This is a special post for quick takes by Jarred Filmer. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
6 comments, sorted by Click to highlight new comments since: Today at 5:57 PM

As we all do from time to time, I imagined what I'd do if through some impossible vagary I was made president of the USA all of a sudden. What would one do differently? How would you perform the kind of rapid sense making needed in such a situation?

The obvious immediate step to me would be to use my presidential powers to assemble a room of people that seem smart, sane, practical, easy to work with and it got me thinking about who would be on that emergency list for you?

One might expect the who's who of the rat-sphere, your Eliezer, Scotst, Zvi's, etc. But who else does your mind immediately go to in this thought experiment? And what qualities are you looking for? And how would you organise them?

The obvious immediate step ...

Yes, seems like an obvious step to me, too. Which makes me wonder, whether people who actually become politicians (not necessarily presidents) do it.

I can imagine that they do, it's just that neither me nor anyone I know is invited to be a member of the "smart group". Either because there are smarter and more competent people available, or because the politicians' criteria for smartness are incompatible with my bubble. Or maybe the invited people are discreet and don't tell me.

I can also imagine that they don't -- that basically, if you have the instinct "there are people smarter than me and their opinions are valuable", you will never make it into the top positions in politics. (In other words, the Dunning-Kruger effect provides an evolutionary advantage in social interaction.)

Or maybe it's something in the incentives, like if you are a politician, your #1 task is infighting and backstabbing, and the outsiders are not going to be helpful there, because you need to spend 24 hours a days doing this to remain competitive. Also, if you are not the party leader, your party leader probably makes all the decisions and your opinion is unimportant; and if you are the party leader, you probably already made so many trades in order to get there that your hands are tied; in either case, the external advice, however good, would be unactionable.

It also seems possible that the politicians are so flooded with all kinds of unsolicited advice, that asking someone for advice is counter-intuitive.

...any successful politician here that could enlighten me?


Sent a few messages on Facebook to actual politicians, I wonder if any will respond. :D The question was not framed "did you assemble a group of smart people" but rather "is it common among your colleagues to assemble a group of smart people".

I also realized that I missed a simple hypothesis: In the thought experiment, if you suddenly become a politician, of course you would need help, because you didn't have time to think about things, and you don't have contacts. But in real life, you rise in politics gradually, so when you get on the top, you have already discussed things a lot, and you are surrounded by people who have also discussed things a lot, so... you probably don't feel a need to find external advisors, because it seems like all the smartest people are already in your bubble, and you meet them regularly. (My hypothesis is that the people in your bubble are not necessarily the smart ones, but that it feels so.)


A response (shortened, anonymized):

I agree [with the idea of surrounding yourself by smart people], and if everyone would do so, we would be in a different situation. Yes, there are many lobbists. Doing my work, I am in regular contact with professional organizations and experts. So far I have achieved only marginal improvements, although big reforms are needed. But there is not enough political will, lack of people and ideas, so as far as I know, only the "maintenance" work is being done. The administration refuses experts; they break the status quo and cost money. They prefer to hire cheaply.

Ha, I never would have thought of just messaging a bunch of politians to sate my curiosity. That reply is fascinating though.

I suppose in the ordinary situation where it doesn't happen suddenly you're also (ideally) running on some kind of policy platform so the gathering of advisors would have to happen well in advance. In that case you don't have the power of the office to summon whoever you want, and maybe there's se kind of loyalty thing to the existing team like you mentioned.

Still curious if you've got thoughts on the top of your head, who's on your list and through what process would you bootstrap to organisation?

Ha, I never would have thought of just messaging a bunch of politicians to sate my curiosity.

It also took me a moment to realize that this option exists. But this is a small country, and there are a few politicians I happen to know personally (especially if we include the municipal level). And the worst case is they would ignore my question, no harm done.

