Shea Levy on why he disagrees with Less Wrong rationality, Part 1

In this podcast, Shea and I tried to hunt down a philosophical disagreement we seem to have by diving into his critique of rationality. We went off on what may or may not have been a big tangent about Internal Family Systems therapy, which I’m a big fan of, and which I think Shea thinks should have more caveats?

Unfortunately, our conversation got cut short because partway through, Shea got a call and had to deal with some stuff. We hope to record a Part 2 soon!

Transcript:

Divia (00:01)

Hey, I'm here today with Shea, who is my Twitter mutual, recently got a job doing software engineering management at Andro and is an objectivist. I think I've, listened to a long podcast with Shea about objectivism on Daniel Filan's podcast, The Filan Cabinet, which I recommend to anyone who's interested in that. We'll see how much we get into that today. I'm sure it'll come up. Our primary topic for today is, Shay sent me a long list of things and one of them was Contra.

rationalists. So I, if you check my Twitter bio, it says I am a rationalist. So I figure there's probably some interesting disagreement here that we can dive into.

Shea Levy (00:38)

Yeah, sure. Thanks for having me. Yeah, I guess with respect to the rationalists, I've been rationalist adjacent for lack of a better term for I guess over a decade now. I originally kind of became aware of it through Ozzy. So Ozzy was on, I don't even remember what the thing, there's something

Some feminist blog kerfuffle and somebody, a blog I followed, Provocracy, recommended looking at Ozzie's thing on this. And then through Ozzie, I then found the rest of the rationalist sphere. I was mainly on rationalist Tumblr for quite some time. I've been socially connected and intellectually engaged with the community since then in various ways. Some of my closest friends are rationalists.

Divia (01:35)

Like they say.

Shea Levy (01:37)

And throughout all that, there's been a kind of general fascination and love for some of the things and then deep frustration and then for some parts of the community, going from frustration into I think something seriously wrong, defective there. I don't know what's interesting to you if you want me to just give my spiel or if you want to talk, I don't know.

Divia (01:57)

Mm-hmm.

So something that comes to mind, I don't know how much this will relate or it won't. The first conversation I remember us having on Twitter, which may not be the first one we actually had, is talking about AI boxing. Does this ring a bell? Yeah, because I think, I don't remember, but I think I was like, okay, who wants to talk about AI boxing or something like that? And you were like, no, this doesn't make any sense because I just would simply not let the AI out of the box. And we ended up talking about it, though then it seemed like-

Shea Levy (02:17)

Yeah. Yeah.

Hey.

Divia (02:35)

This is probably what you're saying all along, but what I eventually understood you to be saying is, unless I think the AI should come out of the box.

Shea Levy (02:41)

Well, so, okay, let's start with that one. So I feel like there are a pieces to it, but let's drill into me personally, if I'm the one in charge of the AI Box. So like, will, will see that there are people you should not put in charge of that. But the fundamental thing that I don't buy in the AI Boxing argument is that there is some, like,

Divia (02:52)

Yeah, you were clear about that. You were not like arbitrary person. That's a good idea. Yes.

Right.

Shea Levy (03:10)

general like key you can turn in anybody's mind to make them believe anything illegitimately like You can I'm not saying I'm like some like you completely like never fall for any scam kind of thing. I'm not saying that like if somebody could trick me about things if I were like in a normal situation where I'm treating them as a like good faith human being that I'm interacting with like I'm certain certainly people can pull something over me I'm not my claim is not that

But my claim is if I know that this is a like potentially hostile intelligence and now let's set aside whether an AI could do that That's a separate concern I have but like yeah, if I know there's a potentially hostile intelligence. I don't understand how it works I'm well aware that it doesn't work like a human being even if I don't know the details of how it does work like there is not some magic sequence of words that it can tell me that

will convince me of something false relevant to letting it out. I just don't buy that and none of the arguments I've seen suggest that it's true and it all really seems to rest on the sort of like determinism, lack of like the impossibility of objectivity that like, like the AI could present a like picture that is an illusion.

It could present something that looks like something, but it's actually, you know, it looks like X, but it's actually Y and could confuse me visually. And I think there's this idea implicit and sometimes explicit, but implicit in the argument that like, reason is like that, that there's some like reason level illusion possible. I just, that's just not how reason works. So like, if I knew what the situation was, I know it's an AI in the box. I know that it's extremely dangerous.

I think it's possible I would let it out, only under the circumstance that it actually convinced me properly and like, because it was right, not because it came up with a clever argument. Like, I would know, I would set up controls beforehand and then reevaluate those controls if I thought they were being breached. Like, there would be a number of things that I could do depending on the specifics to like, actually confirm that the argument it's making is legitimate or not.

Divia (05:09)

Yeah.

