I am an advocate of democracy through sortition. I am also employed as a structural dynamic finite element analyst.
John Huang
Karma: 117
We’re not even capable of aligning of governments and corporations to humanity. How aligned is the US federal government? How aligned is the EU? how aligned is China?
We’re not capable of aligning the most powerful entities.
Moreover EA seems disinterested to be in aligning any of these powerful entities to humanity. EA funds little to nothing in for example, improving democratic decision making, which IMO the only viable alignment strategy. The obvious first step in alignment with “humanity” is to bother to even find out what humanity wants. That demands collective preference evaluation. And there already are already existing techniques to do so, little which interests either EA or AI advocates.
IMO if you were serious about alignment with humanity, you would be spending exorbitant amounts of alignment research on the lower hanging fruit, nation states and corporations, which presumably are less powerful than super intelligence. But you can’t even align the mere human, good luck with the superhuman. AI alignment will be impossible as these groups align AI with their own interests.
But please prove me wrong. Please show me a stronger commitment to democracy, to ensure that any entity can be aligned to “humanity”.
Let’s imagine you solve the “alignment problem” tomorrow. So? Exactly who did you solve alignment for? AI aligned to the interests of Elon Musk, Donald Trump, or Vladimir Putin? Or AI aligned with Peter Singer? Or AI aligned to the interests of Google, Meta, TikTok, or Netflix? Or is it alignment with the Democratically determined interests and moral values of the public?
We’ve never even solved the “alignment problem” with humans either. The interests of Google might be opposed to your interests. The interests of Vladimir Putin might be opposed to your interests.
But of course, seeing who is funding AI alignment research, I’ll readily bet that the goal is for AI to be aligned with the interests of tech companies and tech billionaires. That’s the goal after all. Make AI safe enough so that AI can be profitable.
>inherently uncontrollable, and thus not a tool.
If AI is an uncontrollable God, then alignment is impossible. Alignment to me implies some sort of control. Uncontrollable superintelligent AI sounds like a horrific idea. There’s no guarantees or incentives for God to solve any of our problems. God will work in mysterious ways. God might be cruel and merciless.
So even if we do succeed in preventing the creation of God, then that means we still need to do everything else EA is concerned about.
The reason is that AI is at best a tool that could be used for good or bad, or at worst intrinsically misaligned against any human interests.
Or alternatively AI just isn’t solving any of our problems because AI will just be a mere extension of power of states and corporations. Whether moral problems are solved by AI is then up to the whim of corporate or state interests. AI just as well IS being used right now to conquer. The obvious military application has been explored in science fiction for decades. Reducing the cost of deployment of literal killer robots.
Obvious example, look how the profit motive is transforming OpenAI right now. Obvious example, look how AI is “solving” nefarious actors’ abilities to create fake news and faked media.
There is no theory that our glorious AI overlords are going to be effective altruists, or Buddhists, or Kantians, or utilitarians, or whatever else. As far as I’m aware AI may just as likely become a raging kill all humans fascist.
In my opinion democracy is more likely to be utility maximizing compared to the alternative, oligarchy. In the status quo, the funders provide the moral weights. If your goal is utility maximization, small numbers of funders are more likely to have deviant moral weights compared to the median weights of the public. Their deviancy is less likely to capture maximally satisfactory policy.
A membership-driven democracy is more likely to have moral weights aligned with the rest of the public. Membership implies multitudes and therefore diversity, which can be a huge advantage in decision making. Having dozens of decision makers is assuredly more diverse than the singular vision of a single funder.
Like it or not, funders are also biased towards their self interest and may avoid otherwise effective policies. And perhaps this drives the other primary reason to use democracy—it is possible strategy towards building EA funding that relies less on the big funders and more on smaller donors. Democracy allows small donors to exert influence whereas the status quo is set up for large donors to exert influence.
I think you forget the biggest reason to use democracy—aligning the moral values of some entity with the moral values of its constituency.
Whether you like it or not, people’s moral values are different, even in EA. Some put much greater value on animal welfare than others for example. There is no universal or absolute way to say that yes, “my values system” is objectively better than yours.
Democracy is a way to align an organization’s moral values with its membership through aggregation. Democracy tends to satisfy more people than less through majority rule.
Your perspective on democracy in contrast is more in tune to Madisonian or Schumpeter-ian justifications of democracy.
In my opinion elections are a mediocre, inefficient leadership selection tool.
Because elections demand participation from many people, they are very expensive in terms of opportunity cost.
Elections demand that the candidates engage in marketing and campaigning. Candidates who spend time on otherwise wasteful activity (campaigning) are more likely to win.
Because the effective value of voting is usually negative (ie it costs more to participate than you get out of it), participation is oftentimes either rare or mediocre.
