Hello,

I'm Felix Felicis. I named myself after the drinkable manifestation of Plot Armour, this strange ability for heroes to get away with outrageously improbable plans over and over again because of authorial fiat, thus allowing them, and the audience, to escape the difficulty of utilitarian calculations and actual rational thinking (not strawman-rational thinking like Spock always does).

I was sent here from Less Wrong, looking for a community that shared the ideals but which had a curated forum system rather than the incredibly questionable Karma System.

I've also got a fanfic I would like to shill for a friend, which I bring to you because I suspect you might quite like it. It could be summed up as "Ponies + MMORPG + AI => The Singularity => ????" What part of the forum is most appropriate to post and discuss it?

There's also another topic I'd like to tackle, that of the Pirate Pary and Liquid Democracy, as well as the ides of adhocracy. Which would be the most adequate subforum to do so?

Hi, Felix! It's good to have you here!

Felix Felicis wrote:I was sent here from Less Wrong, looking for a community that shared the ideals but which had a curated forum system rather than the incredibly questionable Karma System.

I look forward to hearing more about your ideals. It's good to have like-minded folk here, but it's also good to have some healthy disagreement and dissent!

Felix Felicis wrote:I've also got a fanfic I would like to shill for a friend, which I bring to you because I suspect you might quite like it. It could be summed up as "Ponies + MMORPG + AI => The Singularity => ????" What part of the forum is most appropriate to post and discuss it?

I've heard of that before; but haven't gotten around to reading it yet. "General Discussion" makes the most sense for talk about it.

Felix Felicis wrote:There's also another topic I'd like to tackle, that of the Pirate Pary and Liquid Democracy, as well as the ides of adhocracy. Which would be the most adequate subforum to do so?

I look forward to hearing more about it. "Applied ethics" is the go-to subforum for almost all politically-related topics.

I find our lack of activity depressing.

Felix Felicis wrote:I find our lack of activity depressing.

This is actually a surprisingly active forum. Give it a little bit of wait. Usually people have to take time to think through what they want to say in response, if anything, which can take a long time. And people are busy this time of year.

It might help if we had a few dozen active members rather than the current dozen or so... Maybe we could recruit? There's a whole bunch of Facebook groups with ongoing utilitarian conversations if you're interested in any of that!

Or maybe you could search through our archives and find more things you want to offer your thoughts on!

Which groups would those be?

And I'm finding the archives quite interesting.

Welcome. I do wish there was more activity here too but I value the activity we do have. I think this is the best morality forum on the web. Recruiting is a good idea but I think you can stimulate more activity by posting more yourself. The more you post, the more likely our threads are to show up in someone's Google searching. Also, the more you post, the more others post in response, etc.

Felix Felicis wrote:Which groups would those be?

There's the Utilitarianism group, but I'm pretty sure it's private and you need to be invited (we can hook you up!). There's also a Effective Altruists group, a Effective Animal Activism group, and a Reducing Wild Animal Suffering group.

peterhurford wrote:

Felix Felicis wrote:Which groups would those be?

There's the Utilitarianism group, but I'm pretty sure it's private and you need to be invited (we can hook you up!). There's also a Effective Altruists group, a Effective Animal Activism group, and a Reducing Wild Animal Suffering group.

I didn't know about the utilitarianism group. I just sent a request to join.

Who actually runs that util group? We really need to open it!

I also keep meaning to post a poll there asking what realistically would persuade more of them to have their discussions here. It seems like a total waste to have them on FB where once they become inactive they're gone for good, so it should be a moral imperative for in-depth discussions to be held here instead

Moral imperatives are not universally compelling.

By the way, what' the general opinion here on Eliezer Yudkowsky?

Also, any bronies?

Felix Felicis wrote:Moral imperatives are not universally compelling.

Well no, but a) who claimed they were? I'm just interested in persuading some >0 proportion of the FB group to come here, which doesn't entail persuading all of them, just as many as possible, b) Eliezer's argument seems to be (from a skim) just that no argument however logical won't persuade everyone, which doesn't seem particularly related to normativity, and c) I don't actually believe in moral imperatives, hence the winky.

By the way, what' the general opinion here on Eliezer Yudkowsky?

Don't think there is one. He gets quoted approvingly quite often, but most of us would probably take issue with the vagary of his not-quite-utilitarian ethics. Views about his actual effectiveness are empirically contingent and we don't all have the same information, so they'll diverge a lot.

