I really wish you wouldn't use AIPAC as your go to example of a single issue pac, given the number of people who (mostly for "Jews control the world" reasons) vastly overestimate the power of AIPAC (it's neither especially influential nor an especially big spender compared to other pacs in its class).
My source said it was extremely influential compared to other PACs in its class. I am basing this on their claim, but feel free to link me to evidence that they're wrong.
I imagine that the apparent importance of a particular lobby group is going to look enormously different depending on where you're sitting. For example, when I was involved in local/state politics in California c. 2008-2015, I heard essentially nothing about AIPAC. The big dogs I kept hearing about over and over again were SEIU on the Democratic Party side and Charlie Munger, Jr. on the Republican Party side.
Any PAC that is consistently on one side is dramatically weaker than "mercenary" PACs like AIPAC.
SEIU only has influence when they spend money. Because AIPAC is occasionally willing to pour money on either side of the aisle, they also have influence in all the races where they don't spend money, by employing the threat of funding the other side. There are a lot of politicians that are afraid of AIPAC flooding the donations to their future challenger, and thus keep their mouth shut about certain subjects even if AIPAC never spends any money on their races.
The other political organization that used to be disproportionately powerful compared to the amount of actual money spent was NRA. This was when NRA was entirely willing to fund the democrat if the republican they were facing was wavering even a little bit on gun rights. As NRA has become more and more just a part of the republican party extended universe, their influence has waned.
Washington Post columnist Dana Milbank used to write a hilarious annual column about the extraordinary lengths AIPAC goes to each year at its convention in D.C. to impress members of Congress with its wealth and fervency.
Well, consider that during the pro-Israel lobby's annual conference yesterday, a fleet of police cars, sirens wailing, blocked intersections and formed a motorcade to escort buses carrying its conventioneers -- to lunch.
The annual meeting of the American Israel Public Affairs Committee has long produced a massive show of bipartisan pandering, as lawmakers praise the well-financed and well-connected group. But this has been a rough year for AIPAC -- it has dismissed its policy director and another employee while the FBI examines whether they passed classified U.S. information to Israel -- and the organization is eager to show how big it is.
Reporters arriving at the convention center yesterday were given a list of "Food Facts" for the three-day AIPAC meeting: 26,000 kosher meals, 32,640 hors d'oeuvres, 2,500 pounds of salmon, 1,200 pounds of turkey, 900 pounds of chicken, 700 pounds of beef and 125 gallons of hummus.
Another fact sheet announced that this is the "largest ever" conference, with its 5,000 participants attending "the largest annual seated dinner in Washington" joined by "more members of Congress than almost any other event, except for a joint session of Congress or a State of the Union address." The group added that its membership "has nearly doubled" over four years to 100,000 and that the National Journal calls it "one of the top four most effective lobbying organizations."
"More," "most," "largest," "top": The superlatives continued, and deliberately. In his speech Sunday, the group's executive director, Howard Kohr, said the "record attendance" at the conference would dispel questions about AIPAC raised by the FBI investigation.
"This is a test, a test of our collective resolve," Kohr said of the "unique challenge" presented by the FBI probe, "and your presence here today sends a message to every adversary of Israel, AIPAC and the Jewish community that we are here, and here to stay." (The official text has two exclamation points after that sentence.) ...
Second-largest for direct contributions to candidates (https://www.opensecrets.org/political-action-committees-pacs/top-pacs/2024), plus very widely reported much stronger donor list than anyone else (as Scott states). I'm not sure how much has changed in the past few years I imagine AIPAC is much much weaker than it used to be, partially because the salience has gone up dramatically which always weakens single-issue groups in the area, and partially because "I'm the best friend of Israel / No I'm the best friend of Israel" is no longer the basic description of political campaigning on the topic. But I don't think that, for example, Mearsheimer randomly decided to be anti-Semitic.
Scott is claiming that AIPAC is "orders of magnitude" more effective than other PACs, which seems like a stretch if their only real advantage is more hard-money funding due their donor class composition and more effective bargaining technique. We're still looking at ~2% of US lobbying spend.
How many other Jewish organizations are there that will follow AIPAC's lead on an issue or a campaign relevant to Israel? There are enough major Jewish organizations that there's an influential "Conference of Presidents of Major Jewish Organizations."
How many big donors to candidates or universities listen to AIPAC? The top 100 political donors in the U.S. are about half Jewish.
The top 100 billionaires are about 1/3rd Jewish. My impression is that Jewish billionaires (e.g., Larry Ellison, lately the world's richest man) tend to be more Zionist than average Jews, just as rich men in Alabama tend to be really into the U. of Alabama Crimson Tide football team. Israel is to a lot of rich Jews what the Notre Dame Fighting Irish is to a lot of rich Catholics.
What fraction of the prestige press is Jewish? When The Atlantic made up its list of the top pundits about 15 years ago, it was about 48% Jewish.
I read the Jewish press a lot so I'm more aware than most people of just how much money and influence Jewish Americans have earned over the last 150 years. It's not just a "trope," it's hard numbers.
Since October 7, 2023, Jewish support has been more up for grabs than it has been for decades. Anybody who wishes to play a role in American public life can't afford to underestimate what Jews have earned for themselves.
I think the big issue is that AIPAC and Israel are following the NRA playbook (and to some extent Planned Parenthood/ACLU) where instead of being single issue pacs that try to play nice with both sides but care deeply about the issue become partisan actors (AIPAC is still bipartisan but less so).
It's interesting that exhibiting more loyalty to a given party dramatically reduces the actual payoff to your org, but there you have it. The NRA doesn't seem to follow the same playbook so much any more, though.
There's objective data available on that question. You can get top donors from OpenSecrets.org. The ethnicity of big donors is not all that hard to find from Wikipedia, Jewish magazines, and the like.
Among the top 100 donors to political candidates in the 2020 election, whether Republican or Democrats, people of Jewish ancestry gave 61% of dollars. In 2024 among the top 100 (a group that somewhat overlaps the top 100 in 2020), however, there was a 6% decline in dollars given by big Jewish donors to Democrats and a 129% surge in dollars given by big gentile donors to Republicans (with Elon Musk accounting for about one-third of that increase). So, the Jewish share of dollars donated to candidates by the top 100 fell from 61% to 39% in 2024.
For details and methodology, see my 2025 column at:
Jews very likely give smaller shares of total political contributions of all sizes since they make up about 1/3rd of the top 100 names in the 2019 Forbes 400.
Among campaign contributions from the top 100 donors in 2024, 68% of the Top 100 money going to Democrats came from Jewish givers, while 31% going to Republicans came from Jewish givers.
Those shares were down somewhat from 2020 and 2018.
I don't think it's Good for the Jews if we try to say that a thing that, as far as I can tell, everyone with any experience in the area agrees is true, is anti-Semitic (or at least in some gray area of "I won't call it anti-Semitic but I'll gesture in that direction").
I think there are plenty of things you can say that recognize AIPAC's influence without getting into Jews-run-the-world territory. E.g. Israel has historically been popular even with non-Jews, and it's a low salience issue, both of which raise its effectiveness.
And now that the circumstances have changed, you have things like moderate Dem Seth Moulton *returning AIPAC money*.
Last time I checked in 2011, since 1901 Jews had comprised between 21% and 26% of the Laureates in each of the three hard science Nobel Prizes (physics, chemistry, and medicine) despite making up only about 0.2% of the world's current population.
My general opinion is that Jews _are_ strikingly rich and influential ... because they've _earned_ their riches and influence.
And also, having been persecuted worldwide for literal millennia, they've learned to stick together pretty well. Of course, this history also means that when any outsider mentions this obvious fact, by default it's taken as an attack, not as a compliment.
I've read Jewish-American and Israeli periodicals more than most, and they _love_ enumerating Jewish billionaires and making other lists quantifying Jewish power and success. The New York Times, in contrast, doesn't usually find that news fit to print.
Often, I get asked accusingly: "Why do you want to _know_ about who has clout in your country? Huh?"
Well, I like knowing data. And I like knowing how my country works.
I am a U.S. citizen or lawfully admitted permanent resident (i.e., green card holder).
This contribution is made from my own funds, and funds are not being provided to me by another person or entity for the purpose of making this contribution.
What would the legality be of donation swapping? Let's say I'm not a US citizen, but I'd like to donate. So I find a US EA who doesn't want to donate, and say "If you donate X to this politician, I'll donate X to a charity of your choice, that you would otherwise have donated to".
In this case, the US citizen is giving the money, and I haven't actually provided them any funds...but I'm not sure if this is a strong legal defense or not.
This is almost certainly illegal for both you and the other person under existing laws (straw donors) although whether you will be prosecuted for it or the existing laws will be found unconstitutional depend heavily on the party in power.
IANAL, but I don't think this is legal. The relevant law can be found here, https://www.law.cornell.edu/uscode/text/52/30121 and criminalizes "direct or indirect" donations. My viewpoint is that if you are making a contribution with the intent of causing another person to make a contribution that you are prohibited from making, that is the kind of indirect donation that this law targets.
Seems weird that it's illegal for a foreign citizen to donate to a PAC, but it's also legal to have a PAC whose explicit and open reason for existence is to advocate for the interests of a foreign country.
That rule feels very correct to me. If PACs are supposed to represent the “voice of the American people”, it’s entirely appropriate that Americans get to decide what they do for others using their government, and it feels quite inappropriate for outsiders to use the same mechanism.
You won't need a legal defense since the only way the government finds out is if one the other guy snitches on you, and he has no incentive o because he's just as guilty.
Indeed, I'm not in the US, but if I were this sounds like it would be *by far* the most effective way I could spend $7k to reduce the chance we all get turned into paperclips or similar.
I feel like the most obvious rallying point here is to go absolutely nuclear on anybody that supported defunding PEPFAR. If there's one single program everybody at every level of EA agreed was very good it was that one right?
Although it would be nice to have a generic EA PAC, for various reasons the system is more set up for single-issue PACs, and the people I know are prioritizing AI - partly because they think it is more important, partly because it is faster-evolving, and partly because the people who care about it have more money (since many of them are in AI themselves).
This goes through lots of information about politics (like what sorts of pressures representatives are amenable to, and how party-line votes will be) that I don't know.
PEPFAR is a much less controversial program to support - "save babies dying in a painful way in countries that will never in any way rich the US" is a pretty easy sell. AI controls is much harder - "cripple a major source of economic growth that also happens to have highly salient military applications because of some hypothetical threat. China also seems to heavily invest in this btw."
For what it's worth, I think AI doomerism is silly and unfounded (yes I've read the book). So if you disagree with me on that, then at least take rallying around PEPFAR to establish credibility then go for AI research as advice.
> "cripple a major source of economic growth that also happens to have highly salient military applications because of some hypothetical threat"
The economic growth is roughly as hypothetical as the AI doom scenario atm (stock market gains are not generalising outside the top 10 US tech firms, and we know that generative AI is having all kinds of negative impacts on entry-level employment and student competencies.) Military applications I'll give you, sure, but that doesn't seem a million miles removed from the doom scenarios.
Well, if the AI bubble pops, the crisis that follows will certainly not feel hypothetical, for the entire economy. Current projections are "very bad, but perhaps not quite 2008 level bad".
PEPFAR used to be extremely bipartisan and even now I don't think there's a strong anti-PEPFAR position in Congress so much that nobody is really crossing the Trump administration on their destruction of it.
Well, it would be an issue PAC, trying to get support for a specific issue. Congress doesn't need to wildly buck its trend of surrendering to the President on everything, just on one issue he probably doesn't care that much about, for it to be successful.
Has Trump ever said anything about PEPFAR one way or another? I won't be surprised if he doesn't even know what it is. As I understand, it basically got in the way of Musk's chainsaw, without anybody specifically targeting it, then reinstated in a crippled state and promptly faded from notice.
I have a few concerns that I'm curious if you share:
1. AI accelerationists are probably going to be at least as much money as the EA AI safety people, so its possible you're going to just end up in a stalemate. This might still be worth it if you think absent your actions the accelerationists will just stomp, but it probably won't see the total victories M.A. has achieved.
2. Crypto and (until recently) Israel were pretty low salience issues in politics. Given that we are one disappointing Gemini demo away from a recession, AI probably will be high salience for the forseeable future. Nobody cares about PEPFAR on the other hand. It was basically canceled so JD Vance could impress a bunch of teenagers on discord. So a smallish, dedicated group could probably make a big difference at the margin.
He did say "until recently". And yeah, up until certain MENA migrant groups inveigled their way into the coalition of the fringes in the last few decades, most people in the US either didn't care much about Israel and or were mildly positive toward it.
PEPFAR would have been a great choice 2 years ago, but now it cuts on party lines. AIPAC was so successful because they got both Republicans and Democrats to support it. Now it’s unfortunately politicized. If anything I think a PEPFAR super PAC would be a good idea in 3 years, when it’s not as connected to the incumbent running for office.
I dont know if partisanship is a huge issue here. Theres always going to be a streak of "foreign aid bad" populism in both parties, so even if it becomes a partisan issue you can still have a lot of influence in the primaries
But who in congress had a pivotal role in defunding it? I thought elon defunded it while marco rubio continued to insist it wasnt defunded and congress largely demurred. There may have been GOP members who cheered elon on twitter, but not in a congressional capacity. Also getting rid of them wont get pepfar back.
That’s accurate. But it’s become a matter of “common knowledge” in online spaces that it has been defunded and is gone now. But it’s not. PEPFAR is supported by just about everyone in Washington, on the record anyway.
The joke was that both villains, in movies from 1985 and 1978 respectively, separately plotted to destroy San Francisco (or California more generally) by various plans involving triggering earthquakes.
Luthor's primary residence is in Metropolis, while Zorin lives in France and also a blimp.
Effective altruists should donate their money to GiveWell recommended charities, like the ones against malaria & parasitic worms, instead. That's what I do.
GiveWell only claims to analyze the best places for global development donations. They make no claim to have considered issues like AI safety, and many of the people involved in GiveWell themselves make AI safety related donations (or donations to animal welfare, or other causes that GiveWell doesn't look at).
Although it would be nice to be able to avoid politics entirely, politics is also extremely very important, and it doesn't seem ideal to leave it entirely to the industry billionaire lobbyists. I don't think the usual "D and R donations cancel out and it's all a wash" argument applies when an industry is trying to spend hundreds of millions of dollars to make politicians serve its naked interests.
It means you already know you're in a tug-of-war with someone. Someone whom you know to have plenty more money. I guess if you could get someone far richer on your side you could convince them they couldn't outspend you, like the Soviet Union in an arms race with the US. Elon Musk is supposedly the richest person, and played a role in founding OpenAI due to an interest in such risk, so maybe you have some hope. But that would be hope from appealing to him (or a set of people with a combined comparable amount of wealth).
When I read that piece when it came out, it convinced me to disengage from politics, which in retrospect was a great choice. Looking at my peers that did go into politics at the time, they are now full blown MAGA Trump loyalists. I definitely think you gave good advice then.
That said, times change, and this issue is not Republican Democrat, but an internal war in Silicon Valley that could kill us all. I’m going to need to reread both pieces and contemplate where is the best place to make my annual donation.
"Lots of people work very hard and raise $10 million for the Democrats. Lots of other people work very hard and raise $10 million for the Republicans. Now the Democrats and Republicans are at exactly the same position vis-a-vis each other as they were before the effective altruists got involved, but we have wasted $20 million that could have gone to healing the sick or feeding the hungry. "
Indeed. It is, realostically, an arms race normal people have no hope of winning. And
That is simply false because you donating to something will not magically cause more people to come out of the woodwork and support your opposition. Also fighting someone in zero some contest is better than just rolling over and letting them win. There is a reason that countries at war don’t just cut the defence spend to 0. Even though both countries are burning resources could spend on something else. Unless you can do a credible deal for both of you to reduce your defence expenditure, it doesn’t make sense to unilaterally step back.
> Also fighting someone in zero some contest is better than just rolling over and letting them win.
It's 2025, and I do wonder about that. Is it really always better? Decades of low-grade pain, waste and slow decline because the idiocy just. won't. die and keeps attracting more and more richer and richer followers; vs one term of total capitulation, just turn the other cheek, lean in hard, bend over backwards, get it all out of our system at once, until the effects are utterly impossible to ignore, so we can finally all agree never to raise the matter again, and when some newborn sweet summer child who did not live through the idiocracy does bring it up the scars and smoking ruins are /right there/, proof positive of where it all leads, all the veterans make it very clear why it's utter bunk until the idea is firmly buried again, like the apocryphal monkey ladder experiment...
...and then I wake up and remind myself that people can stay irrational longer than the world can stay unburned.
