If it's worth saying, but not worth its own post (even in Discussion), then it goes here.


New Comment
482 comments, sorted by Click to highlight new comments since: Today at 10:52 AM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

Gwern Facts Thread

Because we already have an Eliezer Yudkowsky one and this website is awesome.

Found in Yvain's blog post:

Doesn't this mean that I must be wrong about its excellent safety profile? No. See for example Gwern's research on the subject. About half the people reading this paragraph are going to say "Wait, don't the FDA and the entire decision-making apparatus of the United States government have more data and credibility than one guy with a website?" The other half of the people know Gwern.

Unlike Odin, Gwern has plucked out both his eyes for wisdom, knowing the value of double blindness.

I think we have a winner.

I went to the library and it was empty. They said Gwern stopped by for a quick lookup.

Come on, Gwern deserves more than a favorable comparison to the FDA.

I know several people who have more credibility than the "FDA and the entire decision-making apparatus of the United States government", at least when it comes to drugs. Not because I know so many cool folks, but because drug regulation is a paramount example of government irrationality.


Gwern's reality marble, Unlimited Essay Works, is the original, of which Unlimited Blade Works is a mere copy.

I went on an interview with Google. They told me that if I was hired, I'd be working on a unique innovation. When I asked what it was, they told me "We want to make an app that will search this guy named Gwern."

They went to upload Gwern's brain. They said they couldn't do it, but they were glad that someone had made most of the internet redundant.

In the right-hand navigation bar of the site, there is a "Tags" box which lists the most popular post tags. This box has a feature whereby the tags are rendered in a font size proportional to their frequency of occurrence.

Over time, the attribution of the "sequence_reruns" tag to sequence rerun posts has made this feature inoperative: because it's the single most used tag, and no other tag even comes close in frequency, every tag is now rendered at precisely the same font size except for "sequence_reruns".

You deserve at least 24 upvotes [http://lesswrong.com/lw/ein/open_thread_september_1530_2012/7g25] for pointing this out.
I had no idea it had already been reported, but I see it's been corrected now (ETA: nope! I was confusing Main and Discussion). Updating in favor of being a squeaky wheel.

I had no idea it had already been reported, but I see it's been corrected now.

No, it hasn't. There are two different tag clouds, one for Main and one for Discussion. It's the Discussion one that faces the problem.


You should edit your original comment to indicate this, so that other people don't think it has been corrected.

Strictly speaking it's not precisely the same font size. A few, such as "rationality" and "meta" appear to be one size bigger.
Some tags such as “meetup”, “psychology” or “rationality” do appear to have a slightly larger font size, in my browser at least.

Today my 16yo son asked a classmate for a confidence interval on his grade at their latest assignment, after giving one of his own. That's after all 3 of my kids attended a workshop I gave (mixed in with adults, 20 total) on calibrated probability statements and scoring.

Never mind that he gave a 90% interval and missed (due to getting full marks), I'm inordinately proud of him for actually applying the lessons.

Less-than-sincere modesty is probably the culprit for missing the interval.

Intelligent social signaling is another possible explanation. He's 16, after all. On the other hand, anyone who's giving out or requesting confidence intervals at age 16 is probably not too concerned with social signalling, or else is really bad at it.
It's a perfect social signal. It's costly in that it shows his nerdiness to everyone, but the actual level of nerdiness is impressive to everyone who values nerdiness.
. . . or maybe it's just the manifestation of Impostor Syndrome.
What I was about to write. See also Yvain underestimating his exam results [http://lesswrong.com/lw/7o7/calibrate_your_selfassessments/], people giving very low confidence in the calibration question in the last survey [http://lesswrong.com/lw/8p4/2011_survey_results/] but actually getting it right, etc.
Not inordinately.

I just want to double check something with LWers.

Incest among adults is also sex between consenting adults. At least some such relationships are happy ones. Most arguments against incest are arguments where the bottom line is already written since they are made by people who just don't want to admit they are plain grossed out by it. Not only the motivation, but many of the arguments are basically the same as arguments in favour of homophobia. If the person has an identity centred on fighting "bigotry" cognitive dissonance hilarity ensues.

Bonus round: Arguments against incestuous couples having children is a fundamentally eugenicist argument. Applying it like a consequentalist results in concluding many other kinds of couples should be discouraged just as much (perhaps even with imprisonment since that is the price of discovered incest in many countries) or incest being legalized.

German incest couple lose rights ruling

The ECHR said the main basis of punishment for incestuous relationships was “the protection of marriage and the family”, and because it blurs family roles.

It also noted “the risk of significant damage” to children born of such a relationship.

Incest among adults is also sex between consenting adults.

True, but that's largely a noncentral fallacy.

Most arguments against incest are arguments where the bottom line is already written since they are made by people who just don't want to admit they are plain grossed out by it.

If I'm grossed out by it, why am I watching lesbian incest hentai? :-)

Not only the motivation, but many of the arguments are basically the same as arguments in favour of homophobia. If the person's has an identity centred on fighting "bigotry" cognitive dissonance hilarity ensues.

Agreed. But not all the arguments are basically the same. Some of the arguments are more like the "teachers shouldn't date their students" argument and the "psychologists shouldn't date their patients" or even "50-year olds shouldn't date 20-year olds" argument, and reflect on the likely unhealthy effects of power-imbalance.

The power-imbalance in intergenerational incest is obvious. In intragenerational incest it can of course be significantly less clear; especially in cases like the German couple where the siblings only met during adulthood.

Bonus round: Arguments against ince

... (read more)
I should emphasise this is written in a way to highlight some of the cognitive dissonance I saw in how reasonably intelligent people responded to the story, accepting arguments they would be outraged to hear in a different context. If you've read my comment history you probably know that I approve of eugenics (encouraging some people to have children while discouraging others based on their genetic material). Also I'm sceptical of the coherence of the concept "consent" and think power imbalances can be features not bugs when it comes to humans.
Wait - skeptical of the concept of consent? Like I-consent-to-pay-you-money-in-exchange-for-your-car consent?
I think he's referring to things like the age of consent, where the legal definition of "consent" in some jurisdictions might not cover some things many reasonable people would call "consent".
I'd be willing to bet a small amount that he's talking about one person being dominant over another, rather than dubiousness about age of consent laws.
Right, and excessive+indiscriminate killing power might be a feature not a bug when it comes to weapons, as it might give you peace through deterrence instead of just more slaughter. This doesn't imply that nukes aren't horrifically dangerous and don't have the potential to fuck things up for a long time/permanently. And power imbalances can also be horrifically dangerous and do lasting, pervasive and insidious damage to innocent people. That both categories are here to stay doesn't mean that we'd be wise to get less paranoid about them or relax our vigilance. EDIT: also - That you're against a blanket taboo on "eugenics" doesn't mean that you wouldn't literally kill to to prevent an imminent return of Hitler's "eugenics" cluster, right? Well, of course the difference between you and the mainstream is that you aren't blinded by the "Ancient Lurking Evil" meme of Hitler and don't let it affect your risk/benefit assessment. But you have zero evidence that the meta-category of "power imbalances" contains no Hitler-level lurking horrors, and mountains of 2nd-hand evidence to the contrary! I mean, look - that class of Bad Things is something that every single variation of feminism - some of them being at each other's throats - agrees to be a clear and present danger. Certainly much feminist thinking is fallacious, cranky or in bad faith - but seeing such uncommon, wide-ranging consensus should call for a thorough self-update. Also, I don't see why the concept of "consent" has to be coherent in order to be valuable and useful. Plenty of taboos and moral injunctions that we see are incoherent. And yet many of them (take the American centrist mainstream as an example: "extrajudicial execution is always an atrocity when ordered by a state official, less condemnable when done by soldiers or insurgents", or "preach respect for the law, but stall the enforcement of some laws' letter and spirit") you probably wouldn't want to tinker with!
Nope you've caught me red handed, I totally want to resurrect Hitler and am creating a secret army in my Antarctic base for him to command. There are much much better sources for arguments against power imbalances than feminism, why didn't you pick those? But yes power imbalances can be dangerous and open the field up to terrible abuse, I assumed this goes without saying. I wished to emphasise that certain kinds of power imbalances can be desirable. Indeed it doesn't! But it does mean it isn't universally valid and applicable. I think consent is best understood as relatively strong evidence about a persons preferences. This is a stronger argument than it may seem to the average LWer.
Um... it seemed like too much work, so I intentionally pointed at a source with below-average reputation on LW, then directed attention at how that source handles its case in an unusually reasoned, consistent manner. Which should imply that evidence for it must be plentiful and come easy. ("Patriarchy: so obvious and oppressive that even a feminist could see it!" Sorry.) Also yeah, don't worry, I didn't really assume that you abandoned all prudence here and just looked for something illiberal to say, in order to signal cool metacontrarianism. I have a considerably higher opinion of you :) The problem might simply be that I often argue with your stuff from some weird idiosyncrasic position, while you might do the sensible thing for open debates: write with the average LWer opinion in mind and direct much of your reasoning at it - which might make your points look too skewed to me. You actually want the original? Man, you're too late by far, maybe if you hurry up you could grab a cheap 4th-order Hitler clone with blueprints at some EvilCo sale. P.S.: fun fact, Chesterton criticized feminism because he felt that it was contributing to the destruction of an older, better familial order... that is, Matriarchy in all but name [http://www.gutenberg.org/files/1717/1717-h/1717-h.htm#2H_PART3]! If, like me, you ever felt sick after reading the stereotypical amoral PUA shit about gender, reading him is an antidote; gets the sleaze right out. Chesterton was certainly masterful at opposing any ever-modern "misanthropic" creed. I'm not saying I'd really endorse his arguments, but they're a delight to contemplate.
What do you mean by this? Is feminism disfavored here? If so, in what way?
Sorry, I should've clarified. Make that "publicized modern feminist activists", as some LW readers believe them to be dogmatic and epistemologically unsound, or even unproductive for their own cause. Feminist ideas as such - like all the gender-sensitivity stuff - are widespread here.
Ah, okay. Thanks for the clarification!
Freudian slip?
If incestuous desires are common (certain people think they are at least...), having a harsh prohibition on them might cause a lot of guilt even if those people wouldn't actually go as far as to mate with their relatives. So trying to get rid of the prohibition might still be somewhat valuable. Incest themes are quite common in porn.
That might be in part because guilt and shame can act as huge turn-ons for nearly everyone.
That seems a bit unfair. Why should they have to take particular care about a small-ish increase in risk just because some people are freaked out by them?
It's even more unfair to carelessly bestow genetical deficiencies on one's children. http://en.wikipedia.org/wiki/Patrick_St%C3%BCbing [http://en.wikipedia.org/wiki/Patrick_St%C3%BCbing] - "The older two children suffer from severe physical and mental disabilities. The third child was born with a heart condition but is healthy after undergoing a heart transplant" Actually given further information in that page, and after learning how the woman was still a minor at the beginning of their relationship, I withdraw my hypothetical "applauding" anyway; though at least on 2004 the man finally did undergo a vasectomy. Downvoted: I don't see how you could have legitimately misinterpreted my words to mean that people being freaked out is the reason I offered for their need for birth control.
Lots of downvotes. So do we think there's a good case for coercive eugenics here then? That's got a sort of Schelling-pointy feel to it to me, so that I wonder if prohibiting incest might actually be the lesser of two evils. If a case can be made for coercive eugenics here, where else can it be made? And how does the incest ick-factor influence the arguments? Do people without siblings feel the ick-factor? Or do you need personal experience of the Westermarck effect?
If we're comparing two "coercions" (coercive prohibition of incestuous couples from having children) (coercive prohibition of incestuous couples from having sex), the former seems to be the lesser coercion by far, and the more easily justified. So is it "coercive" part of "coercive eugenics" that really bugs you, or is it that you have an ick-factor against eugenics altogether, voluntary eugenics too?
Whenever I want to trigger a moral intuition that can't be justified by any moral system that doesn't just expressly prohibit it by fiat, I use an example that triggers incest avoidance.
The most common intelligent argument I've seen against incest is "power imbalance!" which in the case of your news story looks like a case of the noncentral fallacy.
In principle, a society could frown upon parent-child incest but not upon incest between siblings, but that's not what we see, so I don't think that's a good explanation.
I think that, however society should choose to treat parental incest, for the sake of consistency and coherency it could be compared to parents giving children legal but strong drugs - whether performance-enhancing or recreational. Obviously any medical expert who's just doing a job and is afraid of being sued would "advise" against either if consulted. Clearly, parents don't have total sovereignity over their children, and most of the "decent" parents wouldn't ask for it anyway while most of the abusive parents would love it. On the other hand, clearly the vast majority of people are hostile to the idea of thorough, case-by-case state intervention, a social worker ordering parents where exactly to draw the line, etc - both for political and pragmatic reasons. But still, there are obviously parents who would, in good faith and with good intentions, want to introduce their child to sex or certain drugs. In such cases, not only are their preferences being unfairly violated, they might be right about it being safe and worthwhile for their child. Is there any way at all to filter those benign cases out from deliberate abuse, dangerous carelessness, etc? I can't think of any. (Discriminating against sibling incest is just as senseless and barbaric as discriminating against functioning drug users or homosexuals, IMO. Siblings should certainly be able to enter the complete, standard, state-sanctioned civil union with all its trappings - whether we rename it from "Marriage" to something else, as some liberals and libertarians propose, or not.)
I have no problem with non-productive incest. I also think that severely malformed embryos should be discarded before they have a chance to develop into a disabled person, effectively resolving the main objection against incest. On the other hand, I do not feel bad about the current incest laws, they seem to function well as Schelling fences.

So, Yvain posted a blog post recently. I was disappointed. I'm posting about it here because I'll have an easier time following a conversation about my thoughts here than in livejournal comments. I will note that he claims the post is, at most, 60% serious, but that seems at least ten thousand times too high.

A major supporting claim is that if modafinil were legal, it would become expected, and everything would be harder to match the increased ability of humans to be productive.

So the religious people flunk out, everyone else has to work much harder, and in the end no student gains. Arguably future patients might gain from having better trained doctors, but I think this wildly overestimates the usefulness of the medical education system.

A parable:

In the Old Country, the people once did not know of iodine. It was not illegal, but only a very specific kind of geek would eat dried seaweed carried long miles on the backs of beasts and men. One day, a stranger came to the village, preaching of this mysterious substance, claiming that its consumption would make all men cleverer.

The elders convened and discussed this 'boon,' if you could call it that. If one man is cleverer, he profit... (read more)

What is his main proposition? He has a model of the world in which enormous amounts of energy and money are being spent running a rat race where the satisfaction only comes from winning it, not from running it, and meanwhile there are numerous places where just a small fraction of that energy and money could be spent, creating great and lasting benefits. His proposition is that in the current situation, modafinil is known mostly to a minority which includes people working on some of those important neglected matters, but if modafinil becomes as well known as Prozac or Viagra, its main consequence will be that the rats in the rat race will all run faster, with no net benefit.

Your comments imply that you disagree with this model, but you need to say where and why.

I think that Yvain's thoughts on the matter are poisoned by working in a poisoned field. Would doctors be better if they studied 16 hours a day, instead of 10? Some, but not much. Perhaps people would live a bit longer- but better for everyone to adopt intermittent fasting than to slightly improve the quality of doctors.

But why only give modafinil to studying medical students, and not those who hold lives in their tiring hands instead of books? Given the hideous prevalence of medical errors, and their known association to fatigue, I would far prefer a doctor chemically warded against fatigue to one without such armor.

(I might agree that financiers all turning to modafinil would not noticeably improve the world, and make them worse off- but, truly, he made his example doctors?)

Few engineers, scientists, or programmers that I know would give voice to the complaint that others might work harder. Their whole fields are suffused with positive externalities. When the other groups in my field discover more truths, I am enlightened by their work. When an engineer designs a better device, I am the richer for it. When a programmer writes more and better code, the world hums along more smoot... (read more)

Given the hideous prevalence of medical errors, and their known association to fatigue, I would far prefer a doctor chemically warded against fatigue to one without such armor.

No, the new equilibrium would be 96-hour shifts, with doctors to their physical limits and making as many errors (modulo differences in attention at constant fatigue induced by modafinil, if any).

[comment deleted]
A lot of code is written to win arms races, not improve the world. Online ads, algorithmic trading, the defense industry...
Arms races are strong driver of world improvement.
[comment deleted]

That's a very long winded way of objecting to Yvain's model of the American economy as largely zero-sum games (eg. poker). If the village is a static economy with fixed output... Then sure, modafinil is fairly questionable. But this story is a way of asserting it is not with hypothetical examples.

Of course, it's not obvious that iodine is necessarily a good thing. Malthusian models come to mind: if intelligence has no effect on subsistence wage, then it can have no effect on per capita wealth and so any effects are redistributional, and if you want to argue it's a good thing you need to appeal to extra things like quality of life... which actually probably would affect subsistence wage since now you don't need so much wages, your quality of life has been improved. Intelligence might come with a one-time increase in wealth, which of course simply causes the population to expand and that the temporarily-increased-per-capita-wealth will eventually fall back down to equilibrium as people reproduce more. :)

"It was a bit sloppy essay of Yvain - cool idea, kinda weak execution" is what I might say if he had posted it to Main instead of his blog.

Agreed. That is the heart of my objection, but if I simply say "the economy is not zero-sum!" then those that agree with me will agree with me and those that disagree with me might not see why. I do wish that I had thought to use the reversal test [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7jke] as my example.

I disapprove of this thread on the principle that people should be able to idly speculate on their own blog without being harangued elsewhere.

I disapprove of your use of parables to smuggle in your economic hypotheses, rather than arguing for them competently and clearly.

I disapprove of your commentary, because I agree with wedrifid here:

(Claiming to have) mind read negative beliefs and motives in others then declaring them publicly tends to be frowned upon. Certainly it is frowned upon me.

I disapprove of this thread on the principle that people should be able to idly speculate on their own blog without being harangued elsewhere.

Is this meant to apply just to LessWrongers? Because it seems kosher to discuss and critique blog posts generally in open threads.

On second thought, you make a good point. The problem wasn't Vaniver bringing it up, the problem was me not putting clear muflax-like epistemic state warnings on my blog.

I disapprove of your use of parables to smuggle in your economic hypotheses, rather than arguing for them competently and clearly.

Very well.

First, people prefer longer lives to shorter ones.

Second, just as it is difficult to think of goods that are only absolute, it is difficult to think of goods that are only positional. The used car provides $4,500 in transportation value; the Ferrari provides $50,000 in transportation value.

Third, many professions create durable value and large positive externalities. 25% more lawyering or 25% more derivative trading may not have obvious positive benefits, but 25% more programming or 25% more engineering or 25% more science obviously do. Crunch time may be 20 hours a day instead of 16, and so the programmers have just as little time to themselves, but the product will actually be superior, which seems like a Pareto gain.

