Ep 162: Brian Chau Built an AI That Uncovers Fraud — It Just Found a $21 Billion Racial Wealth Transfer

0:00I'm exposing government spending. I'm exposing misconduct. I'm exposing possible fraud. The crazy exanthropic guy, he was part of this left-wing NGO. They're going from one version of Doom to another. We run these long-running AI queries, 24, 48, 72 hours, connecting the dots between all of these financial payments. You won an IOI gold medal as a competitive programmer. The people that I did, I allied with. Some of them founded billion-dollar companies. Some of them made a theoretical breakthrough. And here I am in the dirty news business. We're excited to have you here to break your latest story.

0:36Brian, what did you uncover? Let's highlight a few other recent investigations. They funded the journalist writing the article, the publisher, the subject of the article, and all three experts quoted in the article. There are people who will complain about this, but I think this kind of transparency will be embraced as just like a fixture of reality.

1:04Brian Chow is the youngest American to win an IOI gold medal, one of the top programmers in the world at a young age. A lot of his friends have gone on to build companies worth tens of billions of dollars, have solved important math and physics problems. He's taken a different direction. He's an investigative journalist who started Effort.News. Why is he doing investigative journalism? Well, it turns out there's new possibilities with AI. They can look at government data, can look at what's going on in our society and find all sorts of crazy stories, tens of billions of dollars of grift, all sorts of schemes none of us have heard about.

1:35Brian's getting a huge amount of attention for his journalism, which you hear about the one we're going to break today on the episode. Really excited to have with us today, Brian Chow of Effort.News. Brian, thanks for joining us. Thank you for having me. Brian, you won an IOI gold medal as a competitive programmer. You were in forbidden courses at UATX a while ago. That's how we originally met you, our university's program a few years ago. Welcome to the show. Tell us a bit about your background. Where are you from? Yeah. So it's always interesting when people ask me this because I have very few memories of my childhood, but I was born in San Jose.

2:07When I went to high school, my parents moved me to this magnet school in Canada called Wolburn Collegiate Institute, which was known for two things. It It was known for producing something like 20 % of Canada's IOI medalists, this like big international programming contest. And it was known for doing something similar with robotics. So I did that. I did that program, ended up being the youngest person in North America to win an IOI gold medal. How old were you when you won? I was 15. Very good. Wow. And then you went to the University of Waterloo for a bit in machine learning, but then you left to work on AI policy in DC.

2:45Tell us about that. Yeah, it was this very crazy time because when I was studying and then going into machine learning, it was not a political issue at all. It was just like, oh, it's just something you did. It was kind of like, you know, writing software. It was like, you know, writing like Postgres, something like that. And it was not a particularly political issue. But then around 2023, it really went into the spotlight or 2022, I think, really went into the spotlight first with Biden's AI executive order, which all sorts of crazy stuff. And then with ChatGPT and that entire release into the public sphere.

3:24And at the time, I had just been writing like a political theory blog about like Leo Strauss and Carl Schmitt and all these like all these thinkers that people have that people read about in theory that like went viral for all sorts of other reasons. and I ended up being in a place where I was talking a lot increasingly about machine learning and what was happening and this increasing wave of pre-censorship that was happening in DC. And at one point, I was just getting really frustrated and I wrote out like my entire plan of how I think an organization in DC to fight against this stuff would work.

4:05And fortunately, some people actually offered me money to run this organization. And what was that called? wildly successful. The organization was called Alliance for the Future. We ended up doing a lot of work in repealing the Biden executive order, but also putting together the directives that would make sure that that entire apparatus that it created in the executive branch would end up being repealed and the equity mandates being removed and all these other mandates being removed. And what did you learn about journalism when you were running Alliance for the Future that led to Effort News?

4:37Because right now you're doing just amazing work we're going to talk about today in investigative journalism. How did that come about? Yeah, I think the biggest takeaway was from the other side of the desk was that more than 80 % of news stories are solicited. A source comes to you, they say, I want you to run this story. And I actually don't blame most journalists for doing that. That's like a lot of the time, that's a real newsworthy story, right? That's a real source. That's a real document. It's not entirely fake, but it distorts what people see. And what we ended up doing now with Effort News is querying all of these massive government databases using AI.

