« The Super Happy People (3/8) | Main | Interlude with the Confessor (4/8) »

February 01, 2009


I have a question on bias related to the current economic mess:

We've seen many claims that we were and are in a credit crisis (meaning the non-financial sectors of the economy were being extraordinarily choked by a lack of credit availability), many of which are backed by antidotes (though this is to be expected about anything the popular media reports on). We've also seen fewer claims that the credit crisis story is false or greatly exaggerated, mostly backed by Fed data and graphs (these claims go back to October, but current data series seem to support the same conclusions as were reached back then). Note that we are having a recession is undisputed; its the causes of the recession that I'm wondering about.

What is a rationalist to believe? Because the credit-crisis skeptic claims seem to be backed by actual data, I am siding with them. But the support for the credit crisis claim is overwhelming, and makes me think the data might not be telling the whole story. The warnings of the crisis were eerily reminiscent of the Iraqi WMD warnings, and political actors obviously have their own agendas. On the other hand, libertarian-minded pundits might want to deny widespread credit market failure.

What do you and Eliezer think of Jon Elster's work?

I think maybe we need a tutorial series on expected utility theory. Searching for "expected utility" on this site, I find many posts that depend on the concept, but nothing that actually explains the idea, it's motivation, history, alternatives/competitors, why is it the "standard" theory of rationality on this blog, etc.

For example, there's a post on the Allais Paradox, but it points to http://cepa.newschool.edu/het/essays/uncert/vnmaxioms.htm for the definition of Axiom of Independence. How many people actually read that page? The encoding on it is completely messed up. I'm seeing the axiom defined as:

(A.4) Independence Axiom: for all p, q, r Î D (X) and any a Î [0, 1], then p ³ h q if and only if a p + (1-a )r ³ h a q + (1-a )r.

(Turns out to be incompatibility with Firefox. I just tried it in Internet Explorer, which worked better. Still, I think an introduction that focuses on developing intuitions, instead of the math, would be useful.)

How would you incorporate "unknown unknowns" into probabilities?

"Violence is the last refuge of the incompetent" is a stupid statement.

It is the last refuge of the **competent** who normally try everything else they think might work first.

Violence is often the **first** refuge of the **incompetent**.

I was happily eating meat one day and I realized that killing sentient things might be the kind of thing society takes in stride but is actually really bad, like slavery used to be. I should do something like a crisis of faith regarding whether or not it's okay.

The question at the heart of the matter is: How do we morally evaluate animal affairs? So here are my intuitions, and hopefully you can help me refine them, or change my mind.

1. If you light a cat on fire, that's really bad. If you press a button to instantly vaporize an unsuspecting cow, that's morally neutral. If you step on a snail, no biggie.

2. If you keep a chicken in such a small cage that it can't turn around, that's bad. If you neuter a dog, that's better but still a little bad. If you make sure an un-neutered dog never gets to interact with a bitch (ensuring he can't have sex and puppies), that's morally neutral.

3. If an animal is already dead, the act of eating it is morally neutral. (I expect even most vegetarians will agree.) Even if the animal is a human.

4. If you pet your dog, that's good because he likes it (and we're ignoring the side effect of you liking it).

5. Imagine there is an Animal Planet which is home to large populations of all the different animals from contemporary Earth in various ecosystems, but with no humans. It would be morally *good* to instantly vaporize Animal Planet because putting all the suffering animals out of their misery is good, while killing all the happy ones is neutral.

6. If you wirehead an animal, it has the same moral value as any other orgasmium. And I think orgasmium's existence is morally neutral.

So when an animal exists, goodness is some function of its happiness that increases while the happiness is within the animal's natural range, and subsequently drops to zero.

The point at which the animal's existence is morally neutral is around "somewhat happy". As you move left from that point, it monotonically decreases without bound. And even while you're within the animal's commonly experienced levels of pain, your trough in the graph is already deeper than the peak is high.

That is my best guess right now. What do others think?

I've always wondered why there appear to be no females in the list of contributors to Overcoming Bias. Is it solely a self-selection issue, and is there some not-so-obvious type of bias involved?

I have a question for Eliezer and Robin. What, in your opinion, is the most rational thing to do for those who live outside of US and wish to sign up for cryonics?

Not too surprisingly:

"Study: Learning Science Facts Doesn't Boost Science Reasoning"

Notice of Impending Doom;

From the transhumanaist POV our existential situation is grim. I must again point out the harsh facts:

To analogize, imagine three ‘grim reapers’ armed with high-powered rifles, called CANCER, HEART DISEASE, and STROKE, sitting on a cliff top taking random pot-shots at us, humanity, crowded in the valley below. The rate of firing doubles every 8 years each of us are there, and more and more of us are being picked off at random by these snipers every day.

The windows of opportunity to ‘make a difference’ are closing down quicker than many realize. Aging shuts people down faster than they think. Between 35-40 years of age, the human body and mind really starts to take substantial ‘performance hits’. Yes, there are exceptions to the rule, but they are mentioned precisely because they *are* the exceptions. Most can no longer perform at the top –level (either physical or mentally) past 40. That’s when aging really starts to accelerate.

At age 50, pathologies of aging start to pop up – any age past 50 is considered a risk factor for cancer, stroke and heart disease, they are very rare before this age, but after this age the frequencies shoot up.