Fourth, phase changes have effects that are difficult to anticipate. A world that moved at startup speed- where more people were massively productive and focused- could be far more glorious, delightful, and pleasant than our world. It is difficult to imagine just how miserable conditions were when society was liquid, rather t... (read more)

You are overestimating the value of reasoning by metaphor and the extent to which your metaphors are shared by others. When I take a pot of water and heat it, it becomes gas. If I seal the pot and keep heating, it won't become plasma. It will blow up in my face. See, a metaphor!
It would seem so, and I will try to adjust my style from here on out. Writing was easier when most were a step or two removed from the farm.
I don't think I understand the metaphor here.
Haiti is miserably poorer than America, in large part because of its people and its institutions. Not just in the sense of physical goods, but in most of the things that make life grand, and the things that make life annoying. Similarly, we are poorer than the future will be- again, because of people and institutions. (Technology- as in, knowledge about reality and devices that make clever use of that knowledge- is the result of people and institutions.) Importantly, this is not just in the sense of physical goods. It is one thing to compare a McMansion to a comfortably sized home; it is another to compare the sort of life lived by someone who lives in a world where they can buy a customized continent to someone who lives in a world where they can buy a McMansion. And so, in light of those changes, to look at a spark that could ionize our gas and say "but we'll just be running in circles [http://en.wikipedia.org/wiki/Ion_gyro_frequency#Frequencies] faster!" seems to miss the point. No, when every manager is a clear-headed executive, commercial organizations will be better run and more pleasant to deal with, and the sorts of things we can do will go from great [http://en.wikipedia.org/wiki/Steam_engine] to fantastic [http://en.wikipedia.org/wiki/Plasma-enhanced_chemical_vapor_deposition]. What does it matter that the yachts will be longer and the quays more crowded with them?
Thank you.
9Scott Alexander11y
Thank you :) EDIT: Actually, see here [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7jxf]
I disagree with your disapproval. While perhaps one wouldn't want to be "harangued", it is entirely appropriate to comment on publicly-available texts, and the open thread here is a perfectly acceptable place to do so.
It would have been legit if there was a link posted to the blog.
There is one. And no edit marker on the comment. Confused.
Can you point me to it? I control +F'd it for "less", "wrong", and "Vaniver" and found nothing.
I link to Yvain's livejournal from my comment on LW, but I don't link to my comment on LW from a comment on Yvain's livejournal, because I don't have a livejournal account and am not interested in making one. I didn't look into it very hard (maybe they let you post comments anonymously?) because it wasn't clear to me which option was more polite.
I'm not super opposed to your not posting a link to the livejournal page. I just think it would definitely unquestionably have been legit if you did that, whereas it's about 3% shady the way you did it right now, to give a super rough estimate of the level of shadiness I feel coming off of that.
The emphasis (and hence the use of the word "harangued" over neutral variants like "discussed" or "criticized") was on the inappropriateness of Vaniver's repeated emotional appeals and status attacks against Yvain. Forum switching is a well-known trolling technique that dates back to Usenet, and indeed possibly further.

Yvain has seen the misery of Haiti and India firsthand; but it seems only with his eyes.

I very specifically mentioned potential First World outlays to Third World countries as exceptions to my point. For example, I said:

There may be useful indirect actions, like advancing technology, increasing tax revenue that can be spent on useful absolute goods, and increasing the amount that flows as charity to the Third World (emphasis added)

Other than that, my entire argument was based on the "happiness follows economic growth up to a certain point, then stops" argument that has been mentioned here so many times before. That means a parable talking about how great certain interventions could be for the Third World is irrelevant; the post was very specifically and explicitly aimed at the First.

(I also think the benefits from lack of iodine deficiency are a lot less siphon-away-able)

The "60% serious" number may indeed be too high, though. I meant it to signal that I thought the argument was correct in all of its main points, but probably falls apart because the increase in productivity would produce very small benefits rather than no benefits, and "very small benefits" multiplied by the entire economy still end out pretty huge, especially if some of them end out in the Third World through the indirect methods I mentioned earlier.

The other organ I was looking for was not the heart but the head. Why are some people poor, and others rich? We run on our golden treadmills faster and longer, and what do we get out of it? Something, it would seem; the indigent in America do not eat mud to feel something in their belly. Happiness! A life's value is not denominated in smiles. How does satisfaction relate to economic growth? One day, a stranger came to the village. He carried with him a curious dried herb and sack of seed. The herb's leaves, he claimed, could be brewed into a soporific tea. Those that took it he would sleep twelve hours a day, instead of eight. The elders again convened to consider the stranger's tea. If one man took it, that man would get less done- but if all men took it, then one man's loss would be balanced by the other's. Many in the village were fond of their dreams, they said to each other, and so the weed seemed a boon. When they brought the tea before the village, many nodded at the wisdom of the elders, but one farmer, so poor he had to pull his plow himself, balked. "If you shorten my day," he said with despair, "then I must shorten my fields, for there are only so many days in the year one can plow, and my poor feet can only move so quickly." A woman spoke next. "Sixteen hours of spinning buys me four fish; enough to feed myself and my three children. If I can only spin for twelve hours, then I will only get enough cloth for three fish- and which of my children would you have me not feed?" The elders did not answer, but then one of the elder's sons spoke. "I already pay for candles to make my day longer," he said, "as the sun does not give me as many hours to read as I would like. If you shorten my day, then you shrink how large my mind may grow, for there are more books out there than a lifetime of reading, and yet I would read as many as I can." A singer was next, her mellifluous voice carrying easily across the village square. "I enjoy my dreams as much as the n
It's not denominated in dollars either, and if I had to pick one word to stand for humans' terminal values it would be much ‘closer’ to “happiness” than to “economy”.
If every spinster drinks the sleeping tea, less cloth will be made, but people will need it just as much. Thus cloth will become more precious, and people will be willing to pay 4/3 of the old price. The kids won't starve.
Instead, someone else goes unclothed. More generally, if everyone drinks the tea and produces only 3/4 as much, everyone, on average, will be 1/4 poorer. Price movements only affect how the poverty is distributed. (Of course, they also affect what new resources are tapped, what new inventions are made, how hard people will work during their reduced hours, how existing resources are redistributed among their uses, and how effort will be redistributed among the different productive activities, but that is going beyond the purpose of the parable.)
Yvain's premise is that the country is warm, so people only make clothes to show off their wealth, ability to sew, and taste in fashion. Someone decides "I was already reluctant to buy those expensive rags, now they're just too expensive" and joins the ranks of streakers.
So is demand for cloth elastic, or is it not?
That's the parable of the broken window [http://en.wikipedia.org/wiki/Parable_of_the_broken_window].
How many fish does the fisherman catch in an hour?

if modafinil were legal, it would become expected, and everything would be harder to match the increased ability of humans to be productive.

I tend to agree that this is a silly argument, especially given that it can be applied to coffee as much as to modafinil, so we better ban coffee, lest those allergic to it be at a disadvantage.

Or indeed to any technology. You may think you are better off using a combine harvester instead of a sickle, but actually it just shifts the expectation of how much grain you need to produce.

Yvain says in his posts' comments that coffee doesn't work, as tolerance builds up. This seems disputed. But why not ban coffee? Because, like alcohol, it's now too ingrained in our culture. But if it wasn't - preventing headaches, irritability, concentration troubles, and the expectation that everyone can pull all-nighters? Fuck yes.
So, the world would be a better place if people like me (who drink butter-coffee everyday) had to give up their favorite health food or risk jail time? Consider me skeptical.
Does it not work with decaf?
Coffee may not work to generate more virtual hours of productive time in the long run but that doesn't mean that it's use in time shifting sleep requirements etc isn't still of net benefit.
Coffee may be too Near to discuss; I suggest a different even-older expensive teeth-staining addictive stimulant plant popular in social gatherings [https://en.wikipedia.org/wiki/Khat].
I've been thinking about qat a lot these past few days, so I'll tap out of the Far mode discussion. Just this: my problem with coffee is that people are often given too much work, which they require coffee and similar stimulants to accomplish. (Witness: programmers' love for soda; project deadlines at university.) Qat doesn't seem to have that problem. It does have another problem: if you don't want coffee, it's usually socially acceptable to drink another hot beverage (though if you don't want tea either you're kinda screwed), whereas qat lacks an alternative.
Given the third world's fondness for tobacco (eg. apparently China is now the largest and growing tobacco market in the world), isn't chewing tobacco an alternative?

It bothers me that no one is applying a reversal test here. The paper even calls out intelligence augmentation as the prime example!

I'm inclined to trust Bostrom's well thought out paper on the matter, but I'd be curious to hear opposing views.

9Scott Alexander11y
I might endorse a certain very specific reversal test. If I could choose between the current world except that freethinkers are at a significant disadvantage relative to everyone else, versus a world with a four hour workday but we all had to sleep four hours more per night so we still had the same amount of free time, plus our economy was at the same level as in the 1990s [http://squid314.livejournal.com/332176.html]... ...then actually I would choose the current world, because the four hours more sleep per night would also apply on the weekends and so totally disrupt the balance, which I hadn't thought of at all in the original post. So never mind.
This assumption that all the change in the amount of waking hours would go towards increasing (or decreasing) labour time is suspect. I mean, why couldn't people keep the current ratio, work 50-hour workweeks and get 14 additional hours of leisure time per week? The rich get better yachts and everybody has more fun.
Vaniver has, now [http://lesswrong.com/r/discussion/lw/eqa/open_thread_october_115_2012/7jl7]. EDIT: and shminux had already done so [http://lesswrong.com/r/discussion/lw/eqa/open_thread_october_115_2012/7jjy].
Although I'm unsure of the etiquette of posting about personal blogs on other sites, I was also disappointed with the blog post in question. It was the first time that Yvain wrote something I disagreed with after reading his post in full and digesting in. I've often disagreed with him before reading it, but he usually persuades me. This post seemed to rely on the principle that having more spare time is a positional good, with which I disagree strongly. Essentially, giving everyone another four hours of awake, productive time, is the same as extending your life by four hours for each day you are alive [and you do so in good health - extending the human lifespan from 80 to 95 might or might not be a good thing, but adding years to your healthy, productive life, seems a good thing] Yvain's claim seems to be that 100% of the extra four hours will be diverted into work, but to me that's a) almost certainly not true [the figure would be more than 0% and less than 100%], b) not a bad thing. a) It seems very likely that, given that our day is > 0% work and > 0% leisure, an extra four hours a day will add more than 0 hours of leisure. b) If all med students get more studying done, it's far from obvious that the net result is a bad one. I assume that there is some value to med students' knowledge of medicine [okay, for anatomy courses this might not hold]. If, say, Apple workers work 50% more, then we stand to get better and faster Macs]
What might convince me that Modafinil is a bad thing would be if a lot of people actively disliked the time they spent working. I personally assume most people roughly like or are neutral towards their jobs and mainly want to work shorter hours because it gives them more time for things outside of work, but I'm almost certainly generalizing from the example of me. If Yvain had made this argument I would understand more about where he was coming from and why.
In some professions saying that you "love you work" is a signal of a good employee. So I would expect some dishonesty in self-reporting. How could we ask the question to reduce this signalling? I imagine only silly questions like this: Imagine that for some external reasons your workplace must be closed for two weeks. During those two weeks you will receive your normal salary, and those two weeks will not be taken from your holidays. How does this message make you feel? a) awesome! b) mildly happy c) neutral d) mildly sad e) depressed On the second thought, is this question really silly, or does it show our true preferences? And the silliness is merely a reflection of dissonance between our professed values and real values.

This doesn't quite answer the question. I would be very happy if my place of work were closed and I could do fun things for two weeks. My objection to working isn't that work is unpleasant; it's that there's a high opportunity cost [all the fun people I could hang out with, the great books I could read, etc]. A better question is "imagine you are asked for your employer to take part in an experiment where you instead have your brain turned off. Your body ages by eight hours, but your brain experiences it as "you step into the office, then step out".

It retains the silliness but solves the opportunity cost problem.

You are right about the opportunity costs. The work is actually not bad -- it's the idea of all the things I could have done in the same time that's driving me crazy. Your question is better (although it does not contain learning during the job, which is important too).
[comment deleted]
I think it genuinely shows our preferences. Of course, I also would be fairly neutral about this (and would probably seek other work for those two weeks).
[comment deleted]
Doctors don't take courses in modafinil, or anything, so I'm not sure what you expect them to base their advice on besides the FDA prescribing guides like http://www.erowid.org/smarts/modafinil/modafinil_provigil_prescribing_info1.pdf [http://www.erowid.org/smarts/modafinil/modafinil_provigil_prescribing_info1.pdf]
A pharmacist might be a better bet.
[comment deleted]
In general, if people are able to get more done in a day, would this give an advantage to the makers or the takers? I don't have a general principle which gives an obvious answer.
By advantage, do you mean subtracting their benefits from each other, to see who benefits more? Or do you think it might be a case where one group is actually made worse off, and another group made better off?
I was thinking about subtracting their benefits from each other. More hours in a day and more focused attention might have been able to give the financial industry even more ability to cause havoc.
I don't think the benefits are measured in the same denomination, and so it's somewhat difficult to subtract them (which agrees with you not having a general principle to give a clear answer). Would more politics get done with more hours in the day? Perhaps. More trading? Certainly. With the whole world more productive, there will certainly be more to skim off the top- but what percentage do the takers take? In aggregate terms, it looks like the makers win more, but in per capita terms, it looks like the takers win more. But that's just looking at income distribution- who gets more of the positive externalities? Here, the per capita seems only slightly better for the takers, and the aggregate far better for the makers.

As long as I'm demanding that LessWrong provide me with the answers to my personal problems, I find myself becoming more and more misanthropic as time goes on. I genuinely like only about five people out of everyone I've ever met, two of whom are family. I feel like almost everyone else is borderline homogeneous, originality seems extremely scarce and I'm bored whenever I try to talk to most people.

Context: I'm in college and not making friends. This is largely because I don't drink or follow or play in sports, I think. I'm bad at small talk. It's also because I'm unhappy with lots of what's perceived as normal around here (eg the subtle dehumanization of women).

I don't really know what to do. I believe humans are social animals and that I'd be happier with friends, but at the same time I really don't like any of the people who I talk to here. Any social advice at all would be useful for me, and anything that deals with the specifics of my situation doubly so. Misanthropy is obviously bad, but I don't know how to transition from my dislike of most people to becoming friends with them, nor am I positive that it's the right thing for me to do in this situation.

I had very few (physical) friends in college and even fewer now. I find that I get enough social interaction online and with my family (I'm married). Of course everyone is different but you may not need as many friends as you currently seem to think.

If I may offer some advice: Be careful not to rationalize social anxiety with "they are homogeneous, they dehumanize women, they aren't as original as I am, they bore me". That's externalizing an internal problem.

There are people of considerable intellectual caliber who have no qualms engaging in random small talk (a required skill in many career paths), and you'll only find out who they are once you get past that barrier.

No simple solution, but nosce te ipsum applies.

How should I distinguish between these types of people? Is there a way that doesn't require me making small talk with lots of people who I don't like?

Perhaps start by actively distinguishing between "people I actively dislike" and "people who I don't actively dislike, and am assigning the dislike label to based solely on my prior that I dislike most people".

Also, in regard to inauthenticity, do you regard making small talk as inauthentic, even if you are saying true things? For example, is it inauthentic to pay someone a compliment if you honestly believe the compliment, but are only making it as a way to start a conversation and find out whether you like them? If yes, I suggest you taboo "inauthentic" and explain why you don't like that approach. I suspect that exploring that label more generally may be fertile ground.

More generally, do you have a problem with people who are not bothered by inauthentic conversation, but also are happy to have authentic conversations? If so, I suggest asking whether this is an area where you should work to cultivate tolerance of tolerance.

To distinguish these people, I would ask what sorts of conversations you consider authentic (again, taboo that word!), and think about what sorts of authentic conversations are easier to start up than others, and what sorts of sett... (read more)

You need to find a (physical world) subculture to get involved in.
What are the people you like like? You say you're becoming more misanthropic; did you use to like more people? What were those people like? Do you have an internal narrative about why you don't like them anymore? On Enjoying Disagreeable Company [http://lesswrong.com/lw/2a5/on_enjoying_disagreeable_company/] is my post on liking people on purpose.
I used to like more people and to just be able to go up to people and talk to them. When I was very young I was extremely outgoing. That stopped sometime during elementary school, I don't really remember when exactly, but it was because I was naive and trusting and people would take advantage of me (stealing stuff from me, copying homework, pranks and "jokes"). I moved to a different town in middle school and was pseudo popular for a year, in that everyone was nice to me and would talk with me. I lapsed back into idealism, and then ended up having no friends again because no one really liked me, they just liked being associated with the novelty that was the new kid. High school was a gradual process wherein I became less and less popular up to the beginning of my senior year, when I began to regain ground. In college I'm isolated. The people I like are simultaneously independent free-thinkers and compassionate. There's tension between the two, but it produces interesting people. My favorite person in the world is my little brother who is one year younger than me, he is hilarious in a highbrow intellectual way and always able to find my blind spots and more factually knowledgeable than me, so he corrects me. (My intellectual strength is that I'm good at understanding how different concepts interact and at generating strategies for argumentation. It's not that he totally dominates me in intellectual discussions, but that I move the discussion forward and he stops it from moving towards the wrong areas.) He shares most of my values and traits except that he's a harder worker, simultaneously better and worse at social things, and he's less selfish. He's ridiculously awesome. I like people less because social norms have grown more complicated as I grew older and I prefer authenticity, I think. Also, the less time I spend socializing, the less knowledge I have about social norms, and there's a feedback loop. Additionally, I think many social norms are morally wrong and

My first idea is to ask your brother for advice - he probably has some friends, and if he's good at correcting you in a way you can appreciate, he might be able to figure out what's wrong on your end and help you fix it.

Additionally, I think many social norms are morally wrong and I'm not willing to engage in them.

Can you be more specific? Different subcultures use different social norms. There might be one compatible with you.

other people would still dislike me unless I engaged in the kind of behavior that I hate.

I think you're underestimating human heterogeneity. The fix for this is to meet many different people, not engage in the kind of behavior you hate, and not bother hanging out with anyone who is put off after you learn that they were put off. You are not overwhelmingly likely to run out of people unless you live in the middle of nowhere. (Do you?)

I think you're underestimating human heterogeneity. The fix for this is to meet many different people, not engage in the kind of behavior you hate, and not bother hanging out with anyone who is put off after you learn that they were put off. You are not overwhelmingly likely to run out of people unless you live in the middle of nowhere. (Do you?)

I strongly second this. The people who are like you and who you would like most likely also hate the behaviors you don't want to engage in. By not engaging in them, you may alienate the people you dislike, but you'll make yourself more interesting to the people you actually do want to hang out with.