5:18Because there was this promise since the Nixon era, basically, that sunlight would be the best disinfectant. That you would have all of these public disclosures, legal requirements, and that would push all the information to the public. and then the people would know. But it didn't happen because people didn't have the effort. People didn't have the time. People didn't really look into these documents. So no one's putting in the effort to take in all these like messy PDFs and databases that are public. So you're putting in the effort. That's why the name, huh? Yeah. Yeah. That's why it's effort news.

5:50And we run these long running AI queries, 24, 48, 72 hours, putting together, connecting the dots between all of these financial payments. So there are large government databases in HHS, with Medicare and Medicaid, with defense spending, all of these large central databases of how the U.S. spends its money. And when we queried these records, it was originally the original idea was that we were ramping up to something like audit preparation that we'd be able to have these big aggregate statistics. And that would really teach us something about how the government works. But what we found was actually a lot more extreme than that.

6:34I started out by just looking for some wild cases. I essentially put the prompt. I had done some optimizations for the search process, both in terms of cost and in terms of how much it was searching. But the question that I asked it was essentially, find me newsworthy items from the HHS grant table. find grants that expanded a lot in terms of absolute or relative value. And then it found all of these stories that I started asking around to a lot of journalists on areas like immigration or DEI, like the biggest news areas, basically in the entire country, in the world. And it found brand new stories that no one had ever heard of.

7:18So people were getting back to me. They were saying like, I have no idea what this is. And at that point, my job was just on the floor. I thought, oh, the crown is totally in the gutter. I just need to do this 24-7. I love it. And why was this not possible a few years ago? I mean, it seems like this is something that AI has just gotten better, right? Like what's AI able to do now that it wasn't before? Yes. Part of it is the long context breakthroughs. Part of it is that you can just run the LLMs in these, what are called, you know, long time window tasks. Something like, you know, two or three days or even longer.

7:53And that's been useful in all sorts of areas, right? That's been useful in proving these open problems in math. That's been useful in doing drug discovery. And it's also useful here. But actually, it's interesting to me, I theorized that this kind of company would exist for basically like three or four years from now. So I thought this would materialize a lot sooner than it actually did. And I think the reason why it didn't is that the skill sets to pull off this kind of thing, they just very rarely come together where, I don't know, it just seems like that a lot of people who are good at the technical side are kind of okay with slop writing.

8:38And a lot of the people who are good at writing don't really get the technicals. That's right. It's two very different circles as well, in my experience, of people who can win an IOI gold medal and then people who care about HHS, wasting tens of billions of dollars on some griff scheme that the Biden administration set up. For whatever reason in my life, you're probably the only person, actually, in those two specific diagrams, which is great. And so I think we all have kind of unique interests and passions. And because you have both, you can do something that no one else in the country is doing, it seems.

9:10Yeah. The people that I did IOI or U.S. Nationals with, some of them founded, you know,$20 billion companies. Some of them, you know, made a theoretical breakthrough in how fast you can run Dijkstra's algorithm. And here I am, you know, in the dirty news business. I love it. I love it. And I want to ask you, Brian, do you think of effort news as nonpartisan? I mean, media is so tribal today. Obviously, you found probably more things that the left did recently. They were the ones in charge you're watching recently. Are you going to look into things the right's doing as well? Like, how do you handle this?

9:47Yeah, we publish everything that's fit to print. And we published an investigation, for example, on Freedom Fuel, which was this like Trump connected set of gas stations. I think that we broke a lot more stuff than was previously reported on that. But it just kind of didn't really go anywhere. In general, our philosophy is that there's all this information hiding in plain sight, and what might come out of it might just be that reality has a bias. There was this phrase that became popular around the Vietnam War of there's no such thing as a pro-war war documentary. The idea was you show people the footage and everyone, or at least 99 % of people are going to be like, oh, I don't want to be in the Vietnam War anymore.