Shea Levy (05:28)

And so like, yeah, that's basically my view. like, don't think that's something, I don't think it's, certainly not everybody in the population is true of, but I don't think it's something like distinctive to me that like, I've got some special skill here. I think like there's a certain degree of like objectivity and self-awareness and understanding of how, you know, how reasoning and how like persuasion works.

that you want to reach that point, like there isn't this magic button that a known adversary in a known adversarial context can compress.

Divia (06:04)

Yeah, ultimately I think you convinced me that that was basically right.

Cause I had, so as people may know, I guess I should say what AI boxing is. And many years ago, Eliezer Yudkowski was, and of course like the whole thing in some ways is moot because his argument was, this was about an AI safety argument where he was like, look, AIs are dangerous. And then at the time people used to be like, but we could put the AI in a box and not connect it to the internet. And then that would not be dangerous. And he was like, okay, but people would let it out. And people were like, I don't think so.

And he was like, I'll tell you what, I'll do an experiment and I'll role play the AI and then the person will let me out. And he did do that a number of times. I don't know if it was 100 % of the times he tried it. It might've been, but it was certainly a bunch of them. And he didn't say exactly how he did it because he was like, well, if I say, then people will be like, well, that exact argument would have worked on me, it was in the point and.

Shea Levy (07:01)

Yeah.

Divia (07:03)

And then I tried it too. I tried it with a friend of mine and in fact, he let me out as well. And so I was sort of like, okay. Yeah, and as it is, people of course have made AIs and immediately put them on the internet. like no AIs are in boxes as far as I know. Like maybe some hypothetical future AI could be in a box, but that's not really what people are up to anyway. But like it is interesting to me because I, because the first time I heard about LAO's are doing it, I was like, well, how could that possibly be right? I can't imagine anyone doing that. And then I was like, all right, I thought about

I think of some ways maybe I could do it and then I did it. But I guess it never really occurred to me, I think until talking to you, which seems like a pretty big oversight, that maybe some of those people should have let the AI out of the box. That there aren't actually zero cases where people could let, and of course it could be that it didn't more like he tricked them and, but yeah, there was this other possibility.

Shea Levy (07:53)

Yeah, I'm curious, you okay sharing publicly what you did or maybe if you could share privately after it?

Divia (08:02)

I just don't think it was, I'm like, I don't think it was that interesting. I tried to role play the AI and say the most persuasive things I could say, which in fact were me being like, yeah, you could like run these safety checks on me or whatever. But then I was like, but of course like I could fake those. And I think your point is like, well, maybe not though. Like couldn't necessarily fake that stuff.

Shea Levy (08:19)

Let's take an example that I don't think this is how you would solve the AI safety issue if there's an AI safety issue, but it's something that like if

Like, let's say it's an AI in the old mode that we, before we had weights and things where we thought we were going to write it in code, right? And we had a, AI, what the AI gave you was a formal specification, which is relatively concise that like a mathematical, that a like formal theorem prover could understand. And then a proof, and then like, you'd obviously have to like inspect the proof to make sure it's not like taking advantage of a bug in the theorem prover. Maybe you would give the same proof in multiple theorem provers and all of that.

Like that kind of thing, like there are certain aspects of that. Now I don't know if that would be enough. That wouldn't be enough by itself, but there are certain aspects of that. Like I would be very convinced. Yes, this code that is validated has certain really important properties that are relevant to safety. so like it, so yeah, like the, the, the level of thing wouldn't be like, you know, it would have to be a fairly robust.

piece and like, part of it would be like, if I'm, if I'm already somebody concerned about AI safety, part of what the AI in the box has to convince me of is like, it basically has to like, if I like, let's say I buy like LA's basic framing around like friendliness and value alignment, like it needs to solve the problem. It needs to solve the value alignment problem and then convince me it's solved it. And if it does, right, if it's super intelligent and this is a real problem and it's or prove that it's not a problem, right? But like.

Divia (09:47)

Mm-hmm.

Shea Levy (09:57)

not just, it wouldn't suffice to just say, you know, here's a list of common sense things that you can check and if you check these things, it's probably fine, right? Like, that's not the bar, but like, I know that's not the bar. And so, you know, I wouldn't let it out. Yeah.

Divia (10:12)

Yeah. Okay. So with that as some stuff in my context, you walk me through your anti-rationalist spiel?

Shea Levy (10:20)

Yeah, I mean, so I think it's a love hate thing. So it's not purely anti, I mean, yeah. I guess the first thing I'll say is like what drew me to the rationalists originally and still does is outside of objectivism, it's not all of objectivists, but outside of objectivism, it's the only group of people where a significant part of the subculture

exists and is driven by people who care about what's right and what's true and in a way that ties in with like modern understanding of economics and scientific development and things like that that like takes industrial scientific civilization seriously and add value even if there might be some issues with it. And again is like dedicated to understanding what's true and following that where it leads. So there are obviously non-rationalists who have that.

Divia (10:54)

Mm-hmm.

Mm-hmm.