I'm probably one of the more critical, for various reasons:

i) I find CEV quite tiresome - I can't see anything in it that isn't a reinvention of preference utilitarianism, which seems significantly inferior to hedonistic util and has been around for decades in any case.
ii) Many of Holden Karnofsky's criticisms of SIAI seem to apply directly to him, esp his not having actually achieved anything demonstrable.
iii) He's at the very least a poor diplomat, who seems to have alienated many of the very AI researchers and analytic philosophers who he needs to be winning over if his views are correct, with his open disdain for academia.
iv) I find his writing style (can't think of quite the right adjective) obtuse. Reminiscent of Tycho Brahe from Penny Arcade, in that he superficially seems both readable and erudite, but that having read arbitrary amounts of his writings I've never felt like I've learned much. In which case, his readability might actually be harmful, if it causes a lot of otherwise productive people to spend a lot of their time reading stuff for little benefit.
v) His very popularity makes me wary - there are a lot of intelligent people writing a lot of important things, but none seem to have attracted such a devout following, which makes me suspect he has charisma more than veracity to thank for his success.
vi) His Timeless Decision Theory (and come to that the very concept of decision theory) has never seemed at all persuasive to me - it doesn't seem to serve any purpose which a more elegant combination of expected value calculation with one's own form of utility plugged doesn't achieve more elegantly.
vii) He generally seems far far more confident than any rational human being should be in his own views - despite writing constantly on that very subject.
viii) He seems to basically ignore population ethics, disregarding (as far as I've seen) for eg the idea that death by AI might be a good thing, either because of negative-utility futures or because the AI might be (or might eventually become) a genuine utility monster.
ix) I can think of numerous other people who write on similar subjects but do so more convincingly to my mind, and certainly more rigorously (eg Alastair Norcross on theoretical ethics, John Broome on population ethics, Toby Ord on various cost-benefit analyses, Nick Bostrom on existential risk - Bostrom even though he makes some fairly basic errors of analysis and style), so it's not a choice between him or nothing
ETA x) His handling of the Roko affair

Also, any bronies?

Well, that's expanded my vocabulary. Not sure how often I'll be able to wield the new tool, though...

Arepo wrote:Who actually runs that util group? We really need to open it!

I also keep meaning to post a poll there asking what realistically would persuade more of them to have their discussions here. It seems like a total waste to have them on FB where once they become inactive they're gone for good, so it should be a moral imperative for in-depth discussions to be held here instead

Agreed. I know it's been talked about on the group in the past, and it reminds me of Brian Tomasik's "Why I Prefer Public Conversations".

IDeally, I'd want everything done here on Felicifia because the UI is better. But I think there are some utilitarians there that are wary of Felicifia over disagreements about whether animal welfare and donations matter. I think this is all for the best though, because I don't want us to be an echo chamber.

Ryan Carey is the guy who runs that place. He's also a Felicifia admin.

~

Felix Felicis wrote:By the way, what' the general opinion here on Eliezer Yudkowsky?

I basically agree with what Arepo wrote about Yudkowsky, though I do like LessWrong and awful lot and think Yudkowsky has a lot to contribute to philosophy. I think reading his work was quite helpful in becoming an atheist utilitarian.

~

Felix Felicis wrote:Moral imperatives are not universally compelling.

You'd be surprised at the number of anti-realists here.

Though I do want to add, as seen in my sig, that I think Yudkowsky captures fairly well my current tensions with utilitarianism over what exactly it means to be happy and what exactly we're supposed to optimize. I'm unworried because it doesn't seem to matter outside obscure hypotheticals, but when they come to wirehead all of us, I might be concerned.

peterhurford wrote:

Arepo wrote:Who actually runs that util group? We really need to open it!

Felix Felicis wrote:Moral imperatives are not universally compelling.

You'd be surprised at the number of anti-realists here.

What's an anti-realist?