10:35This is not going well for us. And I think it's the same thing where if you actually show people the financials, if you show people the documents that the government mandates itself to publish, actually very, very few people would agree with this. And then there are people who are benefiting from it, or there are people who have disproportionately succeeded in the old media system who will complain about this. But I think we are actually shifting into a completely new equilibrium where this kind of thing will be, this kind of transparency will be embraced as just like a fixture of reality. It will no longer, it will maybe be considered partisan in the short term, but will not be considered partisan in the long term.

11:21I love it. Well, somebody who thinks there's massive amounts of waste in defense spending and HHS and so many other areas and so much script. I love this. And I think we're excited to tell our listeners, we're actually going to break some news with you today. So let's do it. All right. We're excited to have you here to break your latest story on HEI health equity incentives. It's called how COVID era financial engineering created a$21 billion racial wealth transfer. Brian, what did you uncover? What is this? Yeah, this is this fascinating story about how health policy or policy more broadly works, because it was this sweeping wave around the era of COVID and BLM that came together, right?

12:05It was the medical and it was the racial stuff. And the net output of that is that there are these programs across the country, including in many red States, which was totally surprising to me, that pay providers to reduce the health disparities. And what that means is essentially that the more providers provide disproportionate care to Black and Hispanic people and implicitly less care to White and Asian people, the more they get paid by these incentives. And that's a total of a$21 billion pool across all these state programs, almost all of which are taking federal Medicaid and Medicare dollars and using them to create these incentives for providers to selectively deny or grant care based on race.

12:55And this is absolutely crazy. So they're focusing on race. I mean, I think we were in a healthy place, I think 20 or 30 years ago, where we were focusing less on race and you treat people based on their character and who they are and you don't focus on their race. And they went all the way back the opposite direction so far as to say we're going to give you more money if you treat more people of a certain race and less people of another race is blatantly illegal, isn't it? Yes. So under the plain text of the Civil Rights Act, this is illegal. Actually, at the time, this was a debate that came up because there were critics of the Civil Rights Act who thought, well, this could be used preferentially.

13:32This could be used against white people, especially. And the authors of the bill insisted that actually this applied in every direction. You could totally sue for anti-white discrimination. And in practice, that wasn't enforced for a long time. In 2025, that changed. The DOJ announced a revision to its guidance, which included that for basically the first time, they would begin enforcing these judgments and enforcing the letter of the law in favor of people who are discriminated against from any race. And in 2025 and 2026, they had settlements in two cases like this, or they had reviews in two cases like this, where they dealt with a health program in Minnesota and a loan program in Illinois.

14:22They opened up a review and forced those programs to remove the racial requirements. So I think something like that could happen here, where DOJ Civil Rights Division under Harmeet Dillon could step in, could open a review with all of these programs, and could, first of all, impose these requirements. And if those requirements aren't met, actually work with the Center for Medicare and Medicaid services to cut off the federal funding, because all of these programs are receiving Medicare and Medicaid funding, and then using this to create these incentives for what is in effect racial discrimination.

15:04This is great. Well, we're friends with Dr. Oz, who runs CMS. We should talk to him about this. I have a question for you on this, Brian, because you've been studying these things. Why does it seem like the left is so much more competent at setting up these kind of grift schemes than the right? Is it stopping them? Is it in balance? Is it just my impression being on the right that I see the left doing them more? Are they actually doing them a lot more? It was really competent. And that's the history that we dive into in this article where these programs, even though it seemed like at the time they were all springing up around COVID and BLM and this was It's like sudden, like big paranoia, you know, social cascade.

15:44In fact, the groundwork and the reports have been laid for more than 10 years. Obamacare created these funding programs for health incentive research, which, of course, they put out these grants. They paid, for example, the Rand Corporation, they paid a bunch of other NGOs to put out these reports of like, oh, this is how you should do racial incentive programs. And then those reports then paid the salaries of a bunch of activists who all advocated for those things. And Massachusetts actually brought in an organization of 100 different of these NGOs sort of advocate for these programs, which eventually passed and became the second largest of these programs next to California in the country.