I don't know how many people do the things I'm interested in. I joined a political science club, which seems good so far. I haven't encountered any other social things I'm interested in though. I need to get more information about what activities are going on in my area, and I should probably expand my areas of interest also. Yay LessWrong gives me momentary confidence and hope for my social future!
Most clubs go out of their way to get more recruits, in my experience. Jugglers and Capoeiristas both like to put on demos and hand out flyers. If there's a student center where you are they probably have info on more clubs/hangouts you can go to also.
Interesting. I know him well enough to know that he would dislike the same people who I'm currently disliking, but I think that for whatever reason he might know more about how to find interesting and intelligent people. As of now, he has more friends than me. We were roughly equal during high school. His social role when he's in groups is generally to be slightly quieter than average, but then to fire off witty and sarcastic one-liners at certain times. My social role is nothing, I find it hard to function when I'm not problem-solving or analyzing. I didn't really have friends in high school so much as people who weren't actively rude to me and who valued my input, to be honest. I should probably figure out a gimmick and stick with it, like what my brother does, the problem is that this feels inauthentic to me. His comes to him naturally whereas I don't really seem to have any inherent social role. Drinking and making jokes about sex. Self congratulatory behavior and bravado. Inauthenticity in general. I'm uncertain whether everyone is really like this, or whether they're just signaling that because they're insecure college freshmen boys and that's stereotypical behavior and they're scared of being an outsider. I think it's probably some of both insofar as they're internalizing these norms because they find the internalization of these norms advantageous. I hope it will calm down soon if it is primarily signaling, but I don't think that will actually happen because the underlying factors will still exist and will actually be intensified by this internalization. I expect it will wind down once there's an external incentive to be responsible or at least to be perceived as responsible, but that will probably take at least a couple years. This college is too small for legitimate subcultures to exist. I thought that small class sizes would be a benefit, but I never considered that it would caused increased pressure for conformity, which it seems to have done. That s
I only know a handful of people who I could fairly sum up as having "social roles" in the same way you describe your brother as having. This could be a deficiency on my end, or I could know weird people - or this could be an inadequate model of how social interaction works, and my bet is on the last thing. Have you considered making friends with girls? There will probably be less (though still some) of the things you list among girls, depending on what you mean by "inauthenticity". (What do you mean by "inauthenticity"?) Have you considered making friends with upperclassmen or socializing with professors you like? Why do your friends have to be your age? Just how small is this college? Mine had like 400 people and there were types, if not outright subcultures. Is there a game store? Those often host gaming events. Do your friends need to be in-person? This one could be a problem. Are there any known ways around or through it that are relatively easy to exploit? Be there to do something else too, and focus (verbally) on that thing (while striking up conversations, of course).
Honestly not sure how. I've never really ever made friends "on purpose" with people in general. That's probably a lot of my problem. Then there's more issues involved when I have to deal with girls, because I have to deal with gender roles or different expectations or whatever. I'm not intrinsically opposed to the idea. My issue is that I don't know how to: 1. Become friends with people unless I interact with them a lot, and that's not really happening. 2. Become friends with girls specifically, I assume the issues there with getting to know someone will be even more challenging. You're dealing with a social wreck here, basically. I also don't think girls tend to be very authentic at my age, but it's not as though they'd be worse than the guys. I don't know how to make friends with people I don't interact with on a more or less daily basis. My friendships have always just "happened", I've never actively pursued them before. 600 in my grade, 2000 something total. Maybe I'm just wrong here and am unobservant. I should check for game events in my area, I guess. There's not one on campus or anything. Preferably. I don't think I could make a very good friendship via the internet. Not sure, I need to fight it. Just about everyone arrives to events in groups and I don't know how to to strike up conversations.
So I'm getting the sense that you were only restricting the demographics of potential friends by default, which is good, since it means there's more space to look in than you thought. Professors in particular might be good if you mostly want to have intellectual conversations! Show up to office hours, and have intellectual conversations with them! There's "authentic" again - what does that mean? My best guess right now is "not wrapped up in signaling" in which case - well, you're gonna have a bad time. Humans do that. (Though I begin to suspect that you're oversensitive to it and may be seeing more of it than there is.) But maybe you mean something else. Why not? Go up to someone. Ask them a question ("do you know if the food here is any good?", "can I borrow a pen?", "is the line for tickets?") or pay them a compliment ("awesome t-shirt!", [laughter at a joke they just made], "your presentation just now was fantastic, my favorite part was [x]!") or stand near them and their group (without being followy if they try to leave) and pick up on something someone in the group says when there's enough of a break to do it ("yeah, Communism would only work for nonhuman aliens", "that's funny, when I was in Japan I didn't see any kaiju at all!", "cool, so snakes don't even have ears? Can they sense vibrations?"). Or the classic standby of: Stick out your hand. "Hi, I'm [chaosmosis]! What's your name?"
The people who like drinking and sports are the most prominent in many colleges, but it doesn't mean that they're the only ones around. I had the same problem as you my first year at college, and mainly solved it through three factors, in order of importance: 1. Making friends in math classes. 2. Going to a few student-organized clubs. 3. Blind luck. Whatever strategy you decide, if you happen to find just one or two friends that also don't have too many friends, you can then try everything you try together and it will be much easier. I realize this might be terribly unhelpful advice.
This is how I feel when parties are going on: http://www.theonion.com/articles/area-man-going-to-sit-quietly-in-darkened-bedroom,29831/ [http://www.theonion.com/articles/area-man-going-to-sit-quietly-in-darkened-bedroom,29831/] . There's one going on tonight, so this is particularly apt.
Most people are terrible. It's a lot of work to sort the awesome people from the terrible people. I've had good luck using "gamer" "geek", "queer", and "kinky" as labels that tend to more reliably apply to interesting people or people I'm happy to get along with, but your mileage may vary. Every single one of my room-mates plays or used to play Magic the Gathering, for example. Not making friends with the random people around you in college who are into drinking, sports, and dehumanizing women is, in my mind, a good sign. You shouldn't force yourself to try and make friends with people who don't share interests with you, or at least are interesting to talk to. Try talking to the people you see who are actively weird.
I tried that a week ago. I now have this kid who just might be repressing some homosexuality following me around whenever he sees me (it's the repression that I have a problem with, it manifests itself by him scaring the hell out of me and talking about obscure mythologies and creepy myths all the time, he seems to have serious psychological problems. He's one of those kids who no one will talk to, so he gets creepier and creepier the more he's left in isolation, and then a feedback loop happens. I talked to him out of pity and regret it). I like the advice, in general. Are there additional filters I can apply?
Not exactly a filter, but if you make friends with a person who is awesome and who seems to have a lot of friends you should try to hang out with them as much as possible. Most of my friends end up being from friends I already had.
People who are in bands or theater tend to be fun, I like most people I've met who play Capoeira, Women with piercings and dyed/shaved hair tend to be more fun to talk to, and most people I've met who are in math programs or math graduates are awesome. As always, your mileage may vary on this sort of advice. You said in another comment that you like gaming. Local game stores often have websites where they post information about which days of the week they encourage people to come in and play various kinds of games, from boardgames to minis to TCGS. Some even have pickup roleplaying groups you can just drop in and out of. I definitely recommend googling {Your Town's Name} + game store, or looking at clubs run by your college. Even if your college seems too small to support subcultures they probably exist anyway. I think it's also helpful to look outside your immediate age group for friends. Many of my friends and most of the people I like best are several years older than me, because when I was meeting them around age 18 or whatever i found everyone around my age intolerable. Similarly, though you probably don't want to hang around high schools you shouldn't necessarily dismiss someone because they're younger than you. This will probably make it harder to date though.
Friends are a much higher priority right now. Thanks for the good advice.
You're welcome. Having gone from basically no friends to quite a few I feel like I owe it to past-mes to help em out
When I was in college, I once thought that I didn't enjoy drinking. Turns out, I didn't enjoy drinking with people who I was not friendly with (and I had poor social skills and thus few friends). But I really only learned that after following the equivalent of your excellent advice.
Hang on. Most people are really nice. Most put a confident facade over a good nature. Most are a bit lonely, unsure of their own value, and mainly worried about how other people see them. Most are full of interesting thoughts that they are shy to express in front of strangers. Most young people are idealistic to the point of charming naivety. And the women. The ones being dehumanized. Who are they hanging out with? The evil dehumanizers, or the self-righteous nerds, full of anger, staring sullenly and lustfully at them from the corners?
Terrible is hyperbole. Most people, even though they're nice, or secretly have interesting thoughts or whatever feel good stuff you say is true about them, are not going to be fun for me to hang out with. Since when did I recommend being a lustful sullen staring cornernerd?
You are reading a little more judgment into the post that I think is intended. The women (and people generally) that are going to be enjoyable to spend time with are not hanging out with the hyper-masculine jocks. There's no shame in noticing that, and picking social groups accordingly in order to try to find social companions. Particularly because the jocks are particularly poor at being reflective about their own social skills and the social skills of others.
I don't mean this snarkily, but have you considered drinking or taking up a sport? Low level sport (where everyone's a bit rubbish and no-one takes it seriously) is superb fun. Obviously, you'll be terrible at it, but if you find a club that's short of people and loses all the time anyway you'll probably be more welcome than you think. And just by taking part you'll get much better at it. It might change your life. And if you're embedded in a society where social life revolves around alcohol you'll miss out on a vast amount of the fun and happiness that comes with being human if you don't join in. You don't have to overdo it. Just try having a glass of wine with someone you like one day and see how it goes. I am a pretty nerdy guy, but if I had to relive my life without ever drinking alcohol or playing cricket or rowing or playing rugby I just think I'd probably not want to bother. ( I am unbelievably bad at rugby. ) If it helps encourage you, for a long time I coached novice rowers for King's College (part of the University of Cambridge). Occasionally we'd get a hopelessly non-sporty introvert turning up wanting a go. Some of these guys were so shy they could hardly speak. And they were often the people who enjoyed it the most, and became most committed and most likely to come back year after year. I'm not going to lie to you, with one exception they never became any good. But they all became much better than they had been, and seemed to enjoy the process, and are some of the people that I most enjoyed coaching. It helps of course that rowing is actually technically complex and I could talk to these people about how best to turn energy into momentum and how it is that a boat can balance even though its centre of gravity is above its centre of flotation and so on. I think one of the reasons that rowing is so popular amongst the sciency types at Cambridge is that it is a sport that you can think about in terms of physics. But it really doesn't matter what the
I like sports, specifically basketball, which I'm decent at. But I dislike almost all sports people, who are the ones who are drinking the most and doing all of the things that I dislike. There are probably good people in intramurals activities, but I don't want to be a teammate to the bad ones to get there.
Non-competitive (i.e. no traveling to tournaments) sports are very likely to have a different (and possibly more receptive-to-you) culture than varsity sports (or their college equivalents).
I take it you don't have a LW meetup near by. Do you think you could start one?
Having friends seems more or less like a prerequisite, I'm also not confident about my ability to lead a group like that. It might be a good long-term goal.
LW meetups don't have to be large or formal events. See Starting a Less Wrong meetup is easy [http://lesswrong.com/lw/43s/starting_a_lw_meetup_is_easy/]. You could even write in the meetup post that it's going to be highly informal, to set expectations.
Nope! If you can send emails to people saying "There will be a meetup on X day, with Y activity," you're most of the way there. Seriously. If there's anyone in a 100-mile radius who is interested in meetups, making them happen is not hard, and in fact is probably easier than many other social activities. You could make a post on LW to gauge interest in your geographic area :)
Seconding drethelin's advice in general. I feel very empathetic towards your situation, and I did not like most of the people I met when I first went to college, and yet I made some serious long-term friends there. Also, for fear of not having a "social role", just make sure there's something memorable about you. Like a cartoon character, have some element of style that is unique and not obviously negative - so that people can see you and think "There's that kid with the X again", and you blend into the background less. If X can also signal desirability to a relevant social group, all the better. Something that already seems cool to you, so it doesn't set off your "inauthenticity" alarm. Bow tie? You're the hero in your own story. Don't worry that most of the people you run into are NPCs - that's normal. Keep looking and you'll find the 3-5 other PCs to join your party.

As a non-mathematician, I enjoyed an old blog post from Dick Lipton, Guessing the Truth, which lists a bunch mathematical conjectures incorrectly expected to be true, along with their resolutions. There's also a great comment by Terry Tao on kinds of evidence for mathematical conjectures. For example:

Attempts at disproof run into interesting obstacles. This one is a bit hard to formalise, but sometimes you can get a sense that attempts to disprove a conjecture are failing not due to one’s own lack of ability, or due to accidental contingencies, but rather due to “enemy activity”; some lurking hidden structure to the problem, corners of which emerge every time one tries to build a counterexample. The question is then whether this “enemy” is stupid enough to be outwitted by a sufficiently clever counterexample, or is powerful enough to block all such attempts. Identifying this enemy precisely is usually the key to resolving the conjecture (or transforming the conjecture into a stronger and better conjecture).

Real numbers are not "real". (Inspired by Imaginary numbers are not real, an elementary introduction to Clifford Algebra I came across a long time ago).

I find it a bit funny that people tend to think of real numbers as "real" numbers, as opposed to, say, imaginary numbers, which are not only not real, but also not "real" in a way a Realist would use the word. The paper above even takes pride in not using i in calculations. There is also an occasional discussion in philosophy papers and online of the wave function in QM not being "real" because it uses imaginary numbers.

I find it funny because real numbers are no more "real" than any other numbers. Even the set of all integers is not very "real", as basically everything in the Universe is finite, due to the cut-offs at various scales, such as the Planck scale and the age of the Universe, and whenever you try to disregard these cut-offs, things tend to blow up in your face.

One can potentially consider finite integers as the most "real", given that they correspond to discrete objects we can see, count and calculate. The rest are simply useful mathematical abstracti... (read more)

I've never thought of real numbers as any more real (in the non-mathematical sense of the word) than other numbers, and I've been peeved by popularizations which use “real” and “imaginary” without making it clear that they're using them with a specific technical meaning (e.g. stuff like “special relativity has shown that if space is real time must be imaginary, and vice versa” -- yeah, they do have squares with opposite signs (though modern notation uses real 4-vectors and a non-positive-definite metric), but that's not how a reader would be most likely to interpret that sentence).
Agreed. I have no beef with the term 'complex' for the complex numbers. It's the 'real' for the others, and the 'imaginary' for the new stuff, that I mind. I wonder if a very short treatment of abstract algebra should be given in high school, right before you get to complex numbers. Might reduce the number chauvinism and help with the illusion of number realism.
Never heard this term before :) Maybe in an AP-level course? The high-school math is pretty instrumental, focused on solving problems and passing tests. Actually, I think this is probably best covered in a relevant college-level philosophy course.
I've taught a few people about the complex numbers, by stepping through expanding the naturals with the introduction of negatives to make integers, fractions to make rationals, irrationals to make reals, and finally (the 'novel' stage for my audience) imaginary numbers to make the complex numbers. I emphasise the point that the new system always seems weird and confusing at first to the people who aren't used to it, and sometimes gets given a nasty name in contrast to the nice name of the old system (especially 'imaginary' vs 'real' and 'irrational' vs 'rational') but the new numbers are never more or less worthwhile than the old system - they're just different, and useful in new ways.
Lots of numbers correspond to things. It seems arbitrary to say that integers are real numbers, because they correspond to discrete objects we see, but rotations aren't numbers, even though they correspond to transformations we can make on an object.
I more or less agree with you that the terms are inappropriate and often actively misleading. In relation to this, I've often wondered about the usual justification for the postulate that observables in quantum mechanics must be represented by Hermitian operators. The justification is that the expectation values of our observables must be real numbers. Doesn't this sort of justification stem from an intuition that real numbers are somehow more real than imaginary numbers? The attitude seems to be: Imaginary numbers can certainly play a theoretical role in physics, but when it comes to describing our immediate empirical situation, quantification must be in terms of real numbers -- it makes no sense for a measurement outcome to be described by a complex number. Maybe this doesn't really represent the crux of the justfication. The argument might be that quantum mechanics needs to approximate classical mechanics at the macroscopic level, and classical mechanics only involves real numbers. Since observables correspond to measurement procedures, and measurement procedures are macroscopic, observables must have real expectation values. This argument seems fine if you genuinely think of observables as representing measurement procedures, but if you think of observables as representing measurable properties (this seems to be a common perspective) then these properties need not be macroscopic, in which case it is unclear why they are constrained by a requirement to replicate classical mechanics. I'm not questioning the claim that observables must be Hermitian, just suggesting that the usual justification isn't really that great. Maybe we should provide other arguments for this claim when teaching QM, in order to avoid entrenching the "real numbers are more real" fallacy. For instance, one could argue that the quantities we measure in QM experiments (energy, momentum, spin) are quantities that are conserved if certain symmetries hold (I think one might be able to construct a
I'd work backwards. As I said, "finite integers as the most "real", given that they correspond to discrete objects we can see, count and calculate". If you accept that, instrumentally, we can only ever see a count of something, like the number of notches on a ruler, or marks to the left of the needle of some gauge, or the number of clock ticks before a certain event, then you can extrapolate backwards in your logic to include negative numbers, fractions, rationals, reals and complex numbers as convenient abstractions. But you have to leave a path forward in your models to eventually end up with the number of notches on a dial. If your observables are not Hermitian, you cut off this path from models to counts.
If that's your view, though, I don't see why you think real expectation values are better than complex ones. If our measurements have the structure of integers, and our expectation values are real, then in our model of measurement we're going to have a map from the real line (or some segment thereof) representing possible expectation values to the integers representing notches on your dial. This map will presumably associate an element of some partition of the real line into finite regions with each integer. So this map must be part of your model of measurement in order to connect your integer measurements with the continuous space of expectation values allowed by the theory. But if you're relying on some such map anyway, why can't the expectation values be complex? You could construct a map from a partition of the complex plane to the integers. More specifically, your claim that the ultimate measurement results must be integers doesn't rule out the possibility that each measurement result gives a pair of integers (instead of a dial moving along a single dimension, think of one that can move in two dimensions). For such measurements, it may even be more natural to think of a map from the complex plane to the measurement outcome, with one of the integers associated with the real part and the other with the imaginary part. As long as you allow for this possibility, I don't see the motivation for restricting observables to Hermitian operators. Surely it isn't a requirement of instrumentalism that the measurements we make must only be one-dimensional.
Yeah, that would be a dangling requirement. That would be an interesting counterfactual universe, where you need at least two numbers to make sense of anything. I cannot quite imagine how this would work, but maybe this is a limit of my imagination. In the Universe we are currently stuck with, with the classical world described by classical mechanics based on real numbers, and the quantum world manifesting itself only through classical measurements, any useful model must ultimately lead to a set of readings off some dial, each of which should make sense separately.
I fail to see the contrast.
The drama caused in other sciences due to not sorting authors alphabetically is truly depressing.

Economics also has the tradition of ordering authors alphabetically. And economists with earlier-letter surnames end up having more successful careers, quite possibly as a result of that tradition.

And yet the effect isn't strong enough for economists named Zweibel to change their names to Aardman?
If the tradeoff is between either some acknowledged bias or a great deal of drama, I choose the former.
There are possibilities for less biased orderings which are also less drama-prone. For example, choose a day 1, then use alphabetical order, but advance the first letter of the alphabet for each a day after day 1. Today is the alphabet for author ordering starts with A, tomorrow with B, and so on. If that still introduces a bias, then perhaps the alphabet should start with the same letter on consecutive days, but alternate between going forwards and backwards.
In the article Hanson proposes a simpler method.
If you want an more-or-less unbiased but deterministic way to do this, you could sort the authors by whose birthday is closest (in either direction) at time of publishing. This additionally makes it so the precise date doesn't matter too much. Making it closest upcoming birthday would be simpler, but if a colleague's birthday is one day before you it kind of sucks. But probably the random-order idea, as suggested by the article, would be even easier.
The reason I wanted something deterministic is that I wasn't convinced that scientists would generally trust something that looked random with the stakes being somewhat high. When I think about the amount of scientific fraud, I'm not even sure that they should trust each other. My alphabetical scheme isn't ideally random-- it gives an advantage to authors whose names begin with unusual letters. Thanks for the information that date of publication can be somewhat foggy.
In some sciences there are established rules for the role of the principal investigator [http://en.wikipedia.org/wiki/Principal_investigator] (first author) and the last author. In the life sciences at least it's often enough to be the head of a laboratory whose equipment was used to get on a publication, no additional input required. Why would you want to hide that information, or does your proposal encompass only all the middle authors / "contributors" (often well over a dozen)?
Mu; I don't want to hide that information. The cultural traditions (calling them "established rules" is a stretch when there's so little agreement about what they are) surrounding authorship hides them in a lossy format while causing at least thousands of scientist-hours of drama a year. If authors want to explicitly flag down who did what, they can do it somewhere in the article proper. A possible counter-argument would be that employers want to see, e.g., how many papers a given research has been the PI on, without reading the article proper. The current system already doesn't work for that use case; the status fighting over first authorship has created the ridiculous notion of "co-first-authorship" and false declarations that "Dr. X and Dr. Y contributed equally to this project."
The best part [http://sciencecareers.sciencemag.org/career_magazine/previous_issues/articles/2010_04_16/caredit.a1000039]?
I loved those things when I was a kid.