To look at recent history for example, how effective was voting for leadership selection? For example, did you know that the Democratic Party actually had primary elections in 2024 where Joe Biden overwhelmingly won candidacy? Months later, Biden was declared insufficiently mentally competent to serve.
In my opinion, the more effective way to select leadership democratically is using a randomly selected leadership panel.
Select by random from membership, say, 10 to 25 panelists.
These 12 panelists will be charged to read resumes, perform candidate interviews, make a final hiring decision, set term limits, and perform annual performance reviews.
This technique is otherwise known as “sortition” or “lottocracy”. It is vastly more efficient than an election. Imagining an election with 100 members, imagine each of the 100 members uses 2 hours of their time to make a decision. That’s a net of 200 hours of opportunity cost. With sortition, 10 panelists can devote 20 hours each for the same opportunity cost (10 x 20 hours = 200 hours). Because information gathering and fact finding is a serial task, one person devoting 20 hours to a decision can be far more effective than another devoting only 2 hours.
Sortition more efficiently uses the time of membership and produces higher quality decisions by orders of magnitude.
You could argue that sortition is un-democratic. However, philosophers have associated sortition with democracy for literally thousands of years since the time of Socrates. Understanding that democracy is about co-equal governance, sortition preserves political equality by giving participants equal probability of being selected to serve.
My interest is in improving democracy. I believe things like Trump, or Netanyahou, or Erdogan, or most democratic backsliding is a sign of democratic incompetence.
I think most people have little to no long-term vision for the question, “What would an advanced government of the future actually look like?” How much better could the world become if governments were smarter and more capable and just produced vastly more utility for people?
There is one thing I think is a strong contender for a superior future government. It’s called sortition. The premise is simple. In elections everybody participates. In sortition, a random sample of the public is chosen to participate. The benefit of sortition is scalability. Randomly chosen people, compensated or forced to participate, can engage in politics at enormous timescales compared to the average voter. More time to ask questions. Resources to become informed. The ability to seek and hire expertise. I elaborate about alleged benefits here.
Of course there’s a lot to do. First we need to prove the hypothesis. We have reasons why we think this is good, next we need to actually go out and test it. That takes a lot of money and research. This research will add evidence in its feasibility and capability. After that, if testing determines this thing to actually be good, more money is needed to campaign in favor of it.
Trump assuredly will not be the last authoritarian to arise out of liberal elections. If you value people’s freedom, if you wish to maximize utility, we should be looking for better things out there.
As far as why things like sortition would stop Trump, competent governments that are able to make their citizens feel content and satisfied with government performance, do not tend to appeal to tyrants for aid. Moreover if sortition is actually an effective way to organize people, it also might be an effective way to organize the Democratic Party.
Another powerful feature of sortition is its potential ability to create Democratic Legitimacy without going through the bureaucracy of government. For example, a Citizens’ Assembly can be potentially created through private funding, or through a referendum. A Citizens’ Assembly could be used as a presidential candidate selection system, and could delegitimize Trump or any other un-endorsed candidate.
Even if this is a “long-termist” project, the resources needed to say, test sortition, or launch a Citizens’ Assembly, are only in the millions of dollars. The economic benefits could be immense. Imagine a government that’s only 5% more efficient at increasing utility… that’s hundreds of billions of dollars of value per year. If the will was there, testing could happen immediately and we’d have results out before the end of Trump’s term.
Of course this won’t happen, not because it’s infeasible, but because there’s not yet funding, because sortition is mostly unheard of, because the idea hasn’t reached the ears of funders. Or if it has, the funders have just written it off for unknown reasons. Tractability is a typical excuse I hear, yet I’m not sure how sortition is any less tractable than any other long-termist project out there.
In my opinion by limiting your question to “defending liberal democracy” you are also vastly limiting potential remedies. The current resurgence of authoritarianism takes advantages of weaknesses of liberal democracy. “Defense” of democracy necessitates making the system smarter and more resilient to these authoritarian takeovers.
“Defense” may necessitate a transformation of the status quo.
Exactly what makes liberal democracy weak? One reason is poor collective decision making. Examples of bad collective decision making include:
Democratically controlled states inability to grow. Anti-growth and high cost-of-living policies of California, New York, and other blue states have encouraged millions of their residents to move to Red states. This foot-voting is a clear indicator of the incompetence of Blue-state policies, and simultaneously hands authoritarians like Trump more power.
Obviously the election of authoritarian Donald Trump who hopes to end democracy as we know it.
Incompetent internal party decision making systems that elected Joe Biden to be the presidential candidate in 2020, whose utter unpopularity, inability to engage in modern-style social-media driven policies, and senility was a large contributing factor to Democratic Party losses in 2024.
Whereas EA has funded some pro-approval voting initiatives in the past, it has not:
Engaged in any small scall randomly controlled trials to validate claims made.