16:30It's like a massive activist scheme where they planned it for years is what you're saying. And how did you uncover this to write about it when no one else has been talking about this? We just asked for racial discrimination in Medicare and Medicaid grant tables. So there's a federal Medicare and Medicaid reimbursement centralized database. And we had the infrastructure open and we just launched these like big open queries. And that's something where I think investigative journalism will completely change. because with the total amount of effort that humans can put in, even the maximum amount, even if you're working 24 hours a day, you could not go through the entire table.

17:13But for us, we can have thousands of misses. The AIs will pass over the misses and all of the hits are manually reviewed, but we can have like a thousand misses for every hit and we can exhaustively search the table. And this is what we found. I love it. Let's highlight a few other recent investigations. You recently published an investigation into the United Kingdom's rapidly expanding counterterrorism program, but it's not actually going after actual terrorists. What's going on in the UK? Yeah, this was fascinating. So the thing that we really added is that, so we found that of the identifiable motivation, so of the terrorist attacks where a motive was identified, 97 % were some form of Islamist terrorism and 3 % would be, even by the most expansive definition, some form of far right or right wing terror.

18:08But in practice, that enforcement is done completely differently because there are multiple reports now of of these programs that were focused on targeting what they would call the extreme right wing terrorism. And we can actually go into what they would consider extreme right. I think it's very different than what you or I would consider. They consider Douglas Murray extreme right. Douglas Murray is like a mainstream figure who's mocking them. So that's what they don't like. Yeah, it's funny because now he's seen as like the sort of centrist guy, right? He's now in the conflict of people who are more right.

18:45I still don't think, you know, those people are terrorists. But, you know, now in conflict with people who, you know, many people would consider to the right of Douglass in the US. Of course. But he's like, honestly, like I'm a pre-moderate figure. And they're also going after, interestingly, parts of the political canon, including John Locke, Thomas Hobbes and George Orwell. They literally put 1984 on the list of books that made you a suspect for terrorism. It's like this parody of themselves that they're saying that if you if you read 1984, you're a suspect. It's like the ultimate big brother.

19:22And by the way, John. Don't read 1984. for my first time i ever got a b in a class was at stanford uh and it was i i wrote a defense of john lock and the teacher was so angry that i would defend the man by the way who's the core of the enlightenment life liberty and property like a big part of how our civilization works is john lock the fact that's considered for i mean she didn't like it either she was far left but the fact that's considered officially now tied to the extreme right it's crazy yeah and And it is this incredibly disproportionate. I have to interrupt. I got this. I got this prize for a speech I gave like some years ago.

19:57This is John Locke in Latin right here. So I guess by having his picture here. Oh, no, you're part of that extreme right terror. Yeah. They're going to send you to the prevent to prevent. They got to come. They got to come watch me here. Yeah. Oh, my goodness. They'll lock you in the gulags with Douglas. That's actually hilarious. I mean, he's considered extreme, man. Yeah. And it showed this clear policy evolution because originally it started as a counter-Islamic group. And it started that way because that was who was committing the terrorism. And that didn't change. And even with the massive disparity in enforcement and how they're looking into these people, we had these amazing statistics in the article.

20:53if you take the ratio of deaths the ratio of deaths caused by you know what they would call the extreme right but you know some some form of right-wing you know vaguely right-wing terror those three people compared to 97 caused by islamic terror but they were very offended by the right-wing people which to them is almost like dying when they read it and they get angry maybe So that's 32 to 1. And on the other hand, the arrests were around 2 to 1. So two terrorist arrests for Islamic terror compared to, or sorry, these are people who are currently still in jail. Two people motivated to commit terrorism or have terrorist charges who are motivated by Islamic terror compared to what they would call extreme right terror.

21:43Which I'm picturing like a guy in Tweed reading an 18th century book. We can get that up on the screen. So UK is losing its mind. The West is in a rough place right now. Let's do one more of this story. Speaking of the crazy stuff going on in the West, I think one of your biggest stories to date that I've seen a lot over the last week online is how effective altruism bought the media. So effective altruism is something that's been funded by a lot of people, especially Dustin Moskowitz, who's a big character on the left, a former friend of mine. And AI doomers, it seems you're finding. These are people who want people to be very afraid, want people to be against AI or to be scared of it.

