« Disagreeing About Doubt | Main | The Uses of Fun (Theory) »

January 02, 2009

Comments

I was reading this article about J.D. Salinger and his reclusiveness, and wondering what the heck he's been doing for the last forty years. The article postulates that Salinger may have become demoralized by the question: "How do you make art for an audience, or a critical establishment, too crass to understand it?"

But wouldn't Salinger's art be taken more seriously primarily because of his fame, and compounded further by his mysterious retreat from the public eye? If Salinger suddenly were to release a new novel tomorrow, does anyone doubt it would be not only an instant bestseller, but taken quite seriously by said critical establishment, being (presumably) the product of a forty-year incubation period of one of the definitive writers of the era? In fact, it's hard to imagine how you could possibly attract more attention to your work.

The lasting residue of fame is well-known (see the Rolling Stones' continued existence, for example). I got to wondering: if you had some important, but highly obscure, complex, and subtle message to get across (artistic, technical, or otherwise), in what way is fame a precondition to propagating this message? Should we all be trying to get famous before we can really succeed in spreading important ideas? You may have discussed this before (I am new here) but I'd love to hear your thoughts.

I wonder if Eliezer would write a (short?) post on morality similar to his "What I Think, If Not Why" post. It would be nice to have a summary of his views without having to read 10 different posts and try to figure out what's his position exactly.

"Should we all be trying to get famous before we can really succeed in spreading important ideas?"

Fame is something of a chimera. No one is universally prominent; we have to talk about to whom one is prominent. I was starstruck when I met Eliezer, but if I were somehow to meet (say) Jennifer Anniston, I wouldn't care. I don't know what she looks like. If fame is a prerequisite for doing important work, then we're all screwed, because status-striving is an unwinnable game: unless you're literally Jesus Christ, there's always someone with more mindshare than you, and if you spend all your energy trying to claw your way to the top (of what?), how will you have time for your Great Work?

There can't be any threshold of prominence above which people care what you think because, as a general rule, no one cares what you think. Focus on the Work.

To whom would we wish the FAI to be friendly? To each individual human being alive now, to each individual now alive or going to be alive, to the Species, or to a combination of these?

I would like to see prediction markets in security of cryptographic primitives. It's a good match for the problem because:

* Some of them are high impact events
* The result is usually unambiguous
* They have low but non-negligible probability, like whole percents to tens of percents
* They have low to mid timescales, like years to decades
* There's a lot of data, like proofs of security against particular attacks, attacks on limited number of rounds, historical record of similar primitives, etc. but this data is very difficult to interpret even for professional cryptography researchers.

Anybody wants to start such markets on Intrade or somewhere else?

FYI, Good and Real has made it into the virtual remainder bin and is going for 5 bucks.

Robin, how rational did you raise your kids to be? Did you tell them about Santa? Did you pretend to not know what the boys were slinking off to do at 13?

Gonna be extending my New Day and concentrating on a project helping shape a rather new non-artificial intelligence, so thank you all for the good times & good reading.

And gents, please don't make anything go foom or boom for a few decades. I'm really looking forward to this.

"To whom would we wish the FAI to be friendly? To each individual human being alive now, to each individual now alive or going to be alive, to the Species, or to a combination of these?"

My favored alternative: it should be friendly to me personally. As I am not a psychopath, the AI ought to know it should give to other people according to my selfless desires. One of my current desires is that I not be deluded, even if I want to delude myself in the future, which ideally would mean that my ideas of who deserves stuff are unbiased. That doesn't mean that I can't just be selfish and take all I want, but I imagine I would restrict myself to what I deserve, as it is possible to be happy that way.

"To whom would we wish the FAI to be friendly? To each individual human being alive now, to each individual now alive or going to be alive, to the Species, or to a combination of these?"

My favored alternative: it should be friendly to me personally. As I am not a psychopath, the AI ought to know it should give to other people according to my selfless desires. One of my current desires is that I not be deluded, even if I want to delude myself in the future, which ideally would mean that my ideas of who deserves stuff are unbiased. That doesn't mean that I can't just be selfish and take all I want, but I imagine I would restrict myself to what I deserve, as it is possible to be happy that way.