I disengaged in 2015, reengaged in 2022, supported Nikki Haley, had a blog for a couple months, realized that was a mistake, took down all my posts in the blog, and am back to being disengaged. Even my blog was designed to get more people on the Right to become Non Partisan Civil Servants, but there is no appetite on that from the Right on Substack. They would rather tear down the institutions entirely than try to compete in them, and the current president, who I didn’t vote for, is happy to do that.
Even during that phase, I kept my donations going to pandemic preparedness research, and never donated to a politician. Also, Greek and Roman history is cool, and I’ll never apologize for having cool ascetics even if they signal that I’m on the right.
It WOULD undo PACs, yes, but it would also abolish the freedom of the press as generally understood if any state can decide that, say, the New York Times, as a corporation, may not do business in that state.
Do you have a specific example of country you think has what, say, the median American would consider a free press where, sticking with the Citizens United case, if a book was being published critical of a candidate by an organization not granted state imprimatur, that the government could stop publication, citing "campaign finance regulation"?
I could indeed name several examples of countries where I believe organizations are restricted from publishing and distributing opinions disfavored by the government – Afghanistan, Belarus, China, Djibouti, Eritrea,… – but how would that help? You'd just say those don't count because you had some other country in mind that made your point better.
But okay, that's more concrete: you're suggesting the UK as an example. Even granting that most Americans would consider its press free, I haven't seen an instance of the government imposing prior restraint on publication like what was forbidden in the Citizens United case.
Your claim was that a state could decide the NY Times couldn’t do business there. I think that’s unlikely and a straw man argument. The UK has fairly strict laws governing actions by non-party actors to influence an election, but the tabloids continue to operate.
It happened in this country before Citizens United. The CU case was literally fought over the government banning someone from producing a political documentary.
I dont see how this plan can work as a run around. Those laws would be struck down on citizens united precedent straight forwardly. The whole corporations as a social/legal construct thing is fine philosophically, but is not relevant to 1st amendment rights legally. Persons are not given first ammendment rights by law, they precede all law. Under citizens united, corporate person hood gives them 1st amendment rights. You cant restrict that with either campaign finance law, or corporate charter law, for better or for worse.
It might undo PACs but then something else would spring up in their place. Politicians like and need money for their campaigns, and interest groups want to influence politicians to make laws and govern according to their ideas of what is best.
So if we (interest group that wants to lobby for all public spaces to be painted magenta) can only donate X amount directly as individuals because of pesky laws, and we can't found a PAC, then some other loophole about "getting a bunch of us together to hand over $$$$ to persuade politicians to pass laws about painting all public spaces magenta" will be found (e.g. "we are all close personal friends of the candidate in this constituency and their family and are related to them by marriage and other ties, and we are giving them wedding, birthday, and Christmas presents of nice fat cheques to spend as they wish on whatever they wish, no strings attached, all perfectly legal, oh what a coincidence that Candidate's cousin owns a paint shop which will supply all the magenta paint for the public spaces").
Very interesting. It does, indeed, seem strange that no-one has tried this before. My instinct is that this is the kind of situation in which everyone assumes that, if it worked, everyone else would already be doing it. But ultra-ambitious risk-taking maverick billionaires (i.e. the people best suited to using this strategy) don't seem like the kind of people to make that particular mistake.
Minor proofreading things:
> is a little bit paranoid and will truly accept that their safe seat is safe
I think there should be a "not" or "never" in this clause.
> and occasionally crossed skirted the border of illegality
It's probably just a case of everyone being pretty happy with the status quo and not wanting to rock the boat by outright buying politicians. The country is already as business friendly as you can reasonably expect a country to be.
The $20 dollars on the sidewalk/everyone before is an idiot; claim is wrong the vast majority of the time, but sometimes it isn't.
The greatest breakthrough in human history came from putting bread mould onto wounds and it was only widely used in the 1940s, wheeled suitcases were only introduced in the 70s.
I think the fact that it happened with bitcoin demonstrates the reasons why it doesn't necessarily happen with other things.
Let's compare bitcoin to oil. Bitcoin is pretty much the optimal case for the value of buying politicians. There's a significant number of very rich people whose fortune is almost entirely tied up in bitcoin. The value of bitcoin is very dependent on US government policy, to the extent that there's plausible policy approaches the US Government might actually take which would 100x or 0.01x the value of bitcoin. And a lot of politicians haven't already made their mind up about bitcoin, so they're relatively easy to influence. Therefore it's easy for a single individual to spend a dollar influencing politicians and expect to get a greater-than-$1 return on it.
Compare to oil. There's a lot of money in oil, somehow, but it's pretty diffuse. Oil companies are ultimately owned by a bunch of people, most of whom own a lot of other things and are not single-mindedly exposed to the oil industry. Furthermore it's not clear that plausible US Government policy is going to have a huge impact on the fortunes of oil companies one way or the other; maybe you make a bit more money here or less money there, but ultimately oil is just going to keep on making money until it doesn't. If I'm some rich oil guy then it's not clear to me that spending money on politicians is going to make me personally richer... maybe it will make the oil industry collectively richer but now we have a collective action problem, unlike the bitcoin case where a single rich bitcoin guy can expect to make money by buying politicians just on his own.
That's a good point, I hadn't thought about the ways in which influencing crypto regulation specifically has a much higher ROI than lobbying for other industries.
I'm still slightly surprised that no eccentric billionaire has used this strategy to indulge a personal hobby-horse without expectation of turning a profit, but I suppose this would require a policy issue lying in the rare overlap between 'the average politician has no strong opinion on this' and 'someone is willing to spend hundreds of millions of dollars on this'.
I'm reminded of Phil Sokolof who spent his own money to successfully campaign McDonald's into changing from tallow to vegetable oils for frying, in a misguided attempt to save people from saturated fats. Instead leading to the proliferation of trans fats, likely killing quite a lot of people. Also ruining the taste of fries forever.
What is an example of a plausible policy approache the USG could take which would reduce Bitcoin's value by multiple orders of magnitude? I inquire because crypto has thus far seemed robust to government bans, e.g. in China, and am curious as to what other vulnerabilities are known/feared.
I think a US Government ban would be a lot more effective than a Chinese Government ban. (A Chinese Government ban was pretty much priced in from the start; of course the Chinese are going to ban it, they ban everything)
That said, I'm only wildly guessing. I don't claim to be able to accurately predict the effect of random events on the bitcoin price; if I were then I'd be incredibly rich.
That's easy - ban the exchange of cryptocurrency for US dollars, and create private right of action penalties against any financial institution with US assets or branches that provides it in violation of the law.
Cryptocurrency would still be available overseas, but it would be like prediction markets where being cut off from the huge, rich market full of would-be speculators in the US heavily constrains it. Both of them desperately needed the money of US investors, speculators, and suckers to really get going.
This is an excellent point. The corresponding point to this is that unlike oil (where you have a few oilmen and many oil buyers) the vast majority of people (not just politicians) have no opinion whatsoever on the topic. It's the perfect lobbying opportunity.
And there was special legislation introduced to break up the monopoly of oil companies like Standard Oil. John D. Rockefeller probably could have bought and sold the politicians of his time, and the corporation did make large donations to particular political campaigns:
"Although it always had hundreds of competitors, Standard Oil gradually gained dominance of oil refining and sales as market share in the United States through horizontal integration, ending up with about 90% of the US market."
"Large sums had to be spent quickly, and Hanna energetically built a businesslike campaign. ... But as the campaign began operations, and began them on a huge scale, money was short. Hanna initially spent much of his time in New York, where many financiers were based. He faced resistance at first, both because he was not yet widely known on the national scene, and because some moneymen, although appalled at the Democratic position on the currency issue, felt Bryan was so extreme that McKinley was sure to win. ...Reports of Bryan support in the crucial Midwest, and intervention by Hanna's old schoolmate, John D. Rockefeller (his Standard Oil gave $250,000), made executives more willing to listen. After a gloomy August for the campaign's fundraising, in September, corporate moguls "opened their purse strings to Hanna". J. P. Morgan gave $250,000. Dawes recorded an official figure for fundraising of $3,570,397.13, twice what the Republicans had raised in 1892, and as much as ten times what Bryan may have had to spend. Dawes' figure did not include fundraising by state and local committees, nor in-kind donations such as railroad fare discounts, which were heavily subsidized for Republican political travelers, including the delegations going to see McKinley. Estimates of what Republicans may have raised in total have ranged as high as $16.5 million."
It's more that until recently, there were de facto limits on how much TV time/etc you could buy with money (only three networks, finite ad space, etc), and limitations on campaign finance used to have more teeth.
I'm also not totally sold on its effectiveness. The crypto money spam worked because Democratic politicians don't actually have that strong of opinions on cryptocurrency regulation, so a ton of money floating around was enough to sway them in a more positive direction (and of course the Trump people and associated Republicans are all too happy to have a new way to accept payments).
>It does, indeed, seem strange that no-one has tried this before.
The detached object-level analysis in the post shouldn't make anyone forget that "this" is "buying politicians", which is (was?) widely frowned upon and considered immoral and bad.
If an industry is outright buying politicians, people or the media would notice it and criticize it, and the damage to the industry's reputation might outweigh the expected gains. This is the Occam's razor explanation of why the almonds industry wasn't buying politicians, I think.
AIPAC is an obvious exception because until very recently it was de facto impossible to criticize it. Even in this comment section you can still see how any criticism is immediately accused of being just another "Jews control the world" conspiracy theory. Even mild criticism by someone who is literally world-famous for his honesty and objectivity and who's Jewish himself.
I'm not sure about why this is changing now, but crypto might have been an exception because regular billionaires still somewhat depend on the good will of their business partners and employees for their status, while for a crypto billionaire it's probably easier to just ignore public opinion.
The big hole in this theory is that before reading this post, I personally hadn't heard of the crypto influence in the 2024 elections at all. But for now, I'm chalking this up to me being uninformed and not even American.
>The detached object-level analysis in the post shouldn't make anyone forget that "this" is "buying politicians", which is (was?) widely frowned upon and considered immoral and bad.
I'm frankly surprised that the "dive bombing" tactics Scott describes in order to bully politicians into line is not very illegal, even in the US.
They probably did try it before, which is why we have laws around campaign finance etc. Think of things like the Teapot Dome scandal, or the various scandals that plagued Ulysses S. Grant's administration. We call such attempts bribes, because they were crude essays in the art.
I get that you're Jewish, but uh, maybe it isn't a good idea to use an antisemitic dogwhistle in a post, even if it's a joke? Especially when said post has an activism component? Just saying.
Edit: Nevermind, apparently mentioning AIPAC at all was way worse. Well, your loss!
Is it really still an antisemitic dogwhistle? It used to be one, but (1) everyone knows what it means now and (2) the actual antisemites aren't using it anymore.
To me it read like jokingly using a period appropriate slur when writing an essay about a historical period -- the thing's been drained of all its dogwhistling power and relevance, so it can be used as a joke.
It would still be a bad idea to call black people negros in a piece complaining discussing the consequences of the Civil Rights Act. If you don't care about optics, that's fine, but Scott does have a reason to care in this case, so...
I think we're just getting different emotional voltages from seeing it -- maybe it's stupid of me to generalize, you're right that this is important. But also, like you mentioned -- Scott is a Jew. Can't he reclaim it like the black people did with Negro? Weren't there Civil Rights Leaders addressing each other like "my fellow negros"?
He doesn't make it his entire identity. But again, I don't care whether this is "right" or not. This is about optics. And given that (((Middle Eastern democracy supporters))) have greater-than-average disposable income, it makes little sense to alienate them in a post like this. Especially when you might tarnish the reputation of others by association.
I think that was reclaimed ten years ago. Half the Jewish commentators on Twitter used to use it, although I think it's dropped out of popularity lately.
I thought the general principle is you can get away with using an ethnic slur if you’re a member of the group it applies to, so African American rappers can use the N word and Jewish bloggers can put brackets round their name.
Triple-parentheses around a name, e.g. (((Name))), is a convention among some antisemites for indicating the names of Jews. As Scott himself says in a different reply to the same parent comment, he's clearly using it in a joking/reclaimed way.
In many ways, the crypto community today are analogous to Arab tribesman in the decades before muhummad, slowly growing in power and influence on the imperial periphery.
The religious angle of that analogy does make a lot of sense. Finally a religion that can have a truly immutable Word of God! Until someone finds the right exploit in the smart contracts, that is.
Marc Andreessen is as close as I think modern society has to the “man of lawlessness” or false prophet which is Biblically asserted to prepare society for the coming of the Antichrist.
"Specifically, he suggests the antichrist would be a “luddite who wants to stop all science”, referencing Thunberg, Eliezer Yudkowsky, and Marc Andreessen.
'My thesis is that in the 17th, 18th century, the antichrist would have been a Dr Strangelove, a scientist who did all this sort of evil crazy science. In the 21st century, the antichrist is a luddite who wants to stop all science. It’s someone like Greta or Eliezer.
It’s not Andreessen, by the way. I think Andreessen is not the antichrist. Because you know, the antichrist is popular. I’m trying to say some good things about Andreessen here, come on.'"
The crypto industry is merely a massive grifting enterprise. The only advantage crypto has it can be used for regulatory arbitrage; i.e. a bank has to verify your identity and take measures to ensure you aren't using your account for crimes. The miners and validators that support the blockchain have to do none of that.
All the Biden administration tried to do was prevent people from using cryptocurrency to break laws; the problem is they did not go far enough. We allow miners and validators to approve transactions between unknown counterparties; bankers would be prosecuted for doing that.
Crypto has many other advantages beyond regulatory arbitrage - for example, it makes it harder for the government to debank people, and substitutes for financial infrastructure in countries that don't have it. See https://www.astralcodexten.com/p/why-im-less-than-infinitely-hostile .
(EDITED TO ADD: Trying to donate to Bores, somehow two different credit cards and Paypal have both failed, both over ActBlue and over the phone, in ways that the credit card company cannot explain to me. Sure wish I could use crypto for this.)
I don't think so. I think of regulatory arbitrage as requiring a careful permit to register a bank, but you can register a crypto bank without a permit so it's cheaper but less safe. I think of debanking as something like the government freezing the accounts of protesters during a protest.
Depends somewhat on how you draw the lines but both of these seem like "crypto can let you do things the government might otherwise stop you from doing" which can be good or bad depending how much you agree with the government.
These are functions banks could provide if we loosened restrictions. US banks could provide financial infrastructure anywhere crypto could if we removed regulations that make it costly to provide accounts to people who live in other countries .
Conversely the government can easily expand restrictions on crypto providers to make it almost impossible in the US to do what Scott desires.
Where did the Biden administration exceed reasonable bounds? Non KYC prosecutions involved securities laws. Perhaps in some cases the scams and lies fell outside of securities laws
That's unfair; it's not all grift. There's also collecting transaction fees from people who want to move money in ways that would otherwise result in law enforcement taking an interest, like drug deals, funding sanctioned foreign regimes and so on.
typo: "Any politician who’s survived long enough to matter is a little bit paranoid and will truly accept that their safe seat is safe" -> "...never truly accept..."
> Just as Substack bloggers may reload their browser again and again watching the likes and restacks come in, so politicians will reload their campaign metrics panel watching the flow of donations. [...] They don't even necessarily have some future campaign they're saving it for. They're just addicted to fundraising.
I do think in some sense this is one of the root issues of politics, in much the same way that people claw each other's eyes out to win a couple of elections even if they never actually do anything with that power. People are great at maximizing quantifiable things (funds, Congressional seats) and will spend their whole lives doing it without actually accomplishing any underlying goal, like a stereotypical rich workaholic who never takes a vacation.
Are you at all worried that your "Too Much Dark Money In Almonds" post might've directly or indirectly affected the political strategy of billionaires like Andreesen? Maybe there was a light taboo around billionaires buying elections which your post helped erase?
Also why isn't Eliezer Yudkowsky tweeting in support of a donation to Bores? Does he think it's another futile effort which isn't worth bothering with? I'm long-term unemployed with a chronic illness, so $7K is a lot of money for me.
Then why has the almond industry not responded? Seems like there must be political muscle there that they are waiting to flex, with all those billions from nut lovers.
To be honest, with the title of this post, I half-expected it to be almond growers versus data centres competing for water resources in California. Perhaps that is how we will restrict the growth of AI - pump it (literally) into the growth of tree nuts instead!
No, I cannot imagine that the market in knowledge is so inefficient that it would take a blogger saying "billionaires could spend more money on politics" to make billionaires realize they could spend more money on politics, if this were actually a very good strategy.
Bores is a moderate whose proposals are extremely far away from anything Eliezer thinks might be helpful. I also think it's reasonable for some people to have a norm against tweeting begs for money (I was nervous about blogging a demand for money, which was why I tried to compensate people with a useful post about the political situation).