Arepo wrote: I find CEV quite tiresome - I can't see anything in it that isn't a reinvention of preference utilitarianism, which seems significantly inferior to hedonistic util and has been around for decades in any case.
ii) Many of Holden Karnofsky's criticisms of SIAI seem to apply directly to him, esp his not having actually achieved anything demonstrable.
iii) He's at the very least a poor diplomat, who seems to have alienated many of the very AI researchers and analytic philosophers who he needs to be winning over if his views are correct, with his open disdain for academia.
iv) I find his writing style (can't think of quite the right adjective) obtuse. Reminiscent of Tycho Brahe from Penny Arcade, in that he superficially seems both readable and erudite, but that having read arbitrary amounts of his writings I've never felt like I've learned much. In which case, his readability might actually be harmful, if it causes a lot of otherwise productive people to spend a lot of their time reading stuff for little benefit.
v) His very popularity makes me wary - there are a lot of intelligent people writing a lot of important things, but none seem to have attracted such a devout following, which makes me suspect he has charisma more than veracity to thank for his success.
vi) His Timeless Decision Theory (and come to that the very concept of decision theory) has never seemed at all persuasive to me - it doesn't seem to serve any purpose which a more elegant combination of expected value calculation with one's own form of utility plugged doesn't achieve more elegantly.
vii) He generally seems far far more confident than any rational human being should be in his own views - despite writing constantly on that very subject.
viii) He seems to basically ignore population ethics, disregarding (as far as I've seen) for eg the idea that death by AI might be a good thing, either because of negative-utility futures or because the AI might be (or might eventually become) a genuine utility monster.
ix) I can think of numerous other people who write on similar subjects but do so more convincingly to my mind, and certainly more rigorously (eg Alastair Norcross on theoretical ethics, John Broome on population ethics, Toby Ord on various cost-benefit analyses, Nick Bostrom on existential risk - Bostrom even though he makes some fairly basic errors of analysis and style), so it's not a choice between him or nothing

i)What's hedonist util?
ii)Well, we have people like Carl Sagan, Douglas Hofstadter, Feynman, Dawkins and so on, on whose actual academic achievements I couldn't possibly comment, but whose ideological work in interpreting scientific discoveries and spreading those interpretations has been invaluable. I think EY is a much more effective teacher (or even preacher) than researcher.
iii)He seems to have improved his diplomacy skills considerably in the last few years, but if it hadn't been for his brazen arrogance, his work wouldn't have been as stimulating and impactful for me as it would have been if he'd been more prudent and moderate. Wet blankets don't get you fired up. Also, universities are terrible learning institutions and it was time someone went out and said it, especially someone who stands to lose from doing so.
iv)Maybe if you're already familiar with the concepts, but I had to learn everything from scratch, and it literally turned my world upside down. If I had been presented with more straighforward, condensed material (like he's been doing in his latest sequence, which aims to be more academic and emotionally neutral), I would have dropped it, bored out of my skull.
v)That seems like fallacious thinking. What bearing does the reason for his success have on the actual merits of his teaching? Does an author need to be impopular and obscure for you to trust that there is no foul play?
vi)What do you mean, "the very concept of decision theory"? "expected value calculation"? "one's own form of utility"? "elegance"? I'm not being difficult for the sake of being difficult, but I really want to make sure that I understand you fully.
vii)If confidence in one's own expectations correlates well with the rate at which reality meets them, what's wrong with them being strong?
viii)What's population ethics?
ix)Never heard of them. Which might be the point; if it weren't for him, I would never have found out about this site, or about those authors, and "utilitarianism" might still mean "John Stuart Mill" in my head, which is very strongly tied to the tag "boring platitudes that completely miss the point", instead of "a sense of morality that bypasses mere common sense and actually tries to apply ethical considerations to their final consequnces".
x)What's the Roko affair?



Also, please go easy on me with the jargon. Do keep in mind that I'm a newbie at the "utilitarian" tribe? There's been other groups I tried to join with repelled me because of a divide in common knowledge that we did not share, that they assumed I should know, and that they looked down on me for not coming to them with that knowledge prepackaged. If I'm going to go through a similar experience, I'd like to be warned in advance; those tend to leave me with a horrible taste in my mouth.

There's also one thing that's been bothering me. Ever since I've started spening time on the Anglosphere internet, there's been a conspicuous drought of French, Spanish, German, Italian, Japanese, Indian, Chinese... names, when it comes to authors referenced. It's not just that they work in English-speaking academic institutions; they also have anglo-saxon sounding names. There's also the unbridled contempt for "continental philosophers" which strikes me as strange. It's almost like it's not just a lanugage barrier problem, but outright deliberate dismissal of anything not produced by and within the Anglosphere. If this is true, it doesn't really strike me as the most efficient way to gather knowledge or spread ideas.

To clarify, I'm not out to harm Eliezer's mission or prevent him from achieving most of his goals. You asked what our views were, so I gave you mine, which in a nutshell are that he's probably a force for good, but not so much that I would support him or his work over a number of other high-expectation causes. I'll clarify a couple of things below, but I don't think a long discussion of where Eliezer fits on the goodness spectrum is worthwhile.

Felix Felicis wrote:
i)What's hedonist util?

Sorry, I assumed you were more familiar with util. Hedonistic util(itarianism - sometimes just HU) is just the subset of util that defines 'happiness' (ie positive emotion) as utility, rather than the satisfaction of preferences (ie preference utilitarianism or PU). Pretty much all utilitarian thought can be partitioned into either HU or PU, though there are a number of other clarifiers one can add, which I won't go into here.

v)That seems like fallacious thinking. What bearing does the reason for his success have on the actual merits of his teaching? Does an author need to be impopular and obscure for you to trust that there is no foul play?