Via John Baez: Mathematics for theoretical physics, a 700-odd page self-contained reference to all the maths you supposedly need to have some idea what contemporary theoretical physicists are talking about.

I saw this pop on arXiv, but I doubt it will be very useful to anyone who doesn't know most of the stuff in it beforehand. The exposition is pretty terse, and there are well over 1000 stated theorems.
Thanks. I'll take a glance at it when I have some spare time.

Course Builder is some software Google used for making an online course, and now it is available, also with free webhosting if you don't have too many students.

Has anyone tried this? Is anyone interested in this? Are there other easy ways to make and publish online lessons without paying money? I did not have time to explore this product deeply yet, but it was easy to download and install.

Could this be useful to provide some CFAR lessons online? Convert the Sequences into Videosequences? ;-)

Yes [http://lesswrong.com/lw/ek8/which_questions_about_online_classes_would_you/], I've been looking into it in relation to this [http://lesswrong.com/lw/ecf/open_thread_september_115_2012/7cug] project. If you'd like to join our little informal research group feel free to contact me via PM and share your email address. :) A few but they seem kind of sketchyware and I haven't tried them.

Free TEDYouth Event in NYC for High Schoolers

Taken from the TEDYouth event description page:

Held annually in New York City, TEDYouth is a day-long event for high school students that includes live speakers, hands-on activities, demonstrations and an opportunity for the youth attendees and speakers to connect. TEDYouth coincides with more than 100 self-organized TEDxYouthDay events happening worldwide over a 48-hour period.

This year’s TEDYouth conference will be held on Saturday, November 17th, 2012, at the Times Center in Manhattan, from 1pm-6pm.

More th

... (read more)

I'm finding the resources on akrasia that I've encountered on this site to be inadequate. I need help.

I usually have problems being motivated with big goals at all, but I've finally triggered one (unwarranted immediate attraction to someone, which I would like to use as a convenient hack to make myself work out and actually put some effort into my school studies). Hopefully, I'll be able to capitalize on that and start to implement good habits.

Links. I need them. Please?


What is the most effective chain of thoughts that a theist can make, that will make him realise that there is no God? Efficiency could be measured in the number of thought steps. I'm especially interested in references to articles that considers question.

What is the most effective chain of thoughts that a theist can make, that will make him realise that there is no God?

"People not in my tribe are sexy and cool. I want to be like and/or mate with those people. I believe I have a sufficient chance of successfully joining and gaining status within the tribe with sexy and cool people in it. I will now change my signalling beliefs."

This doesn't raise the sanity waterline at all. Unless the person is part of a very dangerous Jim Jones style of religion, what's the point?
Yes it does. I think you need to read it again a couple of times and maybe a few OvercomingBias posts. It makes no difference to the point whether the religion is dangerous, admirable or even the literally correct and the path to eternal bliss. Thoughts about tribal affiliations typically matter more than abstract reasoning when it comes to this kind of belief.
Maybe I misunderstood. It sounded like you were suggesting that theists be baited with a honeypot (mate or group) that they'd like to be attractive to. The pressure to be liked would cause them to abandon their beliefs. I'm not saying this can't work, but it's a transformational pressure that works equally well in both directions. The person isn't better informed at the end of it, they're just trying to fit in with a group that happens to be more accurate. (If atheism highly correlated with sexiness, maybe this would pull more people out of religion than in, but, in fact, religions are better at being organized into tribes with status anyway) But, I think, given your response, that I've misread you. Can you correct me?
It'd work, sure. Doesn't sound like a good use of my time.
Your statements have the implied premise that having true beliefs raises the sanity line even if the process of reaching those beliefs is not correlated with the correctness of the beliefs. I prefer "raising the sanity line" to refer to increased usage of processes correlated with creating correct beliefs. Your hostile reaction to this understanding doesn't exactly advance the ball on figuring out how to actually achieve either goal.
I didn't use that premise. But I suppose it could be true anyway.
Given the context [http://lesswrong.com/r/discussion/lw/eqa/open_thread_october_115_2012/7jhy], I understood you to be saying that way to persuade theists was to demonstrate that atheists have higher social status. Regardless of the sociological truth of that assertion, we both know that "high-status" is not correlated with "correct beliefs." Did I misunderstand you?
I said something kind of similar to that so I'll accept it for the sake of the argument without implying any endorsement of the claim. I actually doubt this is true, that is there probably is such a correlation---the world is unfair like that. But putting that aside even if I assume the claims you make in the parent are true it still doesn't mean I had any particular premise (or conclusion) about the sanity line. I think you resolved the ambiguous 'this' in palladias's comment in a different direction.
Yeah, but probably it's a very weak one.
I'm pretty sure that at least some theists are sincere, rather than using belief-as-attire.
Most people are usually sincere and it remains the case that the most significant influence on most people's core ideological beliefs are social factors. That's how people work, for better or worse.
We can carry belief-as-attire slightly further and suggest that the totally sincere and unfaked embarrassment someone might feel about being caught pantsless in public is still governed by social factors. Your belief can functionally be attire without you recognizing it, much in the same way that someone in a suit just seems "well-dressed" even though they're not better dressed in an objective sense than say, someone in a traditional robe or animal skins.
I think a chain of feelings would be much more useful. As they say, you can't reason someone out of a belief they didn't reason themselves into.
Effectiveness should be measured by results. So you would have to look at deconversion stories. The problem of evil witnessed on a personal and intense level has done it for some. Historical scholarship on the origins of their religion has done it for others. For others again, materialist science leaving ever-shrinking gaps for God to be in.

Why I defend scoundrels, an awesome essay by Yvain. Seriously how can his blog be so good? I find myself linking to it all the time.

Virtualization. I think if you are virtualized (uploaded to a computer, or copied into a new brain), you still die. I keep running into people on here who seem to think that if you copy someone, this prevents them from dying. It seems that I am in the minority on this one. Am I? Has this been thoroughly debated before? I would like to start a discussion on this. Good idea / bad idea tips on presentation?

I think the LW consensus is that the copy is also you, and personal identity as we think of it today will have to undergo significant change once uploads and copies become a thing.

Contemporary people are more or less completely bamboozled by the whole topic of minds, brains, and computers. It's like in the early days of language, when some people thought that reality was created by a divine breath speaking the true names of things, or that the alphabet existed before the universe alongside God, and so on. Language was the original information technology that was made into an idol and treated like magic because it seemed like magic. The current attitudes to computers and computation are analogous, except that we really can culture neurons and simulate them, so we are going to be creating hybrid entities even more novel, in evolutionary terms, than a primate with a verbalizing stream of consciousness (which was a hybrid of biology and language).

What is the computational paradigm of mind? Often this paradigm floats free of any material description at all, focusing solely on algorithms and information. But if we ask for a physical description of computation, it is as follows: There is an intricate physical object - a brain, a computer. Mostly it is scaffolding. There are also non-computational processes happening in it - blood circulating, fan spinning. But amo... (read more)