I don't think if you are financially unstable you should spend a difficult-for-you amount of money on AI policy.
If the market in knowledge is so efficient, why even bother writing that post? The information was already known.
Are you at all familiar with the psychological literature around anchoring and framing effects? You reframed political donations from the common view ("way large") to a new view ("way small").
The space between the things that the people setting policy know (in this case, politicians, lobbyists, billionaires, etc), and that everyone else knows, is the entire reason that journalism is possible.
It's not even like I invented this argument! I was commenting on a paper saying the same thing! The idea that the market is so inefficient that commenting on a decades-old debate about why it looks like a $20 bill is left on the ground would lead someone to think "oh, cool, there's a way to take over the entire US government at minimal cost to me, I never was interested in this before, but I guess now this blog post has implicitly given me permission to do it" is just not how any of this works, and worrying about whether maybe it is is a good way to go from truth-seeking to being so paranoid that you never write about anything at all.
Seems like a bit of a false dichotomy. How many billionaires view themselves as "people setting policy"? Seems like a rather fluid category.
"It's not even like I invented this argument! I was commenting on a paper saying the same thing!"
Most academic papers are only read by a few academics, correct? I would guess you are responsible for far more exposures to this argument than anyone else. It's not about inventing arguments, it's about spreading them.
"is just not how any of this works"
Blog posts and memes have little influence on US policy and governance you say? This take is at least a decade out of date. Andreesen appears to be a memelord par excellence. And he is an ACX subscriber, if the Substack UI is any indication.
"go from truth-seeking to being so paranoid that you never write about anything at all"
Again, false dichotomy. Consider the least convenient possible world where you did, in fact, inspire Andreesen. What would the correct update be in that world? Maybe it might look something like: Do a trial run of an "80/20" red-teaming process for your posts, which aims to delay publication by at most 10 minutes, and filter out the posts which have the largest negative expected value. Do it for a month or two, then spend a few hours on a retrospective regarding whether it is worth continuing, or should even be expanded.
Thousands of people subscribe to your blog. Maybe you should take that a little more seriously?
"I would guess you are responsible for far more exposures to this argument than anyone else."
Scott looms large in his social niche, but I'm a non-academic (and recent ACX subscriber) who'd heard of that paper's findings before I read Scott's old blog post today. Let's not Basilisk ourselves into treating any discussion of bad political outcomes as an infohazard.
I do think it’s fair to say that Scott’s old blog was, in its heyday, easily the most influential widely-read source of “media” within Silicon Valley. It does not seem outside the scope of probability to assert what Ebenezer has here.
I'm confused how the push to donate to Bores meshes with what you said about the strategies that work, which wasn't "donate to the long-shot candidate who is already on your side" but "donate to the people that will win, so that you have influence on their later decisions"
I think that given that Andreessen is trying to create common knowledge that he will fight people who support AI regulation, it is especially useful to try to create common knowledge that we will support people who supports AI regulation. This is just my guess though, I am trusting the experienced political operatives.
> But also, AIPAC fights hard. If some random Congressman is anti-Israel, AIPAC will swoop down on their race in Middle Of Nowhere, Missouri and pour $10 million into electing their opponent. By now everyone knows this, and the mere threat of AIPAC action is enough to keep politicians in line.
If so, how do we account for several prominent politicians, including several Congresspeople and the current front-runner for the NYC mayor's race, being out-and-proud antisemites and vocal Hamas supporters? It would sure be nice if AIPAC had the power to reduce these human filth to political irrelevance, but that doesn't appear to be the case.
I think it's pretty obvious why AIPAC is not able to single-handedly defeat Zohran Mamdani, but why they can make the difference in some kind of close Congressional race.
Also, it seems like an extraordinary claim to say that several Congresspeople are “out-and-proud antisemites.” I can think of only one individual at most who might sort of qualify for that claim.
This is still less than almonds, but Michael Bloomberg put in $1 billion of his own wealth into his campaign. Tom Steyer, $342 million.
However, total political spending in 2024 was $20 billion, which is more than almonds. People believe that money has too much influence on politics, and I'm inclined to believe them, although I can't explian why.
I think the difference between Bloomberg and Andreessen is that Bloomberg was only able to spend his money on one thing (electing Bloomberg), which people didn't really want, but Andreessen can focus it on the most winnable races between people who are already both popular.
Bloomberg gives a lot of money to other candidates as well. He was the biggest donor to Democratic candidates in 2024 with $61 million according to OpenSecrets.org.
That saturation point might be higher than it initially appears. Elon Musk was 10 points underwater at the time and very publicly associated w the effort. That certainly did some damage.
Then how is AIPAC so effective? This just doesn't add up. I hear people complain about money in politics all the time but then I hear stories like this all the time too. I also remember the Freakonomics guys studying it systematically and concluding that money doesn't really impact election outcomes. I genuinely don't know what to think about it.
I'm inclined to think that it generally doesn't matter BUT if you happen to bet big on the winner by chance then you'll have an outsized amount of influence over him. Maybe the crypto people just happened to get lucky last year.
AIPAC is effective in many races because there is/was fairly broad support in the US for Israel (maybe not so much anymore among younger folks), and among both parties. It's an easy crowd to work, and you can see that when they actually try to unseat someone more challenging - like the failures to unseat AOC in her district.
Isn't AIPAC effective because US politicians fear it for more personal reasons than for electoral ones? If you're positive towards Israel, but lately have been having doubts, perhaps you refrain from making certain comments because you don't want to be *seen* as being anti-Israel. It's not that there was a period where a huge load of politicians switched from being pro-Israel to anti-Israel, and then AIPAC bought a load of campaign ads and these candidates all lost. It's that AIPAC's looming prescence keeps the least pro Israel of pro-Israel candidates in line, so that doesn't happen in the first place.
It prevents wavering, and keeps an existing status quo of pro-Israel politicians acting in unison. Its influence over anti-Israel/Zionist politicians and voters should be minimal, so it remains vulnerable to sudden generational swings. AIPAC is a creature of a particular era in politics.
This also explains the shift in strategy as well. AIPAC announced their Super PAC in late 2021, right around when public support for Israel was tanking amongst the left. Of course, that was still nothing compared to what would happen afterwards, so I guess they chose a good time to switch gears.
Katie Porter is a weak candidate and probably would lose anyway (I expect her to lose the Governor's race regardless of crypto spending).
You can collect a few scalps as a PAC but if you screw up you negatively polarize the issue and can undermine long term progress (probably less of a concern for short-termists in tech in both AI/crypto).
Preventing action is easier than doing things in the US system.
I think you may have swapped hard and soft money in the section that starts "Safe-seat Congressmen want more hard money." Per my friend that works in political fundraising, "soft money fundraising is also just an expectation of the job [...] there's only so many ways chuck schumer can spend a check that gets written directly to him, and he doesn't really need that much money, so instead he gets someone to write a bunch of checks to a bunch of different groups all doing similar work." --which is the soft money, generally for a broad group of candidates not one specific guy.
> The tiny scale of US political spending is dangerous insofar as it means that one or two billionaires willing to go all-in can distort the national landscape. But it also makes it possible to oppose them.
I’m not sure that it’s true. The billionaires discussed in the post, it seems, aren’t spending for ideological reasons, but because they expect a clear ROI on this. $260m spent on crypto lobbying for a +100% return on BTC means it was just « good business ». The same is not true for people invested in the other side of the debate, but who can’t « cash in » on a win. I feel like the only way for non-billionaires to counter that influence is to fund the PAC to end all PACs that will finally lead to getting money out of politics.
I’ve thought for a while that the first and only “AI safety” law will be one to shield Sam Altman and co from liability. This will be followed by a gigantic tax cut to “incentivize innovation.”
> The only reason it's remarkable is that for some reason which I still haven't figured out, nobody else - not the oil industry, not the firearms industry, not the defense industry - ever tried this before.
If you haven’t read it already, i recommend Robert Caro’s LBJ series. The oil industry DID try this, to great success. LBJ got obscene amounts of hard money for his Senate race (in those days, envelopes of cash) from men like Herman and George Brown, who said something to the effect of “if i’m going to back a politician, i go all the way”.
This paid off massively. Leland Olds, phenomenally respected head of the Federal Power Commission and public power advocate, was ousted over flaky communist allegations, and the FPC’s natural gas profit caps were quietly thrown out.
The reason oil doesn’t do this anymore is because:
- oil is not regulation constrained
- oil is not a gold rush anymore.
“Go all the way” political horse-backing only makes sense in a gold rush. A different question is: why didn’t tech do this earlier?
Was tech particularly regulated? Even if you're in a gold rush, it doesn't make sense to waste money on bribes if there isn't much room for improvement. AI and crypto are different, seeing as there's a sizable amount of public backlash to both, making preventative measures reasonable.
If you own a mature company in a mature field, it almost always makes sense diversify into new fields with higher risk/return anyway. AFAIK most older companies and/or individuals in energy industries have by now substantially diversified into renewables already. You can check the list of biggest oil companies; the owners are by now mainly extremely diversified institutional actors like The Vanguard Group and BlackRock.
Once you've done this, backing a politician to safeguard the interests of the mature field is very mediocre value: It's unreliable, it makes you unpopular, and if you're smart you're already playing both sides, so why bother?
No, you do it the other way around; You back politicians who want to do investments into new and/or popular tech and scrounge up as much of that as possible. The Vanguard Group and BlackRock are primary leaders in ESG investments, for example. This way, you get everything: You have some safe income stream from a well-understood, reliable industry, you get some play money for free to invest into a new risky tech that will turn into real money if it works out (and if not, no worries, it wasn't yours anyway!) and you're popular (well, more popular at least than if you just stuck to the mature field).
The last time ANWR was opened, under Trump’s first administration, they held a leading claims auction that was only subscribed to by one single small Alaskan oil company. The larger oil companies gave it a wide berth.
This is principally because large oil companies saw it as more trouble than it was worth, mostly because they would inevitably face years of maximally motivated environmental groups trying to save their white whale through litigation. This process would almost certainly raise the silence of ANWR in the public eye, discrediting these oil companies to the public (as protecting ANWR is relatively supported albeit low-salience); more importantly, any litigation is almost sure to drag on beyond 4-8 years, after which you have a Democratic administration which is very likely to cancel your lease.
That is to say, large oil companies would probably only drill in ANWR if the political climate had changed to the point where they could be reasonably sure that they could sign a large lease while avoiding much of the litigation or the threat of cancellation. That would require a shift in the Overton window, which may have indeed occurred in recent years.
Reading about Rockefeller and how he got so rich, the government broke up Standard Oil because it was too powerful and ended up (unintentionally) making him even richer, since he held shares in the new companies that were founded out of breaking up Standard Oil, and they all shot to the moon.
So sometimes what is intended to happen results in the exact opposite. And oil companies have a lot of public mistrust due to environmental disasters, so - like the tobacco industry - they need to have the *appearance* of being squeaky-clean and not up to any shenanigans. "Why yes, I *am* in the pocket of the polluting industry that is killing the planet" is not a reputation any politician wants nowadays.
This shift into politics seems quite directly counter to the advice from the original blog: "Guided By The Beauty Of Our Weapons", "Beware Systemic Change", general sentiment of trying to play cooperative games. And fair, maybe indeed "this time is different", but the post doesn't really go deep into this. It doesn't explain the tradeoffs. It doesn't actually argue back and forth, it doesn't try to steelman "why is this a bad idea".
For example, isn't it better to throw this money at finding common ground with pro-AI people politically? What are their main arguments and lines of attack? How does this compare to the usual advice of funding EA causes? What is likely to happen if the fight escalates? Who is likely to have more resources? Etc, etc. Just a donation link and a call to action.
One might naively hope that the same strategy wouldn't be as effective for AI as it was for crypto and Israel, since there aren't as many rich AI lovers as there are rich crypto lovers and rich Israel supporters.
In other words, maybe there are more crypto millionaires who love crypto than there are AI millionaires who love AI.
But this is very optimistic and probably false hope.
Wow. This makes me wish the media was a bit more functional! I see there's been a bit of reporting on this, but not a huge amount, and it's a really big story. I'm glad Scott is aware, but this is the sort of thing that the media should be there to observe, and there just isn't very much.
I would model the millionaire's decision to spend or not spend money on politics to be a function of how much of their profits they expect to save (i.e. by avoiding a tax increase, or by getting a promised tax cut).
MONEY SPENT = (TAX RAISE - CURRENT TAX LEVEL) * PROBABILITY OF ACTUALLY IMPLEMENTING A TAX RAISE.
With this model, then it makes sense that they would actually not spend "too much" money if they were to belief that the actual likelihood of them getting a tax raise is low.
In the case of the oíl, finance, weapons industry... there seems to be few realistic chances to raise taxes on them, let alone to destroy those industries. So they spend little.
Another problem is a classical free loader problem.
In the case of crypto (or AI), it was very possible for those industries to be regulated to death (which should be treated in the model as a tax raise that takes away all their profit), so it makes sense that they are "donating" more money than ever
While EAs may have overlearned the lessons of the 2010s, this effort suggests they (we?) have underlearned the lessons of the 2020s. What are those lessons?
- Plunging tons of money into political campaigns does not, in fact, guarantee success, the somewhat limited evidence presented in this post notwithstanding. In fact, plunging tons of money into political campaigns can actually generate enormous amounts of unwanted attention.
- This unwanted attention can fundamentally alter the landscape for workaday EA organizations actively trying to influence policy in the nation's capital to such an extent that some such organizations actually withdraw from previously successful attempts at policy influence since EA relationships become seen as toxic.
- The story "a bunch of weird nerds are obsessed with AI and are spending tons of money to regulate it" is a much more tasty news peg than "rich guys are spending money to influence politics." In every context, the former will generate more news stories and indeed more negative news stories than the latter.
- Lobbying AGAINST something (more regulation of crypto!) is many multiples more cost-effective than lobbying FOR something (more regulation of AI!). This is a stylized fact rather conclusively well-supported in the empirical political science literature — and also one we can see throughout the 2020s.
- THE BETTER CAPITALIZED PARTY DOES NOT WIN EVERY POLITICAL BATTLE. Boy, this is a big one. In crypto and Katie Porter we have an existence proof of a certain kind of political activity.
This is really where I want to hammer on it. Is it the opinion of the AI Safety community that expending $1b on lobbying is the maximally cost-effective way to influence AI regulation? One comes away with the opinion that very little research has been done on the policymaking process.
I don't think the for/against distinction is clearly meaningful. Suppose we said we were only lobbying AGAINST federal pre-emption of AI regulation (a cause that we indeed don't like, and which is our main target)? Or suppose we said that the crypto people were lobbying FOR the GENIUS act and the Strategic Bitcoin reserve?
Given how thin this line is, I am willing to trust the experienced political operatives who have chosen this strategy, rather than your theory that although sure, spending lots of money worked amazing for our opponents, it will mysteriously backfire for us, and therefore we should let our opponents outspend us 100-1 and win easily.
It is, in fact, a distinction in the poli sci literature — as is the inconsistency of value in outspending. It is by no means an established fact that big spenders win elections or get their policy priorities implemented as a matter of course. As for the former issue, the distinction here is between "maintaining the status quo" and "making change":
Again, the issue here is what's valuable on the margin — trying to compete in this arena is *enormously expensive* and this doesn't just need to be *a good idea* — it needs to be *the best idea on the margin.*
"Plunging tons of money into political campaigns does not, in fact, guarantee success, the somewhat limited evidence presented in this post notwithstanding. In fact, plunging tons of money into political campaigns can actually generate enormous amounts of unwanted attention."
Yeah, I definitely think the Carrick Flynn election result burned them on pushing for any more directly political campaigns (I'm sorry, I mean no disrespect, but this story always makes me laugh since it was so clearly the triumph of idealism over going with the reality on the ground). That, plus SBF being the one shoving money into such backing and similar donations and the subsequent black eye that gave the movement.
Wow, you've really crossed over into "Jews pulling the strings to control society" conspiracy land. Honestly, this post should be the last nail in the coffin of the rationalist movement. Rationalism has a lot of good criticisms of other ideologies, but as a positive theory in its own right, it's time to let go.
Every immoral ideology merges with antisemitism eventually; and for a philosophy that sees itself explicitly as the antidote to conspiracy thinking, seeing you transition so smoothly into embracing the oldest conspiracy theory is a clarifying moment. What was it in the end? Did you just fall for the avalanche of defamation and blood libels? Or was it reactionary hatred of Trump?
Are you aware that Scott is a member of the "pullers of the strings to control society"? It's a lighthearted self-effacing joke which made no conspiratorial insinuations and just stated obvious public information, relax. If anything it's poking fun at the conspiratorial-minded.