22:22They're funding fake anti-AI articles. Tell us about how you uncovered this. Yeah, this is a fascinating story, the extent to which they fake it. Because there was this Guardian article where, first of all, they pay the Guardian. So there's two foundations, Open Philanthropy or Coefficient Giving. They renamed the NGO because it got so unpopular from Open Philanthropy to Coefficient Giving. But that's Dustin's NGO and then Jan Tollen's NGO, Survival and Flourishing Fund. And they're kind of downstream grantees. They funded the journalist writing the article, the publisher, the Guardian, who published the article, the subject of the article and all three fake experts quoted in the article.

23:09So they had six people involved in the production of the article, all every single one of whom was funded by the same ecosystem. And there is this total of$40 million being spent so far that we know of. It's probably much more being spent to pay off influencers, to pay off all sorts of YouTube channels, Curse Gazette to pay off like TikTok creators, things like that, in addition to the salaries of all these journalists. And one other big thing that you might have seen was the Time 100. So the Time 100 is really the Time 42 because 58 of the 100 profiles were bought by one of these open philanthropy funded programs.

23:54Yikes. Yikes. I think that's the one that left out a bunch of famous people. Yeah, yeah. Jensen's not in the top 100 most important people in AI. Yeah, the creator on video who's running the most important company there is not. It's totally crazy. So this is what's called astroturfing, right? Because they're just trying to create a movement. Did you want to get into the crazy like ex-anthropic guy too? Yeah, this is the most recent thing that just came out recently where he went to work for there for like six weeks. And then he seems like he's, and then he said, oh, these companies are going to destroy the world.

24:25And then all the politicians are all at once like saying, OK, we have to regulate this. Tell us about this guy. What do you found out? Yeah. So he was part of this left wing NGO, which whose organizers were like ex-Hillary Clinton staffers, ex-Extinction Rebellion, which is like big like climate doom version. Right. So they're going from one version of doom to another. Was this big prison abolition person in the UK. um so there and and very funny there was a sympathetic profile of them in the guardian where they said uh they were this ngo was trying to hack democracy this was this was this was a positive thing um this is a wild time oh and you know this this actually takes everything full circle do you know what's the name of the ngo what's the call the name of the ngo is unironically called Newspeak.

25:18Oh gosh. Yeah. At long last, we have invented Newspeak from Orwell's 1984. We finally done it. We finally actually created it. Do these people see themselves as somehow they're just like, know better for the world what we're going to do and they're just going to trick us and like astroturf us and push us in a direction? What's going on here? Yeah. I mean, it's hard to project yourselves into their mind but i had a really weird experience with the eas um i remember in 2022 uh tyler cowen recommended i go to one of these ea conferences just to just kind of like know know what the space is right i don't think he was necessarily like endorsing them in any way so you know not not blaming him but um he recommended that i go to these conferences just to just see what what was happening and there was such a strange culture of like saying one thing to me and then like admitting they were lying and then like going on to say another thing as if i should absolutely trust them there was this big term that went around over and over which is this term called info hazard and an info hazard um it sort of morphed into the term malinformation um but it's basically something that's true that would be bad if the public found out.

26:35This is definitely what the Biden White House believed as well, right? Because you saw them pushing Facebook to censor things that were true, but they didn't want the public to know. It's a very kind of dangerous, sick thing in my mind that there's rather than stand for truth and beauty and good, which is like the classical foundation of our civilization, it's like actually they want to push society in some way. So therefore they want to hide certain truths and distort things. Yes, it's the replacement of the classical liberal conception of the Leviathan, which goes back to Hobbes, with the Lovecraftian conception of the Leviathan or Lovecraftian conception of the cosmic horror, which is forbidden knowledge.

27:13Actually, if you know this information, it is a curse upon you. You will turn into a fishman or something like that. I don't think Biden unironically believes that you will turn into a fishman. But that was the implicit message was that if you had this forbidden information, if people were able to know, both in the context of social media and in the context of Biden's executive order, which mandated things like equity or diversity representation in AI, if you knew what was really going on with scientific information, other types of information, if you had access to the truth, that would somehow turn you into some kind of racist or some kind of fascist, that was their metaphysics.