets. We're going to leverage that to then destroy, manipulate, and control your sovereignty, and we'll be independent of you.

Seek to 12:35

And what you now need to do is make that autonomous agent much more informed and give it context and allow it to do a job much more intelligently. So meaning, you take a brain, you give it arms and legs and limbs and eyes, etc. But then you have to train it to be a lawyer or be a customer service rep or be a sales agent or what have you. And to do that next step, you need a ton of contextual information. And so the people that control that today, the large systems of record, what I said before and what I'll double down on now is that those guys hold an incredibly special place in the ecosystem if they do it right.

Seek to 12:09

I mean, look, I think that SaaS products can offer their own agents, and they can offer their own AI and co-pilots and all that. But I think the reality is, is that most of the action with agents is going to happen outside those products and simply want to interact with those products and basically get the data from them, use them as the system of record, as the source of truth, and then write actions back to them when appropriate. And I just think that founders should lean into that trend instead of fighting it. Because if you fight to stay in control and say that only our agents are the way to interact with our system, then people will just move to, they'll leave.

Seek to 24:21

China can do that too, and I guarantee you they are. I don't think this is going to give us a decisive advantage in the AI race. It's going to annoy. It's going to create annoyance. It's going to create friction. And how bad you want our relationship with them to be? Do you want to risk starting another trade war? Look, I'm not against restrictions when I think they're going to pack a punch. For example, I'm really glad that the first Trump administration limited the export of EUV lithography machines to China. That was all the way back, I think, in 2019 So that was a really important decision. And so, look, I think targeted strategic controls make sense. I would just make sure that this one actually meets that bar.

Seek to 1:09:06

They have been trying to essentially figure out how to do EUV lithography since we put the export controls in 2019 No breakthrough has happened. There's nothing more important to their economy than solving that scientific problem and being able to do that. So to me, I feel like one of the most special features about our system is we do have people competing. This sort of free market approach innovation is a huge feature of our ecosystem and one that I think is an important reason why we have all these breakthroughs.

Seek to 27:25

In the spring, the president put fentanyl tariffs on Mexico, Canada, China. They've all come to the table to help and partner with the US government to end this scourge on our people. We're seeing fentanyl deaths drop. Because of the good efforts of China, we made a good faith move and decreased their fentanyl tariffs by half, down to 10 percent. So, that's been national security.

Seek to 7:17

Well, I don't think it's against the administration. I actually think it's against the American people. And it will be, again, as I said, it will be a hit to national security. And the revenues aren't the focus here. The revenues aren't the focus. The revenues can be replaced, but all the things that President Trump has been able to do using tariffs on the national security side will be jeopardized.

Seek to 11:21

Let's keep moving through the dock at Bessent's new bond strategy. Got criticized by his mentor, Stan Druckenmiller. Last week, 30-year treasury hit a 19-year high, 5.3 percent. You've been talking about this a whole bunch, Friedberg, so I'm interested to get your take on it. In response, Bessent doubled the Treasury Department's long-dated bond buybacks from 2 billion to 4 billion. On August 19th, CNBC reported Bessent is considering ramping up bond buying, and Fox Business said he wants to put the fear of God into traders that are shorting long-dated bonds. So a lot of interventionism going on here.

Seek to 33:32

The tactical answer to the question that you asked before is, why is he in the market buying bonds? It's called yield curve suppression, meaning when you buy bonds, you bid up the prices and the implied yields go down, which allows Scott to then finance at a lower borrowing cost because then if yields are 5.1 versus 5.3, and then now you're issuing bonds, they're marked against that 5.1 print versus 5.3. So the yield suppression is trying to help what Friedberg talked about, which is just this refinancing tsunami that is coming. Okay, but look, just take a step back for a second. I think the very big issue is that Congress is ultimately responsible for spending. Congress is completely unable to do anything to get these budget deficits under control. It is consistent. It's not a Democrat nor a Republican problem. It is a congressional problem, meaning both sides consistently spend way more than they should.

Seek to 41:02

California's public employee retirement system, so CalPERS and CalSTRS, they provide the retirement benefits to roughly 3 million California public workers, and there's roughly a trillion dollars of capital in those two investment funds that are meant to support those retirees. They've been making about 7% a year compared to the S&P, making 11% here, and the current accounting estimates that they're going to be short by some estimates, 250 to 300 billion, by other estimates, as high as a trillion dollars in the years ahead, in paying out the benefits that they're legally obligated to pay to public employees as they retire. And you can't just change those benefits. There's a state supreme court case that's made that known, that you can't go in and rescind benefits that you've promised someone. So you are stuck with that liability.

Seek to 43:46

Well, what that has allowed us to do because of the legal limitations that you've mentioned is, and I'm not saying it's perfect, but this is just legally and politically what we were able to get to, is the best outcome we were able to get, is that we are on a long glide path of paying off all of the unfunded liabilities for the tier one employees, all of those older employees. It's a 20-year process. By the early 2040s, San Jose will have cleared the debt. Our general fund is going to be flush. We're going to be increasing staffing and service levels. And you'll start to feel that here much sooner. We are actually roughly at peak cost for unfunded liabilities today because we took the medicine and now we'll start that slow glide path where each year there'll be a little more room in our general fund because we've actually bit the bullet and took this on.

Seek to 47:22

The tactical answer to the question that you asked before is, why is he in the market buying bonds? It's called yield curve suppression, meaning when you buy bonds, you bid up the prices and the implied yields go down, which allows Scott to then finance at a lower borrowing cost because then if yields are 5.1 versus 5.3, and then now you're issuing bonds, they're marked against that 5.1 print versus 5.3. So the yield suppression is trying to help what Friedberg talked about, which is just this refinancing tsunami that is coming. Okay, but look, just take a step back for a second. I think the very big issue is that Congress is ultimately responsible for spending. Congress is completely unable to do anything to get these budget deficits under control. It is consistent. It's not a Democrat nor a Republican problem. It is a congressional problem, meaning both sides consistently spend way more than they should.

Seek to 41:02

Part of where the money's gone, though, is really just a sprawling bureaucracy that when we, in good years, have more money, we create new programs, we add headcount. The state's population has stayed flat over the last six years. Spending is up 75 percent, as you point out. And headcount in state employees is up, I believe, over 20 percent. So we're adding more state workers, we're pouring more money into public programs that aren't starting from the premise of what is the outcome we need and how do we most efficiently get there. I mean, it's time for California to go through an exercise of zero-based budgeting and say, what are the outcomes we need?

Seek to 49:10