(((Scott Alexander))) is pulling on the strings to control society by writing about what he thinks about issues of importance on his popular blog Astral Codex Ten. A lot of smart and influential people read the blog and might be influenced by the ideas they see published there and also by the ideas of other people publishing blog posts engaging with (((Scott's))) posting. Also I heard that he publicly gave grant money to a bunch of people and organizations who he thought were doing good or interesting work in the world, which makes him kind of similar to (((George Soros))).
I suspect this is second- or third-hand influence, and that Scott learned about PACs from a source that is generally reliable except for the lean towards specific dangerous tropes (or a source that themselves learned from such a source), leading to the distortion being presented as fact. This is the kind of thing that will often happen to people who are not themselves at all antisemitic.
(An alternative hypothesis is that this presentation of AIPAC/Jewish donors is actually accurate, but when you're in a society that spontaneously generates "Jews control everything" conspiracies regardless of circumstances, your bar for sufficient evidence for that should be *very* high.)
I think it is unlikely that Scott is a self-hating Jew.
What do you say to "AIPAC is, in fact, Jews pulling the strings to control society"? If it's an organisation set up to defend Jewish interests and the method of action on that is influencing elections?
I don't think it's anti-Semitic to go "yes, this is a Jewish organisation promoting Jewish interests". If every criticism of anything Israel, Israel-supporting organisation, or Jewish actions gets met with "anti-Semitism! you're trying to send me to the gas chambers!" then eventually it will be crying wolf and people will take *more* notice of those who want to say "psst, lemme tell you about some shady stunts these guys are pulling".
With crypto, I'd keep in mind that Trump won the popular vote by less than 2% against the unpopular last minute replacement for a historically unpopular incumbent. The Democratic base is now very against crypto and a lot of Silicon Valley stuff, and if Harris had won, they might be in a bad situation now. It's not clear that spending all that money actually is some unbeatable strategy.
My sense is that the Musk buying Twitter thing really did help but the situation is a bit sui generis.
Also re AI, they tried to add a "states can't regulate AI" provision into their law earlier this year and it generated backlash and got dropped.
Now _that_ would cost a lot in political donations!
California water rights got grandfathered in in 1915. Increasingly, California's hereditary water baron families intermarry so they are getting fewer in number but richer.
Somebody brought up the idea of water property rights reform to Gavin Newsom. He agreed it would a good thing to do, but said that no way was _he_ going to lead a political campaign against the Water Powers That Be.
Maybe the reason oil companies dont buy politicians in the direct sense is that they find it more effective to buy issues via the economy. Oil supports lots of jobs etc, which is why there are no shortage of pro oil politicians.
This makes crypto well suited, maybe uniquely suited for buying politicians directly. Theres tons of money in it without supporting many jobs somehow (kinda the point of the anti crypto position, dont interpret that as me being pro or anti)
I hear AI is supporting the global economy these days. That might be more important than their PAC, but by running both, I guess we wont get clean data
Maybe we can get some good AI regulations but the odds are the regulations will just slow the industry down without doing much good, I think the potential harms of a future AI are to broad to be able to affectivley pin down in legislation
AIPAC was not founded as a PAC; it only became one in 2021. It has historically disbursed a smaller dollar amount than many industry lobby groups and quite a few actual foreign countries lobbying. Compare their 10s of millions to China dropping 1/2 a billion dollars, and the Saudis 1/3 of a billion. Until October 7th, iirc AIPAC was smaller than most ag sector lobbyists.
AIPAC gets its way far less often than other domestic lobbies, especially on the issues it has considered the most important. They completely failed to prevent the JCPOA, which American Jews felt was a bad deal. In fact, they failed so badly that they got replaced by J Street within the Democratic Party, and J Street is at best a fringe organization of American Jews.
Now, AIPAC says they’re successful, and of course they want to say so because who would donate to an ineffective lobby? And antisemites say that AIPAC is successful, because without that conspiracy they might have to reckon with the idea that most people don’t like to see Jews killed.
But are they successful? What specifically have they accomplished in the last twenty years? Thirty? Forty? AIPAC claims they got the development aid and military aid for Israel. Did they? Or were there clear strategic reasons that America chose to make those interventions - such as stabilizing Israel in the Yom Kippur War, and securing Israel enough that they’d go along with risky diplomatic plans such as peace with Egypt or the JCPOA - where America was just following its own interests as it saw them? Those three events account for, if memory serves, more than half of all aid to Israel period, including development aid. AIPAC is pushing on open doors, at eye-watering cost, and thinks that makes them kingmakers.
And if AIPAC *isn’t* the kingmaker they like to say they are, then repeating their narrative can play into the hands of the blood libel of Ilhan Omar and co.
I wish you (and Americans more generally) would consider the implications of posts like these - describing moneyed interests who support often unpopular positions (Cryptocurrency, Zionism) outright sinking hostile candidates with attack ads on unrelated issues, while funding both sides of competitive races - and reconsider their reflexive defenses of the American political system, or their analysis of the country as a democracy.
I think it will go to their heads, they'll try spending $500 million in 2026, and the results will be really disappointing for them. Congressmen obviously don't like having tens of millions in attack ads dumped against them, but ultimately Porter was a weak candidate in a race against a far more popular, high-profile Democrat who ultimately won it (Schiff), and on an issue where most of the Democrats don't have strongly held opinions on the topic.
As I pointed out in the 2019 post, "that's what we spent on almonds, so it's not a lot of money" is bad reasoning because just because we think of almonds as insignificant doesn't mean that the amount is small.
I decided to look up https://data.ers.usda.gov/reports.aspx?ID=4057 which shows almonds at 5.8 billion, but that's "cash receipts", not the size of the industry or how much Americans spend. Still, there are a lot of other crops in the table and I should be able to get relative sizes without having to compare apples to oranges (except literally). Almonds beat out every single one of the 32 "vegetables and melons" (though potatoes are close) as well as all 31 other items in the "fruits and nuts" category except for grapes.
Almonds are about half of wheat and beat out sugar cane and sugar beets combined, cotton (though only for 2024), and tobacco. They're still nothing compared to animals or feed crops, and nowhere near the military or health care, but why Scott thinks almond spending is insignificant is beyond me. Something that spends as much as we spend on almonds is a big deal.
The table also seems to give single lines for things that are both consumed by humans and used as animal feed. Separate googling claims that about half of wheat is used for animal feed and 90% of soybeans. In 2019, when Scott made his original post, grapes were less than almonds, and almonds were more than now (in real dollars), which means that Scott managed to choose an "insignificant" example that was the highest of all food crops.
This discussion is such a narrow view of the issue. The real damage to American politics (which is now washing into international politics) was the Citizens United decision. The real question here is not, “how effective is political spending?”, but “Would American politics be better if it was not completely dominated by campaign finance?” Since I’ve seen estimates that Congresspeople spend 2-3 days per week just raising money, the answer is probably “yes”, just from a functional point of view. And since multiple studies have shown that those policy measures that are implemented are much more likely to be those that are aligned with the interests of the rich, the answer is probably “yes” from the perspective of a functioning democracy.
Why is Marc Andreessen with his puny $2B even relevant? During the 2024 campaign, I have heard (not sure if this is 100% true) that Elon Musk has spent about $250M to support Donald Trump. That's not even mentioning him buying Twitter. There seem to be a lot of people who could spend way way more than Andreessen even has at all.
Also, I know that the author is very concerned about AI alignment and regulation, but aren't there some... more pressing issues about the American democracy, the ones which warrant our attention? Or is that too small in the face of AI?
Assuming someone or a lot of people do this and start outright buying politics generally considered left-wing leaning (to be concrete: ai safety, regulations in general, things that are bad for profit of business head figures) isn't this like opening the flood gates? The people on the opposite hhave far more money and so far didn't use it to buy politics, except in case of silicon valley they kinda did.
What happens if every billionaire starts this and no one has any illusions anymore if you can or can't buy democracy anymore, because obviously you can and you did?
Its probably an irrelevant fear because the arms race already kinda started but still if everyone goes full on plutocracy it seams doubtful to me if this is really a net positive. On the other hand we're already headed there so what is one more push and trying to change direction of the plutocracy with that push ¯\_(ツ)_/¯
I find it mind-boggling that the conclusion of this post isn't "and we should do X and Y to limit the influence of political donations in US politics, here's how you can contribute" and instead the conclusion is "and we should get in on the money-burning game, here's a guy on our tribe's side, give him your savings".
Why are Americans so primed to believe that massive political donations are an inevitable fact of life?
But the US differs from Europe/UK in degrees. Even in Europe election campaigns cost a lot of money. Someone has to fund that. That someone gets a lot of say.
EA should make a PAC that threatens african leaders with Uncle Sam's Big Stick/Denial of Carrot if they don't use their Rolls-Royce and MiGs budgets to buy anti-malarial nets and dewormers to everybody who needs them.
It would be cheaper and more effective than EA paying for them directly and trying to distribute them in the middle of nowhere.
Wait a second - that Katie Porter story? So she *already* got slapped down hard in an election?
And her takeaway from that was *still* “How would I need them in order to win, ma’am?” for the governor of California election? 🤦♀️
I now wonder how the hell she ever got elected to the House of Representatives in the first place! (I can well believe the mashed potato story and the rest of it, but how did someone this clueless ever succeed in an election at all?)
You need to include in your model that /the battle does not stay won/. At best, it stays won for four years. So the right way to think about the cost of getting a policy you want isn't a one-off payment of $X million; it's an ongoing expense of $X/4 million per year - and this not for a certain purchase, but to shift probabilities - increasing over time with at least inflation but also how annoyed your political opponents get with not having their way as time goes on, for as long as you need the policy in place.
Certainly the billionaires will be factoring this into their ROI calculations.
1-Other industries know how far their money can go in politics vs how far it can go doing something else. Example: Oil can price a congressman vs down-payment on a new vessel.
Crypto doesn't actually produce anything, crypto billionaires can buy more coin and/or make the coin they have more valuable.
Maybe they did the math and realized they're at a point where buying more is not going as far as making it a bit more valuable/preventing it from crashing. Maybe right now politics is their blue ocean.
Maybe it's the same for AI, they've reached a point where their money in the actual business only goes so far, because they keep burning money. Maybe dumping this money into politics means AI and crypto are showing some concern with a bubble.
2-Other businesses have long bought their way into politics and the economy, AI and crypto are the nouveau riche and just now are buying their way into politics. Maybe what they're spending now is not so different from what other businesses have spent in the past, or total spent over the years, adjusted for inflation.
I'd like to separate the issue of AI safety from whatever opponents of AI mean when they say "regulation" (all while making light of the safety issue). At least in online discourse and elsewhere, "AI regulation" seems to mean something between taxing, neutering, or banning of various AI tools, with some "think of the children" thrown in for good measure.
I'm sure the PACs will be inclined to fight both forms, but today's AI majors seem likely to acccept a form of safety-as-regulatory-capture, and I suspect they're less likely to go after a vaguely worded "test before you ASI" bill than a "stop AI from stealing our water" bill.
I need to admit that my honest reaction to this post was "come on, Scott, don't feed Moloch." I realise that figuring out a way to stop the bad thing from happening, rather than co-opting the bad thing for your own purposes, is really difficult to even figure out intellectually and even harder to orchestrate (all of which needs more time than the co-opting, besides), so I'm not upset or even want to particularly stop you. Instead, I sincerely wish you well (even though I don't have full agreement with even the cause). But I do hope you reconsider the whole approach.
Heck, even if you don't actually do another approach, I'd be super interested in knowing what you can come up with to solve the underlying issue. You've already thought about it a lot, so I think you could come up with some pretty interesting approaches. So there's at least another article that could come out of this? :) I'm hoping.
Leading The Future still only has 2% as much money as the almond industry.
No, not exactly. The almond industry does not 'have' $12B, it has revenue totaling that. This figure comes from the sales of Almonds and products mostly in American supermarkets. a lot of that number is Supermarket profit, and groceries are a notoriously low-margin business. Shippers, packers, processors and truckers get most of the rest. In the produce industry you are lucky if 10-20% make it back to the grower. Thanks for mentioning produce, though!
Voters in the UK should be worried too. Imagine a trillionaire who owns a social media company. He is best friends with the leader of one of the parties, and wants his party to be elected.
Campaign finance laws are really strict in the UK (£11,000 per donor, per year; total donations in 2021 were £50m). What's to stop this trillionaire from buying the election? Musk doesn't even need to give any money to Farage. He can just buy all the ads on YouTube and Facebook — and he already owns X.
Americans have never been interested in British elections, but for some reason, we now have Evangelicals campaigning about abortion laws, plus Presidents and Trillionaires speaking up for their preferred candidates. What happens next?
I really wish you wouldn't use AIPAC as your go to example of a single issue pac, given the number of people who (mostly for "Jews control the world" reasons) vastly overestimate the power of AIPAC (it's neither especially influential nor an especially big spender compared to other pacs in its class).
My source said it was extremely influential compared to other PACs in its class. I am basing this on their claim, but feel free to link me to evidence that they're wrong.
Did your source mention the importance of "legislative subsidy" for lobbying? Probably less relevant for AIPAC than more domestic lobbies.
I imagine that the apparent importance of a particular lobby group is going to look enormously different depending on where you're sitting. For example, when I was involved in local/state politics in California c. 2008-2015, I heard essentially nothing about AIPAC. The big dogs I kept hearing about over and over again were SEIU on the Democratic Party side and Charlie Munger, Jr. on the Republican Party side.
Any PAC that is consistently on one side is dramatically weaker than "mercenary" PACs like AIPAC.
SEIU only has influence when they spend money. Because AIPAC is occasionally willing to pour money on either side of the aisle, they also have influence in all the races where they don't spend money, by employing the threat of funding the other side. There are a lot of politicians that are afraid of AIPAC flooding the donations to their future challenger, and thus keep their mouth shut about certain subjects even if AIPAC never spends any money on their races.
The other political organization that used to be disproportionately powerful compared to the amount of actual money spent was NRA. This was when NRA was entirely willing to fund the democrat if the republican they were facing was wavering even a little bit on gun rights. As NRA has become more and more just a part of the republican party extended universe, their influence has waned.
Washington Post columnist Dana Milbank used to write a hilarious annual column about the extraordinary lengths AIPAC goes to each year at its convention in D.C. to impress members of Congress with its wealth and fervency.
From https://www.washingtonpost.com/archive/politics/2005/05/24/aipacs-big-bigger-biggest-moment/48af6c09-c5e7-46e3-8ed8-87c788546c7b/
AIPAC's Big, Bigger, Biggest Moment
May 24, 2005
By Dana Milbank
How much clout does AIPAC have?
Well, consider that during the pro-Israel lobby's annual conference yesterday, a fleet of police cars, sirens wailing, blocked intersections and formed a motorcade to escort buses carrying its conventioneers -- to lunch.
The annual meeting of the American Israel Public Affairs Committee has long produced a massive show of bipartisan pandering, as lawmakers praise the well-financed and well-connected group. But this has been a rough year for AIPAC -- it has dismissed its policy director and another employee while the FBI examines whether they passed classified U.S. information to Israel -- and the organization is eager to show how big it is.
Reporters arriving at the convention center yesterday were given a list of "Food Facts" for the three-day AIPAC meeting: 26,000 kosher meals, 32,640 hors d'oeuvres, 2,500 pounds of salmon, 1,200 pounds of turkey, 900 pounds of chicken, 700 pounds of beef and 125 gallons of hummus.
Another fact sheet announced that this is the "largest ever" conference, with its 5,000 participants attending "the largest annual seated dinner in Washington" joined by "more members of Congress than almost any other event, except for a joint session of Congress or a State of the Union address." The group added that its membership "has nearly doubled" over four years to 100,000 and that the National Journal calls it "one of the top four most effective lobbying organizations."
"More," "most," "largest," "top": The superlatives continued, and deliberately. In his speech Sunday, the group's executive director, Howard Kohr, said the "record attendance" at the conference would dispel questions about AIPAC raised by the FBI investigation.
"This is a test, a test of our collective resolve," Kohr said of the "unique challenge" presented by the FBI probe, "and your presence here today sends a message to every adversary of Israel, AIPAC and the Jewish community that we are here, and here to stay." (The official text has two exclamation points after that sentence.) ...
I've also found AIPAC to be a lot more pro than pretty much everyone else in politics.
Second-largest for direct contributions to candidates (https://www.opensecrets.org/political-action-committees-pacs/top-pacs/2024), plus very widely reported much stronger donor list than anyone else (as Scott states). I'm not sure how much has changed in the past few years I imagine AIPAC is much much weaker than it used to be, partially because the salience has gone up dramatically which always weakens single-issue groups in the area, and partially because "I'm the best friend of Israel / No I'm the best friend of Israel" is no longer the basic description of political campaigning on the topic. But I don't think that, for example, Mearsheimer randomly decided to be anti-Semitic.