Shea Levy (11:18)

but there's no other like community or like movement that has that as a distinctive characteristic. Other than objectives that I found. Historically, yes, but I mean, so this gets into some of the like philosophical cultural things I put on our list, but like, I think there is a general element in our culture that

Divia (11:18)

Sure. Community.

other than objectivism, you're saying.

Shea Levy (11:45)

put pits, idealism against reason and science. And I think the rationalists are one of the few groups of people who on an implicit level have combated that. And I'm emphasizing on an implicit level because this is where the bad side comes in, is I think the explicit ideas in so far as you guys have explicit ideas that you're kind of are coheering around, anyway.

Divia (11:57)

Mm-hmm.

Yeah, we sort of do, we sort of don't.

Shea Levy (12:14)

explicit AIS or co-hearing around that that are common, think ultimately undermine that interest. like, don't think... No, no, no, even setting aside safety because so like, think, I think some of the safety stuff is a consequence of some of the things I'm thinking about. But aren't concerned about safety have these issues. And it's basically like, I was thinking about this before before you got on the call and like, the there's a sort of like,

Divia (12:22)

Do you mean the AI safety stuff? Okay. Okay.

Okay, but you don't mean that.

Shea Levy (12:44)

What is the, I have a view on like what the fundamental issue is, but I think the high level it's, and like the sort of, the high level, the surface level of it is a kind of overly, there's an acceptance of a, like a view of the world that,

takes too seriously contemporary views on ethics and the nature of science and reasoning and then doesn't grapple with the issues that those ideas have. And so insofar as the thing that attracts me is people who take ideas seriously and want to actually get to the bottom of the issue, the thing that bothers me is that those of you who are more like,

explicitly philosophical don't seem to be bothered by the issues with your philosophical framework. And those of you who aren't just get letters stray by the issues. like the biggest sort of surface issue, which I don't think is the fundamental, but the biggest surface issue here is the ethics. I think part of the problem is that there is no consistent ethical framework. There's some, there's a sort of modal, there's a modal framework, but there's not, and like that's part of the problem. Part of the problem is

Divia (14:05)

That's true, yeah. I would agree with that.

Shea Levy (14:12)

not seeing these philosophical ideas as needing to be an explicit integrated system. you guys, tents at the level of the kind of culture you're trying to build, a broad tent doesn't work. Like, you need to actually have... It doesn't work for intellectual and value alignment. Like, insofar as your goal is to be an intellectual movement or even a social movement trying to build a...

Divia (14:26)

What is it doesn't work for what?

Shea Levy (14:41)

different way of living within the broader culture, which, know, part of it, like the rationalist project is not a coherent project. It's not a specific thing. So like we're sort of like working backwards from like, what are, you know, if you were to say like, what is distinctively rationalist or not? And like anybody would have legitimate quibbles with anything I'm including here. But if I talk about like, like the consequentialism and I'm going to use the AI terminology, but this is not really an AI thing. This is before even without the AI issues, this is the,

Divia (14:50)

Sure.

Shea Levy (15:12)

orthogonality thesis or, sorry? Let's take the weak historical, the human version of reason is a slave of the passions, basically like your, or to put it maybe in more rationalist jargon, like.

Divia (15:13)

Yeah, I don't buy it for what it's worth. I don't buy the orthogonality. Or sorry, there's the weak version and the strong version I could get into the weeds, but I basically don't.

Shea Levy (15:34)

You have your preferences and maybe you have preferences about your ultimate values, perhaps. It's not ultimate values, it's terminal values is the way you put it, right? And those are maybe for some strains of rationalists, they're evolved in completely. For some, it's sort of a mixture of evolution plus whatever culture you grew up in and experience or whatever. like those are not things you reason about. They're things that you have and then you use reason to achieve. And maybe you have

subsidiary values that you kind of reason yourself into causally that support those ultimate ends but like those are given in some way and then there's a so that's that I think is wrong, but then there's an additional layer on top of it which is and Obviously the ones most of us have and if you're one in this sort of stronger biological camp you say this is like genetic is a sort of like some some flavor of altruism Is sort of like we we were built in

It's built into us to care about other people and to, know, there's, Eliezer has this short story and I wish I knew what it was called, but it's like, it's like about some alien species that isn't at all altruistic and like they're, it's not three worlds collide, no, no, no, it's a different one. It's like, basically, a big part of the story is one of the like aliens evolves an altruistic impulse.

Divia (16:48)

Do mean three worlds? Is this part of three worlds collide? No. Okay.