Evaluate other voting reform alternatives such as Condorcet methods, multi-member districts with Single Transferable Vote, evaluate the possibility of proportionately representative systems, etc
Evaluate the potential for “Deliberative Democracy”
Evaluate the potential of Citizens’ Assemblies as used throughout the world in Ireland, France, UK, Canada, etc.
Evaluate the possibility of more direct democratic alternatives such as sortition where decisions are chosen by lottery.
As far as bothering to test and validate potential reforms:
Lobby labor unions to implement improved decision-making reforms.
Lobby political parties and organizations to implement improved decision-making reforms.
Lobby HOA’s to implement improved decision-making reforms.
Fund research into modeling and simulation of improved decision making with respect to mechanism design, economics, and social choice theory.
Lobbying smaller organizations may be more effective than attempting to run a full city or state referendum campaign for change, and build evidence whether interventions are actually effective or not.
In my opinion pure voting system reforms (such as approval and ranked choice voting) have low probability of success because they don’t tackle the core decision making problems of liberal democracy. Voting system reforms might improve aggregation, but if a majority of voters simply have incorrect information, they will still arrive at incorrect decisions. The only reform I’ve found that tackles the problem of voter ignorance is sortition, which I’ve linked above. In short, you can improve the decision making capacity of voters by paying them, and giving them enormous resources to arrive at better decisions. And the only way to scale such a process is by reducing the number of participants through a fair democratic lottery. As far as organizations advocating for this, it includes: Ofbyfor, Democracy Without Elections, INSA, Assemble America, BANR, etc. Full disclosure, because I believe sortition has the greatest likelihood of success, I volunteer in some of these organizations.
Ultimately we need a system that its own citizens believe in. People like Donald Trump succeed because citizens believe that the status quo is so bad, Trump is a valid alternative. Authoritarians succeed when liberal democracy fails its citizens. A successful defense of democracy demands improvement of democracy.
No, I’m requesting EA actually take the importance of improving democratic decision making seriously. Even if no action was able to stop these 2025 cuts, do you actually think “it’s over”? What about 2026? What about 2028? What about 2050? America is going to continue to make just stupid decisions until enough people get together and change the dumb way the system makes its decisions.
Moreover the second article isn’t about approval voting, I’m not sure how the only thing you got out of deliberation was approval voting.
If people in America were serious enough about improving democratic decision making, is it conceivable a reform could have stopped Trump? Imagine a new and improved Democratic Party was able to clearly demonstrate its ability to govern. Imagine a California government that was actually sufficiently competent to build high speed rail and more and more residential to attract more people into its borders. Instead Californians are fleeing because of rising costs.
Imagine an improved Democratic Party primary system that could elect a younger candidate that wouldn’t have grown senile by 2024.
Are these things *possible* within a small time frame? They certainly are. Trump himself demonstrates how quickly norms can be changed.
What’s wrong with US democracy isn’t just Trump, it’s an incompetent opposition party that people hate so much they’d rather trust something like Trump.
Finally yes, you mentioned approval voting. Would that ever be enough? Why are you putting all your eggs in just this one basket? IMO it’s a clear sign of EA’s myopia and lack of engagement with election theory, to ignore what is out there such as Single Transferable Vote, condorcet methods, and STAR voting. Even in this small niche of election reform in my opinion EA is far behind the theory.
>If you’re not proposing electioneering, what exactly is the program that you are suggesting could have prevented these USAID cuts?
“When should you have planted the seeds to grow a tree”? Just last year is a bit too late to grow a strong and capable democracy able to resist a tyranny.
A better year might have been 2016, when we were better understanding what the stakes were. That gives you 10 years. Or people have been complaining about the downfall of democracy since Occupy Wall Street. That’s 17 years (And people have obviously been complaining about democracy for far longer than that). But the next best thing might be now.
Throwing money at Biden/Harris 2025 is a method of last resort, particularly when it seems that money is highly ineffective in high-profile, money-saturated presidential campaigns.
Now let’s imagine that Trump actually does succeed in turning America into a dictatorship. Does that mean all hope is lost? No, there’s plenty of other countries where democracy can be strengthened.
>This forum might not be a bad place to start?
Plenty of ideas have been posted and ignored. I posted something for example on sortition which I’m a big fan of. Crickets. Neil Dullaghan made a great post about deliberative democracy here. What came of that?
Now maybe my idea is utter shit. OK sure, strikes and gutters. The silence is much more annoying.
In my opinion, attempting to electioneer in 2024 by pumping money towards your preferred candidate, has little to do with democracy. It’s kind of the opposite. You’re engaging in oligarchy, trying to buy power with money, to attempt to save what democracy you have left. You’re not actually addressing the problems that led to the current crisis. As I said, mitigation and reaction.