I guess the question is "is AIPAC far more effective than the National Realtors Association or Beer Wholesalers Association?"
AIPAC goes to great lengths to get powerful people to believe it is powerful
It sounds kind of anti-Semitic to pooh-pooh AIPAC's claim to be powerful by asserting that the most important Jewish lobby is notoriously deceitful.
Scott is claiming that AIPAC is "orders of magnitude" more effective than other PACs, which seems like a stretch if their only real advantage is more hard-money funding due their donor class composition and more effective bargaining technique. We're still looking at ~2% of US lobbying spend.
How many other Jewish organizations are there that will follow AIPAC's lead on an issue or a campaign relevant to Israel? There are enough major Jewish organizations that there's an influential "Conference of Presidents of Major Jewish Organizations."
How many big donors to candidates or universities listen to AIPAC? The top 100 political donors in the U.S. are about half Jewish.
The top 100 billionaires are about 1/3rd Jewish. My impression is that Jewish billionaires (e.g., Larry Ellison, lately the world's richest man) tend to be more Zionist than average Jews, just as rich men in Alabama tend to be really into the U. of Alabama Crimson Tide football team. Israel is to a lot of rich Jews what the Notre Dame Fighting Irish is to a lot of rich Catholics.
What fraction of the prestige press is Jewish? When The Atlantic made up its list of the top pundits about 15 years ago, it was about 48% Jewish.
I read the Jewish press a lot so I'm more aware than most people of just how much money and influence Jewish Americans have earned over the last 150 years. It's not just a "trope," it's hard numbers.
Since October 7, 2023, Jewish support has been more up for grabs than it has been for decades. Anybody who wishes to play a role in American public life can't afford to underestimate what Jews have earned for themselves.
I think the big issue is that AIPAC and Israel are following the NRA playbook (and to some extent Planned Parenthood/ACLU) where instead of being single issue pacs that try to play nice with both sides but care deeply about the issue become partisan actors (AIPAC is still bipartisan but less so).
It's interesting that exhibiting more loyalty to a given party dramatically reduces the actual payoff to your org, but there you have it. The NRA doesn't seem to follow the same playbook so much any more, though.
https://www.astralcodexten.com/p/tech-pacs-are-closing-in-on-the-almonds/comment/168560511
There's objective data available on that question. You can get top donors from OpenSecrets.org. The ethnicity of big donors is not all that hard to find from Wikipedia, Jewish magazines, and the like.
Among the top 100 donors to political candidates in the 2020 election, whether Republican or Democrats, people of Jewish ancestry gave 61% of dollars. In 2024 among the top 100 (a group that somewhat overlaps the top 100 in 2020), however, there was a 6% decline in dollars given by big Jewish donors to Democrats and a 129% surge in dollars given by big gentile donors to Republicans (with Elon Musk accounting for about one-third of that increase). So, the Jewish share of dollars donated to candidates by the top 100 fell from 61% to 39% in 2024.
For details and methodology, see my 2025 column at:
https://www.takimag.com/article/semitical-thinking/https://www.takimag.com/article/semitical-thinking/
Jews very likely give smaller shares of total political contributions of all sizes since they make up about 1/3rd of the top 100 names in the 2019 Forbes 400.
Gave *similar shares, perhaps?
Among campaign contributions from the top 100 donors in 2024, 68% of the Top 100 money going to Democrats came from Jewish givers, while 31% going to Republicans came from Jewish givers.
Those shares were down somewhat from 2020 and 2018.
I don't think it's Good for the Jews if we try to say that a thing that, as far as I can tell, everyone with any experience in the area agrees is true, is anti-Semitic (or at least in some gray area of "I won't call it anti-Semitic but I'll gesture in that direction").
I think there are plenty of things you can say that recognize AIPAC's influence without getting into Jews-run-the-world territory. E.g. Israel has historically been popular even with non-Jews, and it's a low salience issue, both of which raise its effectiveness.
And now that the circumstances have changed, you have things like moderate Dem Seth Moulton *returning AIPAC money*.
Last time I checked in 2011, since 1901 Jews had comprised between 21% and 26% of the Laureates in each of the three hard science Nobel Prizes (physics, chemistry, and medicine) despite making up only about 0.2% of the world's current population.
My general opinion is that Jews _are_ strikingly rich and influential ... because they've _earned_ their riches and influence.
And also, having been persecuted worldwide for literal millennia, they've learned to stick together pretty well. Of course, this history also means that when any outsider mentions this obvious fact, by default it's taken as an attack, not as a compliment.
I've read Jewish-American and Israeli periodicals more than most, and they _love_ enumerating Jewish billionaires and making other lists quantifying Jewish power and success. The New York Times, in contrast, doesn't usually find that news fit to print.
Often, I get asked accusingly: "Why do you want to _know_ about who has clout in your country? Huh?"
Well, I like knowing data. And I like knowing how my country works.
Contribution rules say:
I am a U.S. citizen or lawfully admitted permanent resident (i.e., green card holder).
This contribution is made from my own funds, and funds are not being provided to me by another person or entity for the purpose of making this contribution.
What would the legality be of donation swapping? Let's say I'm not a US citizen, but I'd like to donate. So I find a US EA who doesn't want to donate, and say "If you donate X to this politician, I'll donate X to a charity of your choice, that you would otherwise have donated to".
In this case, the US citizen is giving the money, and I haven't actually provided them any funds...but I'm not sure if this is a strong legal defense or not.
I don't know. The closest case seems to be https://scholarship.law.unc.edu/cgi/viewcontent.cgi?referer=&httpsredir=1&article=1007&context=ncjolt . For now I would not do this, because I think the ethical grayness outweighs the chance of gaining something.
For reference, VOTE swapping is legal.
This is almost certainly illegal for both you and the other person under existing laws (straw donors) although whether you will be prosecuted for it or the existing laws will be found unconstitutional depend heavily on the party in power.
IANAL, but I don't think this is legal. The relevant law can be found here, https://www.law.cornell.edu/uscode/text/52/30121 and criminalizes "direct or indirect" donations. My viewpoint is that if you are making a contribution with the intent of causing another person to make a contribution that you are prohibited from making, that is the kind of indirect donation that this law targets.
Seems weird that it's illegal for a foreign citizen to donate to a PAC, but it's also legal to have a PAC whose explicit and open reason for existence is to advocate for the interests of a foreign country.
That rule feels very correct to me. If PACs are supposed to represent the “voice of the American people”, it’s entirely appropriate that Americans get to decide what they do for others using their government, and it feels quite inappropriate for outsiders to use the same mechanism.
You won't need a legal defense since the only way the government finds out is if one the other guy snitches on you, and he has no incentive o because he's just as guilty.
Indeed, I'm not in the US, but if I were this sounds like it would be *by far* the most effective way I could spend $7k to reduce the chance we all get turned into paperclips or similar.
I feel like the most obvious rallying point here is to go absolutely nuclear on anybody that supported defunding PEPFAR. If there's one single program everybody at every level of EA agreed was very good it was that one right?
Although it would be nice to have a generic EA PAC, for various reasons the system is more set up for single-issue PACs, and the people I know are prioritizing AI - partly because they think it is more important, partly because it is faster-evolving, and partly because the people who care about it have more money (since many of them are in AI themselves).
Which do you think would be better? A PAC focused on PEPFAR or AI?
This goes through lots of information about politics (like what sorts of pressures representatives are amenable to, and how party-line votes will be) that I don't know.
PEPFAR is a much less controversial program to support - "save babies dying in a painful way in countries that will never in any way rich the US" is a pretty easy sell. AI controls is much harder - "cripple a major source of economic growth that also happens to have highly salient military applications because of some hypothetical threat. China also seems to heavily invest in this btw."
For what it's worth, I think AI doomerism is silly and unfounded (yes I've read the book). So if you disagree with me on that, then at least take rallying around PEPFAR to establish credibility then go for AI research as advice.
> "cripple a major source of economic growth that also happens to have highly salient military applications because of some hypothetical threat"
The economic growth is roughly as hypothetical as the AI doom scenario atm (stock market gains are not generalising outside the top 10 US tech firms, and we know that generative AI is having all kinds of negative impacts on entry-level employment and student competencies.) Military applications I'll give you, sure, but that doesn't seem a million miles removed from the doom scenarios.
Well, if the AI bubble pops, the crisis that follows will certainly not feel hypothetical, for the entire economy. Current projections are "very bad, but perhaps not quite 2008 level bad".
Wouldn't a PAC focused on PEPFAR, just be one of many pro-Democratc PACs?
PEPFAR used to be extremely bipartisan and even now I don't think there's a strong anti-PEPFAR position in Congress so much that nobody is really crossing the Trump administration on their destruction of it.
So it would mostly just be an anti-Trump (and eventually anti-Vance) PAC, which feels like a saturated market.
Well, it would be an issue PAC, trying to get support for a specific issue. Congress doesn't need to wildly buck its trend of surrendering to the President on everything, just on one issue he probably doesn't care that much about, for it to be successful.
Has Trump ever said anything about PEPFAR one way or another? I won't be surprised if he doesn't even know what it is. As I understand, it basically got in the way of Musk's chainsaw, without anybody specifically targeting it, then reinstated in a crippled state and promptly faded from notice.
I have a few concerns that I'm curious if you share:
1. AI accelerationists are probably going to be at least as much money as the EA AI safety people, so its possible you're going to just end up in a stalemate. This might still be worth it if you think absent your actions the accelerationists will just stomp, but it probably won't see the total victories M.A. has achieved.
2. Crypto and (until recently) Israel were pretty low salience issues in politics. Given that we are one disappointing Gemini demo away from a recession, AI probably will be high salience for the forseeable future. Nobody cares about PEPFAR on the other hand. It was basically canceled so JD Vance could impress a bunch of teenagers on discord. So a smallish, dedicated group could probably make a big difference at the margin.
Stopped reading at "Israel is a low salience issue in politics".
The Only Democracy In the Middle East?™️ Low salience?
He did say "until recently". And yeah, up until certain MENA migrant groups inveigled their way into the coalition of the fringes in the last few decades, most people in the US either didn't care much about Israel and or were mildly positive toward it.
PEPFAR would have been a great choice 2 years ago, but now it cuts on party lines. AIPAC was so successful because they got both Republicans and Democrats to support it. Now it’s unfortunately politicized. If anything I think a PEPFAR super PAC would be a good idea in 3 years, when it’s not as connected to the incumbent running for office.
I dont know if partisanship is a huge issue here. Theres always going to be a streak of "foreign aid bad" populism in both parties, so even if it becomes a partisan issue you can still have a lot of influence in the primaries
But who in congress had a pivotal role in defunding it? I thought elon defunded it while marco rubio continued to insist it wasnt defunded and congress largely demurred. There may have been GOP members who cheered elon on twitter, but not in a congressional capacity. Also getting rid of them wont get pepfar back.
Is PEPFAR canceled? I thought they reauthorized it and it’s been operating until the expected congressional reauthorization next year.
That’s accurate. But it’s become a matter of “common knowledge” in online spaces that it has been defunded and is gone now. But it’s not. PEPFAR is supported by just about everyone in Washington, on the record anyway.
This is news to me. I have updated my views to be that San Fransisco must be destroyed.
Feel free to start a SuperPAC to that effect, although I think competing SuperPACs will probably outraise you.
There should be a Dominant Assurance Contract contingent on destroying Godzilla should he surface there.
I dunno, Max Zorin and Lex Luthor both have pretty deep pockets.
Don't they have residences in SF too?
The joke was that both villains, in movies from 1985 and 1978 respectively, separately plotted to destroy San Francisco (or California more generally) by various plans involving triggering earthquakes.
Luthor's primary residence is in Metropolis, while Zorin lives in France and also a blimp.
Well I've got good news for you about SF...
Effective altruists should donate their money to GiveWell recommended charities, like the ones against malaria & parasitic worms, instead. That's what I do.
Please provide the reasoning behind your assertion.
GiveWell analyzes the best places for donated money. Those places avoid the political tug of war. https://www.overcomingbias.com/p/rope-tugs-are-not-charityhtml
GiveWell only claims to analyze the best places for global development donations. They make no claim to have considered issues like AI safety, and many of the people involved in GiveWell themselves make AI safety related donations (or donations to animal welfare, or other causes that GiveWell doesn't look at).
Although it would be nice to be able to avoid politics entirely, politics is also extremely very important, and it doesn't seem ideal to leave it entirely to the industry billionaire lobbyists. I don't think the usual "D and R donations cancel out and it's all a wash" argument applies when an industry is trying to spend hundreds of millions of dollars to make politicians serve its naked interests.
It means you already know you're in a tug-of-war with someone. Someone whom you know to have plenty more money. I guess if you could get someone far richer on your side you could convince them they couldn't outspend you, like the Soviet Union in an arms race with the US. Elon Musk is supposedly the richest person, and played a role in founding OpenAI due to an interest in such risk, so maybe you have some hope. But that would be hope from appealing to him (or a set of people with a combined comparable amount of wealth).
I think the best argument against this are what you wrote a decade ago https://slatestarcodex.com/2015/09/22/beware-systemic-change/
When I read that piece when it came out, it convinced me to disengage from politics, which in retrospect was a great choice. Looking at my peers that did go into politics at the time, they are now full blown MAGA Trump loyalists. I definitely think you gave good advice then.
That said, times change, and this issue is not Republican Democrat, but an internal war in Silicon Valley that could kill us all. I’m going to need to reread both pieces and contemplate where is the best place to make my annual donation.
I quote:
"Lots of people work very hard and raise $10 million for the Democrats. Lots of other people work very hard and raise $10 million for the Republicans. Now the Democrats and Republicans are at exactly the same position vis-a-vis each other as they were before the effective altruists got involved, but we have wasted $20 million that could have gone to healing the sick or feeding the hungry. "
Indeed. It is, realostically, an arms race normal people have no hope of winning. And
That is simply false because you donating to something will not magically cause more people to come out of the woodwork and support your opposition. Also fighting someone in zero some contest is better than just rolling over and letting them win. There is a reason that countries at war don’t just cut the defence spend to 0. Even though both countries are burning resources could spend on something else. Unless you can do a credible deal for both of you to reduce your defence expenditure, it doesn’t make sense to unilaterally step back.
> Also fighting someone in zero some contest is better than just rolling over and letting them win.
It's 2025, and I do wonder about that. Is it really always better? Decades of low-grade pain, waste and slow decline because the idiocy just. won't. die and keeps attracting more and more richer and richer followers; vs one term of total capitulation, just turn the other cheek, lean in hard, bend over backwards, get it all out of our system at once, until the effects are utterly impossible to ignore, so we can finally all agree never to raise the matter again, and when some newborn sweet summer child who did not live through the idiocracy does bring it up the scars and smoking ruins are /right there/, proof positive of where it all leads, all the veterans make it very clear why it's utter bunk until the idea is firmly buried again, like the apocryphal monkey ladder experiment...
...and then I wake up and remind myself that people can stay irrational longer than the world can stay unburned.
That’s an argument for not being the one who starts the arms race. The post is saying that the other side has already started it.
I find it interesting that you disengaged from politics but kept the greco-roman bust avatar. Also, what's up with the blog?
I disengaged in 2015, reengaged in 2022, supported Nikki Haley, had a blog for a couple months, realized that was a mistake, took down all my posts in the blog, and am back to being disengaged. Even my blog was designed to get more people on the Right to become Non Partisan Civil Servants, but there is no appetite on that from the Right on Substack. They would rather tear down the institutions entirely than try to compete in them, and the current president, who I didn’t vote for, is happy to do that.
Even during that phase, I kept my donations going to pandemic preparedness research, and never donated to a politician. Also, Greek and Roman history is cool, and I’ll never apologize for having cool ascetics even if they signal that I’m on the right.
Hey, I got no beef with greco-roman aesthetics. I think going scorched-earth on the institutions is the only realistic strategy at this point, though.
Recently, I've read of a movement to undo Citizens United by state-level action: https://www.americanprogress.org/article/the-corporate-power-reset-that-makes-citizens-united-irrelevant/ If this succeeded, wouldn't it undo the PACs and erase this concern? Could be another route.
It WOULD undo PACs, yes, but it would also abolish the freedom of the press as generally understood if any state can decide that, say, the New York Times, as a corporation, may not do business in that state.
Has that happened in other countries with campaign finance regulation?
Do you have a specific example of country you think has what, say, the median American would consider a free press where, sticking with the Citizens United case, if a book was being published critical of a candidate by an organization not granted state imprimatur, that the government could stop publication, citing "campaign finance regulation"?