Shea Levy (17:02)

But the alien species is kept technologically down because they're all egoistic in a very, I think, very short-sighted, dumb way. think Eliezer's not taking egoism seriously in this story. But the idea is that each, it's very genetic, selfish gene type thing of each individual cares about itself and then cares about its relatives.

based on how closely genetically related they are. And so there's sort of like planet wide warfare constantly. there's like very limited technological development. then the humans come along and they're extremely technologically advanced. And anyway, one part of the story is there's one sort of like genetic mutant in the tribe who's developed altruism. they're like, humanity has like adopted this alien into humanity. And like part of that story though, is that like,

Like the alien doesn't believe the humans that they're being altruistic. They think it's like some kind of lie. They're like, how could you possibly evolve this? And like the human is like, yeah, well, we, best guess is a sort of like lucky evolutionary path where like we evolved altruism in a way that could have been exploited, but wasn't. And then it got widespread enough that then it was able to sort of sustain itself. And then that's sort of what.

But like, anyway, a key part of the story is that like altruism is clearly in this view, like at least locally opposed to like this flourishing life and maybe it's only if you get lucky. Anyway, but the reason I get into that is like, there's clearly a view amongst many rationalists that not only is altruism my personal terminal value, but it is in fact like a human wide,

Divia (18:35)

Yeah.

Shea Levy (18:51)

thing that many humans, if not all, have baked in. And that being baked in is not questionable and is kind of a foundation of ethical reasoning. So first of all, I find that depressingly parochial because not all human beings have And it's only historically, especially historically. If you look at like

Divia (19:11)

because lots of people don't behave that way or even profess it. Yeah.

Shea Levy (19:17)

The first people who explicitly talk about ethics in the Western tradition, it's the ancient Greeks. there's a wide variety of ethical stances that they put forth, but basically none of them are explicitly altruistic in this way. You have elements of things that look like it, but nothing like, the goal was your own perfection and happiness, That is the goal of most of these virtual, ethical, sorry.

Most of these ethical theories. That's the goal of most of these ethical theories. And like some of them think, okay, well, that's not really possible. Happiness is not possible, but at least we can reduce pain. But again, it's reducing your own pain. It's not like reducing pain for the people at large. And I think it's a very Christian thing that comes up. So like that's the sort of most service thing. And I can get into like where I think that comes from. But like I think that is...

Divia (20:06)

Mm-hmm.

Wait, okay, so.

Shea Levy (20:12)

that's the area where like it ties into EA, but even before you get to like the weirdness of EA versus cultural altruism, it's really the altruism part. yeah, anyway, sorry, I've been talking.

Divia (20:24)

Yeah, okay, no, this is interesting. Let me see if I can back up and maybe try to summarize and try to see what... So unfortunately, I do have like a small sinking feeling that we might not have as substantive a disagreement as I hoped, which... But yeah, okay, I think we will too. Because I'm not really willing to defend the orthogonality thesis, like values are sort of arbitrary things that came from evolution or evolution plus culture. Like I will not defend that thing.

Shea Levy (20:38)

think we'll get there. I think we'll get there if we keep cooking.

Divia (20:54)

Okay, I will say a little bit, this doesn't feel like a central point, but like about the Greeks, some could be like, okay, but clearly they did care about some people somewhat, right? That weren't just themselves, not necessarily just for their end.

Shea Levy (21:08)

Yeah, but nobody doesn't, right? There is nobody who's claiming that like, it's the stupidest, stupidest straw man of egoism to say that an egoist literally doesn't have friends, literally doesn't care about anybody, right?

Divia (21:11)

Yeah.

Okay, yeah, yeah, no, and I don't want to spend time with that straw man, but I guess I'm like, but what, I think I don't fully know what you mean by altruism here.

Shea Levy (21:33)

So I mean, think the basic idea here is that like, who is the proper beneficiary of like, one way to think about, yeah, so one of the big questions that I any moral system should answer is like, moral principles give guidance to your human action. The question is like, who ought to benefit?

And like there's other questions like who ought to benefit from your actions and like the altruist stances like fundamentally not as a means to some other end not as a part of some bigger thing but fundamentally like the standard of value is the good of others now there's the utilitarian version where it's not the good of others it's the good of everybody including you but like you're one still one piece of it and I like I don't think that distinction is very important like it's

Divia (22:20)

But it's.

But is that what people are saying that rational... is this really part of rationality?

Shea Levy (22:38)

So I think it's, again, like, can quibble on what is and is not exactly part of rationality, and I'm happy to sort of have that discussion, but like, there's a strong, I think there's a strong, they wouldn't say that explicitly, so like, that's, I think for everything, yeah, okay, that's a point, and I'll flag that a lot of the things I'm gonna claim, like, many of them aren't explicitly endorsed in that term, but like,

Divia (22:45)

Well...

Yeah, that's what I'm asking about. I'm like, how can you see that? Like what are some maybe less obvious ways that you can see this as more part of the rationalist worldview?

Shea Levy (23:07)

So I think the, I think you see this in the embrace of utilitarianism, of like, what we're trying to do is do the most good, right? What does that mean? It doesn't mean build the best life for yourself. It means like, create a universe where happiness across the whole population or the whole, you know.

intelligent light cone, right, is is maximized in some sense and you're trying to, I'm like that's not everybody.