It's evidential thinking ('fallacious thinking' doesn't really exist except in the mind of philosophers, IMO. An argument either has evidential significance or doesn't, and appealing to discrete fallacies is unlikely to demonstrate which). If a boring person persuades several people of their views, it's more likely that their prominence is due to their views being well-grounded (yes there are other reasons, but that's still one of them) than if a charismatic person does. For eg, I find Peter Singer, at least in a traditional sense, quite uncharismatic, yet he's still had a huge influence in person as well as in writing, which (if I didn't know much about him) I'd think spoke well of his views. If someone is particularly exciting, it should reduce our expectation of the accuracy of their message.

vi)What do you mean, "the very concept of decision theory"? "expected value calculation"? "one's own form of utility"? "elegance"? I'm not being difficult for the sake of being difficult, but I really want to make sure that I understand you fully.

Decision theory is not my puppy, and one of my main complaints about it is its lack of definition, so I suggest just Googling that one. The expected value calculation of a decision is the sum of (each possible outcome * the probability of that outcome). So a 100% chance that I will give you £10 is worth the same amount (of money) to you as a 2% chance of me giving you £500 (assuming the only other possibility is you get nothing).

'One's own form of utility' relates to eg PU vs HU - so my approach as an advocate of HU would be to value a decision based on its expected hedonic value, and decide whether to do it based on the expected hedonic value of the alternative decisions I have available to me (in practice I'm not unselfish enough that it's so straightforward, but that's the basic idea. I can use a similar calculation while thinking selfishly if I replace 'HU' with 'my own happiness').

By elegance I suppose I just mean parsimony.

vii)If confidence in one's own expectations correlates well with the rate at which reality meets them, what's wrong with them being strong?

It's not clear how much Eliezer's expectations do correlate with reality. What successful scientific predictions has he made?

viii)What's population ethics?

It just refers to ethical questions addressing entities that don't currently exist but will/might in various conceivable futures. So it's not separate from consequentialist ethics, it's just an area of concern that they might address.

ix)Never heard of them. Which might be the point; if it weren't for him, I would never have found out about this site, or about those authors, and "utilitarianism" might still mean "John Stuart Mill" in my head, which is very strongly tied to the tag "boring platitudes that completely miss the point", instead of "a sense of morality that bypasses mere common sense and actually tries to apply ethical considerations to their final consequnces".

I think you're underestimating counterfactual thinking - ie what the world would have been like had Eliezer not written all this stuff.

There would still be plenty of stuff written about rationality and utilitarianism on the internet, Felicifia would almost certainly exist in some form, possibly more visibly since its current state is dwarfed by LW, and there would have been a gap in popular explanation of statistical reasoning that someone else might have filled. Meanwhile, millions of intelligent people's working hours would have been saved, possibly those people would have had less of a sense of moral or probabilistic clarity, but have had a lot more time to direct into things like actual scientific research etc.

I'm not saying that would be a better world, just that it's far too easy to think 'I can see some benefits that I've gained from X, therefore I'd be worse off had X not happened'. I recently read something about this bias, which I've now forgotten, annoyingly - it's an interesting subject.

x)What's the Roko affair?

I read about it here: http://rationalwiki.org/wiki/LessWrong# ... s_Basilisk (I don't particularly recommend the article in its own right, but it gives you some context about the string it referenced that you might want to read before following the reference, esp if you do find yourself agreeing with Yudkowsky in general.

Also, please go easy on me with the jargon. Do keep in mind that I'm a newbie at the "utilitarian" tribe? There's been other groups I tried to join with repelled me because of a divide in common knowledge that we did not share, that they assumed I should know, and that they looked down on me for not coming to them with that knowledge prepackaged.

Not a problem - like I said, I just misestimated how familiar you'd be with all this stuff from your OP.

There's also one thing that's been bothering me. Ever since I've started spening time on the Anglosphere internet, there's been a conspicuous drought of French, Spanish, German, Italian, Japanese, Indian, Chinese... names, when it comes to authors referenced. It's not just that they work in English-speaking academic institutions; they also have anglo-saxon sounding names. There's also the unbridled contempt for "continental philosophers" which strikes me as strange. It's almost like it's not just a lanugage barrier problem, but outright deliberate dismissal of anything not produced by and within the Anglosphere. If this is true, it doesn't really strike me as the most efficient way to gather knowledge or spread ideas.