I think if you’re so sure in your convictions you could name examples. Do you think UK tabloids don’t attack candidates?
I could indeed name several examples of countries where I believe organizations are restricted from publishing and distributing opinions disfavored by the government – Afghanistan, Belarus, China, Djibouti, Eritrea,… – but how would that help? You'd just say those don't count because you had some other country in mind that made your point better.
But okay, that's more concrete: you're suggesting the UK as an example. Even granting that most Americans would consider its press free, I haven't seen an instance of the government imposing prior restraint on publication like what was forbidden in the Citizens United case.
Your claim was that a state could decide the NY Times couldn’t do business there. I think that’s unlikely and a straw man argument. The UK has fairly strict laws governing actions by non-party actors to influence an election, but the tabloids continue to operate.
It happened in this country before Citizens United. The CU case was literally fought over the government banning someone from producing a political documentary.
I dont see how this plan can work as a run around. Those laws would be struck down on citizens united precedent straight forwardly. The whole corporations as a social/legal construct thing is fine philosophically, but is not relevant to 1st amendment rights legally. Persons are not given first ammendment rights by law, they precede all law. Under citizens united, corporate person hood gives them 1st amendment rights. You cant restrict that with either campaign finance law, or corporate charter law, for better or for worse.
It might undo PACs but then something else would spring up in their place. Politicians like and need money for their campaigns, and interest groups want to influence politicians to make laws and govern according to their ideas of what is best.
So if we (interest group that wants to lobby for all public spaces to be painted magenta) can only donate X amount directly as individuals because of pesky laws, and we can't found a PAC, then some other loophole about "getting a bunch of us together to hand over $$$$ to persuade politicians to pass laws about painting all public spaces magenta" will be found (e.g. "we are all close personal friends of the candidate in this constituency and their family and are related to them by marriage and other ties, and we are giving them wedding, birthday, and Christmas presents of nice fat cheques to spend as they wish on whatever they wish, no strings attached, all perfectly legal, oh what a coincidence that Candidate's cousin owns a paint shop which will supply all the magenta paint for the public spaces").
Very interesting. It does, indeed, seem strange that no-one has tried this before. My instinct is that this is the kind of situation in which everyone assumes that, if it worked, everyone else would already be doing it. But ultra-ambitious risk-taking maverick billionaires (i.e. the people best suited to using this strategy) don't seem like the kind of people to make that particular mistake.
Minor proofreading things:
> is a little bit paranoid and will truly accept that their safe seat is safe
I think there should be a "not" or "never" in this clause.
> and occasionally crossed skirted the border of illegality
You only need one of crossed/skirted.
It's probably just a case of everyone being pretty happy with the status quo and not wanting to rock the boat by outright buying politicians. The country is already as business friendly as you can reasonably expect a country to be.
The $20 dollars on the sidewalk/everyone before is an idiot; claim is wrong the vast majority of the time, but sometimes it isn't.
The greatest breakthrough in human history came from putting bread mould onto wounds and it was only widely used in the 1940s, wheeled suitcases were only introduced in the 70s.
And even as late as the 90s the leading design for a wheeled suitcase had two wheels on a single corner and always wanted to tip over.
Penicillin is the greatest breakthrough in history??
Yes.
I think the fact that it happened with bitcoin demonstrates the reasons why it doesn't necessarily happen with other things.
Let's compare bitcoin to oil. Bitcoin is pretty much the optimal case for the value of buying politicians. There's a significant number of very rich people whose fortune is almost entirely tied up in bitcoin. The value of bitcoin is very dependent on US government policy, to the extent that there's plausible policy approaches the US Government might actually take which would 100x or 0.01x the value of bitcoin. And a lot of politicians haven't already made their mind up about bitcoin, so they're relatively easy to influence. Therefore it's easy for a single individual to spend a dollar influencing politicians and expect to get a greater-than-$1 return on it.
Compare to oil. There's a lot of money in oil, somehow, but it's pretty diffuse. Oil companies are ultimately owned by a bunch of people, most of whom own a lot of other things and are not single-mindedly exposed to the oil industry. Furthermore it's not clear that plausible US Government policy is going to have a huge impact on the fortunes of oil companies one way or the other; maybe you make a bit more money here or less money there, but ultimately oil is just going to keep on making money until it doesn't. If I'm some rich oil guy then it's not clear to me that spending money on politicians is going to make me personally richer... maybe it will make the oil industry collectively richer but now we have a collective action problem, unlike the bitcoin case where a single rich bitcoin guy can expect to make money by buying politicians just on his own.
That's a good point, I hadn't thought about the ways in which influencing crypto regulation specifically has a much higher ROI than lobbying for other industries.
I'm still slightly surprised that no eccentric billionaire has used this strategy to indulge a personal hobby-horse without expectation of turning a profit, but I suppose this would require a policy issue lying in the rare overlap between 'the average politician has no strong opinion on this' and 'someone is willing to spend hundreds of millions of dollars on this'.
> no eccentric billionaire has used this strategy to indulge a personal hobby-horse
At least one HAS. His "hobby-horse" is "Open Societies," and has been quite influential.
I'm reminded of Phil Sokolof who spent his own money to successfully campaign McDonald's into changing from tallow to vegetable oils for frying, in a misguided attempt to save people from saturated fats. Instead leading to the proliferation of trans fats, likely killing quite a lot of people. Also ruining the taste of fries forever.
On the other hand, at least making it so that I can eat McDonald’s fries.
What is an example of a plausible policy approache the USG could take which would reduce Bitcoin's value by multiple orders of magnitude? I inquire because crypto has thus far seemed robust to government bans, e.g. in China, and am curious as to what other vulnerabilities are known/feared.
I think a US Government ban would be a lot more effective than a Chinese Government ban. (A Chinese Government ban was pretty much priced in from the start; of course the Chinese are going to ban it, they ban everything)
That said, I'm only wildly guessing. I don't claim to be able to accurately predict the effect of random events on the bitcoin price; if I were then I'd be incredibly rich.
Grey market BTC activity continues in China, four years post-ban:
https://www.forbes.com/sites/digital-assets/2025/03/31/chinese-people-continue-buying-more-bitcoin-despite-strict-bans/
Need an action be priced in if it is not terribly effective? The Chinese New Year holiday price movements are also fairly well-established.
That's easy - ban the exchange of cryptocurrency for US dollars, and create private right of action penalties against any financial institution with US assets or branches that provides it in violation of the law.
Cryptocurrency would still be available overseas, but it would be like prediction markets where being cut off from the huge, rich market full of would-be speculators in the US heavily constrains it. Both of them desperately needed the money of US investors, speculators, and suckers to really get going.
This is an excellent point. The corresponding point to this is that unlike oil (where you have a few oilmen and many oil buyers) the vast majority of people (not just politicians) have no opinion whatsoever on the topic. It's the perfect lobbying opportunity.
And there was special legislation introduced to break up the monopoly of oil companies like Standard Oil. John D. Rockefeller probably could have bought and sold the politicians of his time, and the corporation did make large donations to particular political campaigns:
https://en.wikipedia.org/wiki/John_D._Rockefeller#Monopoly
"Although it always had hundreds of competitors, Standard Oil gradually gained dominance of oil refining and sales as market share in the United States through horizontal integration, ending up with about 90% of the US market."
https://en.wikipedia.org/wiki/William_McKinley_1896_presidential_campaign#Fundraising_and_organization
"Large sums had to be spent quickly, and Hanna energetically built a businesslike campaign. ... But as the campaign began operations, and began them on a huge scale, money was short. Hanna initially spent much of his time in New York, where many financiers were based. He faced resistance at first, both because he was not yet widely known on the national scene, and because some moneymen, although appalled at the Democratic position on the currency issue, felt Bryan was so extreme that McKinley was sure to win. ...Reports of Bryan support in the crucial Midwest, and intervention by Hanna's old schoolmate, John D. Rockefeller (his Standard Oil gave $250,000), made executives more willing to listen. After a gloomy August for the campaign's fundraising, in September, corporate moguls "opened their purse strings to Hanna". J. P. Morgan gave $250,000. Dawes recorded an official figure for fundraising of $3,570,397.13, twice what the Republicans had raised in 1892, and as much as ten times what Bryan may have had to spend. Dawes' figure did not include fundraising by state and local committees, nor in-kind donations such as railroad fare discounts, which were heavily subsidized for Republican political travelers, including the delegations going to see McKinley. Estimates of what Republicans may have raised in total have ranged as high as $16.5 million."
It's more that until recently, there were de facto limits on how much TV time/etc you could buy with money (only three networks, finite ad space, etc), and limitations on campaign finance used to have more teeth.
I'm also not totally sold on its effectiveness. The crypto money spam worked because Democratic politicians don't actually have that strong of opinions on cryptocurrency regulation, so a ton of money floating around was enough to sway them in a more positive direction (and of course the Trump people and associated Republicans are all too happy to have a new way to accept payments).
>It does, indeed, seem strange that no-one has tried this before.
The detached object-level analysis in the post shouldn't make anyone forget that "this" is "buying politicians", which is (was?) widely frowned upon and considered immoral and bad.
If an industry is outright buying politicians, people or the media would notice it and criticize it, and the damage to the industry's reputation might outweigh the expected gains. This is the Occam's razor explanation of why the almonds industry wasn't buying politicians, I think.
AIPAC is an obvious exception because until very recently it was de facto impossible to criticize it. Even in this comment section you can still see how any criticism is immediately accused of being just another "Jews control the world" conspiracy theory. Even mild criticism by someone who is literally world-famous for his honesty and objectivity and who's Jewish himself.
I'm not sure about why this is changing now, but crypto might have been an exception because regular billionaires still somewhat depend on the good will of their business partners and employees for their status, while for a crypto billionaire it's probably easier to just ignore public opinion.
The big hole in this theory is that before reading this post, I personally hadn't heard of the crypto influence in the 2024 elections at all. But for now, I'm chalking this up to me being uninformed and not even American.
>The detached object-level analysis in the post shouldn't make anyone forget that "this" is "buying politicians", which is (was?) widely frowned upon and considered immoral and bad.
I'm frankly surprised that the "dive bombing" tactics Scott describes in order to bully politicians into line is not very illegal, even in the US.
They probably did try it before, which is why we have laws around campaign finance etc. Think of things like the Teapot Dome scandal, or the various scandals that plagued Ulysses S. Grant's administration. We call such attempts bribes, because they were crude essays in the art.
https://en.wikipedia.org/wiki/Teapot_Dome_scandal
I get that you're Jewish, but uh, maybe it isn't a good idea to use an antisemitic dogwhistle in a post, even if it's a joke? Especially when said post has an activism component? Just saying.
Edit: Nevermind, apparently mentioning AIPAC at all was way worse. Well, your loss!
I was shocked when I saw that. Is this intentional outrage bait? It almost seems specifically designed to be screenshotted and posted everywhere.
If it was just an attempt at humor, it was in very bad taste.
Is it really still an antisemitic dogwhistle? It used to be one, but (1) everyone knows what it means now and (2) the actual antisemites aren't using it anymore.
Well yeah, but the fact that everyone knows about it now makes it worse for optics...
To me it read like jokingly using a period appropriate slur when writing an essay about a historical period -- the thing's been drained of all its dogwhistling power and relevance, so it can be used as a joke.
It would still be a bad idea to call black people negros in a piece complaining discussing the consequences of the Civil Rights Act. If you don't care about optics, that's fine, but Scott does have a reason to care in this case, so...
I think we're just getting different emotional voltages from seeing it -- maybe it's stupid of me to generalize, you're right that this is important. But also, like you mentioned -- Scott is a Jew. Can't he reclaim it like the black people did with Negro? Weren't there Civil Rights Leaders addressing each other like "my fellow negros"?
He doesn't make it his entire identity. But again, I don't care whether this is "right" or not. This is about optics. And given that (((Middle Eastern democracy supporters))) have greater-than-average disposable income, it makes little sense to alienate them in a post like this. Especially when you might tarnish the reputation of others by association.
I think that was reclaimed ten years ago. Half the Jewish commentators on Twitter used to use it, although I think it's dropped out of popularity lately.
Anecdote: I thought it was hilarious.
Well, only half my family is Jewish, but I found it rather funny. Similar to the classic "Hitler's lottery numbers" joke. Mensch points awarded.
I did open my eyes wide a little, but hey, if you can't make a joke about your fellows, who can?
I thought the general principle is you can get away with using an ethnic slur if you’re a member of the group it applies to, so African American rappers can use the N word and Jewish bloggers can put brackets round their name.
I looked through and I can’t find the antisemitic dogwhistle. Am I missing something?
> AIPAC's natural constituency, (((Middle Eastern democracy supporters))),
As others have pointed out, it's not really a dogwhistle anymore, everyone knows what it means and it's mostly used as a joke.
Triple-parentheses around a name, e.g. (((Name))), is a convention among some antisemites for indicating the names of Jews. As Scott himself says in a different reply to the same parent comment, he's clearly using it in a joking/reclaimed way.
Surely the AARP is even more influential than AIPAC?
In many ways, the crypto community today are analogous to Arab tribesman in the decades before muhummad, slowly growing in power and influence on the imperial periphery.
The religious angle of that analogy does make a lot of sense. Finally a religion that can have a truly immutable Word of God! Until someone finds the right exploit in the smart contracts, that is.
Marc Andreessen is as close as I think modern society has to the “man of lawlessness” or false prophet which is Biblically asserted to prepare society for the coming of the Antichrist.
Peter Thiel doesn't think so (or does he?):
https://www.theguardian.com/us-news/2025/oct/10/peter-thiel-lectures-antichrist
"Specifically, he suggests the antichrist would be a “luddite who wants to stop all science”, referencing Thunberg, Eliezer Yudkowsky, and Marc Andreessen.
'My thesis is that in the 17th, 18th century, the antichrist would have been a Dr Strangelove, a scientist who did all this sort of evil crazy science. In the 21st century, the antichrist is a luddite who wants to stop all science. It’s someone like Greta or Eliezer.
It’s not Andreessen, by the way. I think Andreessen is not the antichrist. Because you know, the antichrist is popular. I’m trying to say some good things about Andreessen here, come on.'"
I appreciate this because I learned something. How to buy an election.
No, we learnt how to buy an election before 2024.
Typos:
"will truly accept" should be "will never truly accept"
"miniscule" should be "minuscule"
"crossed skirted" should be just one of those words,
@gmail.com? Really? Are DeepMind anti-AI now?
The crypto industry is merely a massive grifting enterprise. The only advantage crypto has it can be used for regulatory arbitrage; i.e. a bank has to verify your identity and take measures to ensure you aren't using your account for crimes. The miners and validators that support the blockchain have to do none of that.
All the Biden administration tried to do was prevent people from using cryptocurrency to break laws; the problem is they did not go far enough. We allow miners and validators to approve transactions between unknown counterparties; bankers would be prosecuted for doing that.
This is false in several ways.
Crypto has many other advantages beyond regulatory arbitrage - for example, it makes it harder for the government to debank people, and substitutes for financial infrastructure in countries that don't have it. See https://www.astralcodexten.com/p/why-im-less-than-infinitely-hostile .
(EDITED TO ADD: Trying to donate to Bores, somehow two different credit cards and Paypal have both failed, both over ActBlue and over the phone, in ways that the credit card company cannot explain to me. Sure wish I could use crypto for this.)
The Biden administration's attack on crypto went beyond normal KYC; see some of the discussion in https://www.axios.com/2025/02/06/debanking-crypto-operation-choke-point-house-hearing
Isn't "it makes it harder for the government to debank people" just "regulatory arbitrage" but with positive affect?
I don't think so. I think of regulatory arbitrage as requiring a careful permit to register a bank, but you can register a crypto bank without a permit so it's cheaper but less safe. I think of debanking as something like the government freezing the accounts of protesters during a protest.
Depends somewhat on how you draw the lines but both of these seem like "crypto can let you do things the government might otherwise stop you from doing" which can be good or bad depending how much you agree with the government.
These are functions banks could provide if we loosened restrictions. US banks could provide financial infrastructure anywhere crypto could if we removed regulations that make it costly to provide accounts to people who live in other countries .
Conversely the government can easily expand restrictions on crypto providers to make it almost impossible in the US to do what Scott desires.
Where did the Biden administration exceed reasonable bounds? Non KYC prosecutions involved securities laws. Perhaps in some cases the scams and lies fell outside of securities laws
> is merely a massive grifting enterprise
That's unfair; it's not all grift. There's also collecting transaction fees from people who want to move money in ways that would otherwise result in law enforcement taking an interest, like drug deals, funding sanctioned foreign regimes and so on.
typo:
> Any politician who’s survived long enough to matter is a little bit paranoid and will [never?] truly accept that their safe seat is safe.
typo: "Any politician who’s survived long enough to matter is a little bit paranoid and will truly accept that their safe seat is safe" -> "...never truly accept..."