@ Z.M. Davis claims: "I don't know what [Jennifer Aniston] looks like."

She looks like this.

I've noticed that extra options can be bad in a different way. For example Braess paradox.

Suggestion: how about having an "Ask Overcoming Bias" component to this site, with questions and answers about rationality and bias, like the below post from November?

http://www.overcomingbias.com/2008/11/leaving-the-fol.html

Similar features on other websites, like "Ask Hacker News" and "Ask Metafilter," are hugely popular, and "Ask OB" posts would require little effort on behalf of Eliezer and Robin. This seems especially appropriate, given that the admins are contemplating a more hands-off and user-generated site. Just a thought.

Question:

Why shouldn't we create an AI with the goal of peacefully and non-fraudulently improving human intelligence (with some required distribution; e.g. not a minority of humans) while preserving human utility functions? Would it be easier or somehow better to create an AI which acts to maximumize human utility functions, without necessarily improving the intelligence of the species?

Or in other words, instead of creating something smart enough to make itself smarter, why not create something smart enough to make us smarter?

This talk of non-person predicates reminds me of all the times in history and in fiction when people have developed non-person predicates. For real-life, look to slavery, ownership of women, the untouchable caste, the creation of a "counter-revolutionary" caste in the USSR and communist China (when the child of a "counter-revolutionary" was a counter-revolutionary by birth), and the treatment of animals. For fiction, it's hard to beat Cordwainer Smith's "The dead lady of Clown-town", one of the earliest SF stories to deal with transhumanist themes.

The creation of non-person predicates has never been a good thing.

It's hard to optimize society and the distribution of resources. The more idealistic the rule-makers are, the more cognitive stress they're subject to, as they try to structure a society where everybody is happy. And they deal with this by constructing non-person predicates. The easiest way to have a decent life yourself, without going crazy from guilt and sympathy for all the screwing-up of lives that is necessary to sustain your lifestyle, is to pick a non-person predicate that lets you focus must of the social goods in a restricted group of people and then call them "people".

I would love it if someone started converting these posts to podcasts. What a great way to start each day.

The first of the 2008 Singularity Summit videos are up! See:

http://singinst.org/media/singularitysummit2008

These are currently large, raw movie files, with no obvious associated licenses - but third parties sticking them up on YouTube seems to have begun already.

Howdy folks! Happy New Year!

I was just writing a book review of "Good Calories, Bad Calories" by Gary Taubes, and I happened to notice the overcomingbias thread on it from last year. (I read overcomingbias on occasion, and I've read or listened to a few other things that various contributors have written, too. I also had a delightful, if uncomfortably mind-expanding, conversation with Hal Finney on a lawn at dusk in 2006.)

Here's my book review of "Good Calories, Bad Calories", in which I quote Hal Finney and Robin Hanson from that thread, and swerve over into the topic of prediction markets at the end, and also momentarily mention life-extension. I would be interested in your comments on the book or on my book review.

Happy new year... ?

Re: podcasts - FWIW, I've archived most of the "Yudkowsky" podcasts I know of here.

Tomasz Wegrzanowski: I would like to see a prediction market in crypto primitives. Also, I would like to see a real prediction market in *anything*. By "real" I mean depth and liquidity and a real scarce currency such as US Dollars or e-gold at stake. Isn't the "Intrade" that you mention just another one of those play-money markets?

Zooko,

No! http://www.intrade.com/

@Zooko

"perhaps surprisingly, the play-money markets performed as well as the real-money markets."

-- J. Wolfers et al

Real money's just a tax hassle, lots of record-keeping - I've found people prefer other kinds of prizes, awarded in public, and oh heavens do they love a leaderboard! They will fight to be displayed in the top 10. Srsly.

So what's the point of a real-money market to you - read the whole paper and see if you change your mind.