Divia (23:37)

And is this what, yeah, like to what extent is this a rationalist position? And again, like I'm not trying to make it, I'm trying to say some ask the substantive version of that.

Shea Levy (23:48)

Yeah, I mean, I think so.

I think insofar as there is a vision of the good, it is not always explicitly this, but it is almost always either something like this or something that, and I think these ultimately bottom out in the same thing, but, or a view that sort of externalizes the good in some way, excuse me, from yourself. it's sort of like, just like, it's not exactly this, but it's like,

It's something along the lines of like you could imagine like what would I have to do to maximize the mass of the earth right like that's something sort of like a Problem you could present and then identify that and like goodness is something like that. It's more complex, but like it's some like physical it's some not physical, but it's some like Characteristic of the state of the world that you are trying to maximize right that you're trying to say like

you ease the, you like, you see this in the sort of like, kind of reasoning about, like, some of the thought experiments about, if you could press a button and, you know, like, and so some magical thing would happen, like, which world is better? And like, as if that's a coherent question. And like, people's... Yes, yeah, yeah, but I mean, so that this is sort of going lower in the fact than epic. But yeah, so like, that, that is, I think, a big part of

Divia (25:16)

Yeah, I do tend to be a conscientious objector to these types of questions. Yeah, yeah, okay. I like your answer.

Shea Levy (25:26)

that kind of thinking about ethics. Basically, okay, here maybe here is the thing I would more strongly defend as central to rationalism, but have a harder time pointing out to like a lot of convincing examples. So there's a mistaken view about what science is and that like what kinds of identifications and especially quantitative identifications are properly considered scientific. And then

Divia (25:39)

Yeah, I got it.

Shea Levy (25:54)

an application of that view to ethics that ethics is a scientific thing. And I agree that ethics is a scientific thing, but if you have this wrong view about what science is doing, then... I'm gonna try to see if I can, like... It's... Okay, so this this view that objectivity, especially in a scientific context, is somehow impersonal. It's somehow like you have to take yourself out of the picture and you have to think about like,

Divia (26:03)

Yeah, so can you expand on the wrong view of what science is doing?

Shea Levy (26:24)

What would, know, sometimes that impersonality surfaces as like, what would convince any rational being? And like they're very clear to say like not just a human being, but like any sort of rational thing like would convince them or what would, you these like veil of ignorance type approaches, like what would convince anybody who doesn't have any particular interests at stake? So they're like starting out to say that that's what scientific reasoning has to do.

And then says, okay, if we're going to make a scientific ethics, if we think that's possible, if we're going to do that, then whatever we're aiming for has to be measurable and identifiable in that way. Yeah, has to be legible in that, but that particular view of what legibility is, right? Like.

Divia (27:05)

to be legible.

Right, legible to some kind of view from nowhere type of thing.

Shea Levy (27:14)

Yes. Right. And then what a lot of people then do is say, but we can't find that. And so what is going to ground this process is what we can make legible is given an ultimate preference, X would be good. But the ultimate preference isn't itself legible or it's made legible in non-ethical ways. Like it's legible because it's what you evolved to have or it's legible because it's, you know, what...

culture has acculturated us to have. like that's where I think the other, that's where I think these, these kind of two paths merge. Like either you think you can succeed at this project fully for ethics and there's some like just intrinsic goodness you can measure with a, you know, a goodness thermometer about the world. Or you say, no, you can't do that.

Divia (28:01)

Moral realism, right? I mean, people maybe disagree about what that means. Yeah.

Shea Levy (28:03)

I'm a moral realist though, so like that's my point. It's a view of what realism would have to mean. I think that...

Divia (28:10)

Yeah, I consider myself a moral realist also, I think. Though again, sometimes when I try to talk to people on both sides of this, they seem to get really different, really varied answers about what moral realism is supposed to mean.

Shea Levy (28:13)

Yeah, okay.

Yeah. Yeah, I so like, guess I say I'm a moral realist because I think that kind of gets at the issue, but it's, think this is an area where the...

the conceptualization that we have, not with, not a, of an object level position, but of like how we categorize the positions, slants the discussion to rule out the right answer.

Divia (28:43)

Mm-hmm.

Yeah, I will say, so in terms of who's saying this explicitly, think, and apologies if I'm misrepresenting anyone, but I think this is Spencer Greenberg's position that rationality is for how you achieve your values. Values can be whatever. I think he calls it valueism. I think you're saying you can see it in Eliot's writing and maybe in Lesserang in general, but I think this is something that Spencer will be like, yes, this is my position. Yeah, okay.