>I think it’s generally okay to place the burden of showing that a cause area warrants further investigation on proponents,
And how can any cause area demonstrate this when you just won’t evaluate it anyways because of your limited evaluative capacity, because it’s not a priority cause area for your organization? Let’s imagine I have a proposal or a white paper. How and where can I submit it for evaluation? Take for example Open Philanthropy. Democracy’s not a cause area with any requests for proposals. Is there any organization accepting proposals?
The cause areas are driven from the top down, as far as I’m aware. Causes outside the org priorities are just not considered at all.
I’m not asking EA to focus singularly on democracy. I’m asking EA to give any resources at all to the cause of democracy. Prove my ignorance wrong. Is any organization in EA involved with democracy at this moment? Is any organization bothering to evaluate potential interventions? What work has been done? What papers have been written? Is there some work saying, “Look, we’ve done the work, yes it turns out democracy has a terrible ROI!” How about you guys? Are you making any consideration or analysis on potential pro-democracy interventions? If you have, I’d love to see the analysis. My search for it, I’ve seen nothing. I hear crickets.
Here’s the thing about evidence. You have to look for it. Is EA bothering to look for it? Is your organization bothering to look for it? Otherwise, you have no idea how tractable it is or is not.
These sorts of cuts highlight IMO the incorrect strategy EA has been on. Whereas the EA space deals with the millions of dollars, US government aid deals with the billions of dollars, orders of magnitude greater funding.
Yet EA’s refusal to engage in the political has created a huge blind-spot. As the United States unironically moves towards authoritarian dictatorship, of course the foreign aid is disappearing, and your cost effectiveness calculations are completely out of whack. How the hell are you going to fill the gap on billions with mere millions?
You wanted to settle for the ease of linear thinking. A particular set of interventions was easy to measure and had more linear responses, so that’s where your funding went. Politics is incredibly messy and the response is extremely nonlinear—you can pour money into politics and see no response, until perhaps one day you can have a huge response, or perhaps not. You didn’t want to deal with the nonlinearity.
So you forgot to think about protecting, or even enhancing, democracy. I suppose protecting democracy just wasn’t tractable enough. And we’re going to be suffering the consequences for thinking democracy just isn’t tractable, even though people had been sounding the alarm for years/decades.
All you can be is just reactive. You’ll react to the destruction of democracy. You’re about mitigation, not prevention or enhancement.
Maybe it is time to think about being pro-democracy. The consequences of forgetting about democracy will cause orders of magnitude more suffering than you have ever prevented.
What’s the Difference between the AI Threat and the Multinational Mega Corporation?
Why do you think improving democracy is intractable? None of the highest priority world problems are tractable.
Risks from artificial intelligence—nobody knows if an AI safety solution is even possible yet that warrants hundreds of millions in funding
Catastrophic pandemics—Preventing/mitigating pandemics is a trillions dollar endeavor. Incredibly costly. How is this tractable?
Nuclear war—Exactly how is this cause area tractable?
Factory farming—Good luck on this cause, especially without the force of an enlightened government to demand change.
Comparing to the toughest problems, how is improving democracy intractable? Of course, tractability needs to be balanced with importance and neglectedness.
Developing strong evidence that some specific reform (ie maybe sortition) could be a real improvement could be done in the millions of dollars range. That could be cheaper than training your LLM. That’s definitely cheaper than fusion power.
What is the value to humanity of learning what kind of governments are best? Even in the short term perspective, the value of an improved government could be trillions of dollars of tax dollars saved. In the long term perspective, every top priority world problem would immensely benefit from enlightened governance.
Sortition as a specific reform might be slightly harder to implement on some political campaign, yet imagine hypothetically sortition yields 10% greater ROI in taxpayer benefits whereas ranked choice or approval voting might yield closer to 0%. Of course we don’t know the numbers, and that’s a huge problem. Ranked choice might be more tractable, yet it also might be mostly useless.
Yet we don’t know, because nobody is doing any testing, there’s no empirics and I bet, there’s no funding.
I’ve written what I think is the most potent possible reform here:
https://forum.effectivealtruism.org/posts/HwoSHayLt4zqqeyun/how-to-make-democracy-smarter
John Huang’s Quick takes
Has anyone done a effectiveness comparison between GiveDirectly vs Community oriented giving, such as Sparks Microgrants?
I think it’s strange to talk about Christianity and then forget about heaven and eternal damnation. It sounds like if you believe in eternal damnation of non Christians, your effective priorities need to be drastically different. Prioritizing the infinite afterlife is infinitely more effective than prioritizing life.
If Christians wanted to give effectively, they would surely continue giving to Christian branded charities that therefore continue spreading the popularity of Christianity to save the most souls. Their priorities are in conflict with secular organizations.