There are a couple of considerations here:

1) The anglophone world has a higher concentration of tertiary institutions, so a anglophones will make up a disproportionate amount of academia.
2) Translation issues are certainly a factor - we're more likely to have read things in our own language, and our friends are more likely to read things we've read, so our bias (such as it is) compounds.
3) 'Continental philosophy' is not 'philosophy produced in Europe', per se. It's a (not well defined) subset of philosophy identified mainly by its lineage (diverting from analytic philosophy between roughly the early 19th and early 20th centuries). Figures like Kant, Hegel, Wittgenstein are sort of pivots marking various points of diversion without being meaningfully called either analytic or continental (though Hegel leaned towards the latter). Obviously there's been a fair bit of discussion between them, but for the most part, after Wittgenstein's time, analytic philosophers tended to engage with analytic philosophers and continentals with continentals.

It's also characterised by being quite obscurantist - it's common for continental philosophers to spend a large part of their career interpreting the writings of other continentals, whereas you don't really get the same phenomenon in analytic phil, which is typically blander and more technical but clearer.

The reason I ignore it is mainly the same as I ignore most (but not all) non-utilitarian analytic philosophy (but more so) - because of its difficulty and the lack of any indication that one gains anything other than personal satisfaction from persevering with it, it has low expected value compared to, say, reading a maths textbook, which is difficult but provides personal satisfaction and actually has demonstrable effects on understanding the world around me, or just reading a novel/watching a film etc, which is not difficult and has similar hedonic benefit.

More anecdotally, it just seems to be full of unpleasant people whom I don't want to dignify with the epithet 'philosopher' - Schmitt and Heidegger were (to differing degrees) Nazi supporters, from what I understand of Sartre and De Beauvoir they were the worst kind of philanderers, who used to sleep with people just to score points against the other regardless of who it hurt, Lacan was close to being a cult leader, Zizek intermittently supports Stalinism when he's not contradicting himself, Terry Eagleton, seems to pick on people who advocate views he doesn't like and write full 'essays' and so on that consist mainly of abuse and belittling their conclusions (see 'Lunging, Flailing, Mispunching', and I think he wrote something similar on Hofstadter's I Am a Strange Loop, which I can't now find). From personal experience also in an English dept, the traditional hub of a lot of continental philosophers, disagreement with their views could get you ostracised and abused, especially if you disagreed with their politics, and I could add to that a few other - albeit not especially serious - personal experiences along similar lines that I can't really discuss in a public forum.

While I'm sure I'm cherry-picking to some degree, and there have doubtless been numerous unpleasant analytic philosophers, I have yet to encounter an unpleasant hedonistic utilitarian, people generally seem more pleasant the closer they are to HU and indeed most of the leading lights of utilitarian philosophy (eg Peter Singer, Toby Ord) really do seem close to being moral saints. (and believe me, the sceptic in me winced at least as much writing that as you will reading it). I actually can't think of an example of a well-known purely analytic philosopher who's famously unpleasant.

If (as I do), you think the main significance of philosophy is in ethics, then unlike in science (where Newton being an excretion of a human being did not make him much less of a scientist) the character of people who espouse a view is quite relevant to how seriously you might take that view, at least a priori. So while I have not read that much continental phil, I feel no more need to educate myself further on it than I do to read the Book of Mormon or Scientology scripture.

I don't think a long discussion of where Eliezer fits on the goodness spectrum is worthwhile.

Honestly, I'm not so much interested in placing EY's Character Alignment as I am in reducing the gap in priors between me and the community here. Yudknowsky's just a starting point to navigate ideas.

PU and HU: Isn't PU inclusive of HU? Isn't it tyrannical to impose happiness on people who would prefer other things over it, such as achievement?

Decision theory is not my puppy, and one of my main complaints about it is its lack of definition, so I suggest just Googling that one. The expected value calculation of a decision is the sum of (each possible outcome * the probability of that outcome). So a 100% chance that I will give you £10 is worth the same amount (of money) to you as a 2% chance of me giving you £500 (assuming the only other possibility is you get nothing).

Well, that seems legit as long as you don't think predictability is useful in and of itself in furthering your goals. Which it sometimes is, and sometimes isn't; isn't that factor quantified as "risk aversion"?

'fallacious thinking' doesn't really exist except in the mind of philosophers, IMO. An argument either has evidential significance or doesn't, and appealing to discrete fallacies is unlikely to demonstrate which

Could you please elaborate on that?

