00:03:20.580Take it as an aside here to sort of frame this for people.
00:03:24.640It's been five minutes and you've already been truth noob.
00:03:27.680I think a lot of people do not realize
00:03:30.620how much the Catholic church upper class,
00:03:34.620the cardinals the pope and and the bishops are profiting from the stripping of the catholic
00:03:41.840church right like as they lower the numbers of people like you're like well if the roman mass
00:03:47.140is more popular why aren't they doing it if they can get a diocese to like super low numbers then
00:03:52.900they can sell off all the stuff in the diocese and make a lot of money and they've been selling
00:03:56.560these very frequently recently to become mosques and what's interesting is they'll take these old
00:04:02.160catholic institutions and the spsx multiple times has raised the money to buy these old you know
00:04:08.640cathedrals and and catholic schools and stuff like that but they refuse to sell to the sspx
00:04:13.240and then they convert them into moths and they and they run off with the money but it's not just
00:04:17.920there it's also that they benefit from the leftist policies through bringing in just endless waves of
00:04:23.480immigration and so if you're wondering like why is this being pushed so hard for by the vatican
00:04:29.460They have written a number of things about, like, no, you need unlimited immigration into Europe and into the United States, is that they are directly, not just financially benefiting from this, but it is their primary source of income.
00:07:44.280This is the out-of-control exponential growth of the grant sizes.
00:07:48.740I was going to draw the exponential curves here, you know, at this rate, you know, the preferred communities program will consume all world GDP by 2030.
00:07:57.780That's a little joke about the AI discourse.
00:08:01.480So this was like right at the end of COVID?
00:29:29.300But it's also true that, like, it creates a number of checkboxes where you can just say, like, actually, these new competitors did not fill out the proper checkboxes. So it is this, like, somewhat procedurally effective gatekeeping mechanism.
00:29:45.520But once that you're right, that once that grant approves is approved, there's very little recourse and there's very little that you can actually do to continue applying oversight or to request further information.
00:30:00.080It's it's this very sort of one way door.
00:30:04.720It's really I just I don't know. It horrifies me. And now that I know that there's this additional way that people are getting a ton of extra money, it really freaks me out.
00:30:12.660so this is fascinating where can people find the article and effort.news what effort.news
00:30:22.840at the time of recording but it will be up at the time of releasing this do you have any other
00:30:30.440fun findings that you can give as like summaries for people that they can dip duper on oh he has
00:30:35.000other fun findings there's a ton of stuff that that we're launching we're launching with between
00:30:40.680five to seven stories depending on how much time i have to to typeset the other stories
00:30:46.680later today but um there is a lot there there is foreign interference by the british government
00:30:54.400pushing censorship censorship laws in the united states there is this one i think is one that you
00:31:00.440guys will also find really delightful were you guys aware of that like degrowth letter that was
00:31:05.280going around a few months ago. There was this degrowth letter. A lot of people made fun of it
00:31:12.220because it was like, the headline was like, the economists have done the math, you know,
00:31:16.600we need to agree with or something like that. It's not the exact quote. But of course, like
00:31:20.500economists were like dunking on this because of course, most economists do not believe that we
00:31:25.460need to degrowth. But this organization was funded by a number of European NGOs who were founded by
00:31:32.740and who received the majority of their money
00:33:32.480With the punchline being that people who were not basically black or brown broadly were being denied, in some cases, like prenatal services because they were not black or brown, which is race-based discrimination with federally funded programs.
00:33:49.160which is super not okay to do like technically like not not legal to do but it's still happening
00:33:55.980so this is against the equal protection clause of the the 14th amendment and the the civil rights
00:34:01.720act but it was not enforced until very recently there is a chance that we might actually get
00:34:06.380something from this because i talk about this in the article that's a policy that's changed very
00:34:12.100recently in the united states department of justice civil rights division it might be fun
00:34:16.900to talk about how i got some of these stories i would really like to like taking large federal
00:34:22.900data sets and then using some modified lm tools really more so from a cost perspective i think
00:34:29.100you could theoretically do this if you're if you're just using one of the vanilla models
00:34:32.880they'll cost you more and take you a lot longer i'm basically saying like find me a story it's
00:34:38.380like fine really wait not even like i want you to look in this data set for this specific thing
00:34:43.880you're just like i don't know do you think do you see anything interesting yep in some ways it's a
00:34:49.260lot more open-minded way to do journalism right i mean yeah but like dude i think a lot of the
00:34:55.140stories will be right waiting because like it's not new information stuff that happened like five
00:35:00.280years ago or more right most of these grants for the first story were issued in 2021 so you could
00:35:05.320have recorded this in 2021 conservative media i mean like like it or not and of course like the
00:36:22.680Yeah, so all these programs take federal Medicaid funding.