Nicholas: fame could be a negative, if you value being misunderstood as an outright bad event. Suppose Salinger released some advanced elaborate novel; the result of 40 years of gestation, it is magnificent but like _Finnegan's Wake_ very difficult to appreciate properly. Being Salinger, it's a major event and will become required reading for the literati and possibly in college courses and in short, many people incompetent to read it will buy and read it. Naturally, such people will misunderstand it. There could well be far more people misunderstanding it than 'getting' it.

I don't really have any good examples for a work which is widely misunderstood; I could mention examples like Gene Wolfe's _Peace_ (few readers figure out on their own that the narrator is a ghost, and so the correct interpretation is completely out of reach for them), but Wolfe is a minor writer for any but a SF critic and _Peace_ not one of his more popular works.

frelkins: Well, the point would be if you're in it for more than smug satisfaction - if you want to actually be richer as a result of your wagering. If all you care about is satisfaction or personal development ('Well, the facts have it - I am biased in favor of my favorite political party's predictions. I'd better work on that.'), then there's no need to register or participate in the actual markets: you could just keep a text file with your predictions of probabilities.

@Gwern

"if you want to actually be richer as a result of your wagering

Making money is only one use of the market, maybe not even the most important.

"I don't really have any good examples for a work which is widely misunderstood"

I do know of a few.

1) "Harrison Bergeron" by Kurt Vonnegut. When I read this quote from the TV Tropes Wiki, I think I finally understood it.

Kurt Vonnegut's SF short story "Harrison Bergeron" is a frightening picture of the future in which the state has taken quite literally the line from the Declaration of Independence about all men being created equal, and mandated federal laws that forbid anyone from being smarter, more attractive, or more physically capable than anyone else; various penalties have been instituted to ensure compliance (e.g. radios that prevent the smart from thinking and weights that prevent the strong from being graceful). Most people see it as a cry for respect for the rights and talents of the individual. Vonnegut originally intended it as a sarcastic caricature of what right-wing pundits and philosophers, like Ayn Rand, thought America would become if the Communists took over. One can easily blame his evocative descriptions for his readers Completely Missing The Point.

2) "The Fun They Had" by Isaac Asimov

Again, I'll just quote TV Tropes, because I was the one who wrote the initial version of the entry.

Isaac Asimov's short story The Fun They Had is about two children who hate the future version of school, which is computerized, individual instruction directly tailored to the needs of each child. They find an ancient book that describes school as it is today, and imagine "the fun they had." Asimov intended it to be ironic; he hated school as a child because the classes were paced for less able students and he did not get along with his teachers. Many people, though, miss the intended irony (having forgotten just how bad school is) and take the story's concluding sentence at face value. It's even appeared in elementary school readers, presumably to get kids to appreciate school...

For more examples, see the TV Tropes Wiki article "Misaimed Fandom."

Vonnegut originally intended ["Harrison Bergeron"] as a sarcastic caricature of what right-wing pundits and philosophers, like Ayn Rand, thought America would become if the Communists took over.

This possibility has occurred to me before, but [citation needed], please?

@Aaron Brown,

At the bottom of the Wikipedia article for "Harrison Bergeron", I found a link to this critical essay, which points out that the rest of Vonnegut's fiction and nonfiction was sympathetic to the Left and critical of the Right, and promises to discuss internal evidence.

Zooko: You can bet both real money and play money on intrade. Using real money is unfortunately a problem because laws tend to treat prediction markets as gambling and enforce all kinds of restrictions and taxes on them.

As long as play money is a scarce resource, and people care about it, play money prediction markets should work pretty much as well as real money prediction markets. They seem to do so, or at least lose of prediction power due to play money is no more serious than lose of prediction power due to people's unwillingness to deal with gambling laws.

I'm feeling suicidal impulses right now. I don't intend to act on them.

The strongest reason I have not to act on them is that killing myself will cause great pain to my parents (and other relatives). However, my parents are both 60 and will eventually die, so that reason will eventually be gone.