"Uncharismatic people who are still listened to are more likely to be listened to on the merits of the contents they release rather than on the form of them." Seems legit, I'm just not sure the inverse is true. There are, after all, many charismatic people competing for our attention. I don't think "popular, therefore suspicious" is a good heuristic, unless you fear that the charisma affects your own ability to judge.

By elegance I suppose I just mean parsimony.

I see. Some people, by "elegance", mean "wit", "brevity", and so on, failing to take in account the hidden complexity, the wealth of assumptions and shared priors, that the witty phrase or short equation or simpl(istic) explanation requires to function. The notion that you are referring to, I for one prefer to call it "simplicity" or "straightforwardness". One of the reasons I abandoned theism was because atheism made moral calculations much more straightforward, and saved me much anguish.

What successful scientific predictions has he made?

You mean original scientific predictions? Because what I'm praising here is the confidence he displays in that particular brand of Bayesian empirism, which is hardly his invention. As someone who's been raised with the understanding that science should be humble in the face of religious authority, this was quite a shocker, especially because of the way he justified it (most of the time, neo-atheists give off, to the external observer, an aura of "arrogance" and "aggressiveness", rather than legitimate, crushing, unwavering confidence, which seemed to me to be a monopoly of the truly-faithful religious).

It just refers to ethical questions addressing entities that don't currently exist but will/might in various conceivable futures.

Sounds counterintiutive. I thought it meant something like sociology, or something to do with contraception, healthcare policies, and demographic control in general. Who gets to live, who gets to die, should we send aid to Africa, etc.

I think you're underestimating counterfactual thinking - ie what the world would have been like had Eliezer not written all this stuff.

I don't disagree with what comes after, but I have trouble with this sentence in particular. The first bit doesn't seem to relate to the second.

Hey Felix. So I share the ideals of the LW community. I think EY is a genius-writer.

I think the Friendly AI research performed by the Singularity Institute makes sense. Basically, you just have to prepare to engage with the arguments.

RyanCarey wrote:Hey Felix. So I share the ideals of the LW community. I think EY is a genius-writer.

I think the Friendly AI research performed by the Singularity Institute makes sense. Basically, you just have to prepare to engage with the arguments.

I find that genius is a rather useless qualifier. It boils down to "intelligent and accomplished person whom I cannot understand". Let's not worship our own ignorance.

By the way, what' the general opinion here on Eliezer Yudkowsky?

I'm not qualified to evaluate his ideas, but I would like to register two stylistic grievances. First, why oh why does he use two spaces after terminal punctuation marks? Never mind, he stopped doing this. He is forgiven. But not totally, since he did it for so long. Second, why does he capitalize certain phrases, such as "Very Serious People"? Is it his version of scare quotes? I find it annoying.

Other than that, he's smart and clever. He could work on his public-speaking skills. He needs to take some lessons from Joe Biden or something, i.e., he doesn't come across as very likeable in the flesh.

EDIT: It is Paul Krugman, not Eliezer Yudkowsky, who uses the phrase "Very Serious People." How'd I mix that up? They're both charismatic leaders of sorts, but still. Maybe Mr. Yudkowsky doesn't engage in idiosyncratic capitalization after all. I could've sworn… So he's off on both counts.

@Arepo: could it be that Utilitarianism is actually a rationalization of kindness, that kind philosophers gravitate towards it, rather than utilitarianism making people kind?

PU and HU: Isn't PU inclusive of HU? Isn't it tyrannical to impose happiness on people who would prefer other things over it, such as achievement?

Not sure what you mean by ‘inclusive of’. Certainly some people might argue that properly understood there’s no difference between the two, but some advocates of each will define their versions in ways that are clearly mutually exclusive.

What you’re arguing sounds a lot like Luke Muehlhauser’s argument that because we can show that people don’t always seek happiness, HU is wrong (or maybe contraindicated, if you want a less philosophically loaded term).

I think it’s a very weak argument, since it’s essentially circular. If we judge against HU on the grounds that it contradicts our preferences, we have already tacitly accepted PU, since we’re making a preference-based value judgement. It’s equivalent to saying PU fails because it sometimes makes us unhappy.

Note also that on these terms, all common forms of PU fall on their own sword. Most advocates of the philosophy support some version more complicated than ‘having any preference satisfied is always a good thing’, to avoid conclusions like giving drugs to desperate junkies or fulfilling a preference that you hold based on a factual error (eg opening a window for you so you can jump out of the third floor of a building you think is burning rather than just pointing out to you that someone’s burnt their toast in the other room, and you’re actually perfectly safe). But clearly – by definition – such nuanced versions of PU conflict with what people actually want, so isn’t it tyrannical to impose on people their own nuanced preferences rather than the ones they actually hold?