> Just as Substack bloggers may reload their browser again and again watching the likes and restacks come in, so politicians will reload their campaign metrics panel watching the flow of donations. [...] They don't even necessarily have some future campaign they're saving it for. They're just addicted to fundraising.
I do think in some sense this is one of the root issues of politics, in much the same way that people claw each other's eyes out to win a couple of elections even if they never actually do anything with that power. People are great at maximizing quantifiable things (funds, Congressional seats) and will spend their whole lives doing it without actually accomplishing any underlying goal, like a stereotypical rich workaholic who never takes a vacation.
Are you at all worried that your "Too Much Dark Money In Almonds" post might've directly or indirectly affected the political strategy of billionaires like Andreesen? Maybe there was a light taboo around billionaires buying elections which your post helped erase?
Also why isn't Eliezer Yudkowsky tweeting in support of a donation to Bores? Does he think it's another futile effort which isn't worth bothering with? I'm long-term unemployed with a chronic illness, so $7K is a lot of money for me.
It is pretty likely Thiel, Musk, Vance, Anderseen and Horowitz read the article given that they were/are fans of Scott.
Maybe a great example of the virtue of silence.
Then why has the almond industry not responded? Seems like there must be political muscle there that they are waiting to flex, with all those billions from nut lovers.
To be honest, with the title of this post, I half-expected it to be almond growers versus data centres competing for water resources in California. Perhaps that is how we will restrict the growth of AI - pump it (literally) into the growth of tree nuts instead!
No, I cannot imagine that the market in knowledge is so inefficient that it would take a blogger saying "billionaires could spend more money on politics" to make billionaires realize they could spend more money on politics, if this were actually a very good strategy.
Bores is a moderate whose proposals are extremely far away from anything Eliezer thinks might be helpful. I also think it's reasonable for some people to have a norm against tweeting begs for money (I was nervous about blogging a demand for money, which was why I tried to compensate people with a useful post about the political situation).
I don't think if you are financially unstable you should spend a difficult-for-you amount of money on AI policy.
If the market in knowledge is so efficient, why even bother writing that post? The information was already known.
Are you at all familiar with the psychological literature around anchoring and framing effects? You reframed political donations from the common view ("way large") to a new view ("way small").
Also you didn't address my point about norms.
Thanks for answering my q about donation.
The space between the things that the people setting policy know (in this case, politicians, lobbyists, billionaires, etc), and that everyone else knows, is the entire reason that journalism is possible.
It's not even like I invented this argument! I was commenting on a paper saying the same thing! The idea that the market is so inefficient that commenting on a decades-old debate about why it looks like a $20 bill is left on the ground would lead someone to think "oh, cool, there's a way to take over the entire US government at minimal cost to me, I never was interested in this before, but I guess now this blog post has implicitly given me permission to do it" is just not how any of this works, and worrying about whether maybe it is is a good way to go from truth-seeking to being so paranoid that you never write about anything at all.
"people setting policy"
Seems like a bit of a false dichotomy. How many billionaires view themselves as "people setting policy"? Seems like a rather fluid category.
"It's not even like I invented this argument! I was commenting on a paper saying the same thing!"
Most academic papers are only read by a few academics, correct? I would guess you are responsible for far more exposures to this argument than anyone else. It's not about inventing arguments, it's about spreading them.
"is just not how any of this works"
Blog posts and memes have little influence on US policy and governance you say? This take is at least a decade out of date. Andreesen appears to be a memelord par excellence. And he is an ACX subscriber, if the Substack UI is any indication.
"go from truth-seeking to being so paranoid that you never write about anything at all"
Again, false dichotomy. Consider the least convenient possible world where you did, in fact, inspire Andreesen. What would the correct update be in that world? Maybe it might look something like: Do a trial run of an "80/20" red-teaming process for your posts, which aims to delay publication by at most 10 minutes, and filter out the posts which have the largest negative expected value. Do it for a month or two, then spend a few hours on a retrospective regarding whether it is worth continuing, or should even be expanded.
Thousands of people subscribe to your blog. Maybe you should take that a little more seriously?
"I would guess you are responsible for far more exposures to this argument than anyone else."
Scott looms large in his social niche, but I'm a non-academic (and recent ACX subscriber) who'd heard of that paper's findings before I read Scott's old blog post today. Let's not Basilisk ourselves into treating any discussion of bad political outcomes as an infohazard.
I do think it’s fair to say that Scott’s old blog was, in its heyday, easily the most influential widely-read source of “media” within Silicon Valley. It does not seem outside the scope of probability to assert what Ebenezer has here.
I'm confused how the push to donate to Bores meshes with what you said about the strategies that work, which wasn't "donate to the long-shot candidate who is already on your side" but "donate to the people that will win, so that you have influence on their later decisions"
I think that given that Andreessen is trying to create common knowledge that he will fight people who support AI regulation, it is especially useful to try to create common knowledge that we will support people who supports AI regulation. This is just my guess though, I am trusting the experienced political operatives.
> But also, AIPAC fights hard. If some random Congressman is anti-Israel, AIPAC will swoop down on their race in Middle Of Nowhere, Missouri and pour $10 million into electing their opponent. By now everyone knows this, and the mere threat of AIPAC action is enough to keep politicians in line.
If so, how do we account for several prominent politicians, including several Congresspeople and the current front-runner for the NYC mayor's race, being out-and-proud antisemites and vocal Hamas supporters? It would sure be nice if AIPAC had the power to reduce these human filth to political irrelevance, but that doesn't appear to be the case.
I think it's pretty obvious why AIPAC is not able to single-handedly defeat Zohran Mamdani, but why they can make the difference in some kind of close Congressional race.
> the current front-runner for the NYC mayor's race, being out-and-proud antisemites
Andrew "those people and their f**king tree houses" Cuomo is actually losing in the latest polls.
Probably AIPAC doesn't care about the NYC mayoral race because mayors have no ability to affect aid to Israel.
Also, it seems like an extraordinary claim to say that several Congresspeople are “out-and-proud antisemites.” I can think of only one individual at most who might sort of qualify for that claim.
This is still less than almonds, but Michael Bloomberg put in $1 billion of his own wealth into his campaign. Tom Steyer, $342 million.
However, total political spending in 2024 was $20 billion, which is more than almonds. People believe that money has too much influence on politics, and I'm inclined to believe them, although I can't explian why.
(data sources: ChatGPT)
I think the difference between Bloomberg and Andreessen is that Bloomberg was only able to spend his money on one thing (electing Bloomberg), which people didn't really want, but Andreessen can focus it on the most winnable races between people who are already both popular.
Bloomberg gives a lot of money to other candidates as well. He was the biggest donor to Democratic candidates in 2024 with $61 million according to OpenSecrets.org.
Elon musk spent 100m dollars to unseat a judge in a Wisconsin special election and lost.
I feel like there's only so many tv ads you can buy before it gets oversaturated and spending more money just doesn't convince people better.
That saturation point might be higher than it initially appears. Elon Musk was 10 points underwater at the time and very publicly associated w the effort. That certainly did some damage.
Then how is AIPAC so effective? This just doesn't add up. I hear people complain about money in politics all the time but then I hear stories like this all the time too. I also remember the Freakonomics guys studying it systematically and concluding that money doesn't really impact election outcomes. I genuinely don't know what to think about it.
I'm inclined to think that it generally doesn't matter BUT if you happen to bet big on the winner by chance then you'll have an outsized amount of influence over him. Maybe the crypto people just happened to get lucky last year.
AIPAC is effective in many races because there is/was fairly broad support in the US for Israel (maybe not so much anymore among younger folks), and among both parties. It's an easy crowd to work, and you can see that when they actually try to unseat someone more challenging - like the failures to unseat AOC in her district.
Isn't AIPAC effective because US politicians fear it for more personal reasons than for electoral ones? If you're positive towards Israel, but lately have been having doubts, perhaps you refrain from making certain comments because you don't want to be *seen* as being anti-Israel. It's not that there was a period where a huge load of politicians switched from being pro-Israel to anti-Israel, and then AIPAC bought a load of campaign ads and these candidates all lost. It's that AIPAC's looming prescence keeps the least pro Israel of pro-Israel candidates in line, so that doesn't happen in the first place.
It prevents wavering, and keeps an existing status quo of pro-Israel politicians acting in unison. Its influence over anti-Israel/Zionist politicians and voters should be minimal, so it remains vulnerable to sudden generational swings. AIPAC is a creature of a particular era in politics.
This also explains the shift in strategy as well. AIPAC announced their Super PAC in late 2021, right around when public support for Israel was tanking amongst the left. Of course, that was still nothing compared to what would happen afterwards, so I guess they chose a good time to switch gears.
A few thoughts:
Katie Porter is a weak candidate and probably would lose anyway (I expect her to lose the Governor's race regardless of crypto spending).
You can collect a few scalps as a PAC but if you screw up you negatively polarize the issue and can undermine long term progress (probably less of a concern for short-termists in tech in both AI/crypto).
Preventing action is easier than doing things in the US system.
I think you may have swapped hard and soft money in the section that starts "Safe-seat Congressmen want more hard money." Per my friend that works in political fundraising, "soft money fundraising is also just an expectation of the job [...] there's only so many ways chuck schumer can spend a check that gets written directly to him, and he doesn't really need that much money, so instead he gets someone to write a bunch of checks to a bunch of different groups all doing similar work." --which is the soft money, generally for a broad group of candidates not one specific guy.
> The tiny scale of US political spending is dangerous insofar as it means that one or two billionaires willing to go all-in can distort the national landscape. But it also makes it possible to oppose them.
I’m not sure that it’s true. The billionaires discussed in the post, it seems, aren’t spending for ideological reasons, but because they expect a clear ROI on this. $260m spent on crypto lobbying for a +100% return on BTC means it was just « good business ». The same is not true for people invested in the other side of the debate, but who can’t « cash in » on a win. I feel like the only way for non-billionaires to counter that influence is to fund the PAC to end all PACs that will finally lead to getting money out of politics.
I’ve thought for a while that the first and only “AI safety” law will be one to shield Sam Altman and co from liability. This will be followed by a gigantic tax cut to “incentivize innovation.”
I have no idea what you are talking about.
The post is about tech PACs and I am predicting what they will lobby for.
> The only reason it's remarkable is that for some reason which I still haven't figured out, nobody else - not the oil industry, not the firearms industry, not the defense industry - ever tried this before.
If you haven’t read it already, i recommend Robert Caro’s LBJ series. The oil industry DID try this, to great success. LBJ got obscene amounts of hard money for his Senate race (in those days, envelopes of cash) from men like Herman and George Brown, who said something to the effect of “if i’m going to back a politician, i go all the way”.
This paid off massively. Leland Olds, phenomenally respected head of the Federal Power Commission and public power advocate, was ousted over flaky communist allegations, and the FPC’s natural gas profit caps were quietly thrown out.
The reason oil doesn’t do this anymore is because:
- oil is not regulation constrained
- oil is not a gold rush anymore.
“Go all the way” political horse-backing only makes sense in a gold rush. A different question is: why didn’t tech do this earlier?
Was tech particularly regulated? Even if you're in a gold rush, it doesn't make sense to waste money on bribes if there isn't much room for improvement. AI and crypto are different, seeing as there's a sizable amount of public backlash to both, making preventative measures reasonable.
Can you explain why it doesn't make sense outside a gold rush? Surely oil wants stuff like climate regulations and ANWR to go its direction.
If you own a mature company in a mature field, it almost always makes sense diversify into new fields with higher risk/return anyway. AFAIK most older companies and/or individuals in energy industries have by now substantially diversified into renewables already. You can check the list of biggest oil companies; the owners are by now mainly extremely diversified institutional actors like The Vanguard Group and BlackRock.
Once you've done this, backing a politician to safeguard the interests of the mature field is very mediocre value: It's unreliable, it makes you unpopular, and if you're smart you're already playing both sides, so why bother?
No, you do it the other way around; You back politicians who want to do investments into new and/or popular tech and scrounge up as much of that as possible. The Vanguard Group and BlackRock are primary leaders in ESG investments, for example. This way, you get everything: You have some safe income stream from a well-understood, reliable industry, you get some play money for free to invest into a new risky tech that will turn into real money if it works out (and if not, no worries, it wasn't yours anyway!) and you're popular (well, more popular at least than if you just stuck to the mature field).
The last time ANWR was opened, under Trump’s first administration, they held a leading claims auction that was only subscribed to by one single small Alaskan oil company. The larger oil companies gave it a wide berth.
This is principally because large oil companies saw it as more trouble than it was worth, mostly because they would inevitably face years of maximally motivated environmental groups trying to save their white whale through litigation. This process would almost certainly raise the silence of ANWR in the public eye, discrediting these oil companies to the public (as protecting ANWR is relatively supported albeit low-salience); more importantly, any litigation is almost sure to drag on beyond 4-8 years, after which you have a Democratic administration which is very likely to cancel your lease.
That is to say, large oil companies would probably only drill in ANWR if the political climate had changed to the point where they could be reasonably sure that they could sign a large lease while avoiding much of the litigation or the threat of cancellation. That would require a shift in the Overton window, which may have indeed occurred in recent years.
Reading about Rockefeller and how he got so rich, the government broke up Standard Oil because it was too powerful and ended up (unintentionally) making him even richer, since he held shares in the new companies that were founded out of breaking up Standard Oil, and they all shot to the moon.
So sometimes what is intended to happen results in the exact opposite. And oil companies have a lot of public mistrust due to environmental disasters, so - like the tobacco industry - they need to have the *appearance* of being squeaky-clean and not up to any shenanigans. "Why yes, I *am* in the pocket of the polluting industry that is killing the planet" is not a reputation any politician wants nowadays.
This shift into politics seems quite directly counter to the advice from the original blog: "Guided By The Beauty Of Our Weapons", "Beware Systemic Change", general sentiment of trying to play cooperative games. And fair, maybe indeed "this time is different", but the post doesn't really go deep into this. It doesn't explain the tradeoffs. It doesn't actually argue back and forth, it doesn't try to steelman "why is this a bad idea".
For example, isn't it better to throw this money at finding common ground with pro-AI people politically? What are their main arguments and lines of attack? How does this compare to the usual advice of funding EA causes? What is likely to happen if the fight escalates? Who is likely to have more resources? Etc, etc. Just a donation link and a call to action.
One might naively hope that the same strategy wouldn't be as effective for AI as it was for crypto and Israel, since there aren't as many rich AI lovers as there are rich crypto lovers and rich Israel supporters.
In other words, maybe there are more crypto millionaires who love crypto than there are AI millionaires who love AI.
But this is very optimistic and probably false hope.
Maybe the AI datacenters coming for the almond industry water will get the nut bros aligned and active 🤔
Wow. This makes me wish the media was a bit more functional! I see there's been a bit of reporting on this, but not a huge amount, and it's a really big story. I'm glad Scott is aware, but this is the sort of thing that the media should be there to observe, and there just isn't very much.
There have actually been lots of good mainstream stories on this, for example https://www.nytimes.com/2025/08/26/technology/silicon-valley-ai-super-pacs.html
I would model the millionaire's decision to spend or not spend money on politics to be a function of how much of their profits they expect to save (i.e. by avoiding a tax increase, or by getting a promised tax cut).
MONEY SPENT = (TAX RAISE - CURRENT TAX LEVEL) * PROBABILITY OF ACTUALLY IMPLEMENTING A TAX RAISE.
With this model, then it makes sense that they would actually not spend "too much" money if they were to belief that the actual likelihood of them getting a tax raise is low.
In the case of the oíl, finance, weapons industry... there seems to be few realistic chances to raise taxes on them, let alone to destroy those industries. So they spend little.
Another problem is a classical free loader problem.
In the case of crypto (or AI), it was very possible for those industries to be regulated to death (which should be treated in the model as a tax raise that takes away all their profit), so it makes sense that they are "donating" more money than ever
While EAs may have overlearned the lessons of the 2010s, this effort suggests they (we?) have underlearned the lessons of the 2020s. What are those lessons?
- Plunging tons of money into political campaigns does not, in fact, guarantee success, the somewhat limited evidence presented in this post notwithstanding. In fact, plunging tons of money into political campaigns can actually generate enormous amounts of unwanted attention.
- This unwanted attention can fundamentally alter the landscape for workaday EA organizations actively trying to influence policy in the nation's capital to such an extent that some such organizations actually withdraw from previously successful attempts at policy influence since EA relationships become seen as toxic.