I am currently 26 years old, male, job-free, and have a horror of paid work. I am currently living with and am supported by my parents. My current life plan is to remain as I am until my parents die and their savings is exhausted, and then kill myself. I fully expect that my parents will live into at least their eighties, though, so that gives me another twenty years in which to do nothing of consequence except play video games and waste my life in front of a computer.

My fondest dream is to become a wirehead. The first time I read about wireheading, I knew that it was exactly what I wanted out of life. I find passages in fiction that describe wireheading to be sexually arousing.

I don't really know why I'm posting this here; I guess it's mostly because I'm lonely and don't feel comfortable talking about this to anyone with the power to affect my life in a meaningful way, such as my psychiatrist.

Doug S., I'm 30, and I've spent my entire life in school as a way to avoid getting a real job. I've spent far more of that time than was healthy surfing the internet instead of finishing my degree -- I'm even doing that right now!

How would your psychiatrist have any power to affect your life in a meaningful way? As long as you don't declare that you intend to harm yourself or others, I don't think a psychiatrist could do anything to you but talk to you.

Hi, Zooko - I too have enjoyed our occasional conversations. As far as Taubes, my main concern is whether he is being completely honest and open in his presentation. Is he forthrightly presenting all of the evidence, both for and against his theories? Or is he cherry picking, selectively presenting only data that supports his ideas? I don't recall much discussion of problems with the bad-carb theory such as, for example, studies showing that people on low carb diets tend to gain back almost all of the weight they lost, similar to people on other diets. I see the book as a one sided polemic rather than a balanced evaluation of the evidence. It is dangerous to allow oneself to be persuaded by this kind of argumentation.

I have been thinking about an Idea Futures market claim on this issue. I have played the Foresight Exchange game for many years. It is play money but as referenced above, studies show that such markets do pretty well. However I have not been able to come up with a good clean definition of what we want to know. From experience with the game, it is common for issues to be resolved in a manner that was unanticipated when the claim was created. Chances are the eventual understanding will not be as simple as carbs=bad, even if the core of Taubes' idea might be considered validated. I would like to hear ideas for a concrete, judgeable claim that would shed light on the issue.

@Hal
" I see the book as a one sided polemic rather than a balanced evaluation of the evidence"

Uh-oh. My impression after reading the book twice, reading some of the better known attempted refutations, and listening to the hour-long rebuttal by Taubes is the opposite. Likewise Dr. Andrew Weil largely endorses Taubes, altho' some may find Weil political and marketing-motivated. I worry we may be on the verge of disagreement.

Because I have the highest respect for you, I will immediately downgrade my previous 90% estimate of Taubes' correctness to 80%.

I myself live on steak tartare, Prosecco, and wear a size 8 dress. My recent exam for a new life insurance policy left the insurance nurse exclaiming positively at my blood pressure and cholesterol numbers. I have a BMI of 22.8. While this shouldn't matter, as just one person's experience, I offer it to you as a proof of my lack of hypocrisy - that is, I bet my own ability to gain life insurance on Taubes.

Can we come to further agreement? I would be most loathe to disagree with you. Or are you content with where we stand now?

@ frelkins (immediately above): @Hal " I see the book as a one sided polemic rather than a balanced evaluation of the evidence"

Uh-oh. My impression after reading the book twice, reading some of the better known attempted refutations, and listening to the hour-long rebuttal by Taubes is the opposite.

Is this the same "frelkins" who wrote this comment about Taubes a week ago?

@retired urologist

"Is this the same "frelkins""

Indeed, I am. Whenever possible I RealName. Please note that I stated "the medical establishment" thinks Taubes to be a lunatic. I never said, nor can that statement be implied to read, that I did. Why?

That statement was in the context of cholesterol numbers, which do in fact appear to remain medical standards - within the boundaries of which low cholesterol numbers are a common "goal." To reject that standard would be, I think it is fair to say, a rejection of the medical establishment.