If you use the concepts of "dying" or "personal identity" in this context, you risk committing the noncentral fallacy [http://lesswrong.com/lw/e95/the_noncentral_fallacy_the_worst_argument_in_the/], since uploading is an atypical case of their application, and their standard properties won't automatically carry over. For example, concluding that an instance of you "actually dies" when there is also a recent copy doesn't necessarily imply that something bad took place, since even if you do in some sense decide that this event is an example of the concept of "dying", this is such an atypical example that its membership in that concept provides only very weak evidence for sharing the property of being bad with the more typical examples. Locating this example in the standard concepts is both difficult and useless, a wrong question. The only way out seems to be to taboo the ideas of "dying", "personal identity", etc., and fall back on the arguments that show in what way typical dying is bad, and non-dying is good, by generalizing these arguments about badness of typical destruction of a person to badness of the less typical destruction of a copy, and goodness of not destroying a person to goodness of having a spare copy when another copy is destroyed. It seems to me that the valuable things about a living person (we've tabooed the "essence of personal identity", and are only talking about value) are all about their abstract properties, their mind, their algorithm of cognition, and not about the low-level details of how these abstract properties are implemented. Since destruction of a copied person preserves these properties (implemented in the copy), the value implemented by them is retained. Similarly, one of the bad things about typical dying (apart from the loss of a mind discussed above) seems to be the event of terminating a mind. To the extent this event is bad in itself, copying and later destroying the original will be bad. If this is so, destructive uploading
Almost everybody starts with the intuitive notion that uploading will kill the "real you". The discussion seems to have been treading the same ground since at least the 1990s, so I don't really expect anything new to come out of yet another armchair rehash. Chapters 9 and 10 in David Chalmers' singularity paper [http://consc.net/papers/singularity.pdf] are a resonably good overview of the discussion. Chalmers end up finding both stances convincing given different setups for a thought experiment, and remains puzzled about the question.
Really? I started with the assumption that uploading wouldn't necessarily be destructive, but people chose to discuss destructive uploading because it simplifies some of the philosophical questions. On second thought, there may also be a bias from science fiction, where promising developments are likely to have a horrific downside.
Yeah, assuming some sort of destructive upload in my comment there, naturally. My assumptions for the initial stance most people will have for the various scenarios are basically: * Non-destructive upload, the initial person remains intact, an exact upload copy is made: The "real you" is the original human, all that matters is whether real you lives or dies. * Destructive upload, the initial person gets knocked out and ground to pieces to make the exact upload copy: "Real you" dies from being ground to pieces, end of story. * Moravec transfer, the initial person's brain gets converted to a machine substrate one neuron at a time: People other than John Searle seem to be OK with personal continuity remaining in this scenario.
Also, embracing the possibility of nondestructive uploads requires us to think about our identities as potentially non-uniquely instantiated, which for a lot of people is emotionally challenging.
Define "dying".
Elements of death: There are a lot of elements to dying and if technology progresses far enough I think we could have incidents where some but not all of them happen. However, depending on what exactly happens, some of these should still be regarded as being just as bad as death. Death of experience * Your experience of the world stops permanently. * This is important because you will never experience pleasure again if you stop experiencing permanently. Death of self * Your personality, memories, etc, your "software pattern" cease to exist. * This is important because other people are attached to them and will be upset if they can't interact. Death of genes * Your genetic material, your "hardware pattern", is lost. Your genetic line may die out. * This is unacceptable if you feel that it's an important purpose in life to reproduce. Death of influence * It becomes impossible for you to consciously influence the world. * This is important because of things like the necessity of taking care of children or a goal to make a difference. Death of body * Your body, or the current copy of your "hardware" becomes unusable. * This is important if your brain isn't somewhere else when it happens but may not be important otherwise. There may be others. Can you think of more?
It's a good list. Now to define "you" and see if an upload fits into the definition and if so, how much of your list applies.
I am uploaded. A copy of my "self" is made (I believe this is the definition of "you" people are using when they're talking about uploading themselves) and the original is disassembled or dies of natural causes. That's all that was done. I'm assuming no other steps were taken to preserve any other element of me because it was believed that uploading me means I wouldn't die. I'll call the original Epiphany and the copy I'll call Valorie. Epiphany: Death of body - Check. Brain was in it? Check. Death of experience - Check. (See previous note about my brain.) Death of genes - Check. Pregnancy is impossible while dead. Genes were not copied. Death of influence - Check. Upload was not incarnated. Death of self - No. There is a copy. Valorie: Death of body - No body. It's just a copy. Death of experience - Doesn't experience, it isn't being run, it's just a copy. Death of genes - Doesn't have genes, a copy of my "self" is being stored in some type of memory instead of a body. Death of influence - Cannot influence anything as a copy, especially if it is not being run. Death of self - No. It's preserved. Conclusion: I am dead. Of course it's not hard to imagine other scenarios where everything possible is copied and the copy is incarnated, but Epiphany would still stop experiencing, which is unacceptable, so I would still call this "dead".
I'm perfectly willing to accept that if you get uploaded and then nobody ever runs the upload then that's death. But if you're trying to give the idea a fair chance, I'm not sure why you're assuming this.
There's one really important detail here. If you get uploaded, even if the copy is put into a body exactly like yours and your genes are fully preserved and everything goes right, you still stop experiencing as soon as you die. Is that acceptable to you?
Okay, I was pretty sure that was your real point, so I just wanted to confirm that and separate away everything else. But to be honest, I don't have a real answer. It's definitely not obvious to me that I will stop experiencing in any real way, but I have a hard time dismissing this as well. One traditional answer is that "you will stop experiencing" is incoherent, and that continuity of experience is an illusion based on being aware of what you were thinking about a split second ago, among other things.
The continuation of experience argument is compelling if you consider my transporter malfunction scenario [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l5a]. That is one situation that would definitely result in a discontinuation of experience. Others which I have discussed with Saturn and TheOtherDave (a wonderfully ironic handle for this discussion) have resulted in my considering other possibilities like being re-assembled with the exact same particles in the same or different locations [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l7c] and being transformed over time via neuron replacement or similar [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l3a]. I decided that being transformed would probably maintain continuity of experience, and being re-assembled out of the same particles in the exact same locations would probably result in continuity of experience (because I can't see that as a second instance), but I am not sure about it (because the same particles in different locations might not qualify as the same instance, which brings into question whether same instance guarantees continuous experience) and I'm having a hard time thinking of a clarifying question or hypothetical scenario to use for working it out. (It's all in the link right there). What's not incoherent, though, is looking forward to experiencing something in the future, yet knowing you're going to be disassembled by a transporter and a copy of you will experience it instead. That, in no uncertain terms, is death. We can tell ourselves all day that having a continuous experience relies on you being able to connect your current thought and previous thought, but the real question we need to ask is "Will I have any thoughts at all?" so the connected thoughts question is a red herring (as it relates only to your second instance, not your first one) and is a poor clarifying question for telling whether you (the original) survived. In coherent terms, what we
Either way, only a copy of you will experience it, because the non-copy of you is trapped in the present and has no way to experience the future. The copy can be made artificially, using a transporter, or naturally as time passes. Why is there a difference?
Why do you think that time copies you?
Well, it doesn't even perfectly preserve the original, so I fail to see what else it could be but a copy. You might argue that for some reason the time-derived copy is more important than an artificial copy, of course, but why?
Why is the time-copy even a copy though? If we call some A a copy of some original B, then we have to have reason to associate A with B (if A and B are paintings, the one is a copy of the other if it closely resembles it, say). What association does EpiphanyA at t0 have with EpiphanyB at t1?
You... don't see a reason to associate future-you with present-you?
Well, I think I persist through time. But you're saying that time makes copies of me, and I'm curious to know why you think those things are copies and not just new (very short lived) people.
I don't think the distinction is meaningful. Possibly we just mean different things by the word "copy"?
I think I should, at this point, just ask for some elaboration on the theory.
Wait, wait, wait. I'm still confused as to why you think that time is copying me. By what mechanism does time create new instances of me and destroy the old ones? At what interval does this happen? Has anyone actually observed this phenomenon or is it just a theory?
I could reverse the question. Why do you think you're the same person at different times, as opposed to being a copy? By what mechanism is a single person carried forward through time? Has anyone actually observed this phenomenon, or is it just a theory? It's not clear to me that those are fair questions, but then it's not clear to me that their reversals are fair, either.
Occam's razor. The theory that I'm being copied and destroyed over and over again doesn't explain anything additional that I can think of, so it's more likely the simpler idea (that I am not being copied and destroyed over and over) is true. Also, not believing that I am being copied does not qualify as a belief. That's just lack of belief in a theory. If you guys believe I'm being copied over and over again, that IS a belief though, and if you want me to agree with it, the burden of proof lies on you.
I think both of you are sorta failing to address (or not addressing clearly enough) the point that objects being "copied" "destroyed" or "persisted" is not really meaningful at the level of physics at all -- like envisioning electrons as billiard balls [http://lesswrong.com/lw/e7f/introducing_simplexity/], it's mapping a concept that's intuitive in one's mind onto the physical world where it does not apply. At the bottommost level of quantum physics that we know of, electrons have no identity. From what I gather to "destroy" an electron from here and "copy" it there is indistinguishable physically (even in principle) from "moving" it from here to there. Those are concepts which are differentiatable in our adapted-via-evolution minds, not in reality. That having been said I don't dismiss your concerns about uploading altogether because we still aren't unconfused enough about consciousness to be able to clarify to ourselves what the fuck it's supposed to do... I would really like to be unconfused about qualia and the nature of existence before I do any uploading of myself.
Yup. Which is why I say it's not clear to me those are fair questions. That said... if in the future two entities exist that are physically and behaviorally indistinguishable from one another, and one of them is me, it follows that either both of them are me, or one and only one of them is me. In the latter case, it seems "me-ness" depends some physically and behaviorally undetectable attribute which only one of them has. Occam's razor also seems to suggest that both of them are me, since the alternative posits an additional unnecessary entity in the system.
I'm interpreting this as difficulty figuring out who the burden of proof belongs to. I think it helps to realize that with each theory there are at least three options: Believe it's true. Believe it's false. Not believe anything. If you say "There's a dragon in my garage." and I say "I don't believe this." I am not saying "I believe there is no dragon in your garage." I'm saying "I don't have a belief about this." Now, I could go in there and inspect everything and conclude that there's no dragon, at which point I'd have a belief that there isn't a dragon. But why should I do this? You might claim next that there's a God in your garage. Then I'd have to go to all sorts of work trying to prove there is no God in your garage. Then you could claim that there's a pink elephant, and on and on. This is why, if you want people to believe something the burden of proof lies on you - you can't just turn it around and say "Well prove that it's NOT this way!" - if that were the rule, people would troll the crap out of us with dragons and Gods and pink elephants and such. Does that give you any clarity in whose burden it is to offer evidence regarding time copying people? No. The additional entity is not unnecessary. The second instance is absolutely required to explain the way you reacted to my teleporter with technical failure argument. I am surprised you didn't update after that by recognizing that there were two separate instances, and I don't know what to do about it. I'm stumped as to why you aren't seeing it this way.
Perhaps you are. That's certainly not what I would be saying if someone said that to me and I gave that reply. Proof in the sense you are discussing here is mostly useful when trying to win debates. I have no particular desire for you to believe anything in particular. The unnecessary entity in the second case is the physically and behaviorally undetectable attribute which only the "real me" has. I don't see any need for it, and I have no idea why you think it's necessary to explain any part of my reaction to any of your hypotheticals.
So your definition of self stops at the physical body? Presumably mostly your brain? Would a partial brain prosthesis (say, to save someone's life after a head trauma) mimicking the function of the removed part make the recipient less of herself? Does it apply to the spinal cord? How about some of the limbic system? Maybe everything but the neocortex can be replaced without affecting "self"? Where do you put the boundary and why?
No. As I mentioned, "This (referring to Death of Body) is important if your brain isn't somewhere else when it happens but may not be important otherwise." If you get into a good replacement body before the one you're in dies, you're fine. If you want to live, a continuation of your experience is required. Not the creation of a new instance of the experience. But the continuation of my (this copy's) experience. That experience is happening in this brain, and if this brain goes away, this instance of the experience goes away, too. If there is a way to transfer this experience into something else (like by transforming it slowly, as Saturn and I got into) then Epiphany1's experience would be continued. If Epiphany1's experience continues and my "self" is not significantly changed, no. That is not really a new instance. That's more like Epiphany1.2. Not sure why these are relevant. Ok limbic system is sort of relevant. I'd still be me with a new spinal cord or limbic system, at least according to my understanding of them. Why do you ask? Maybe there's some complexity here I missed? Hmmm. If my whole brain were replaced all at once, I'd definitely stop experiencing. If it were replaced one thing at a time, I may have a continuation of experience on Epiphany1, and my pattern may be preserved (there would be a transformation of the hardware that the pattern is in, but I expect my "self" to transform anyway, that pattern is not static). I am not my hardware, but I am not my software either. I think we are both. If my hardware were transformed over time such that my continuation of experience was not interrupted, then even if I were completely replaced with a different set of particles (or enhanced neurons or something) that as long as my "self pattern" wasn't damaged, I would not die. I can't think of a way in which I could qualify that as "death". Losing my brain might be a cause of death, but just because something can cause something else doesn't mean it does in
So the "continuity of experience" is what you find essential for not-death? Presumably you would make exceptions for loss of consciousness and coma? Dreamless sleep? Anesthesia? Is it the loss of conscious experience that matters or what? Would a surgery (which requires putting you under) replacing some amount of your brain with prosthetics qualify as life-preserving? How much at once? Would "all of it" be too much? Does the prosthetic part have to reside inside your brain, or can it be a machine (say, like a dialysis machine) that is wirelessly and seamlessly connected to the rest of your brain?
If it helps, Epiphany has implied elsewhere, I think, that when they talk about continuity of experience they don't mean to exclude experience interrupted by sleep, coma, and other periods of unconsciousness, as long as there's experience on the other end (and as long as the person doing that experiencing is the same person, rather than merely an identical person).
Right, it's her definition of "same" vs "identical" that I am trying to tease out. Well, the boundary between the two.
Yeah that has gotten tricky. I've worded the question as "Same instance or different instance?". I've also discovered a stickier problem - just because a re-assembled me might qualify, in all ways, as "the same instance" I am not sure that guarantees the continuation of my experience. I explore that here, in two examples being re-assembled from the same particles both in the same arrangement and in a different arrangement [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l7c]. (scroll to "Scenarios meant to explore instance differentiation and the relation to continuous experience" - I labeled it to make it easy to find.)
As TheOtherDave pointed out, the question is what is, in your opinion, the essence of "self". Clearly it cannot just be all the same "particles" (molecules?), since particles in our bodies change all the time. You seem to be relating self with consciousness, but not identifying the two. That's why I'm asking questions aimed to nail the difference. That's why I asked these questions earlier:
"The essence of self" seems like the wrong question to me. That sounds too much like "What is the essence of your personality?" and that's irrelevant here. What I'm talking about is my ability to experience. We all have an ability to experience (I assume) that, although it may be shaped by our personalities, it is not our personalities. Example: A Christian sees a Satanic ritual. A Satanist sees the same ritual. The Christian is horrified. The Satanist thinks it's great. The reason one was horrified and the other thought it was great is because they have different beliefs, possibly different personality types, different life experiences and possibly even different neurological wiring. What did they have in common? They both saw a Satanic ritual. THAT is the part I am trying to point out here. The part that experiences. It's not one's personality, or beliefs, or experiences or neurological traits. I am saying essentially "Even if personality, beliefs, experiences and neurological differences are copied, this does nothing to guarantee that the part of you that experiences is going to survive." Asking to define the essence of self is not relevant since I'm saying to you "Even if self is copied, this thing that I am talking about may not survive". Here is a clarifying example: Transporter Malfunction Scenario [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l5a]
Note to self: Thinking about motion might be the key to this.
How would you convince someone who thinks instants of experience are real and memories that give instants of experience historical context are real, but doesn't believe in any meaningful process of forward continuity from one instant of experience to another beyond the similarity of memories, to believe otherwise? There's no difference between blinking, taking a nap and being destructively teleported in this stance. It's all just someone experiencing something now, and someone else with very similar memories that include the present experience moment experiencing something else in the future.
Well, that makes the second time you ignored my questions, so I will tap out.
I've noted to self that this seems like a pattern with us, as you have complained about a question being ignored a few times now. Not sure what I should be doing about it when I don't see a question as relevant but maybe I should just be like "I don't see how this is relevant." Don't know how I got the habit of ignoring things that seem irrelevant and moving on to whatever seems relevant but I can see why it would be annoying so I will be thinking about that. Thanks for getting me to see the pattern.
Questions to consider: Would you feel the same way about using a Star Trek transporter? What if you replaced neurons with computer chips one at a time over a long period instead of the entire brain at once? Is everyone in a constant state of "death" as the proteins that make up their brain degrade and get replaced?
The million dollar question: Do I stop experiencing? If I were to be disassembled by a Star Trek transporter, I'd stop experiencing. That's death. If some other particles elsewhere are reassembled in my pattern, that's not me. That's a copy of me. Yes, I think a Star Trek transporter would kill me. Consider this: If it can assemble a new copy of me, it is essentially a copier. Why is it deleting the original version? That's a murderous copier. I remember researching whether the brain is replaced with new cells over the course of one's life and I believe the answer to that is no. I forgot where I read that, so I can't cite it, but due to that, I'm not going to operate from the assumption that all of the cells in my brain are replaced over time. However, if one brain cell were replaced in such a way that the new cell became part of me, and I did not notice the switch, my experiencing would continue, so that wouldn't be death. Even if that happened 100,000,000,000 times (or however many times would equate to a complete replacement of my brain cells) that wouldn't stop me from experiencing. Therefore, it's not a death - it's a transformation. If my brain cells were transformed over time into upgraded versions, so long as my experience did not end, it would not be death. Though, it could be said to be a transformation - the old me no longer exists. Epiphany 2012 is not the same as Epiphany 1985 because I was a child then, but my neural connections are completely different now and I didn't experience that as death. Epiphany 2040 will be completely different from Epiphany 2012 in any case, just because I aged. If I decide to become a transhuman and the reason I am different at that time is because I've had my brain cells replaced one at a time in order to experience the transformation and result of it, then I have merely changed, not died. It could be argued that if the previous you no longer exists, you're dead, but the me that I was when I was two years old or ten y
I used to think this way. I stopped thinking this way when I realized that there are discontinuities in consciousness even in bog-standard meat bodies -- about one a day at minimum, and possibly more since no one I'm aware of has conclusively established that subjective conscious experience is continuous. (It feels continuous, but your Star Trek transporter-clone would feel continuity as well -- and I certainly don't have a subjective record of every distinct microinstant.) These are accompanied by changes in physical and neurological state as well (not as dramatic as complete disassembly or mind uploading, but nonzero), and I can't point to a threshold where a change in physical state necessitates subjective death. I can't even demonstrate that subjective death is a coherent concept. Since all the ways I can think of of getting around this require ascribing some pretty sketchy nonphysical properties to the organization of matter that makes up your body, I'm forced to assume in the absence of further evidence that there's nothing in particular that privileges one discontinuity in consciousness over another. Which is an existentially frightening idea, but what can one do about it? (SMBC touched on this [http://www.smbc-comics.com/index.php?db=comics&id=2705#comic] once, too.)
What do you mean by discontinuities? I have not heard about this.
Sleep, total anesthesia, getting knocked on the head in the right way, possibly things like zoning out. Any time your subjective experience stops for a while.
Actually, I expect that our normal waking experience is also discontinuous, in much the same sense that our perception of our visual field is massively discontinuous. Human consciousness is not a plenum.
Yeah, I was trying to get at that with the parenthetical bit in my first paragraph. Could probably have been a bit more explicit.
Ok are you saying that temporarily going unconscious is the same as permanently going unconscious? Would you assert that because we temporarily go unconscious that permanent unconsciousness is not death?
Temporarily going unconscious is not the same as permanently going unconscious. Whether we temporarily go unconscious or not does not entail permanent unconsciousness being or not being death. Now, some questions of mine: you said "If I were to be disassembled by a Star Trek transporter, I'd stop experiencing. That's death." When you fall asleep, do you stop experiencing? If so, is that death? If it isn't death, is it possible that other things that involve stopping experiencing, like the transporter, are also not death?
We need to focus on the word "I" to see my point. I'm going to switch that out with something else to highlight this difference. For the original, I will use the word "Dave". As tempting as it is to use "TheOtherDave" for the copy, I am going to use something completely different. I'll use "Bob". And for our control, I will use myself, Epiphany. Epiphany takes a nap. Her brain is still active but it's not conscious. Dave decides to use a teleporter. He stands inside and presses the button. The teleporter scans him and constructs a copy of him on a space ship a mile away. The copy of Dave is called Bob. The teleporter checks the copy of Bob before deleting Dave to make sure he was copied successfully. Dave still exists, for a fraction of a second, just after Bob is created. Both of them COULD go on existing, if the teleporter does not delete Dave. However, Dave is under the impression that he will become Bob once Bob exists. This isn't true - Bob is having a separate set of experiences. Dave doesn't get a chance to notice this because in only fractions of a second, the teleporter deletes Dave by disassembling his particles. Dave's experience goes black. That's it. Dave doesn't even know he's dead because he has stopped experiencing. Dave will never experience again. Bob will experience, but he is not Dave. Epiphany wakes up from her nap. She is still Epiphany. Her consciousness did not stop permanently like Dave's. She was not erased like Dave. Epiphany still exists. Bob still exists. Dave does not. The problem here is that Dave stopped experiencing permanently. Unlike Epiphany who can pick up where Epiphany left off after her nap because she is still Epiphany and was never disassembled, Bob cannot pick up where Dave left off because Bob never was Dave. Bob is a copy of Dave. Now that Dave is gone, Dave is gone. Dave stopped experiencing. He is dead.
Ah! So when you say "If I were to be disassembled by a Star Trek transporter, I'd stop experiencing" you mean "I'd [permanently] stop experiencing." I understand you now, thanks. So, OK. Suppose Dave decides to go to sleep. He gets into bed, closes his eyes, etc. The next morning, someone opens their eyes. How would I go about figuring out whether the person who opens their eyes is Dave or Bob?
Well, first, is there a human copier nearby? If not, you're probably Dave. How about this: If you had stepped into a teleporter and pressed the button, how would you know that it killed you?
This is exactly backwards. I recognize a copier because it makes copies. That's how I know something is a copier. If I need to know whether something is a copier before I can decide whether what it creates is a copy or not, there's something wrong with my thinking. I wouldn't, naturally. Of course, if Dave steps into an incinerator and presses the button, Dave also doesn't know that the incinerator killed Dave. Dave is just dead, and knows nothing. OTOH, if Dave steps into a non-incinerator and presses the button, Dave knows it didn't kill Dave. And the way that Dave knows this is that something is standing there, not-dead, after pressing the button, and that something identifies as Dave, and resembles Dave closely enough. This happens all the time... I have pressed many buttons in my life, and I know they haven't killed me, because here I am, still alive. And I expect this is exactly what happens with a properly functioning teleporter. I press the button, and in the next moment something is aware of being Dave, and therefore not dead. It just happens to be in a different location.
Okay, so would you recommend I check under my bed tonight for anything that might make a copy of me and disassemble the original? I need something more to go on. I'm having a hard time not equating this with worrying about boogeymen. Actually, for at least a few seconds, possibly a few minutes, Dave would be screaming in agony and he would most certainly notice that he is experiencing death by incineration. Unless the non-incinerator happens to be a human copier, and Dave did not recognize it at first. Yes, exactly. The original Dave has died in such a way that he didn't even notice. Dave2 definitely doesn't want to think that an exact copy of himself died just a moment ago, and really definitely doesn't want to have to worry that he will need to cease experiencing in order to "go back" to where he came from, so due to normalcy bias, Dave2 declares that the fact that Dave2 exists means that Dave1 never died, and enjoys the confirmation bias that this non-sequitur gives him until he ceases to experience when "loaded" back onto his space ship. That's one insidious death. Two, actually. :p
Indeed! And you should equate it with worrying about boogeymen. It's a silly thing to worry about. The question is why it's silly. I would say it's silly, not because I haven't noticed any boxes marked "human copier" under my bed, because every time in the past that I've woken up I've resembled the person who went to bed so closely that it's been ridiculous to worry that I might not be the same person. Nope. Dave would notice that he's experiencing being incinerated, certainly, if the incinerator were as slow as you describe. But he would not experience death by incineration. He wouldn't experience death at all. Here's how I know: as long as Dave is experiencing anything, Dave isn't yet dead. And if he's not dead, he certainly can't be experiencing death. (nods) Just like his predecessor did the night before when he went to bed, and Dave woke up in his place. But of course, as above, that was too silly to worry about, just like boogiemen. So is this.
Okay, I guess you were trying to say that my concern about being disassembled after being copied as a method of "transportation" is the equivalent of worrying about boogeymen? "OH GOD I'M DYING AHHH!" < I call this experiencing death. Different definitions, I guess. If you want to get technical about it, and talk about death in a solely tangible way, sure Dave isn't dead when he's thinking about that. But Dave is experiencing death emotionally and intellectually. He knows he's in the process of dying, that death is inevitable. He also feels emotional (and, well, physical) pain that amount to an experience worthy of symbolizing death. Maybe it would be more grammatically correct though if I said he is experiencing dying. In any case, I meant to differentiate this from transporter death because with transporter death, Dave believes that he is going to survive the "transportation" and doesn't feel any emotional or physical pain, so there's no knowledge of or suffering about his death. If I offered you the free use of a device that could make a copy of you and put it anywhere you want and cause the current you to be disassembled and dispersed in the surrounding environment, (2-way trip) would you use it?
(shrug) OK, sure. Incidentally, by your definition, many many people walking around today have experienced death. Hell, I've experienced death myself. Anyway, using your definition, if I stepped into what I thought was a molecular disassembler that would kill me, and it disassembled me slowly enough that I experienced the process of being disassembled, I would "experience death" by your definition, and I would know I'd experienced it the same way I know I experience the taste of cheese when I experience the taste of cheese. Later, I would look around the teleport receiver booth and say "Huh. I'm not dead? Cool" and go on with my life. That is, I would have "experienced death" but not actually died, just as many many people do in real life when they wake up after heart attacks, accidents, etc. Assuming that it reliably creates that copy? Absolutely. Far more convenient than airplanes. (By "reliably" here I just mean that I trust it to actually create a close-enough copy, and not to instead create some imperfect copy that does not resemble me closely enough to satisfy my preferences regarding consistency over time.)
Yes. I already know what your bumper sticker in the future is going to say: I break (down) for transporters! Now, say the transporter has a malfunction at the exact fraction of a second between the time when Dave2 has been verified as a complete copy and the time when Dave1 is going to be disassembled. The technician says it's going to take three hours to fix. You go out and catch a movie. After the movie, you go outside and stretch, and you see that it's a beautiful day. You have two options: 1. Go to the transporter and get disassembled. 2. Avoid getting disassembled by the transporter. What do you choose?
I choose #2, of course. More than that... if I arrive at the transporter complex and am told that this is an option, that I can duplicate myself and send one copy to my destination while the other one stays here, I absolutely prefer to be duplicated... no reason for a conveniently timed technical failure. Indeed, I might postpone the trip altogether and spend the next week right here hanging out with myself and having threesomes with our husband and meeting with lawyers to figure out what we do with our funds and material goods. Relatedly, given a button that I know creates two perfect copies and then picks one of the resulting three Daves at random to destroy an hour later, I press it. At the time of pressing the button, I'm indifferent as to which of the three copies gets selected for destruction... they are all me. After pressing the button, one of me goes "Crap! I'm going to die in an hour!" and is unhappy about it, and the other two of me go "Whew! Dodged that bullet!" but feel bad for the third of me. On my account it does not matter in the least which one of the three "was the original me," assuming there's even any way to tell, which there may not be. Now, a question for you. I enter a spaceship traveling to Alpha Centauri in suspended animation, along with all my friends and loved ones. We could have teleported instead, but we've been convinced by your account that this would be suicidal, so we opted for the slower but safer route. While we lie in frozen sleep, the spaceship has a technical failure in mid-flight which reduces the ship and everything in it to constituent atoms. The ship's captain has the option of using the ship's transporter to beam us from the doomed ship to the surface of Alpha Centauri. As far as I can tell, on your account, there's no particular reason why she should do so... either way, we're all going to die. Sure, if she does so some complete strangers will pop into existence on Alpha Centauri, but what has that got to do with
Scenario meant to discover whether the experience of life is valued Okay, so I guess what you're saying here is that what you value about being alive is NOT the experience of life. How do you feel about this scenario: You and your husband are planning to go to a really awesome event soon. Maybe it's the Singularity summit, maybe your favorite rock star is having a concert, maybe it's the birth of a new baby you guys have been wanting for a long time. Imagine whatever sort of event you'd enjoy most. You're really looking forward to it! Then work calls and says "Dave, two days from now, we need you to do this really important job 3,000 miles away from your ordinary work site. We couldn't get you a plane ticket on such short notice, but fortunately we have a transporter." You agree, as it is your job. Now you hang up the phone and your husband comes over, saying "I can't believe we're actually going to have this event soon! Isn't it exciting!" "Yeah, of course!" You say. But something feels wrong. You realize that you are going to be disassembled by the transporter BEFORE the event happens. YOU won't experience the event whatsoever. A copy of you will be there instead. Is this acceptable? I certainly don't want to live a lifestyle where we use transporters to go everywhere and each instance of me only experiences until the next transport. My life would never be long enough to experience any satisfaction. That's reminiscent of Alice in Wonderland's absurd circumstance: "Jam tomorrow, jam yesterday, but never ever jam today." A new instance of me can experience a future event I've been planning for tomorrow, and a past me may have experienced a continuous life before transporters, but most instances of me would just be slaving away during the few hours or days in which they experience, doing things like working or buying groceries, so that other temporary instances of myself can reap the rewards. The instances that do get a reward still wouldn't get to exper
Nope, that's not what I'm saying at all. All of the Daves have the experience of life, and I absolutely do value it, which is why I press the button that I expect to create more of it. No, that simply isn't true. I will in fact experience the event (assuming I can get back from my work assignment in time, or assuming that my employer uses a nondestructive teleporter such that I can both experience the event and do my job). No, sorry, I was unclear. The engine is going to overload in ten minutes, say, and the captain has the choice of transporting us off the ship before it explodes. Which, on your account, is not worth bothering with, since we're going to be just as dead whether she does or not. Sure. Given the choice of telecommuting this way, rather than teleporting my body back and forth, I would probably choose tele-operating a remote body, assuming the experience was comparable. No, not really, especially since you're in the habit of not answering the questions I do ask. Either way, though, no: I think you've created a confusion here that is unresolvable as long as you hold on to your belief that there is some essence of selfness (continuous experience, identity, real-me-ness, whatever) that is undetectable and unduplicatable but somehow still important. Your model creates the possibility that I am not the person I was a moment ago and there's simply no fact about the world that would resolve the question of whether that possibility is actual or not. This seems absurd to me: if nothing depends on it, I simply don't care whether it's true or not; if we insist that that is what it means to be "really me", then I must accept that maybe I'm not "really me" and I'm OK with that.
What motivates you to link personal identity to your specific particles? Any two atoms of the same type are perfectly indistinguishable.
I haven't touched on personal identity - for clarity I'm not equating that with continuous experience nor am I even equating continuous instance distinctions with continuous experience at this point. (I guess I'm interpreting personal identity either like "self" or identity the way it's used in "identity theft" - like a group of accounts and things like SSNs that places use to distinguish one person from another. I'm not using that term here and I'm not sure what you mean by it.). I'm not trying to figure out whether my "self" maps to certain particles. I feel sure that "self" is copy-able (though I haven't formally defined self yet). However, I am separating self from continuous experience (like you can see in my Elements of Death comment [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l4a]). What I am trying to do is to figure out whether the continuous experience of my current instance is linked to specific particles. The reason I am asking that question is made apparent in my transporter failure scenario [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7l5a].
Note to self: "I break (down) / break down / breakdown / brake down / brakedown for transporters!" all get zero Google results. Yay.
Now they don't.
No, temporary unconsciousness is not the same thing as permanent unconsciousness; you perceive yourself to return to consciousness. The tricky part is unpacking the "you" in that sentence. Conventionally it unpacks to a conscious entity, but that clearly isn't useful here because you (by any definition) aren't continuously conscious for the duration. It could also unpack to about fifty to a hundred kilos of meat, but whether we're talking about a transporter-clone or an ordinary eight hours of sleep, the meat that wakes up is not exactly the meat that goes unconscious. In any case, I'm having a hard time thinking of ways of binding a particular chunk of meat to a particular consciousness that end up being ontologically privileged without invoking something like a soul, which would strike me as wild speculation at best. So what does it unpack to? It's actually very tricky to pin down the circumstances which constitute death, i.e. permanent cessation of a conscious process, once you start thinking about things like Star Trek transporters and mind uploading. I don't claim to have a perfect answer, but I strongly suspect that the question needs dissolving rather than answering as such.
It seems like you both die and live. It also seems like there become two different versions of you. If the original is deleted immediately; I don't think you die.
I think there no such mystery about pattern continuation. People just keep confused when the word "identity" come. If you really bother about these things, think in normal cases like you now and tomorrow, and find a flaw in the argument.