00:36:27.260They're all also both facially against the text of the Civil Rights Act
00:36:31.780and against the recent, you know, but historically it hasn't been enforced that way.
00:36:36.540But but also against the recent revisions in the Department of Justice Civil Rights Division rulemaking, which is here is basically saying, like, actually, full protection does apply to all races and you can't.
00:36:49.380So there are some cases previously where Title VI agreements were used to either force,
00:36:58.560essentially force states to open these programs up to all races, which is what happened in the
00:37:02.720past, or to threaten to cut off Medicaid funding if that doesn't exist.0.63
00:37:07.180So there is a remedy legally, which is that the Department of Justice Civil Rights Division
00:37:12.040can actually shut off Medicaid funding if they continue to racially discriminate.
00:37:16.060And there are two cases before where they started that process, once in Minnesota and another time in Illinois, both with these kind of race-based Medicaid rationing programs.
00:37:26.200And in both cases, the DOJ managed to get those programs basically opened up to all races.
00:37:31.240But not the case now. No one has done this.
00:37:33.460Yeah, I don't really know if they know about it. They have not yet responded to my request for comment.
00:37:39.520Maybe they will by the time this is published, but as of recording, they have not.
00:37:43.980so i don't know if they even know about this interesting maybe it's help right you guys get
00:37:50.620more people to know about this know about so they're literally birthing to try to be a civil
00:37:56.240rights violation so they're using medicaid money to try to increase the black birth rate like that's0.83
00:38:02.440that's why providing medical services only to in some cases only to black people in some cases only0.65
00:38:08.240to Black as well as American Indian, Alaska Native, Pacific Islanders.
00:38:13.140So yeah, they're selectively providing medical services based on race.
00:38:16.720And in this case, it's like basically like prenatal services.1.00
00:38:20.700And I think you even mentioned in the article from when I first read it that like one woman
00:38:25.140reported just being turned away when she asked.
00:38:30.380This will be a super interesting case of whether we can have more private rights of action
00:38:34.780based on discrimination, where historically this has been an enormous cash cow for a lot of the
00:38:41.940kind of DEI NGOs, where, you know, they'll sue on behalf of minority groups, mostly
00:38:49.580whether on behalf of people who are Black or on behalf of women. But really, you know,
00:38:54.520the civil rights, the text of the Civil Rights Act, a lot of these suits in any direction.
00:38:58.620This case can prove out whether there can be more of a rights towards discrimination on behalf of other groups or against other groups, and whether those will be rewarded a similar kind of settlement in court.
00:39:14.120So you have this program where, based on our name, seemingly like a white Hispanic lady went in to try to access one of these services and then was denied those services and is now filing a Title VI civil rights lawsuit about this refusal.
00:39:30.580And that stuff is kind of still in court right now.
00:41:08.140Yeah. And the thing about, you know, Harmeet posting about this is he actually, you know, is the main person to do something about this. I think has been one of the most effective members of the Trump administration. This might actually be, you know, I guess I should praise her more if I if I want us actually, you know, the stories I'm breaking to have some consequence.
00:41:29.860This is such an interesting effort, though, because it's both investigative journalism that we've really lost for 10, 20 years, maybe.
00:41:40.280Like, there just really hasn't been much work coming out like this that I would have expected to get.
00:41:46.180And all powered by AI, which is super, super cool, plus your brilliant mind.
00:41:50.740then this additional layer of activism which i guess i mean do you primary expect to enact it by
00:41:56.820asking people for comment and getting other people to do something or is there an additional branch
00:42:01.580of what you're working on where you're actually going to push harder for like to deploy people
00:42:05.700to take action on the information probably the most you know publication out there like the best
00:42:10.880news that is fit to print you know and part of that is also the data presentation i think it's
00:42:15.720like crazy thing for even like i was showing my friend the versell dashboard the other day this
00:42:21.820is very very successful software company used for for fun and hosting for for a lot of sites
00:42:27.100but really it's not just versell it's like every series a or b or later startup has better data
00:42:33.380presentation than new york times it's a very common skill now it's a skill that i have and
00:42:37.800i think that's reflected in this site where really there are a single digit number of like really
00:42:44.780competent data presentation people in the mainstream press. Maybe the most famous is
00:42:48.780John Byrne Murdoch from the Financial Times. He's like a household name
00:42:52.760basically by, I don't want to be mean to John Byrne Murdoch, I think he does good work,
00:42:56.920but basically has become a household name by publishing charts that are
00:43:00.740about the same quality as a Series A or Series B startup. So there's
00:43:04.640plenty of capability to have much better data presentation, actually.