And as far as that goes, I believe my stance on these issues are quite well known.

Fortune, I'm not saying Taubes is wrong, merely that his book is one-sided. I don't recall him offering much evidence against his position. However I read the book slowly, over the course of most of last year, so I may be remembering wrong. Can you suggest some places where he does this? (Unfortunately I have the ebook, so chapter numbers would be more useful than page numbers.)

But not to avoid the real issue, which is judgement of his correctness, here my general practice is to be persuaded by the consensus rather than by my personal evaluation of the evidence, on the theory that I am not an expert on the topic and am in no position to judge. My impression is that Taubes' position is in the minority but is gaining strength. (However I don't think it is safe to extrapolate such gains forward.) I'd welcome a more informed evaluation of any reasonable consensus.

Doug S., no shift in things since June? You seem to have adopted a time horizon about as long as you have lived so far, which provides a fair chance to find options other than the current plan. If nothing else, given your limited requirements, you should be able to survive indefinitely on interest from family savings (with a roommate in a low cost of living area). Medical expenses could be problematic at some point.

frelkins (31 December, 2008): The only person who questions this fixation (with cholesterol numbers) seems to be Gary Taubes.

Indeed he does, as well as the motives and accuracy of the medical establishment. You seem to endorse both Taubes and the medical establishment at once, so yes, I was/am confused by your reaction to Finney. As to Finney's statement people on low carb diets tend to gain back almost all of the weight they lost, similar to people on other diets., I am not aware of any study in which subjects on a particular diet lost significant amounts of weight, and while still following the diet, gained it back. Indeed, I am not aware of any study where subjects strictly followed any diet long-term. What I am aware of, however, is that most people who have been significantly overweight will be so again.

During my internship, our VA hospital admitted obese veterans for a year at a time, putting them on no-calorie diets, while providing all the necessary supplements. An annual weight loss greater than 100 pounds was not unusual. Following dismissal, the average time to re-achieve the original weight was six months. It wasn't because the no-calorie diet developed a different physiologic effect.

Every week or two I run across an article which seems like it might be of interest to readers here. Today I found one in the New York Times: Some Protect the Ego by Working on Their Excuses Early. It discusses "self-handicapping", where people intentionally sabotage themselves before some challenge, so that if they fail, they have an excuse - even though the sabotage will increase the chance they will fail. The article goes into why people do this, when it works, when it doesn't work, male vs female, etc.

Pointers to these stories might well be relevant here, but don't seem to justify a top-level posting, especially not compared to the high quality analyses that Robin and Eliezer typically provide. I wonder if readers would prefer to see such pointers anyway? Or would it make more sense for me (or someone) to bookmark them and then post them in a batch, every month or so?

I would prefer batch posting, since comments are so easy to miss (and many readers apparently never look at them).

Yeah, no change since June, except that I turned 26 in July. I'm feeling a bit better today; like a hit of cocaine, playing some video games and "reading" pornographic stories tends to make me feel better for a while. (Not that I know what a hit of cocaine feels like.)

I don't know if I could live off of family savings; I expect that end-of-life care will exhaust my parent's savings and leave me with a minimal inheritance. To approximately quote my mother: "Nobody can afford a nursing home. If you go into one, first, you sell your house, then you go on Medicare." They say that I'll need an annual income of $30,000 to maintain myself as a single man living alone. Not that I'd particularly want to, anyway.

I believe the world would have been a better place had I not existed (for example, consider the tax dollars that went to pay for my college scholarship, the oil burned to transport me places, etc.) and from a purely egoistic perspective, I myself also would prefer that I never existed. I don't like living in this world and I want out - but I'm not willing to hurt other people in the process. Video games and other fiction lets me leave this world, at least for a little while, so I make do with whatever escapes I can get.

Hal, Marginal Revolution has regular posts called "Assorted Links" with a half dozen sentences with a link to another article. That seems to be a good way of pointing readers to articles that aren't worth a post of their own. If you started making such posts, readers would start sending you links to consider including, so the work is less than it seems.