Shea Levy (29:14)

Yeah, yeah, yeah, that's and like I mean, I think if you where you see elements of this is like People talk about like preferences and preference stability and like this there's a sort of notion the notion of preference and the way That fits into rationalist discourse. I think implicitly bakes in a lot of this right like it's like obviously there's I think there's a sort of like innocent path there because there's like a preference like I prefer

like long video chat discussions about philosophical ideas and my wife very much doesn't, like she would not be having fun right now and I am. And like that's, that.

Divia (29:52)

Therefore, I should like offer it to you and not to your wife and stuff like that.

Shea Levy (29:55)

Right. Yeah, right. So like, should seek it. like, so there's, there's a real phenomenon that people are seeing here. And like, it's also not only like that, but also like, the fact that I have this preference and she doesn't doesn't mandate that one of us is right and one of us is wrong. Right? Like, yeah. yeah, with that, but like, then there's a sort of like, okay, well, that's what all of valuing must be like. And like, really, it's a

Divia (30:11)

Totally. I think everyone would agree. Everyone reasonable would agree about that.

Shea Levy (30:22)

Like I think it's more the better way to think of it is that like there is optionality within a general positive view about what a, not sorry, positive view is getting too epistemic. There's optionality within the realm of like what is a good human life. But like the nature of what a good human life is is not.

downstream of those choices. At the right level of abstraction, it is a fact of reality what a good human life is. Now, there's things to say about that that I cringed with that formulation, but like, first line of the fan, I'll put that there.

Divia (31:03)

Okay, I'm still a little worried that we don't disagree about as much as I thought. So what is something, maybe you have a guess, like what's something that you think we probably disagree about?

Shea Levy (31:13)

I, so I don't know. I think if my guess is there's something in epistemology and maybe, maybe some like metaphysical stuff that we disagree on. that's, that's only because of things around, like

Like, okay, the discomfort I have with internal family systems. We haven't gone into that, but like, I suspect your willingness to...

Divia (31:38)

yeah, we can talk about that. I do like it.

Shea Levy (31:47)

Embrace it despite whatever concerns you might have. I suspect that's downstream of some like Yeah Yeah, yeah, I mean maybe maybe let me start at the bottom then because I think that's that this is maybe like Yeah, okay. So my view and again, I don't think this is every single rationalist. There are definitely rationalists who disavow this explicitly but still

Divia (31:53)

Okay, yeah, let's try it. This is a good trailhead and we'll see what happens at end of it.

It's like with libertarians, they say, you know, there might be two rationalists who agree with each other, but I'm not one of them, right? That's what they say.

Shea Levy (32:17)

Right exactly. Actually libertarianism is definitely a place where I think we would disagree So let's let's so they I think the basic thing is Again, so we have this pre explicit philosophical Value that rationalists tend to share of caring about what's true and what's right and seeing that on through or seeing

Divia (32:22)

That might be also true. Okay, let's get.

Shea Levy (32:45)

some elements of the post-enlightenment scientific and industrial and cultural establishment as reflecting that. Not everything, but like something about the way science has developed, something about the way like our culture has developed, some elements of that in some ways reflect the success of like taking seriously what is true and or enable us to do better at taking seriously what is true.

Divia (33:14)

Okay, I mean, that seems like a pretty narrow statement. Some asked, I'm like, yeah, okay, I think I'm on board with that so far.

Shea Levy (33:15)

So.

Yeah. So, like there are people who either reject modernity altogether, there's like the religious mindset which is like the real religious mindset, not the like, you know, postmodernist, like I'm religious, but like all other religions are fine, but like the religious mindset who thinks their religion is true, takes their religion very deeply seriously, and like thinks like God literally exists and literally did these things. That religious mindset you could sort of say cares about what's true.

Divia (33:27)

True.

Shea Levy (33:48)

but not in a way that embraces any element of modernity. And then there's the other side, which embraces some element of modernity, but then thinks that implies that you are, that, like, taking truth, like, the whole post-rat ethos of, like, taking truth seriously is for squares. Like, that's, that's the-

Divia (33:52)

Right.

That's why I rationalists in my Twitter bio. It's not the only reason. Like, I really do come out of that community, but I'm like, no, I don't think that's true.

Shea Levy (34:12)

Yeah, right. Yeah. So like, I think that that's why I put those two together is because I do both to kind of just anyway, so you have that. But the I think at the bottom,

The rationalists are by and large materialists. They take the success of the scientific revolution and the ability, sort of the reductionists success of basing chemistry on physics and biochemistry on chemistry. then basically they're saying, okay, let's take that fully seriously. We could talk about different variants of materialism, but basically what we care about is physics, things moving around, everything is built on that.

in some cases there's a sort of degree of taking that seriously where they'll say like there's no mental anything and some will say yeah there's mental stuff but they don't grapple with like how their worldview like actually doesn't allow for mental things so that's a very like basic philosophical issue how does that how does that apply in different cases is an interesting question yeah

Divia (35:17)

term by the way that I sometimes use on Twitter called the Materiast Epistemology Industrial Complex.