To resolve the issue we need to somehow look beyond the value judgements of each form and somehow assess them without invoking them. How to do that is a long discussion for another thread.

Well, that seems legit as long as you don't think predictability is useful in and of itself in furthering your goals. Which it sometimes is, and sometimes isn't; isn't that factor quantified as "risk aversion"?

We need to be clear on what we mean by ‘goals’ here. If you have a strictly quantifiable goal, such as ‘maximise utility’, then risk is irrelevant – you follow whatever path leads you to maximal expected utility. If you feel like risk aversion is fundamentally important then you’re actually making it part of your goal – which is then extremely hard to quantify. If you just literally mean it’s useful to achieving your goal, then you’re still ultimately running an expected value calculation, in which predictability is just one of the whole array of factors which you’re having to interpret. But I haven’t seen an example of decision theory helping you do so in a way that isn’t just mathematical analysis + value judgement.

Could you please elaborate on that?

Forget I said it. I have a bad habit of throwing in bugbear issues to conversations which they just complicate.

"Uncharismatic people who are still listened to are more likely to be listened to on the merits of the contents they release rather than on the form of them." Seems legit, I'm just not sure the inverse is true. There are, after all, many charismatic people competing for our attention. I don't think "popular, therefore suspicious" is a good heuristic, unless you fear that the charisma affects your own ability to judge.

I think it’s trivial to infer that if being charismatic makes you more likely to be popular, being charismatic and advocating a popular view makes that view less credible, ceteris paribus, than being uncharismatic and doing so. It might not be that strong an effect though.

I see. Some people, by "elegance", mean "wit", "brevity", and so on, failing to take in account the hidden complexity, the wealth of assumptions and shared priors, that the witty phrase or short equation or simpl(istic) explanation requires to function. The notion that you are referring to, I for one prefer to call it "simplicity" or "straightforwardness". One of the reasons I abandoned theism was because atheism made moral calculations much more straightforward, and saved me much anguish.

The principle of parsimony is usually described that way too – ‘the simplest explanation is usually the best one’. But parsimony is a better defined concept than simplicity.

You mean original scientific predictions? Because what I'm praising here is the confidence he displays in that particular brand of Bayesian empirism, which is hardly his invention. As someone who's been raised with the understanding that science should be humble in the face of religious authority, this was quite a shocker, especially because of the way he justified it (most of the time, neo-atheists give off, to the external observer, an aura of "arrogance" and "aggressiveness", rather than legitimate, crushing, unwavering confidence, which seemed to me to be a monopoly of the truly-faithful religious).

I don’t understand how this answers my question. You seem to be saying his confidence is justified because he’s really confident. He’s picked some pretty strong targets for criticism – scientists, AI programmers etc. And he seems just as cocksure in his criticisms of them as of easier targets. But they use statistical analysis too, and unlike him they often do stuff which produces demonstrable results. Just proclaiming the value of what he understands as Bayesianism doesn’t constitute any kind of evidence in my book.

I think you're underestimating counterfactual thinking - ie what the world would have been like had Eliezer not written all this stuff.

I don't disagree with what comes after, but I have trouble with this sentence in particular. The first bit doesn't seem to relate to the second.

What doesn’t relate to what? As far as I can see what I’ve written here makes sense.

@Arepo: could it be that Utilitarianism is actually a rationalization of kindness, that kind philosophers gravitate towards it, rather than utilitarianism making people kind?

That’s probably a large part of the effect (at least the second bit. I don’t see it as a rationalisation in the sense of concealing ulterior motives). But it doesn’t really change my view – while I wouldn’t discount the possibility that naturally unkind people can just figure out something that’s more coherent about ethics than kind people have managed, I’d expect them a priori to have much lower chance of doing so, since they’re the ones more likely to be seeking rationalisations so they’re less likely to be as scrupulous in their analyses.

(it might not help here that ‘kind’ is a slightly vague or loaded concept, possibly reliant on one’s value system. It’s more helpful IMO to think of the degree to which people are fundamentally interested in themselves and their own wellbeing.)

Pat wrote:First, why oh why does he use two spaces after terminal punctuation marks?

What's wrong with two spaces after terminal punctuation marks?

Felix Felicis wrote:PU and HU: Isn't PU inclusive of HU? Isn't it tyrannical to impose happiness on people who would prefer other things over it, such as achievement?