Has anyone noticed any problems with the comment score below threshold feature? I have my preferences to show all comments, regardless of score, and this comment is hidden from me for some reason.

Yes, and here's [http://lesswrong.com/lw/emo/memes/7hwe] another. On the page of all that user's posts, it does appear, at -4. But the permalink to the post shows it as hidden. Is this a bug or a new feature?
My guess is that it's a bug caused by the new troll feeding penalty where you automatically lose 5 karma if you reply to a comment at -3 or below. But I don't really have any evidence to corroborate that. And I don't think downvoted comments have ever been hidden if on a users page. I'm not sure though, I set my preferences to see all comments as soon as I created my LW account. I have negative associations with things being hidden from me and I'm too curious not to open it up and look anyway.
Huh, ditto, and un-downvoting it didn't make it stop being hidden.

What's the best place for LW feature requests: I'd like to be able to walk up comments all the way to the top comment, using "Show more comments above". As it is currently implemented, there is no way to differentiate between "button does nothing because it only work a certain number of levels up", and "reached the highest level, i.e. the top level comment".

Seconding this. You can get around it by clicking the permalink button to the topmost comment and traveling up further from that but it's still anoying
And I'd like "context" to show (or perhaps have an option to show) the whole tree which includes a comment rather than a single thread.
If you click the permalink on any comment, you get the whole tree descending from it.
You can use the "parent" button. It sits beside the reply button and looks like a bent arrow that goes left and up. When you are at the top level comment the parent button disappears.
Neat; that makes my request more of a bug report than a call for a new feature.

Are you a Bigot? --- a good 5 minute youtue video, it works as introductory level rationality material

A main point is that you're bigoted if you only listen to critiques of people you disagree with (or set critique on authomatic as you read or listen to them) rather than paying attention to their words with attention to those people intend.

I've been looking for a site that offers calibration tests from a farily large bank of questions, but I haven't really been able to find any. I found some resources from the last place this was discussed, but none of the sites had very many questions and most of the questions were very US centric.

Does anyone know of anything else?

So I was musing about 'one man's modus tollens is another man's modus ponens', and about how you would put that in probabilistic terms.

It seems to be applicable to when you have a probability for P(A v B), update on positive evidence to P(A' v B'), but instead of AB' (or same thing, A>A' and B<B').

I'm just wondering what additional stuff you need to get that; nothing mentally pops out for me as relevant.

The way I see it is this: We consider A likely, but B unlikely. Say P(A)=1-a and P(B)=b, where a and b are small. A and B are currently independent. Then we observe that A implies B (i.e. "¬A v B"). We get probabilities P(A|A=>B) = P(A=>B|A)P(A)/P(A=>B) = P(B)P(A)/[P(B)P(A)+P(B)P(¬A)+P(¬B)P(¬A)] = [b-ab]/[b+a-ab] or approximately P(A|A=>B) = b/[a+b] P(B|A=>B) = P(A=>B|B)P(B)/P(A=>B) = 1.P(B)/[P(B)P(A)+P(B)P(¬A)+P(¬B)P(¬A)] = b/[b+a-ab] or approximately P(B|A=>B) = b/[a+b]. So (to first order) we have that observing "A implies B" gives a probability of b/[a+b] for both A and B. So if we're more sure that A is true than that B is false we have a 0.5; both A and B are likely (Modus Ponens). But if we're more sure that B is false than that A is true we have b<a so b/[a+b] < 0.5; both A and B are unlikely (Modus Tollens). To summarise: "one man's modus tollens is another man's modus ponens" occurs when two beliefs that we strongly believe come into conflict. In this situation our final beliefs depend on our relative confidences in our two beliefs. We keep the one we are more confident in, and discard the one we are less confident in. This means that two people who both strongly held those two beliefs could suddenly find themselves in disagreement; this happens when one of them thought that a<b and the other one thought that b<a. EDIT: Note that I used A likely and B unlikely because the standard phrasing of MP is A and A=>B together give B. If we had taken ¬A instead of A we would have the situation where we have two unlikely beliefs, and we suddenly learn that one or the other has to be true. Similarly, if we had had ¬B instead of B we would have had the situation where we have two stong beliefs, and we suddenly learn that they conflict with each other. All these situations are equivalent.

I'm writing my CV now and was wondering whether I should indeed be "as confident as possible" (which basically means, according to some people, that I'm limited to sentences that don't even contain words like "but", "mostly", "although" etc.). Overconfidence is a killer of rationality, and displaying it might signal that you're irrational. I would personally trust much more someone who actively doubts in many things he says, rather than someone who is always confident. However, some people say the opposite.

I was wond... (read more)

Yes you should be as confident as possible.

In interview, you can admit that you used to have flaws, which you identified and corrected, but this is as close as you can get.

Why do you think so? I would personally like more people who are actively talking about their good and bad sides, although I'm not sure if I'd do that in an interview, because it might mean they don't know what appears to be the most effective strategy.
I don't know man, but that weird habit of humans drives me up a friggin wall.
I think, just like you have been recommended to assume the person conducting the interview is not rational, that person has as a default you are not rational. To expand: in general someone filling a job position will have many applicants. They will go through a negative search procedure: looking for ways to quickly discard applications. Since most applicants are overconfident, sounding less confident means you are perhaps less skilled than the overconfident applicants. Furthermore, employees who are confident about their deadlines (and meet them of course) are most valued. To rephrase colloquially: if you aren't confident about your skills/accomplishments/whatever, why should the person making the hiring decision be confident in them, when they know less about you than you do?
As a first approximation, assume everyone you're dealing with is default-level irrational and incapable of recognizing or appreciating rationality. This is true > 98% of the time. Also be aware that even a rationalist in a hiring situation might just interpret your "self skepticism" as attempted tribal-affilation signalling. They are hiring, not looking for beer buddies. Very different thought process.
I've heard that cover letters are not very popular these days, some people are doing away with them and viewing them as just another thing that can get you rejected. Before you put a lot of effort into this, you might want to check around and see if anyone even wants cover letters anymore. I know at least one significant company that does not even accept them.
I'm talking about personal statement. Not sure if this is the same as cover letter, but I do know that they require it. And it appears that mine is going to significantly increase the overall quality of the CV.

Dr Doug works through all the numbers for the UK National Lottery: Mistaken gambling, The secret thing, The Lottery Thing. (And, before: Making a hash of it, Making the future.)

Doug's blog is ace. I very much recommend it!

I have seen people mention two algorithms to decide whether to upvote or downvote a comment: 1) upvote/downvote it if you'd like to see more/fewer comments like that, and 2) assign it a karma score you think it deserves, look at its current karma, and upvote/downvote it if the former is above/below the latter. I've recently thought about a compromise: 3) assign it a karma score you think it deserves, multiply its current karma by a, and upvote/downvote it if the former is above/below the latter. Note that 3) reduces to 1) as a approaches 0 and to 2) as a approaches 1. (I'm using a = 0.5.)

Does this have any obvious drawback that neither 1) nor 2) has?

I wager that most people don't use an algorithm beyond "I feel like upvoting/downvoting this comment", they just click and then explain/rationalize their actions.
Yeah, but still, do they look at the karma score when deciding that?
My first filter is how I feel about the comment, and my second is a check on whether its karma level looks reasonable to me.
Same here. If it is already downvoted, the signal that "this is not a valuable comment" is already there, thus there is less reason (maybe no reason?) for me to add a downvote. Downvoting an already downvoted comment just seems like punishment, which I am not a fan of.
If everyone follows this policy then all it serves to do is discard most of the information that karma is intended to communicate. Comments that would be voted to -1 with voting as it is currently done would be indistinguishable from comments that nearly everyone downvotes. The -1 comment author is left unsure whether on net merely one person disapproved or whether he is making an extreme faux pas. Observers are left with the same information, if appearences matter. The -1 represents something far more significant than it does now. To the extent that punishment is involved at all the punishment has merely been redistributed along with uncertainty.
There's already uncertainty. A comment that 1 person has downvoted will look identical to a comment that 24 people have upvoted & 25 have downvoted. If the system was designed differently, for example by showing how many upvotes & downvotes individually a comment has received, then your criticism would make more sense to me. Please let me know if I'm misreading you.
There is more uncertainty. Significantly more. I was careful to use 'net' so as not to be commenting on what seems to be the distinct issue of displaying up and down votes separately.
Can't speak for others, but my guess is that some do and some don't, and those who do may or may not use the equalization approach 2. Maybe someone should consider making a list of testable models.
I do something a little like your (2), except that I don't downvote comments that I think deserve a positive score, and vice versa.

Recent arXiv pre-prints

arXiv is a well-known preprint server for mathematics, computer science, physics, etc. In exchange for weakening the demands of peer review, it encourages people to share articles at a much faster pace than would be possible otherwise. I've been a long-time subscriber of their RSS feed, which helps me keep abreast of developments in my field. On a typical day, between 100~150 new preprints are submitted, of which I usually find five or six "interesting."

So in accordance with this I have added this week an additional "i... (read more)

During a recent real life encounter I saw something that I am almost certain is a statistical fallacy, and I am trying to find the formal name for it. As the incident involved a political topic I am filling the serial numbers off. Someone pointed out that in population P, a rather nonstandard group, subgroups a and b suffered from (high number)% frequency of untimely death and presented this as evidence that a and b were being discriminated against, without provided the base rate for population P/ the death toll for non a, non b, members of P. Can anyone help me out here?

edited for grammar/clarity

The base rate fallacy [http://en.wikipedia.org/wiki/Base_rate_fallacy] seems like an appropriate name, but in practice it seems like that's reserved specifically for confusing Pr[A|B] with Pr[B|A] in the way outlined in the Wikipedia article.
Appropriately enough, ignoring the base rate of an event is known as the base rate fallacy [http://en.wikipedia.org/wiki/Base_rate_fallacy].

Why wasn't slavery outlawed quickly after the US started? I would expect the free non-slaveholders would vote against slavery, since they wouldn't want to compete with slaves, and they'd outnumber the slaveholders.

I don't know anything about the politics of slavery in the US at the time, but in general, a relevant question is: how strongly did the non-slaveholders desire slavery to be outlawed, as compared to their desires with regard to other issues?

In general, in politics it's quite common that the majority of the populace has a moderate preference to do X, a much smaller fraction of the populace has a very strong preference to not do X, and the desires of the minority win out. For the majority, the issue might not be important enough that they'd change their vote because of it, especially if the politician in question supports other issues who the people feel are more important. For the minority, however, the issue may be important enough to be the deciding factor in whom they vote. So the politican will maximize their votes by doing what the minority wants with regard to issue X, and what the majority wants with regard to everything else. At the same time, if the minority and majority vote for different politicians, then it's beneficial for the elected politicians to barter votes, so that the majority "buys" the minority's support for laws that they might not be able to pass ot... (read more)

On a somewhat related note, non-slaveholders often bought goods that involved slave labor at some point in the process of their production. It's quite possible that they at least thought that freeing the slaves would raise the cost of their tobacco, clothes ect. edited to add This is a bit more speculative, but I suspect that labor was significaly non-fungibal, and in particular that Northerner's didn't consider Southern agricultural jobs to be closely tied to their own labor market.
Also, poor white Southerners supported slavery because they had racial pride.
Why not also vote to prohibit holding capital? People get rich by owning capital, and it's hard to compete with them if you don't. What's the difference? I think you're conflating your ethical views on slavery with what you wish other people would decide for consequential reasons.
Workers compete with other workers, not capitalists. Worker wages are positively associated with the level of capital accumulation in society. In the slavery example, free workers compete with slaves, not slave owners.
People have, but it's not that common. Also, when communism is instituted, it's often by revolution rather than voting. There is a lot of incentive for the rich and powerful to be against it, but it only takes a comparatively small number of people who'd rather be powerful than rich. Were there enough historical examples of this happening and failing horribly to stop this? Were the elite just really good at convincing people it was a bad idea? If so, did their ability to do so correlate with it actually being a bad idea?
Perhaps I should have explained better, but that was sort of my point. You ask why an event didn't happen, and I pointed out that the event seems to be rare. I don't think it requires a circumstances-specific reason. Or, put more simply: people didn't do that because people didn't do stuff like that then. Also, I don't think there are many instances of people voting out the capitalists before the Communist Manifesto, which was published in 1848. That's well after "quickly after the US started". I don't think the timing is entirely coincidental.
Can you give me a circumstances-nonspecific reason? I haven't heard of any, but I don't know much history.
Sure, here are a few off the top of my head, not having done any research. There are strong social norms against taking other people's property. The prevailing culture of the time held this norm. Other people owning slaves has little direct impact on most people. I don't think the economic competition argument was one likely to be known or understood by those competing with slaves, so I don't think they would have made it. People who could make that argument were busy doing other things with their wealth. Status quo bias is generally strong. For circumstances-specific logic, I'd suggest reading up on the period debates surrounding slavery; there certainly were some at high levels. I think the decisions were mainly made on political grounds, and by people who liked the economics the way they were.
For one thing, one of the main groups suporting seperating from Great Britain (especially in the southern colonies) were slaveholders who were scared that the British were going to outlaw slavery (something that they had threatened to do in response to the revolution).
That would explain why they would try to institute a form of government that would not abolish slavery. I'm more interested in how the particular form of government they instituted would not immediately abolish slavery. CronoDAS largely answered that question.
In addition to the other points made in response to your question, a national law abolishing slavery would have needed to pass the US Senate, where each state got two votes, regardless of population. By the time abolition was something that might plausibly have passed the popular vote, the Southern states had formed a unified bloc on the issue. Admission of new states into the Union was explicitly evaluated on the basis of the balance of the Senate [http://en.wikipedia.org/wiki/Missouri_Compromise] until this principle came into too much conflict with the principle of popular sovereignty [http://en.wikipedia.org/wiki/Kansas%E2%80%93Nebraska_Act].
I was thinking state laws. It doesn't seem like a national issue, and it was banned in several states pretty early. The question isn't why the north didn't outvote the south. It's why the south voted against it in the first place.
How are you imagining the US government enforcing the abolition of slavery ca. 1800? Even in a much stronger relative position ca. 1865, it was extremely costly to do so. There was fair less abolitionist sentiment in earlier decades, and in relative terms, the federal government was far weaker and the southern elites far stronger. Attempting to outlaw slavery "quickly after the US started" (I'm assuming a window from about 1790-1810, please correct me if I mis-guestimated) would have been an act of suicide by the central government.
I meant the state governments. I guess saying when by referring to when the national government started was misleading.
As I understand it, people suspected something like the Civil War would happen if they tried that. A lot of compromises were made to forge a single country out of this mess.
In a lot of places, you had to own property to vote. (Also, there were plenty of places where slavery was, indeed, outlawed relatively quickly. Just not everywhere.)
That would seem to explain it. Looking into it more, it doesn't quite seem to fit. According to the Wikipedia page on Jacksonian democracy [http://en.wikipedia.org/wiki/Jacksonian_democracy#The_philosophy], nearly all requirements to own property were dropped by 1850, and the Voting rights in the United States [http://en.wikipedia.org/wiki/Voting_rights_in_the_United_States#Milestones_of_national_franchise_extension] page seems to imply that it ended completely by 1860, but the civil war wasn't until 1861. Was it just that people don't stop it that quickly, and had the South been allowed to leave, they would have outlawed slavery in a few decades? I still would prefer it if I could find something saying exactly when slavery ended and when voting was allowed for non-property owners. Also, I'm curious as to how and when slavery ended in different countries. Unfortunately, my schooling has been somewhat biased.
The property requirements might have kept slavery from being banned while it was getting started, though. As for the whole "not wanting to compete with slaves" thing, that wasn't actually a factor. As I understand it, slaves weren't looked on as competition; rather, they were the people even the poorest whites got to look down upon. No matter how low on the proverbial totem pole a white person fell, they could still feel superior to black people. To quote HPMOR:
People can look down on immigrants too, but that doesn't keep them from getting mad at them for taking their jobs and trying to enact laws to restrict immigration.
"Vote against slavery" is not something that happens in a vacuum. There isn't someone reading minds who realizes the country has a majority of people who would prefer slavery be illegal (even if you ignore the minds of the slaves), and calls a vote to happen on the issue of "should slavery be a thing". In any given district (town, city, country, whatever area), you would have to campaign and convince people both that they would be better off without slavery, AND that this means it should be illegal. There are reasons why people don't form coalitions of 80 percent to rob the 20 percent.
The politicians have incentive to know what the majority of people think, and they are quite capable of asking people. If a candidate makes it their platform that they'll free the slaves, that's basically what the vote will come down to. If it's clear enough, both candidates will have it in their platform, and there won't even be a vote. They seemed to work out that they don't want immigrants, women, and children taking their jobs. It didn't happen until the industrial revolution, but they weren't taking their jobs until then. Why would slaves be different? I know they made it illegal for children to work. I don't remember hearing anything like that about women, though. Also, once slavery started getting outlawed in a few states, you'd think people would be more willing to outlaw in it other states. I remember something about one state mandating that all contracts could be payed in cash (as opposed to gold), knowing full well that their cash was worthless. They essentially forced all the banks to forgive the debts of the farmers.