Shea Levy (35:22)

Yes, okay, so you, okay, think you, okay, so like, if you reject that, if you reject that, and it sounds like you do if you have that kind of like, dismissive terminology for it, like.

Divia (35:28)

Yeah. I think so. Yeah.

Shea Levy (35:35)

In what sense are you rational? like, I think part of the like, centering,

Divia (35:41)

Okay, I do want to answer that, but you were gonna say some stuff about internal family systems. I really do like that.

Shea Levy (35:46)

So okay, let me let me jump to the internal sorry I forgot about that right the internal fandom of systems thing is like my guess and I don't know enough about your view but my guess there's there's an element of like

Divia (35:54)

Yeah, well, we'll figure it out.

Shea Levy (36:05)

Treating yourself and the development in psychology of treating yourself as a third party, a third person object to evaluate, that's like, internal family systems, it's not literally true that there are multiple yous in your head, right? Unless you're a...

Divia (36:24)

Okay, but internal family systems for the record, calls most, it calls a bunch of things parts and then also talks about self.

Shea Levy (36:33)

Okay, so like maybe, right. But okay, maybe part of it is that like the popularization that I've heard of it sounds too much. It's a sort of thing that, okay, I guess maybe I'll let me just say like what I've heard and what makes sense. What seems like is going on to me is that like, it can sometimes be helpful to sort of like, for the sake of a like,

Divia (36:33)

So it's not trying to say they're all the same thing.

Is it fat suit?

Shea Levy (36:59)

thought experiment, like treat certain motivations or thought patterns or something like that as distinct from you as like a separate person or a separate like motivational system that is either distinct from you or distinct from other motivational systems that are then like somehow at play at different times or you need to integrate in and like I can see why that would be helpful to like as a like tool for like

working your way through a problem. But as a literal description of what is going on psychologically, I think it's false in a way that if you like, try to seriously integrate it with your reflective experience, you couldn't hold coherently. Like...

Divia (37:32)

Okay.

Okay, let me try to hash this out. So let me give you my brief description of what I think internal family systems is. Okay, so I believe it was invented by a guy named Richard Schwartz. And I think he got there sort of phenomenologically, like empirically, he was talking to different people. And he noticed that sort of interesting things happened when he asked them certain questions, such as like, and because talking about parts, like this is the thing people just do though. Part of me wants to stay in bed, but part of me is like, nah, you should get up and go to work, with a normal thing to say.

Shea Levy (38:14)

Yeah, yeah

Divia (38:16)

And then he used to be like, okay, well, what is that part of you think about this? like people would go with it and do the answer. And then he started getting some interesting answers when he would ask, for example, questions like, okay, now can you ask that part of you to step aside? Because sometimes they would and there's a little bit of nuance to like, that's not necessarily gonna work. But it is also my experience working with myself and with other people that sometimes this really does happen. And it...

it means something like then people are like, okay, and they do actually sort of set those concerns aside, which doesn't mean they won't come back or anything like that. And he also noticed that when he sort of kept asking people to do it, again, there's some nuance to people, it's not always gonna just work straightforwardly, that he'd be like, okay, what if what's, and people will be like, no, no, this one doesn't feel like a part, this one feels like me. And that's what in IFS he calls self. it characteristically, like it has some qualities, like people experience more curiosity, more compassion, more whatever, whatever.

I've tried to sort of ask other IFS practitioners about this somewhat, like I've taken some classes and things. It seems like people are now like, yeah, well some parts sort of have more self than other parts and whatever, like, which seems kind of right to me. It seems like it's certainly true that people have some perspectives that are not totally integrated at this time. You would agree with that, right?

Shea Levy (39:33)

Yes, well, mean, like, yes, I mean, think that's a, if it's an important one, think that's in like a flaw. Like if it's something like surface level thing.

Divia (39:36)

It sort of should be uncontroversial, I think.

We get, but that's the point of the IFS process is to integrate the things. Like the first time I talked to my IFS therapist that I used to work with, she was like, I don't really have parts like the way I used to have parts. Like it worked basically.

Shea Levy (39:46)

Yeah, okay.

So maybe it's just the metaphor has been to, like, the metaphor of family, I don't know.

Divia (40:02)

I do think it's a terrible name. think it was because someone else was a family systems therapist. And then they were like, we'll call it internally family systems.

Shea Levy (40:07)

Maybe then the only disagreement I have is that adopting a name that doesn't actually accurately describe what you're doing. I guess on the one hand it's interesting and reassuring because now I want you to send me what you think is your best IFS, the best IFS resource. that sounds more like, so the way I would conceptualize the same thing is there are

Divia (40:17)

Yeah.

Hmm.

Shea Levy (40:37)

You can have different thought patterns and subconscious identifications that don't integrate at any given point, which depending in certain contexts, you can like oscillate between or both have both, you know, subconsciously activated at the same time leading to a conflict. And you need to sort of tease out like, which is which is true or more usually it's not which is true sort of like

Divia (40:45)

Mm-hmm. Mm-hmm.