As someone who is a little wary of living in a "utopia" where I'm wireheaded, I sympathize with the idea that "happiness" (or some odd version of it I'm not yet prepared to accept) can be imposed tyrannically. But on the same token, preferences aren't much help either:

1.) How do I make sense of preferences that I have, but don't want (don't meta-prefer) other people help me satisfy them? I want to quench my thirst all by myself, thank you. For some preferences, it would make me happier if (and I would meta-prefer that) you helped me, but for others it would make me sadder (and I would meta-prefer you refrain).

2.) How do I make sense of the idea of creating a preference for the sake of fulfilling it? Would it be a moral benefit to make you thirsty just to offer you a drink? Doing so would satisfy your preference, but wouldn't make you any happier.

3.) How do I make sense of preference satisfying population ethics? What good would it be, if at all, for someone to be born, who wouldn't be born otherwise? What good would it be, if at all, to launch a utilitronium shockwave? Is this just creating beings to fulfill their preferences?

4.) Is the best life one in which all our preferences are satisfied? Would an ideal lifeform be one with just a single preference that is satisfied? What about a life form with 1000 preferences, all of which are satisfied?

peterhurford wrote:

Felix Felicis wrote:PU and HU: Isn't PU inclusive of HU? Isn't it tyrannical to impose happiness on people who would prefer other things over it, such as achievement?

As someone who is a little wary of living in a "utopia" where I'm wireheaded, I sympathize with the idea that "happiness" (or some odd version of it I'm not yet prepared to accept) can be imposed tyrannically.

Have you heard of the experiment (think it might have been casual) where various people are asked if they would want to plug into Nozick's experience machine, with a decent majority saying 'no', and then a separate group are asked if they found out they were in such a machine and they were offered the chance to unplug, many (most?) also declined?

That is precisely why we don't want to be plugged in in the first place; we don't want to find ourselves stagnating in a local maximum, unable to summon the will to move out and explore new options, never knowing if there aren't greater maxima.

Arepo wrote:Have you heard of the experiment (think it might have been casual) where various people are asked if they would want to plug into Nozick's experience machine, with a decent majority saying 'no', and then a separate group are asked if they found out they were in such a machine and they were offered the chance to unplug, many (most?) also declined?

I've heard of it -- see this article for the argument you mentioned (PDF). I don't have any objection to being in a sufficient experience machine -- I was objecting to having my brain rewired to find optimum pleasure from having elbows or counting grass.

(This doesn't count the objection that being wireheaded or in an experience machine would prevent me from being able to prevent suffering here where it exists. Which, come to think of it, is strange because it is a revealed preference that I prefer more to prevent actual suffering than be in an experience machine. I've touched on this before.)

On EY striking at philosophical institutions... it's not just him

I fully agree that philosophy has serious problems - as does every philosopher I've ever asked. The challenge is establishing why *your own* solution to its problems is the one that people should adopt.

For what it's worth I have a lot of sympathy with Muehlhauser's and Yudkowsky's views on what actually is wrong with philosophy. But they're so closely aligned in their views that I don't really take agreement between them as adding much extra evidence to an issue - it would be like thinking voting for Obama became a better prospect if it transpired that Joe Biden thought you should.

It's not just Muehlhauser. LW philosophy appears to unwittingly be part ofa pretty mainstream movement of philosophy.

I fully agree that philosophy has serious problems - as does every philosopher I've ever asked.

Huh. A perpetual "X is the cancer that is killing /b/" type of problem?

Arepo wrote:I fully agree that philosophy has serious problems - as does every philosopher I've ever asked.

There are two very different ways this sentence could be interpreted!

Felix Felicis wrote:It's not just Muehlhauser. LW philosophy appears to unwittingly be part ofa pretty mainstream movement of philosophy.

Sure. I didn't mean to imply that no philosophers would agree with Eliezer's broad position (I know quite a few in Oxford who do), rather that other people write similar stuff better - see the ninth of my criticisms on my earlier list.

I wouldn't agree that Eliezer's philosophy is specifically Quinean naturalism - possibly I'm just misconstruing the subject, but there are a whole bunch of schools of thought that have approached philosophy in quite similar (but often subtly and perhaps importantly different ways) - Sellars The Myth of the Given, modern pragmatists, the Wittgenstein of Tractatus, the logical positivists, fallibilists like Popper etc.

Eliezer's views seem to have a lot in common with them all (inevitably, since they have a lot in common with each other), but as far as I can tell, aren't crisp enough to identify as any particular one.

If you want more philosophically rigorous variations on Eliezer's views regarding existential risk, I recommend reading:
from FHI: Daniel Dewey, Stuart Armstrong, Nick Bostrom, Toby Ord, Anders Sandberg
on LessWrong: current SI researchers Anja and Alex, Luke M, Paul Christiano, Carl Shulman and others...