- The story "a bunch of weird nerds are obsessed with AI and are spending tons of money to regulate it" is a much more tasty news peg than "rich guys are spending money to influence politics." In every context, the former will generate more news stories and indeed more negative news stories than the latter.
- Lobbying AGAINST something (more regulation of crypto!) is many multiples more cost-effective than lobbying FOR something (more regulation of AI!). This is a stylized fact rather conclusively well-supported in the empirical political science literature — and also one we can see throughout the 2020s.
- THE BETTER CAPITALIZED PARTY DOES NOT WIN EVERY POLITICAL BATTLE. Boy, this is a big one. In crypto and Katie Porter we have an existence proof of a certain kind of political activity.
This is really where I want to hammer on it. Is it the opinion of the AI Safety community that expending $1b on lobbying is the maximally cost-effective way to influence AI regulation? One comes away with the opinion that very little research has been done on the policymaking process.
I don't think the for/against distinction is clearly meaningful. Suppose we said we were only lobbying AGAINST federal pre-emption of AI regulation (a cause that we indeed don't like, and which is our main target)? Or suppose we said that the crypto people were lobbying FOR the GENIUS act and the Strategic Bitcoin reserve?
Given how thin this line is, I am willing to trust the experienced political operatives who have chosen this strategy, rather than your theory that although sure, spending lots of money worked amazing for our opponents, it will mysteriously backfire for us, and therefore we should let our opponents outspend us 100-1 and win easily.
It is, in fact, a distinction in the poli sci literature — as is the inconsistency of value in outspending. It is by no means an established fact that big spenders win elections or get their policy priorities implemented as a matter of course. As for the former issue, the distinction here is between "maintaining the status quo" and "making change":
See e.g.:
https://www.jstor.org/stable/41759323
https://www.cambridge.org/core/journals/american-political-science-review/article/abs/organized-interests-and-agenda-setting-in-the-us-supreme-court/7E9792FB9880F9F333DD3F74C7E7DA36
https://direct.mit.edu/rest/article-abstract/100/2/303/58458/Information-and-Legislative-Bargaining-The?redirectedFrom=fulltext
Again, the issue here is what's valuable on the margin — trying to compete in this arena is *enormously expensive* and this doesn't just need to be *a good idea* — it needs to be *the best idea on the margin.*
"Plunging tons of money into political campaigns does not, in fact, guarantee success, the somewhat limited evidence presented in this post notwithstanding. In fact, plunging tons of money into political campaigns can actually generate enormous amounts of unwanted attention."
Yeah, I definitely think the Carrick Flynn election result burned them on pushing for any more directly political campaigns (I'm sorry, I mean no disrespect, but this story always makes me laugh since it was so clearly the triumph of idealism over going with the reality on the ground). That, plus SBF being the one shoving money into such backing and similar donations and the subsequent black eye that gave the movement.
Wow, you've really crossed over into "Jews pulling the strings to control society" conspiracy land. Honestly, this post should be the last nail in the coffin of the rationalist movement. Rationalism has a lot of good criticisms of other ideologies, but as a positive theory in its own right, it's time to let go.
Every immoral ideology merges with antisemitism eventually; and for a philosophy that sees itself explicitly as the antidote to conspiracy thinking, seeing you transition so smoothly into embracing the oldest conspiracy theory is a clarifying moment. What was it in the end? Did you just fall for the avalanche of defamation and blood libels? Or was it reactionary hatred of Trump?
Are you aware that Scott is a member of the "pullers of the strings to control society"? It's a lighthearted self-effacing joke which made no conspiratorial insinuations and just stated obvious public information, relax. If anything it's poking fun at the conspiratorial-minded.
(((Scott Alexander))) is pulling on the strings to control society by writing about what he thinks about issues of importance on his popular blog Astral Codex Ten. A lot of smart and influential people read the blog and might be influenced by the ideas they see published there and also by the ideas of other people publishing blog posts engaging with (((Scott's))) posting. Also I heard that he publicly gave grant money to a bunch of people and organizations who he thought were doing good or interesting work in the world, which makes him kind of similar to (((George Soros))).
I suspect this is second- or third-hand influence, and that Scott learned about PACs from a source that is generally reliable except for the lean towards specific dangerous tropes (or a source that themselves learned from such a source), leading to the distortion being presented as fact. This is the kind of thing that will often happen to people who are not themselves at all antisemitic.
(An alternative hypothesis is that this presentation of AIPAC/Jewish donors is actually accurate, but when you're in a society that spontaneously generates "Jews control everything" conspiracies regardless of circumstances, your bar for sufficient evidence for that should be *very* high.)
I think it is unlikely that Scott is a self-hating Jew.
What do you say to "AIPAC is, in fact, Jews pulling the strings to control society"? If it's an organisation set up to defend Jewish interests and the method of action on that is influencing elections?
I don't think it's anti-Semitic to go "yes, this is a Jewish organisation promoting Jewish interests". If every criticism of anything Israel, Israel-supporting organisation, or Jewish actions gets met with "anti-Semitism! you're trying to send me to the gas chambers!" then eventually it will be crying wolf and people will take *more* notice of those who want to say "psst, lemme tell you about some shady stunts these guys are pulling".
Does anyone know where to donate to Leading the Future? Is there somewhere better to give for individuals opposed to AI regulation?
With crypto, I'd keep in mind that Trump won the popular vote by less than 2% against the unpopular last minute replacement for a historically unpopular incumbent. The Democratic base is now very against crypto and a lot of Silicon Valley stuff, and if Harris had won, they might be in a bad situation now. It's not clear that spending all that money actually is some unbeatable strategy.
My sense is that the Musk buying Twitter thing really did help but the situation is a bit sui generis.
Also re AI, they tried to add a "states can't regulate AI" provision into their law earlier this year and it generated backlash and got dropped.
Man I read the title and I was hoping they were going to do something to rationalize California's water rationing policies.
Now _that_ would cost a lot in political donations!
California water rights got grandfathered in in 1915. Increasingly, California's hereditary water baron families intermarry so they are getting fewer in number but richer.
Somebody brought up the idea of water property rights reform to Gavin Newsom. He agreed it would a good thing to do, but said that no way was _he_ going to lead a political campaign against the Water Powers That Be.
Maybe the reason oil companies dont buy politicians in the direct sense is that they find it more effective to buy issues via the economy. Oil supports lots of jobs etc, which is why there are no shortage of pro oil politicians.
This makes crypto well suited, maybe uniquely suited for buying politicians directly. Theres tons of money in it without supporting many jobs somehow (kinda the point of the anti crypto position, dont interpret that as me being pro or anti)
I hear AI is supporting the global economy these days. That might be more important than their PAC, but by running both, I guess we wont get clean data
Maybe we can get some good AI regulations but the odds are the regulations will just slow the industry down without doing much good, I think the potential harms of a future AI are to broad to be able to affectivley pin down in legislation
AIPAC was not founded as a PAC; it only became one in 2021. It has historically disbursed a smaller dollar amount than many industry lobby groups and quite a few actual foreign countries lobbying. Compare their 10s of millions to China dropping 1/2 a billion dollars, and the Saudis 1/3 of a billion. Until October 7th, iirc AIPAC was smaller than most ag sector lobbyists.
AIPAC gets its way far less often than other domestic lobbies, especially on the issues it has considered the most important. They completely failed to prevent the JCPOA, which American Jews felt was a bad deal. In fact, they failed so badly that they got replaced by J Street within the Democratic Party, and J Street is at best a fringe organization of American Jews.
Now, AIPAC says they’re successful, and of course they want to say so because who would donate to an ineffective lobby? And antisemites say that AIPAC is successful, because without that conspiracy they might have to reckon with the idea that most people don’t like to see Jews killed.
But are they successful? What specifically have they accomplished in the last twenty years? Thirty? Forty? AIPAC claims they got the development aid and military aid for Israel. Did they? Or were there clear strategic reasons that America chose to make those interventions - such as stabilizing Israel in the Yom Kippur War, and securing Israel enough that they’d go along with risky diplomatic plans such as peace with Egypt or the JCPOA - where America was just following its own interests as it saw them? Those three events account for, if memory serves, more than half of all aid to Israel period, including development aid. AIPAC is pushing on open doors, at eye-watering cost, and thinks that makes them kingmakers.
And if AIPAC *isn’t* the kingmaker they like to say they are, then repeating their narrative can play into the hands of the blood libel of Ilhan Omar and co.
AIPAC constantly boasts about how powerful AIPAC is. Is AIPAC blood-libeling Jews?
I wish you (and Americans more generally) would consider the implications of posts like these - describing moneyed interests who support often unpopular positions (Cryptocurrency, Zionism) outright sinking hostile candidates with attack ads on unrelated issues, while funding both sides of competitive races - and reconsider their reflexive defenses of the American political system, or their analysis of the country as a democracy.
I think it will go to their heads, they'll try spending $500 million in 2026, and the results will be really disappointing for them. Congressmen obviously don't like having tens of millions in attack ads dumped against them, but ultimately Porter was a weak candidate in a race against a far more popular, high-profile Democrat who ultimately won it (Schiff), and on an issue where most of the Democrats don't have strongly held opinions on the topic.
As I pointed out in the 2019 post, "that's what we spent on almonds, so it's not a lot of money" is bad reasoning because just because we think of almonds as insignificant doesn't mean that the amount is small.
I decided to look up https://data.ers.usda.gov/reports.aspx?ID=4057 which shows almonds at 5.8 billion, but that's "cash receipts", not the size of the industry or how much Americans spend. Still, there are a lot of other crops in the table and I should be able to get relative sizes without having to compare apples to oranges (except literally). Almonds beat out every single one of the 32 "vegetables and melons" (though potatoes are close) as well as all 31 other items in the "fruits and nuts" category except for grapes.
Almonds are about half of wheat and beat out sugar cane and sugar beets combined, cotton (though only for 2024), and tobacco. They're still nothing compared to animals or feed crops, and nowhere near the military or health care, but why Scott thinks almond spending is insignificant is beyond me. Something that spends as much as we spend on almonds is a big deal.
The table also seems to give single lines for things that are both consumed by humans and used as animal feed. Separate googling claims that about half of wheat is used for animal feed and 90% of soybeans. In 2019, when Scott made his original post, grapes were less than almonds, and almonds were more than now (in real dollars), which means that Scott managed to choose an "insignificant" example that was the highest of all food crops.
This discussion is such a narrow view of the issue. The real damage to American politics (which is now washing into international politics) was the Citizens United decision. The real question here is not, “how effective is political spending?”, but “Would American politics be better if it was not completely dominated by campaign finance?” Since I’ve seen estimates that Congresspeople spend 2-3 days per week just raising money, the answer is probably “yes”, just from a functional point of view. And since multiple studies have shown that those policy measures that are implemented are much more likely to be those that are aligned with the interests of the rich, the answer is probably “yes” from the perspective of a functioning democracy.
Was Eddington secretly a documentary?
Why is Marc Andreessen with his puny $2B even relevant? During the 2024 campaign, I have heard (not sure if this is 100% true) that Elon Musk has spent about $250M to support Donald Trump. That's not even mentioning him buying Twitter. There seem to be a lot of people who could spend way way more than Andreessen even has at all.
Also, I know that the author is very concerned about AI alignment and regulation, but aren't there some... more pressing issues about the American democracy, the ones which warrant our attention? Or is that too small in the face of AI?
The great part about this plan is that you can bribe politicians regardless of what kind of political system you're in!
Assuming someone or a lot of people do this and start outright buying politics generally considered left-wing leaning (to be concrete: ai safety, regulations in general, things that are bad for profit of business head figures) isn't this like opening the flood gates? The people on the opposite hhave far more money and so far didn't use it to buy politics, except in case of silicon valley they kinda did.
What happens if every billionaire starts this and no one has any illusions anymore if you can or can't buy democracy anymore, because obviously you can and you did?
Its probably an irrelevant fear because the arms race already kinda started but still if everyone goes full on plutocracy it seams doubtful to me if this is really a net positive. On the other hand we're already headed there so what is one more push and trying to change direction of the plutocracy with that push ¯\_(ツ)_/¯
I find it mind-boggling that the conclusion of this post isn't "and we should do X and Y to limit the influence of political donations in US politics, here's how you can contribute" and instead the conclusion is "and we should get in on the money-burning game, here's a guy on our tribe's side, give him your savings".
Why are Americans so primed to believe that massive political donations are an inevitable fact of life?
💯💯
I had the same reaction.
But the US differs from Europe/UK in degrees. Even in Europe election campaigns cost a lot of money. Someone has to fund that. That someone gets a lot of say.
Awesome, thanks a lot Scott
EA should make a PAC that threatens african leaders with Uncle Sam's Big Stick/Denial of Carrot if they don't use their Rolls-Royce and MiGs budgets to buy anti-malarial nets and dewormers to everybody who needs them.
It would be cheaper and more effective than EA paying for them directly and trying to distribute them in the middle of nowhere.
Wait a second - that Katie Porter story? So she *already* got slapped down hard in an election?
And her takeaway from that was *still* “How would I need them in order to win, ma’am?” for the governor of California election? 🤦♀️
I now wonder how the hell she ever got elected to the House of Representatives in the first place! (I can well believe the mashed potato story and the rest of it, but how did someone this clueless ever succeed in an election at all?)
Do extra brackets signify irony?
You need to include in your model that /the battle does not stay won/. At best, it stays won for four years. So the right way to think about the cost of getting a policy you want isn't a one-off payment of $X million; it's an ongoing expense of $X/4 million per year - and this not for a certain purchase, but to shift probabilities - increasing over time with at least inflation but also how annoyed your political opponents get with not having their way as time goes on, for as long as you need the policy in place.
Certainly the billionaires will be factoring this into their ROI calculations.
Couple of ideas:
1-Other industries know how far their money can go in politics vs how far it can go doing something else. Example: Oil can price a congressman vs down-payment on a new vessel.
Crypto doesn't actually produce anything, crypto billionaires can buy more coin and/or make the coin they have more valuable.
Maybe they did the math and realized they're at a point where buying more is not going as far as making it a bit more valuable/preventing it from crashing. Maybe right now politics is their blue ocean.
Maybe it's the same for AI, they've reached a point where their money in the actual business only goes so far, because they keep burning money. Maybe dumping this money into politics means AI and crypto are showing some concern with a bubble.
2-Other businesses have long bought their way into politics and the economy, AI and crypto are the nouveau riche and just now are buying their way into politics. Maybe what they're spending now is not so different from what other businesses have spent in the past, or total spent over the years, adjusted for inflation.
I'd like to separate the issue of AI safety from whatever opponents of AI mean when they say "regulation" (all while making light of the safety issue). At least in online discourse and elsewhere, "AI regulation" seems to mean something between taxing, neutering, or banning of various AI tools, with some "think of the children" thrown in for good measure.
I'm sure the PACs will be inclined to fight both forms, but today's AI majors seem likely to acccept a form of safety-as-regulatory-capture, and I suspect they're less likely to go after a vaguely worded "test before you ASI" bill than a "stop AI from stealing our water" bill.
I need to admit that my honest reaction to this post was "come on, Scott, don't feed Moloch." I realise that figuring out a way to stop the bad thing from happening, rather than co-opting the bad thing for your own purposes, is really difficult to even figure out intellectually and even harder to orchestrate (all of which needs more time than the co-opting, besides), so I'm not upset or even want to particularly stop you. Instead, I sincerely wish you well (even though I don't have full agreement with even the cause). But I do hope you reconsider the whole approach.
Heck, even if you don't actually do another approach, I'd be super interested in knowing what you can come up with to solve the underlying issue. You've already thought about it a lot, so I think you could come up with some pretty interesting approaches. So there's at least another article that could come out of this? :) I'm hoping.
I wish they *had* called it AI PAC. Although you'd probably get called racist for caring so much about whitespace.
Leading The Future still only has 2% as much money as the almond industry.
No, not exactly. The almond industry does not 'have' $12B, it has revenue totaling that. This figure comes from the sales of Almonds and products mostly in American supermarkets. a lot of that number is Supermarket profit, and groceries are a notoriously low-margin business. Shippers, packers, processors and truckers get most of the rest. In the produce industry you are lucky if 10-20% make it back to the grower. Thanks for mentioning produce, though!
Voters in the UK should be worried too. Imagine a trillionaire who owns a social media company. He is best friends with the leader of one of the parties, and wants his party to be elected.
Campaign finance laws are really strict in the UK (£11,000 per donor, per year; total donations in 2021 were £50m). What's to stop this trillionaire from buying the election? Musk doesn't even need to give any money to Farage. He can just buy all the ads on YouTube and Facebook — and he already owns X.
Americans have never been interested in British elections, but for some reason, we now have Evangelicals campaigning about abortion laws, plus Presidents and Trillionaires speaking up for their preferred candidates. What happens next?