Fiction writing advice

I am musing about writing fiction. I would like get help on the following questions: 1. What forums to ask the following questions can you recommend? 2. What is the general rule on including real living people in fiction? I'm afraid that it's hard and fast "Don't." Where does it become ok to include real people in fiction? (sometime after they die?) Is it ok if fictional character strongly resembles a living person, but some details are left out so that one can write it off as a coincidence? 3. How many threads to the story can I have? The more the merrier? E.g. suppose that I write a story about a scientist that designs an AI to save the world, but instead bringing the doom upon us all (or vice versa). Would it be good if they are pursuing a romantic interest at the same time? What about solving a major philosophical problem, too? Where do we stop? 4. If I would like to make a story split between two interleaving timelines: e.g. the "future" and the "past" timeline. One chapter is in the "future", then the next one in the "past"; what should be the natural way to arrange the passage of time in those timelines: a) parallel (time goes forward in both threads) or b) diverging threads (time goes forward in "future", backwards in the "past"). Here's the illustration: Book timeline [https://docs.google.com/spreadsheet/ccc?key=0Ajh5Kk2edLrgdERIV1VBMnZQNHQ1OVNMMFF6Tkw4Mmc]
TvTropes has some [http://tvtropes.org/pmwiki/pmwiki.php/Main/TwoLinesNoWaiting] things [http://tvtropes.org/pmwiki/pmwiki.php/Main/ThirdLineSomeWaiting] to say [http://tvtropes.org/pmwiki/pmwiki.php/Main/FourLinesAllWaiting] about that.
The TV Tropes forums [http://tvtropes.org/pmwiki/topics.php], especially the "writer's block" and "world building" sections.
It's pretty common in fiction these days to have a romance plot and a mystery plot running simultaneously. It's probably worth your while to study what cues authors give to help the reader stay oriented. Beta readers are good, too-- if you're going that route, ask them to tell you if they start getting confused, and if so, where. (Second-hand information: beta readers are more useful for identifying problems than for suggesting solutions.) You might need to work on plausibility if you're adding threads, especially if your main characters are deeply involved in all of them. Does your philosophical problem have something to do with the AI? With the romance? Rowling's A Casual Vacancy (recently released, not fantasy, depressing) does an impressive job of tracking the consequences of many goals and consequences through a large cast of characters. I think time could go either way in the past. I assume you're talking about flashbacks where time goes forward in each flashback, and your concern is what order to put the flashbacks in. Off the top of my head, I think the question is what you're trying to build suspense about-- is it revealing a root cause for the whole situation? Flashbacks in reverse order. Or is it revealing the whole sequence of your story? Flashbacks in forward order. I recommend research on the flashbacks problem. I assume someone (among all the books and discussion boards about how to write) has addressed it, but I've never seen it done. Also, how are flashbacks handled in your favorite books with flashbacks? Reading like a reader isn't the same are reading like a writer.
Ok, I guess the more the merrier, the limiting factor being how much of it I can tie into a coherent story. Yes, that's what I was trying to say. I was afraid if flashbacks in reverse order would just confuse the reader: in each subsequent flashback you expect to learn what happens next, but get to know what happened before instead.
I asked about flashbacks on my blog [http://nancylebov.livejournal.com/557453.html], and there are a few comments. The technical thing I've seen writers talk about the most is point of view, which I suppose is easier because it can be addressed at the sentence level, but possibly there should be more about wrangling chunks of story.
There is no such thing as the reader-- it's amazing how much readers vary, even though there are some commonalities. As far as I can tell, the big deal for me is whether a flashback is an interesting story in itself.
Standard practice is to change names and at least some identifying details; if practical it might also be a good idea to get permission first. Writing semi-fictionalized stories based on real people and events has a long and honorable history (see Jack Kerouac or Hunter S. Thompson), but it's wise to tweak identifying features enough that people won't automatically assume you're documenting a true story. I'm not a lawyer, but in most jurisdictions my impression is that you're in a better position to handle possible challenges if the people you're writing about are public figures; libel laws are usually weaker for people in the public eye.

I don't see a Myers-Briggs personality survey anywhere on LessWrong but I would like to make one. I also have predictions, and I think it would be neat to see if I'm correct (predictions below in an unedited comment.)

I am aware that the Myers-Briggs is considered to have inaccuracies - for instance, I've scored different types at different times. I do not feel that this makes it useless but that it reflects the fact that your personality can change due to things like (for me) switching from doing a lot of art and people work (feeler type) to doing more i... (read more)

Personality Type Predictions: The vast majority are introverts, ballpark 90% introverts. Most common type: INTJ NT types > 75% of the population NF types - a handful or none (possibly more than the next type, possibly less) ISTJ - a single digit percentage of the LW population Other guardians and artisans: none or nearly none.
The survey results are in [http://lesswrong.com/lw/fp5/2012_survey_results/], so I am updating this: If you scroll down to "MYERS-BRIGGS" you'll see that there are 436 people in Yvain's selection of results (of greater than 10 people for each type, leaving out a total 3.1% of the survey data). That's what these figures are based on. (The raw data is missing around 10% of the responses due to people wanting anonymity, and the graphic provided to show more detail has some issues [http://lesswrong.com/lw/fp5/2012_survey_results/7xka] so I used Yvain's selection.) * Ballpark 90% Introverts: Correct 371 Introverts (85% of 436) * Most common type: INTJ: Correct 163 INTJs (37% of 436) * NT types > 75% of the population: Correct 371 NTs (85% of 436) * NF types - a handful or none (possibly more than ISTJs) : Correct 51 NFs 436 (12% of 436) * ISTJ a single digit percentage of the LW population: Correct 14 ISTJs (3% of 436) I wasn't sure exactly how I should interpret the somewhat vague "a handful or none" for NF types, but I see that I used enough numbers to be able to do a literal, mathematical interpretation so I chose that method. I had predicted it was possible that there would be more of them than the ISTJs who I had predicted would be in the single digit percents (implying that 10% or more of them wasn't outside the range) and that there could necessarily be no more than 25% of them because it would contradict the NT prediction, so since they were within the numerical bounds, I interpreted this as correct. Another interesting thing to note is that each personality type in the top 98% of LW personality types is in the same order as the type list I wrote here [http://lesswrong.com/lw/eqa/open_thread_october_115_2012/7kjm]. Unfortunately that comment had been previously edited, so whether or not you believe that I did this intentionally will be based on how much you trust me not to lie and what you think the probability is o
INTP Introvert(11%) iNtuitive(38%) iNtuitive Thinking(25%) Perceiving(11)% You have slight preference of Introversion over Extraversion (11%) You have moderate preference of Intuition over Sensing (38%) You have moderate preference of Thinking over Feeling (25%) You have slight preference of Perceiving over Judging (11%) From the below linked test Also this feels like it can't possibly be that useful since many of the questions have different answers in different situations. If I'm up I love being around crowds, if I'm down I hate being around all but a very few people, etc.
Mmm, I noticed this too when I filled out an official MBTI. It probably comes up quite often; I remember the test having an instruction to answer each question with the choice that most often applies to you, even if sometimes it doesn't.
I did the exercise once of filling out an MBTI on a five-point scale and calculating weighted sums, rather than a binary scale. My resulting classification didn't change.
Another INFP here. NF slight preferences, moderate on the I and P (I wish I was an E, but I'm just not). Why did you put J in your prediction of the most common type?
Here is one. INFP. Fairly consistent across tests, with the "N" and the "P" being close to the extremes.
I would not have guessed that. I wonder if some of your personality dimensions fluctuate or are on the border. For me, the E/I fluctuates and so does the F/T. I'm always an N and P. Are you right on the line between T and F? If this test is the one that I remember (the page changed) then I think it gives you percentages: http://www.humanmetrics.com/cgi-win/JTypes2.asp [http://www.humanmetrics.com/cgi-win/JTypes2.asp] No idea about the accuracy, but it's free.
People who know me on lesswrong tend to tell me that I come across very differently in person than I do online. I think they are right although I suspect that my personal interactions with people here (few though they may be) are rather similar to who I interact in person in the 'real world'. I know from experience that acting like a typical INFP in an online environment where INFP is rare is a recipe for disaster---it just doesn't work. I also find that I am best served by rationing my lesswrong interactions and keeping them balanced by interactions with INFP friends (and lovers). Too much dealing with "Js" just gets tiresome. I actually suspect I'll take another hiatus from here soon and get my intellectual stimulation from the textbooks and papers on my to-read queue for a while. I'm very close to the line on T/F, fairly close to the line on E/I.
This [http://www.personalitypage.com/INFP.html] makes INFP sound a lot like Elliezer.
Good point. That's likely to make it harder to discover Fs here. Aww. ): I hope you'll still talk to me.
I've heard that test repeatedly labeled as the "only personality test on the internet that works", but I can't really find many other Myers-Briggs tests.
Proposed Poll: What is your last Myers-Briggs personality type score: * INTJ * INTP * ENTJ * ENTP * INFP * INFJ * ENFJ * ENFP * ISTJ * ISTP * ESTJ * ESTP * ISFJ * ISFP * ESFJ * ESFP These questions are interesting because there are some connections with personal development: Regarding I/E (introversion / extroversion), have you gotten a score near the border between them, or gotten a different I/E result when taking the test multiple times? * I got results near the border (maybe the same result maybe different). * I got two very different results (not near the border, not the same result). * None of the above. (Etc. for the other three dimensions)
I like the idea of asking Yvain to add it to the big survey. That's probably the least obtrusive way, and it'd maximize responses, which you'd need for a decent sample size in each of 16 subcategories.

Any good books on mathematics for software engineers? I've been looking at the best universities in UK, they all have much more mathematics in their degree than what I'm taught.

Also, any good books for probability theory and all the things needed for AI development? I'm doing this course: https://www.edx.org/courses/BerkeleyX/CS188.1x/2012_Fall/about

Edit: These are the programs I've been talking about.

Imperial college: http://www3.imperial.ac.uk/ugprospectus/facultiesanddepartments/computing/computingcourses http://www3.imperial.ac.uk/computing/teaching/u... (read more)

I assume they teach mathematics, because it's useful for software engineers.

Be aware that computer science and software engineering are different disciplines, and don't assume that people who design university curricula are experts on teaching software engineering. You can find top-notch computer scientists at universities, but top-notch software engineers tend to end up in the industry instead of academia.

What kind of problems in particular would you like to be able to solve better? I don't really find myself needing much mathematics in everyday programming. That said, Knuth's Concrete Mathematics [http://en.wikipedia.org/wiki/Concrete_Mathematics] comes to mind as a book on the sort of mathematics used in computer science, and Alexander Stepanov's Elements of Programming [http://www.elementsofprogramming.com/] takes an interesting mathematics-like first-principles approach to constructing programs in a C++-like language (Stepanov came up with the STL for C++). There are also people claiming that category theory can be used as a foundation for software engineering, but I'm not able to point to many convincing examples of a real-world software engineering problem solved neatly using category theory. I could sort of follow some of the bananas and barbed wire [http://academic.research.microsoft.com/Paper/296068.aspx] paper, which constructs CT-ish algebraic representations for basic looping constructs in programs. Actual category theory stuff just puts my brain to sleep by the time it jumps to the third level of abstracting categories into categories with me still without a motivation for connecting the thing to something I can do something with, but Pierce's Basic Category Theory for Computer Scientists [http://www.amazon.com/dp/0262660717] is at least thin and has a title which makes you think you should read it. There's also Conceptual Mathematics [http://www.amazon.com/dp/052171916X/] that explains categories at a college freshman reading level with several examples, which I should probably get back to reading at some point. I could actually follow it, but still came out with no idea what I would actually want to use categories for. For AI and probability, Bishop's Pattern recognition and machine learning [http://research.microsoft.com/en-us/um/people/cmbishop/prml/] comes up a lot. I haven't actually read any of these to the end, though I'm pretty sure I've re
Oxford's course is unusually pure. My freinds reading it don't actually use, you know, computers. They just write their algorithms down on paper.
Are you saying that you have a friend who attends this program at Oxford and they don't do any actual programming? I suspect I'm misunderstanding because that sounds really unlikely.
I think they did at one point - but yes, when I asked them about it, they basically hadn't entered code into a terminal for an entire semester.
"Pure" one word you could use for that teaching strategy. Just what it is a 'pure' representation of is up for debate. "Pure backwards self-congratulatory tripe" would be the cynical description.
I think they're trying to train a Scott Aaronson, not a John Carmack. A Scott Aaronson really does work by not actually using a computer much for anything other than typesetting LaTeX.
And that sounds like a brilliant idea. Most of the problems I have with having being forced to write algorithms on paper at times disappear right there. It's even worse than forcing people to write sentences on paper, given the need for correctness in the details.
Oh, they don't have to write on paper. I just don't know any maths students who do. Handwriting maths is easier than typesetting it.
(At least it is for people who are bad at typesetting.)
Wolfram-style automatic formatting buttons ftw.
Maybe try with a bit less sarcasm? I'm having genuine trouble parsing what you are objecting to, exactly.
There isn't any sarcasm in the grandparent.
You're being awfully cynical lately, and I don't like it. Would you like an internet hug?
I have never respected cheap 'purity' signalling at the expense of practical considerations. Not when I was learning at university, not now. I will always consider the obligation to use paper rather than modern technology to be a bad thing, not an indicator of elite quality. No, you are making me uncomfortable. Please don't ask again.
It's not the teaching strategy, it's the subject matter. See for instance pure maths [http://en.wikipedia.org/wiki/Pure_mathematics].
It looks like you're talking about software engineering, but looking at computer science courses. No, they teach mathematics because it is necessary for computer science. They probably have little care for what is useful to software engineers.
That's because I wanted to see what their programmers study and computer science was the closest I found. I assume that their programmers would study that? And that it's better than study software engineering in a significantly worse university?
It really depends on what field you're going into and what specifically the employers you'll be courting will be looking for. For example, having a degree in philosophy was a major boost at my current job, because there are (oddly) so many CS majors out there with no critical thinking skills. I was going to give various detailed advice here, but I realized I have no idea what the tech jobs are like in the UK. I assume they involve a lot of paperwork.
Maybe, but I think a bigger reason is that it is useful for computer scientists, and these courses aim to prepare you for research as well as for work in industry.

There is at least one post on LW about undergraduate application essays. Instead of writing a similar post detailing my specific circumstance, I am posting on the Open Thread in search of people who would be interested in talking to me/private messaging me about undergraduate application essays. I imagine that I would benefit from reading some successful and unique essays, perhaps about the subjects we discuss on LW. Since UChicago is my "dream school", I imagine I would also benefit from reading successful application essays for their provocativ... (read more)

I would like advocates of TDT, UDT, etc, to comment on the following scenario.

Suppose I think of a possible world where there is a version of Genghis Khan who thinks of this version of me. Then I imagine Genghis imagining my responses to his possible actions. Finally I imagine him agreeing to not kill everyone in the next country he invades, if I commit to building a thirty-meter golden statue of him, in my world. Then I go and build the statue, feeling like a great humanitarian because I saved some lives in another possible world.

My questions are: Is this crazy? If so, why is it crazy? And, is there an example of similar reasoning that isn't crazy?

I think one needs to significantly abstract this example to understand the reasoning at human levels. (EDIT TO ADD: And I also think your usage of the word 'imagine' is confusing because it connotates 'making things up' instead of 'attempt to accurately model in your mind'.) E.g. Let's say you have made a habit of providing a helping hand to strangers. One day you learn that Genghis Khan, in a different time and a different continent, put an end to his butchering because he saw people helping strangers and suddenly took seriously this idea, and this made him reevaluate e.g. his cynicism towards humanity, and whether brutality truly provides happiness. In this sense a part of you, a part of your decision process, the kindness-to-strangers part is responsible for stopping Genghis Khan. Other parts of you (your memories, your sense of identity, your personal history) aren't. Nothing "recognizably" belonging strictly to you, but part of you is 'responsible' nonetheless. -- Or here's a different example, a more scientifictional one. An alien informs the human population that the next day, they'll select at random an adult human to observe secretly for a day from the whole human population. That person will not have to do anything special, just clap their hands once during the day. If they do that the earth will be safe, if they don't clap their hands during the day at all, the earth will be doomed. Next day, three billion people clap their hands, just to be on the safe side. Three billion other people don't -- the chance that they'll be the "one chosen" is only one in six billion afterall, close to nothing. The aliens choose Alice. Alice happened to not clap. The earth is destroyed. My moral intuition tells me that the three billion people who chose not to clap share equally in the responsibility for the Earth's destruction; Alice who got randomly selected didn't decide anything differently from any of the rest of them and therefore is no more "responsible" than any
The "probability" of the imagined world is low, so the opportunity cost of this action makes it wrong. If there was a world fitting your description that had significant "probability" (for example, if you deduced that a past random event turning out differently would likely lead to the situation as you describe it), it would be a plausibly correct action to take. (The unclear point is what contributes to a world's "probability" [http://lesswrong.com/lw/1iy/what_are_probabilities_anyway/]; presumably, arbitrary stipulations drive it down, so most thought experiments are morally irrelevant.)
I'm not an advocate (or detractor) of those decision theories, but the answer that immediately appears to me is to question what drew this particular scenario to your attention out of all possible scenarios. Abstractly, the scenario is that in some possible world, someone doing X prevented disaster Y. For which X and Y should I therefore do X, even if disaster Y cannot occur in this world? Somehow, you obtained the bits necessary to pull from possibility space the instance X = build a golden statue of Genghis Khan and Y = Genghis Khan in another world stops making war. What drew that instance to your attention, rather than, for example, Y' = Genghis Khan, inspired by this monument, wages war even more mightily? Or Y" = to get all this gold, the monument-builder himself must conquer the world? And so on. It's like the type of Pascal's Mugging scenario that gives no reason to expect that particular consequence to result from the action more than any other. A more fruitful question is "should I be the sort of person who does X-ish actions in Y-ish situations?" for various values of X and Y. Here, TDT etc. may give justifications for e.g. cooperation in PD, Parfit's hitchhiker, etc., that conventional decision theories have problems with.

As LW garners more traffic, it also becomes a larger target for (commercial) spam. The wiki in particular seems to lack some basic protection:

See this very recent wiki spam post.

A possible solution would be to require users who edit the wiki to have, say, 1 karma (if the user databases are synchronized).

Also, just from the "recent wiki edits" and its smörgåsbord of sketchy new user names ("IvanosbevfkwwbBohan"), it seems that the user creation process is in urgent need of a good CAPTCHA, which may also help with the first problem.

I don't know if the spambots are fully automatical or human-aided. If they are fully automatical, we could just add a question "What is Eliezer's surname?" The advantage would be people learning to write it correctly. :P
Fully automatic, for the most part.
Just for reference: this has been pointed out at least once before [http://lesswrong.com/lw/8ei/wiki_spam/], and I believe there was a (temporary) fix implemented (but I can't find any reference to it at the moment). But that was almost a year ago now, so it's good to bring it up again.

Old material

Related to: List of Public Drafts on LessWrong

This stuff is obsolete or just plain old, it can still serve as draft material:

Could someone recommend any books on investing that might appeal to a LWer?

Has anybody been using the brainstorming techniques I posted about a while ago? I'd be interested to hear about your results.

Personally, I haven't been using them much since making that post, so I don't really have anything interesting to share. That's a failure on my part, though, not the techniques.