Actually, I once invited a (government-level) politician to a local Less Wrong meetup, and they accepted. This is probably easier than it seems. They are humans, too. Just be a polite person, give an interesting proposal, don't choose the one at the top but rather someone who is like 100th from the top, and don't do it at their most busy moment.

who's on your list and through what process would you bootstrap to organisation?

The case of "becoming a president overnight" is too magical. The less magical option would be becoming an average parliament-level politician, which means having a specific agenda (in my case, that would probably be education), and having a time to prepare. I would try to meet people who seem to be domain experts in my country; the more, the better. Just talk to them, try to extract their experience and opinion, and keep contact with those who seem rational and friendly.

Then, I would probably contact local rationalists I trust, simply because I would expect local knowledge to be useful. I would share with them my data, and ask them to help draw conclusions. (Even better if they would also independently collect data. Because more data is better, and also to cover my blind spots.) I would also ask foreign rationalists who seem to have some domain knowledge in education. Ironically, I probably wouldn't include Eliezer, because he doesn't seem to have domain-specific knowledge here. (Maybe I am wrong.)

But I assume that having a good plan would actually be just a beginning. I expect that the deep state would protect the status quo and throw all kinds of obstacles at me. So I would need domain experts at overcoming these obstacles. Such as lawyers, actually quite specialized ones, and I have no idea where I would find them. (I assume many of the obstacles would have a form of someone pointing out that my idea X is incompatible with some existing law Y, providing zero guidance at how to solve this problem, because their preferred outcome is that I give up on X.) Allies within the bureaucracy, who would notify me about things their colleagues are trying to hide from my attention, for example to set me up for an embarrassing failure. And this is something I couldn't bring from outside.

All things considered, even if I had my position perfectly guaranteed for 4 years, I would expect most of my efforts to fail, simply because people who spent decades working in the system (and benefit from the existing situation) would be able to wear me down by passive resistance and sabotage, and it would not be possible to replace them with new loyal people, even if I had them, because the new people couldn't sufficiently quickly gain full understanding of the existing system, and the things would start falling apart. I have seen people more experienced than me to fail exactly this way.

So at the end, I guess, I would need a rationalist with experience in politics. Do we have any? (And no, I wouldn't call Dominic Cummings, because to me he seems like an expert at making things worse.)

Makes you wonder if training and embedding an army of altruistic rationalist in politics would be a worthwhile endeavour. Although now that I type this that's probably not too far off what 80000 hours is attempting 😄

Given the power of mindkilling, the result could easily be an army of ex-altruistic ex-rationalists in politics. (Which wouldn't necessarily be worse than the current state of politics, it just wouldn't be the expected improvement.)

It's not like a have a better plan, though. I was thinking along the line of "suppose that certain fraction of politicians will be responsible, and will seek advice among the experts... look at what algorithm they use to pick their advisors... and position yourself so that they pick you".

But I suspect the algorithm would be something like "choose the most visible people already working in the domain you want to improve". In which case my advice reduces to "the 'life hack' to improve domain X is to spend your life working in domain X and become successful and famous", which sounds like doing things the hard way and being sufficiently lucky. (Maybe that is the optimal answer, dunno.)

The only point of intervention I see here is that we could notice the people who are doing the right thing, and try making them more visible, e.g. by writing articles about how they are doing the right thing. Which might slightly increase their chances of being picked as an advisor, compared to a person who is doing a wrong thing, but is good at climbing the hierarchy, so from outside seems like an equally qualified expert. In other words, instead of trying to place rationalists into domain X, just find people already in domain X who are relatively more rational than average, and try giving them more light.

Another potentially interesting project would be to create and publish a compilation of "rational policies on everything", and allow politicians to steal the ideas from the book. Let your memes travel farther than you can. The question is whether we could even compile such book. Because it's not just about technical answers, but also choosing your values. Often the choice is not between a policy that is "good" or "bad", but "better for X and worse for Y" and "better for Y and worse for X". Even the obviously bad choices usually have someone who derives some small benefit from status quo.