Does anyone else occasionally suffer from intense blogosphere-related anxiety? You post a comment, only shortly thereafter to be stricken with fear: did I say it right? Should I have said it at all? The vicissitudes of human existence are many, and what seems like a good idea at the time of posting may simply seem embarassing in a different mood. Perhaps instituting a personal waiting period on the timescale of hours would help. A possible downside is that this increases the chances of getting scooped: if someone else posts a similar thought before you, your contribution is rendered useless. On the other hand, if you really think your comment is obvious enough such that you realistically fear being scooped, then maybe it wasn't such a great contribution after all, and you would do better to simply remain silent.

Z. M. Davis: I got a habit of posting only better comments that I write, and discarding even long researched comments if they don't look right in the end. Conscious effort to evaluate a comment with a real possibility of discarding it at least keeps short-term discontent in check most of the time, and in long term you can always say to yourself that the new recognition of past actions as mistakes supports the assertion that you are actually learning.

Looking at this comment, I thought that it's not good enough and almost discarded it, then caught a thought of how interestingly recursive that decision turned out to be, and so posted it anyway.

I suspect that Doug is the way I would have turned out if I didn't have something to protect.

Doug S., I forget if I've already pointed it out to you, but you should check out the blogs Antinatalism: The Greatest Taboo and The View From Hell.

If you're scared of the impact your suicide might have on your family members (like the author of the latter blog and to some extent the former) you might consider an option that would make them believe that you were away but still alive. Maybe if they thought you had joined the French Foreign Legion or a monastery somewhere. If you have friends they might help you out with ensuring that your real fate is not discovered by your parents within their lifetimes (perhaps by concealing your remains somehow).

I've recently started paid office work in software development, and it's not so bad. Because I'm new I'm not given that much to do, so I spend a sizeable amount of time surfing the web. You might even find you'll enjoy it if you get a job (hell, survivors of disabling accidents aren't much less happy than average thanks to adaptive coping).

I suspect that Doug is the way I would have turned out if I didn't have something to protect.

Something to protect. That must be nice.

If I ever have another office job in software development, I will kill myself. (Or quit. Quitting should work too.) The ones I've had have been absolutely horrifying. I suspect that they were atypical, but I'm not thrilled with the prospect of trying again.

Doug, maybe I’m just being optimistic, but do you put such a low probability on a high value post-human future that it’s not worth 80 years of mild depression?

Burger Flipper - good luck with the intelligence project. Whereabouts is this bargain bin? As far as I can see, Amazon UK only has it for £25, far too much for me at the moment. Anyone want to buy it and send it to London? If it's only five merkan bucks I can Paypal that plus any shipping.

Doug - have you tried taking up the ukelele? Changed my life.

Doug S., without touching the larger issues, do you want to want something different (or perhaps another "want" or two down the chain)? I am told that optimism about the value of life can be had in pill form these days.

As L points out, transhumanism may hold true within your lifetime. If you have not found something to protect by then, inexpensive uploading and wireheading should be available. That $30,000/year estimate sounds high. We housed someone who imitated a hikikomori while failing to put her life together; that cost ~$2,600 over six months, and I would be surprised if she spent more than $400 out-of-pocket. Since the $2,600 included a normal share of rent, food, and utilities, I would be surprised if living alone cost five times as much. Although, as I said, medical expenses could be an issue.

@Zubon, Doug S

"pill form"

Nope, prozac doesn't work. Doug's problems appear largely from his attachment disorder. Current findings suggest oxytocin is the chemical that regulates attachment; Doug may not make normal amounts or his receptors may be deformed.

Either way, my suggestion would he seek a prescription for oxytocin nasal spray, which is available in a clinical setting. Such treatment might alleviate his attachment disorder, thus allowing him to form bonds that would give him "something to protect." While he is undergoing treatment, he might also find an open-source project to interest him.

@Eli

The attachment disorder is most certainly why you could never ever be Doug. You appear to attach mostly normally.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31