00:43:09.400And this is the other thing that I'm kind of frustrated by with
00:43:12.580kind of new media companies is number one they don't break stories number two they don't really
00:43:17.880go for a more like high quality publication a lot of them just end up looking like the new york post
00:43:24.200and actually the new york post is an interesting example because they do break stories but they
00:43:28.720also look like the new york post which is really unfortunate i really think that this is the the
00:43:36.280crown in the gutter because right now is a complete change in the scope and possibility
00:43:44.620of what kind of stories can be reported it would take a team of multiple people working for a year
00:43:51.060to report the kinds of stories i'm reporting now so it's not an entire surprise that they've gone
00:43:56.440unreported for a long time that's changing now it's a crown in the gutter and absolutely do it
00:44:01.120I mean, I am doing it, and I think this will be very soon the greatest investigative reporting publication in the world.
00:44:09.560Well, I hope so, and we're glad to have you on, and I hope that it's upstream of a bunch of YouTubers like I to just read the story and then go, ah, you know, that's the way to do it.
00:44:19.660Oh, yeah. You're going to give us so much episode fun.
00:44:21.980ecosystem for a bit like you know we're in the fox girls or the media chat there's like
00:44:26.380an entire you know new ecosystem that's popping up and you guys really i think you guys know more
00:44:33.320about this than me so it'd be interesting to get your thoughts on it the vtuber ecosystem
00:44:39.100play in that's what we do but the interesting thing about this new media ecosystem is it has
00:44:43.980a few people who are sort of top of the funnel and find the stories first and and they're like
00:44:49.720kirsha is an example of one of these she'll do the the app synthesis and then the story flows
00:44:55.900down so if we have more people there then that's more for the rest of us right you know
00:45:00.900is she like uh finding stories from scratch or is she kind of like serving as a bridge between like
00:45:08.160no she thinks to be finding them from scratch so what she does is she goes and she'll like look
00:45:14.780when somebody will be like oh x company was woke and then it was fixed and then she tries to go
00:45:19.380into their public documentation and everything like that to see if they're actually not woke now
00:45:23.860looking up the programs they're doing and stuff like that and that's that's an area of of research
00:45:29.840that you could definitely use more reach is the crazy woke programs that companies are doing
00:45:35.700that large companies are doing and shining a light on those because that's something that i think
00:45:39.600people would find pretty interesting yeah so there was this story i like kind of got scooped on this
00:45:44.960A lot of the data was already public, but independently found the same data of company filings, earnings calls and such dimensions of diversity, equity, inclusion, stuff like that has dropped more than 90%.
00:45:57.780That's, of course, just the overt public statements.
00:46:02.440It's not always reflecting the underlying stuff.
00:46:04.680But that is actually pretty interesting.
00:46:06.880Maybe you guys can put that graphic up as well.
00:46:09.560Taking a lot of these stories, there's like, there's like a pretty long incubation period of before they actually see the light of day. That's something that if I am successful with this company, will absolutely change in five years where it's like, right now, like the Office of Refugee resettlement story, we've had that information, that information has been publicly available for five years, or maybe a little less than five years.
00:46:38.640I think the first grants appeared or like the first grants in this latest
00:48:58.880if you are ahead of ahead of the curve and publish something that then becomes viral then becomes a
00:49:03.980center of attention you do get credit for that but but people just don't do it maybe it's like
00:49:08.940a cost benefit thing maybe they haven't fully priced in ai i'm not entirely sure but yeah it's
00:49:14.320still seems still seems very odd to me it is i think i mean to your point a lot of it can come
00:49:21.980down to good visuals a lot of the stuff that ends up going viral that is more substantive
00:49:26.440goes viral because there's some striking graph that people can share online like you have to
00:49:31.960think about what is someone's motivation to share something online and malcolm's model for that has
00:49:35.940always been well then this is going to reinforce someone's self-image for example like i'm the
00:49:42.820person who shares stories about this or it just validates their self-image like family members
00:49:49.660selling sending us information about studies that show that reading physical paper objects is
00:49:55.620better for your brain or something right and they're like i told you so so you have to think
00:50:00.740about why someone's going to be compelled to share your story and also how much that share on its own
00:50:07.340is going to captivate audiences and if you have a really compelling visual then it is going to
00:50:11.320captivate audiences like i've seen so much of the almond water versus data center water graphic
00:50:17.420floating around recently yeah because it's just so like salient and easy and also for anyone who's
00:50:23.920on the pro data center side like we can so smugly share it and be like so there you go ha ha ha and
00:50:30.400it makes us feel very smug and so i think it comes down to that i think that's why it's so good that
00:50:36.740you're looking at the importance of really strong visuals if you can create visuals that people can0.99
00:50:43.240smugly post and that make them look so smart and well-informed and oh you don't you idiots don't0.99
00:50:49.680realize how serious this problem is especially if it feeds into yeah a lot of it you know a lot of1.00
00:50:54.520it is like the immediate truth nuke right yes something that that i want to obviously that's
00:50:58.940not most of these stories most of these stories are kind of longer investigations using data from
00:51:03.640a time frame of like five to ten periods but something else i want to do a lot more of is
00:51:08.160like the immediate truth nuke one of these stories is kind of like this the data is actually kind of
00:51:12.520old but i have this story on like the the ev dumping theory which knowledge was completely
00:51:19.040fabricated by kind of like the auto lobby and by defense interest by like kind of like
00:51:24.460wait what is this what is the ev dumping theory i'm gonna bring it up all sorts of these stories
00:51:30.240that pop up that are just like clearly fabricated the one that i remember the best because i was
00:51:35.560kind of like directly involved with this one the the 10 000 h100 story so there was this story
00:51:43.140around the time when deep sea came out of course it's like wildly cheaper model than anything we
00:51:47.500was published before. And a bunch of like defense industry types were like very enthusiastic about
00:51:54.620promoting this H-100, 10,000 smuggled H-100s claim. And this claim was completely false.