Who here owns weapons? Pick the highest "level" if you match more than one.


I don't own weapons, but if I ever got the time and money to pick up the practice again, I wouldn't mind owning a longsword.
For those who are interested: The best place to purchase swords online is likely Kult of Athena [http://kultofathena.com/swords.asp], after doing research at the Sword Buyer's Guide [http://www.sword-buyers-guide.com/index.html]. Personally, I own a piece from Zombie Tools [http://zombietools.net/tools/].
I own a bunch of airsoft guns and boffer swords! Also fencing equipment. As far as actual dangerous implements, I have a few knives I mainly use to open things and a retractable baton (a gift from my paranoid father). My parents, on the other hand, have what amounts to a small armory in their gun locker.
May I ask why you're asking. Also, I own a two bows and a large number of arrows.
Because everyone's equipping themselves to win in the intellectual realm, and I was wondering how many were equipping themselves to win in the physical realm.
The sort of victory LWers are looking for, I hope, is winning as a result of a cooperative effort to gain more knowledge. In a sense, it's us ("us" includes all relatively honest seekers of knowledge, not just LWers) vs. the universe. Winning in the physical realm should include some estimate of the odds of needing weapons.
Eliezer [http://lesswrong.com/lw/5gh/avoiding_factual_muggings/42a1]: However, I bet he's unarmed. It's my estimate that this community is pacifistic and avoids the calculation you mention.
That would be far too intellectual.
No gas guns? I don't own anything that could be counted as weapon, but if I did, it would be it. Gas gun is the weapon that I expect to have a chance of actually using without training.
Why is 'assault rifle' above 'hunting rifle'? Are these supposed to be ordinal?
It's more appropriate for combat use. If I were redoing the poll, I would group all 3 into 'firearms'. At the time, I was thinking to try and break out 'incidental' weapons from 'actual' weapons.
At 1 meter, a 'combat knife' is more appropriate. At 750 meters, a 'hunting rifle' is more appropriate.
Yes, though the assault rifle was developed after nations determined most actual combat takes place between 200 and 300 meters.
A major incentive in the design of the combat rifle was a cost-benefit analysis in terms of the expenses involved in the training, equipping, and potential resource loss of soldiers. Better-trained soldiers outfitted with larger-cartridge battle rifles -- even when they are semi-auto only and not select fire rifles, like the M1 Garand -- are more individually effective, for instance, than assault rifles. On the other hand, fielding such more highly trained, effectively equipped soldiers is much more expensive and a greater loss to aggregate military power when they die in the field than the same of less highly-trained, more lightly equipped soldiers. That is, someone who can make full use of a battle rifle out to its ideal engagement range and issued such a rifle is significantly more effective in the field than someone whose skills do not extend past the ideal engagement range for an assault rifle issued such a rifle -- but the former is more expensive, both to deploy and to lose in combat, than warranted by the increase of individual effectiveness, if you treat the value of the soldier's life as nonexistent and only regard the soldier as being equivalent in value to equipment. Of course, this thinking also tends to undervalue the often substantial value of the exceptional case of a single soldier who can account for a far higher number of lesser-trained, lesser-equipped enemy soldiers from longer ranges, because the capability to reliably perform under such circumstances is essentially prohibited by the strategic decision to issue assault rifles by default and only provide combat and marksmanship training out to around 350 meters to the general run of soldiers. It's the classical mistake of focusing on the statistical averages to the exclusion of considering the sometimes overriding value of the exceptional case. The American Revolutionary War was essentially won by the exceptional cases, after all (discounting, for the moment, additional factors such as French
Like Simo Hayha [http://en.wikipedia.org/wiki/Simo_H%C3%A4yh%C3%A4]!
That guy was a scary motherfucker in the Winter War. I don't remember whether the Wikipedia article you linked mentions it, but I seem to recall that a reporter asked him once how he got to be such a good shot, and he said "Practice."
I learned about him from a fun article at Cracked [http://www.cracked.com/article_17019_5-real-life-soldiers-who-make-rambo-look-like-pussy.html].
Thanks. That is an entertaining read.
I'm looking at this list, and I do not know how to identify what you consider the "highest level". If I had to judge by position, it would seem that "I own a combat knife or other melee implement" trumps "I own a pistol", "I own a hunting rifle", and "I own an assault rifle". Is that correct?
I'm not Rain, but the reverse, I believe. Consider everything on the list that applies to you and select the thing that's lowest on the list (which will be the "highest level" in the sense of being either a more powerful weapon or, in the case of a decision between the first two list items, a more powerful stance against weapons). This doesn't quite work with the "other weapon" choice -- if you own an assault rifle and a throwing star, you should choose "assault rifle", not "other weapon". Now that I think a bit more about this, the ranking between, say, pistol and hunting rifle is arguable in the sense that a hunting rifle is a more powerful weapon, but unlike the pistol, its self-defense use is limited to the home (and the zombocalypse).
I phrased my query based on the fact that the moment I start trying to judge by anything other than ordinal position in the survey, myriad possibilities of roughly equal potential suitability come to mind. My first thought was that lower on the list is better, but (like you) I ran into a problem with the "other weapon" option being at the bottom, then I noticed that the "hunting rifle" option came later than the "assault rifle" option which seemed inconsistent with popular understandings of terms like "assault rifle", then I thought about the fact that gun control advocates rate "hunting rifles" as less dangerous than both "assault weapons" (which of course include technically termed assault rifles and handguns of all sorts including "pistols"), and next I realized. . . . . . . so I just fell back on mentioning the first-blush ordinal guess (in case by "higher level" the querent meant more pacifistic ideas or something like that) and the "any other ordering" contrast to make my question simpler. tl;dr summary: Yeah, I thought lower made more sense in some respects, but had second (and third, and more) thoughts as you did, so I just simplified the question.
It was poorly put together, and I can't edit poll options. I'd put firearms above melee, larger guns above smaller.
Okay, thanks. I "voted" for the "I own an assault rifle" option. Given common ideas of what what constitutes a "hunting rifle" (though an AR-15 is one of the "varmint hunting" rifles of choice), I do not own a "hunting rifle", but I own at least one of every other category of weapon mentioned. There are at least two caveats that apply to my case, and one more that probably applies to the way you phrased things. The first caveat for my specific case is that, technically, an assault rifle is defined as a select fire (has a switch or other way to toggle between firing "semi-auto", or one shot per trigger pull, and "full-auto" aka "automatic", or more than one shot per trigger pull) combat rifle that fires an intermediate or carbine cartridge. What I own is a semi-auto (that is, not select fire and not full-auto) AR-15 -- what mainstream media dimwits and politicians who hate scary black firearms often call an "assault weapon" -- which is not technically an assault rifle, though the US military's M16 is a family of assault rifle variants of the AR-15. All AR-15s, including the various M16 variants, fire 5.56x45mm or .223 Remington cartridges (the two cartridge types are close enough to identical than they are essentially interchangeable), an intermediate cartridge. The second caveat for my specific case is that civilian AR-15s are among the most favored "varmint hunting" rifles available -- used for hunting medium-smallish animals often regarded as pests, such as coyotes and rabbits. Putting both of these conditions together, I selected "assault rifle" because I thought you would probably mean a semi-auto AR-15 rifle with a high-capacity magazine to fall within that category, and probably consider "hunting rifle" to include things more like high-power bolt action rifles with wooden stocks. More on that next. For the more general caveat about phrasing, a hunting rifle is in many cases going to qualify as "bigger" than an assault rifle by any technical standard. By d

I found this interesting. I personally don't think it's a paradox, but I think it's interesting that the logic behind it works.


If it's not a paradox, that the logic behind it works should never be interesting. How it works is another story.
I disagree. I don't approach it from the perspective of "it's not a paradox, therefore the logic behind it is interesting". I approach it from the position of "oh, this does work after all, neat, therefore this is not a paradox". I sorta feel like I'm being chided by your comment, too, and I don't like that.

A query about The Dark Knight on Marginal Revolution originally from Brad Allen:

I was watching the Dark Knight on a bus yesterday evening (I’m not sure how familiar you are with the movie) – there was a scene that I thought was pretty interesting to think through, and was curious how you might go about it.

There is a scene where the Joker kills a mob boss, and then gives his 3 subordinates one half broken pool cue – and basically tells them that to live, the other two have to die. You don’t see what happens, but what do you think happens? Is it advantageo

... (read more)
I think the best move would be to back off. They'd be unsure whether to pursue you or to defend themselves against the person closest to them. Mutual uncertainty on their part means they'll attack each other first, instead of pursuing you. Whoever won would get the pool stick, but they'd also be tired and you'd have an advantage. You also would have had the opportunity to observe their fighting style. Needless to say in real life I would just die.
http://en.wikipedia.org/wiki/Truel [http://en.wikipedia.org/wiki/Truel]
It depends on your ability to come up with an alternate force-multiplier to the offered weapon to establish some kind of tactical superiority. If you quickly come up with one, or are at least confident of your ability to do so, the smart move is to induce the others to deal with each other first, then attack the winner from a position of strength after he has been weakened by the initial exchange with his first opponent. Otherwise, pick up the offered weapon; then the ideal strategy is still probably to see if you can get the other two to attack each other before coming after you, perhaps acting as though your only reason for picking up the weapon is to be a "coward" who wishes to avoid a fight altogether -- because, of course, the other two are unlikely to go after the more dangerous opponent first, even though collaborating to eliminate the primary threat before attempting to finish off a (hopefully) weakened remaining foe is probably the winning strategy for them. Of course, the real smart move, if you can get everyone in on it, might be to ensure that all three of you collaborate in a surprise attack on the Joker after giving him the impression he's safe by pretending to initially target each other. That's pretty damned unlikely to happen, though, given the level of trust most likely needed to achieve that kind of alliance without cluing in the Joker and getting yourself killed by him instead of the other two.
In general, it is a smart move to precommit to mutual support against opponents attempting to sow dissension. Of course, how effective this is depends on how reliably individuals in the group can precommit and how effectively they can signal reliable precommitment. (Also, insert obligatory UDT reference here.)

What's the name of the idea that morality is a scalar rather than a binary property (i.e., rather than asking whether A is moral, one should ask whether A is more moral or less moral than B)? I'm pretty sure I recently saw a discussion of that somewhere in a SEP article, linked to from a comment on LW, but I can't find it now -- and I've been searching for a while.

EDIT: Larks nailed it.

[This comment is no longer endorsed by its author]Reply
Scalar is the right word. Scalar consequentialism [http://www.iep.utm.edu/conseque/#SH1c] is a thing. It's possible the comment you're thinking about was one of mine; I've introduced a fair few people to this (IMHO) superior version of consequentialism.
Thank you! That's the one I was thinking of. For some reason, I incorrectly remembered that it was on the SEP. EDIT: Why, when I failed to find that on the SEP, I assumed that I misremembered the name and tried using different search keys, as opposed to suspecting that I misremembered the site and searching Google for the same key?
IIRC, that discussion was in the context of utilitarianism/consequentialism, where “[word A] consequentialism” was the moral system where the action that maximizes expected utility is moral and any other action is immoral, and “[word B] consequentialism” was the moral system where an action is more moral than another if it has higher expected utility, even if neither saturates the upper bound, or something like that. EDIT: on looking at http://plato.stanford.edu/entries/consequentialism/ [http://plato.stanford.edu/entries/consequentialism/], “[word A]” is “maximizing”.
The right way to understand the difference between maximizing and satisficing consequentialism is not that the maximizing version treats morality as a binary and the satisficing version treats it as a scalar. Most proponents of maximizing consequentialism will also agree that the morality of an act is a matter of degree, so that giving a small fraction of your disposable income to charity is more moral than giving nothing at all, but less moral than giving to the point of declining marginal (aggregate) utility. The distinction between maximizing and satisficing versions of utilitarianism is in their conception of moral obligation. Maximizers think that moral agents have an obligation to maximize aggregate utility, that one is morally culpable for knowingly choosing a non-maximizing action. Satisficers think that the obligation is only to cross a certain threshold of utility generated. There is no obligation to generate utility beyond that. Any utility generated beyond that threshold is supererogatory [http://plato.stanford.edu/entries/supererogation/]. One way to think about it is to think of a graded scale of moral wrongness. For a maximizer, the moral wrongness of an act steadily decreases as the aggregate utility generated increases, but the wrongness only hits zero when utility is maximized. For the satisficer, the moral wrongness also decreases monotonically as utility generated increases, but it hits zero a lot faster, when the threshold is reached. As utility generated increases beyond that, the moral wrongness stays at zero. However, I suspect that most satisficers would say that the moral rightness of the act continues to increase even after the threshold is crossed, so on their conception the wrongness and rightness of an act (in so far as they can be quantified) don't have to sum to a constant value.
A related term which I sometimes forget is value commensurability [http://plato.stanford.edu/entries/value-incommensurable/].
I don't think this is what you're looking for, but just in case: The Repugnant Conclusion [http://plato.stanford.edu/entries/repugnant-conclusion/] discusses morality systems quite a bit, so it might mention the article or name the idea you're looking for at some point, though I don't remember it if it does. I do remember that the article was entertaining, at least.

Noah Smith wrote up a humorous piece about the different kinds of economics blog trolls. Where does Robin Hanson fit in?

Funny pictures. Other than that, it looks like a viable case study for Yvain's meditation on superweapons and bingo [http://squid314.livejournal.com/329561.html].
I think he's more of a legendary monster than a species of any ordinary troll, though he probably fits best as one of the scientists.
Actually, these seem to be only macroeconomics blog trolls. Robin Hanson doesn't talk about present-day macroeconomic issues very much, so he's probably not any of them.
I vaguely remember him criticizing macro for failing to actually grapple with the evidence, but I could easily be wrong.


An indepth description of CogPrime's architecture by Ben Goertzel:

http://wiki.opencog.org/w/CogPrime_Overview CogPrime: An Integrative Architecture for Embodied Artificial General Intelligence

I have finally added some improvements to my web page, including responsive design and Atom feeds.

Here is an article (first in a series) about my summer 2012 in USA. This includes the Rationality Minicamp in July (although the first article does not get there yet).


For anyone interested, the Atom feed is: http://bur.sk/en.atom


I am musing about writing fiction. I would like get help on the following questions:

  1. What forums to ask the following questions can you recommend?

  2. What is the general rule on including real living people in fiction? I'm afraid that it's hard and fast "Don't." Where does it become ok to include real people in fiction? (sometime after they die?) Is it ok if fictional character strongly resembles a living person, but some details are left out so that one can write it off as a coincidence?

  3. How many threads to the story can I have? The more the mer

... (read more)
[This comment is no longer endorsed by its author]Reply

What is Science? From Global Warming to Evolution


Presented by Michael Vassar.

What is science? Are science and rationality the same thing? If science was something new, what sort of a new thing was science? I will discuss different ways of knowing, focusing on the differences between the analytic method of the enlightenment and the synthetic method of romanticism (and scholarship classically). These methods should be used together, but in fact their practitioners have been at war since Marx and Rousseau, leading to a schism in Western intellectual

... (read more)
[This comment is no longer endorsed by its author]Reply

Interesting podcast interview, from December 2011 with Michael Vassar

  • first 19 minutes deal with some very basic arguments on existential risk
  • 19 to 31 minutes interesting thoughts on what science is
  • 31 to 37 some discussion of self-awareness in machines, followed by the risks of technology ...

I've been playing around with the poll on isidewith.com. It's a questionnaire on political issues that matches your views up against those of the US 2012 presidential candidates. It's supposed to give you an idea of who you should vote for. I have a few criticisms of the way the poll is designed, but I still think the concept itself is interesting.

Could a well designed poll like this help raise the sanity waterline? Here's what I'm thinking:

  • Most people don't have the time, energy, or incentive to independently research candidates and compare their pos
... (read more)
Mostly, I think that any tool of this sort that somehow becomes sufficiently popular and effective to actually make a measurable difference to US presidential election results, relative to the effects of advertising and blogs and newspaper articles and so forth, will be co-opted by a raft of deliberately biased competitors long before that point. That said, I do think something similar would be useful for local elections. Of course, it would be a lot more work to develop and maintain at that level. Those two facts are not unrelated.

A PDF of the original paper whose title is the origin of the phrase "Nothing in Biology Makes Sense except in the Light of Evolution", by Theodosius Dobzhansky, from The American Biology Teacher, Vol. 35, No. 3 (Mar., 1973), pp. 125-129.

Can anyone help explain this to me? Quantum measurements leave Schrodinger's Cat Alive

I don't have that great of a background in physics. If my understanding is correct, this just turned all of quantum mechanics on its head (if it's accurate). That doesn't seem particularly likely to me. Has anyone else seen this yet, and what do you make of it?

It's more of a quantum computing development. Perhaps a more appropriate title would be "Quantum measurements leave Schrodinger's Cat in an ambiguous state while telling us just how ambiguous the state is".
It's clever, but it doesn't do anything previously thought impossible.
I'm not sure why you think this turns quantum mechanics on its head. Could you articulate what it is about that article that conflicts strongly with your understanding of quantum mechanics? Is it the idea that one can make measurements that don't completely destroy superpositions, or is it the idea that the experimenters could use information garnered from these measurements to drive the system back to its original pre-measurement quantum state?
That. I did not think that was possible. Like I said, my physics background is pretty weak. I've tried reading the quantum physics sequence, but it was really difficult, because it was fairly uninteresting.

Ah, I see. I don't think reading the sequence would have helped you here, because this is a subtle issue that wasn't (as far as I recall) covered in the sequence. In fact, it isn't even covered in most undergraduate courses on QM, so your assumption that measurements must destroy superpositions doesn't indicate a glaring lack of knowledge of QM.

It is standardly taught that the outcome of a measurement is an eigenvalue, which would mean that (at least within a particular branch) the quantum system "collapses" to a determinate state, and is no longer in a superposition. However, this standard story depends on treating the measurement device itself as a classical system, which is usually not a bad approximation.

But measurement devices are quantum systems, and in the late 80s some theorists demonstrated that this fact lets us obtain information about a quantum system without destroying a superposition. The procedure is called "weak measurement", and the basis for it is that there is some quantum uncertainty about the reading of the measuring device itself (uncertainty about the position of a pointer on the device, for instance). One can arrange it so that the measu... (read more)

OK, that actually makes a lot of sense. Thanks!

Interesting blog by someone trying out mindfulness practices for managing stress and anxiety. I'll be following it to see how it goes, as mindlfulness has been recommended to me.

[This comment is no longer endorsed by its author]Reply

Interesting blog by someone trying out mindfulness practices for managing stress and anxiety. I'll be following it to see how it goes, as mindlfulness has been recommended to me.

[This comment is no longer endorsed by its author]Reply

Quantum proofs of classical theorems is a review article about how ideas and techniques from quantum information and computing have been used for proofs in classical computer science. I thought it was pretty fun. Are reviews of "proof techniques" common in math/CS theory? Are they actually useful for researchers or for students in those fields? I really like the idea; even well-developed techniques aren't the kinds of things textbooks emphasize. (At least the textbooks I read tend to focus more on the "content" of theorems and so on--ma... (read more)

In short: at least in mathematics, yes. The kind of papers you're talking about usually figure as communal lore or online preprints. Sometimes these things are written as a kind of propaganda for a relatively new or obscure field of mathematics; see, for instance, "Generatingfunctionology."

My default assumption is that uploading will be [pollid:115]