Mm-hmm.

Shea Levy (41:06)

How do I reconceptualize this issue to understand like they're both identifying something real about the situation, but like not in a way that I can integrate these together. But like I think the biggest question I have with it is with the description you have is to what extent these are like coherent over time. And I guess like maybe maybe the answer is they're coherent over time insofar as you have not done the work and you have like continually acted on one.

Divia (41:08)

Right.

Wow.

Shea Levy (41:35)

or the other side of a given divide.

Divia (41:38)

Yeah, I think sometimes they are, sometimes they aren't. I guess a different thing I would say is, so I believe my IFS therapist when she was like, don't have parts like I used to have parts. I tend to think it's not usually, it doesn't usually make sense to try to sort of resolve every polarity fully. Like if it were free, sure, but it's not. And so I guess I think that there's some shifts that I think of as more like, I mean, nothing's always worth it.

but closer to like overdetermined that it's worth it. That if some people's perspectives are like really sticky and unresponsive to evidence and wrong, then there's something I think very good about trying to free that up and get things like moving in a way where they can't integrate. But then I think like given, I don't know, like given my practical constraints, I think it makes sense for some things to be sort of persistently represented in more like a polarized way, but like that's pretty fluid.

Like a friend of mine, and I know it's not the same with groups, but as an analogy, I have a friend of mine who sometimes talks about like, I don't know, he would be involved in planning workshops. And sometimes like the one person who was always like, we should have more structure. And the other person who's like, nah, we should talk to each other more. And there's some impulse to be like, okay, let's like really get on the same page about this. Let's put you in a room and like resolve it. And I think sometimes that can be right, but given practical constraints, I think sometimes it just makes more sense to have one, to have the group kind of split and you argue it out and like, whatever.

Shea Levy (43:06)

I do think the group versus individual thing is... No, no, I know, know, know what I'm saying. I guess I think the analogy fails, I guess. This is my...

Divia (43:10)

I said this analogy. Yeah.

Okay, I think it doesn't. I think it makes sense if there are certain patterns in me that are like, that are more attuned to some things. You can make it more concrete.

Shea Levy (43:23)

Can you like, are you able to give a question? Yeah.

Divia (43:28)

Yeah, let me see. I, I don't know. I remember, so this is like a parts type description of my life. can critique it if you want. So about a little more than a year ago when I was moving, I was pregnant. It was just like a lot going on. And I remember was talking to my friend and I was like, yeah, some part of me is like sort of screaming at me that this is not okay. And we're like, okay, well, some like assumption that like, well, that must mean there's something better than that, right?

And I sort of sit down and try to like introspect. I'm like, okay, I think this part of me is sort of like, look, it's not my job to figure out what the, what thing would be better here. It's just like locally, like this part of your mind really doesn't like it. So it's like a request from allocation of more resources towards making it not like this. And ultimately like, I think this, you know, this move to introspect and whatever, it's sort of like an integrative move. But the part where some part of me is like, this is more like, I'm in charge of looking at this part of your mind and being like, are things okay over here? And it's like, no.

That seems fine to me.

Shea Levy (44:25)

So do you think there is actually some part of your processing which is the same thing that is caring about that same issue? That's the piece, I get in the moment.

Divia (44:38)

issue. I don't know about issue, but I think there's certain contours. Like I think I have different types of, I could try to be concrete, but it's hard. Like I think I have different types of processes and that that makes sense because like, I don't know, I, for example, like I certainly have visual processing, right? Like do you use a really low level example? I think sometimes it's true that like my visual processing is kind of tied.

because I've been looking at really fine details or something. And so I think there is a relatively persistent part of my mind that's like, hey, part of you is like, give the eyes a break right now. Or something like other things like that.

Shea Levy (45:11)

Yeah, that's how I'm on board with I'm on board with I think the place that I'm less clear on is like,

different persistent carers about or attendance to different values. That's where I am like, yeah, so like definitely different kinds of processing. Like sometimes I can think like do like deep abstraction or like, you know, very like thorough, like thorough reasoning through a problem. And sometimes a different process is like very generative thinking of just like,

Divia (45:28)

Mm-hmm.

Shea Levy (45:50)

coming up with ideas that might solve something and that's not the same thing as checking all the boxes, right? Those are different. Completely unborn with that.

Divia (45:52)

Yeah.

And I think there will persistently be different. Like, I don't know, maybe someone somewhere could integrate them, but it doesn't seem like a good idea to me. Right, they're not in conflict. Though they could be locally in conflict at some time, potentially.

Shea Levy (46:02)

they're not in conflict. They're not integrated. that's why I wouldn't call it.

I mean, think it's more, the conflict is not between, it's not like those processes aren't conflict. The conflict is which of those processes should I use? It's not