00:52:01.260The original source for this claim, I actually kind of respect that he did this, which is Dylan
00:52:06.160from Semi-Analysis. He had kind of like a tweet, like, oh, there's 10,000 H-100s or something.
00:52:13.820And then he like later published a full article kind of like retracting that claim saying like, oh, what I actually meant was 10,000 H100 equivalents, including hardware that was legally purchased from NVIDIA and allowed on the US, which is obviously a different claim.
00:52:30.980But I actually respect, I don't really blame Dylan for this.
00:52:34.300I think what probably happened was that he had a source inside like state or defense that just straight up lied to him.
00:52:40.540And he got caught and he didn't know that, like, people from defense or state would just straight up lie to you and then go, you know, promote these completely false claims everywhere.
00:52:51.340Anyways, we now know that that absolutely did not happen.
00:52:54.600This was just, like, fabricated by the defense industry.
00:52:58.900So EVs are way, what, more environmentally friendly than?
00:53:03.160No, so that was related to the DeepSea H100 stuff, which I was kind of, like, directly involved in.
00:53:08.380i was kind of like um what's the point ai stuff with the admin at the time what's the plot uh0.87
00:53:14.140this is a similar thing happening with evs right it's it's the same like fucking export control0.54
00:53:18.920people but what's the point like what's what were they trying to argue or just oh so go go on no no0.96
00:53:27.040like what were they trying to argue versus what's the reality i did like the to the average person
00:53:32.100here they don't know the reality is that that deep seek made a number of algorithmic improvements
00:53:36.300mostly that made their model a lot cheaper to train and run they were arguing that you know
00:53:40.800there were all these smuggled h100s these smuggled like ai specific oh so they were saying they were
00:53:45.300getting really good machinery okay it was this kind of like mild like conspiracy drained take
00:53:50.820that like actually deep seek was trained on all of this hardware because there is no way that they
00:53:55.560trained they trained their model on like limited hardware that they were legally allowed to
00:53:59.720purchase but turns out that actually they they did train it on the legally purchased hardware
00:55:34.780I mean, we have 100% import taxes in the US on EVs, so nothing.
00:55:40.640Yeah, it's 127.5 now, and they're doing a different ban.
00:55:44.880You have this super interesting, you know, maybe not interesting, but like funny thing.
00:55:49.140They had to keep upgrading the tariffs because it kind of sounds crazy.
00:55:52.880But, like, there was this theoretical model where, like, even if you had 25% tariffs, the Chinese cards would still outcompete the American ones.0.62
00:56:01.000Which, if you're coming from the software world, it's like, oh, 25% off your margin, maybe that's doable.0.97
00:56:06.420But if you're coming from manufacturing, that's fucking crazy.0.82
00:56:09.240That's how much of a difference in technology there is.0.92
00:56:21.900And even if you count all of the differences in financing laws, which we could reform our financing laws at home, but even if you count those as de facto subsidies, it is roughly 7% of the price difference between Tesla and BYD.
00:56:37.340Tesla is, of course, the most efficient American company, which is around 2.3% of the total BYD price.
00:56:43.360So even if you're, oh, it's 1.2% for direct subsidies.
00:56:46.700And even if you're touting all of this, like, financing reform stuff that we could totally
00:56:51.060implement in the US, that's, like, 2.3% of the total price.
00:56:55.260So it is this, like, massive truth nuke where there's the story that was being passed around
00:57:00.880all the time, this, like, cult about Chinese EVs of, like, oh, they're massively subsidized.
00:57:05.220That's why we can't compete with them.
00:57:06.760It turns out that's completely untrue.
00:57:08.520Most of the, most of the savings are based on, like, supply chain integration.
00:57:12.000And generally, there's some improvements compared to even Tesla, right?
00:57:16.720Of course, the gap will be even bigger than this for the non-Tesla.