Past 60, the human body is ‘springing leaks’ faster than the Titanic, even if pathologies are not overt, there is cellular damage popping up all over the place, and survival rates plunge. Past 60, your continued survival is highly precarious.

Improvements in life expectancy have not changed these facts, the improvements are due to reductions in infant mortality only.
Transhumanists need to direct their attentions to life extension or they will not live long. Please, consider redirecting your priorities.

Thank you.

Liron, I'd say your intuition breaks down around #5. I love meat but I would consider it wrong to kill a healthy animal and not eat it. Personally I place killing animals for food slightly evilward of neutral. The counterweight I use to justify my continued consumption:

1) The animal wouldn't exist in the first place without a consumer lined up to eat it. I apply the golden rule: I'd rather exist, lead a decent life and be eaten in my prime than never exist. So I do support free range practices and the like.

2) It's a critical component of my diet. I don't digest fruit and legumes well, I'm allergic to soy and lactose intolerant, and I don't feel well after too much carbs in general. My diet is essential spinach and meat =D My quality of life drops substantially when I stray too far from that diet, and I value my life more than an animal's. I do prefer to eat 'stupid' animals though, like fish. This point is pretty weak though, I'm sure I could find some alternative diet if I really tried.

A counterpoint to my first justification might be to ask "So if you were an alien belonging to a civilization that had domesticated humans for consumption you would feel justified in eating people in order to guarantee their continued existence?" To which I would have to reply... uhh, er. Hm?


"no females"

Dr. Rebecca Roache has contributed more recently than many of the men on the contributors list. The rest of us potential female contributors are probably still working out our optimal cheating strategies for the Gale-Shapley stable marriage problem. But thanks for your concern.

The gaming-oriented webcomic Order of the stick features Vaarsuvius: an elven wizard who has recently been allowing a psychological trauma to interfere with his/her ability to aid the order in their quest. Vaarsuvius has been making lots of obvious rationalizations.

Today, Vaarsuvius is faced with a more pressing concern than the end of the world, and has something to protect. This will require him/her to either face and resolve said psychological trauma, make a deal with a devil (Imp), or do both.

Any OB readers in Tokyo and want to hang out, drop me a line: daniel dot burfoot at gmail dot com

Liron: I tend to disagree pretty strongly here. I worry about the psychological factors that lead in the direction of negative utilitarianism and wonder if they are the result of a broadly Christian culture with some ambivalence regarding the virtue of causing pleasure but none regarding the sinfulness of causing pain. I would consider it unclear whether neutering a dog painlessly is even a little bad, pretty clear that eliminating a generic cow life painlessly but without replacement is significantly bad, and very likely that vaporizing animal planet is very bad, possibly significantly worse than it naively seems which is FAR worse than neutral. The low value of orgasmium has little to do with it's place on a continuum and everything to do with being sphexish, simple, and in a sense lacking an experiencer. Agonium and orgasmium are equally neutral, and are in fact the same thing with different attached lisp tokens.

What, in your opinion, is the most rational thing to do for those who live outside of US and wish to sign up for cryonics?

There's at least KrioRus.

Agonium and orgasmium are equally neutral, and are in fact the same thing with different attached lisp tokens.

I think this is watering down the concept a little further than is necessary: orgasmium must be something a little more interesting than just a piece of paper on which someone has written "this paper is very happy right now," otherwise I question whether the concept is worth discussing.

What of all the people who were never conceived?

Is it OK to destroy humanity to prevent the death of all the unborn humans if the current human population is significantly smaller than the total amount of humans that will be born and will subsequently die provided humanity isn't destroyed now?

If I happen to develop a sentient AI program on my home computer, is it evil to run and terminate it repeatedly? What would constitute death for a sentient computer program anyway?

Ethics Report on Autonomous Military Robots:


I worry about the psychological factors that lead in the direction of negative utilitarianism and wonder if they are the result of a broadly Christian culture with some ambivalence regarding the virtue of causing pleasure but none regarding the sinfulness of causing pain.

This doesn't sound like a plausible explanation of my own intuitions toward negative utilitarianism; for one thing my intuitions aren't much more strongly "negative" when it's about other people than when it's about me. I worry that it might be a case of people simply having persistent strongly different preferences (that may lead to different policy prescriptions), caused not by one side committing cognitive errors but by differences in personality factors like introversion and neuroticism. I don't know though.

Sincerely not steven? It's my pretty strong impression that most people endorse the idea that they would rather live a life with much pain and much pleasure than one of little pain and little pleasure, but that they are unwilling to endorse this for others. This phenomenon is particularly stark regarding parents, many of whom would, it seems to me, be ashamed of children so spiritless as to live the sorts of milkwater lives that the parents nonetheless encourage.

A life with little pain and little pleasure sincerely sounds far more appealing to me than a life with much pain and much pleasure, but I'm not sure how to calibrate what "much" means in that sentence.

Related to post on abstract - concrete, far - near:

People procrastinate when asked to think in the abstract


Here is my page on Expected Utility Maximisers:


Does anyone know of any good introductionary books on probability theory and mathematical statistics written in a more casual and reflective way, more akin to popular science than heavy textbook?

gaffa: Understanding Uncertainty, by Dennis V. Lindley.

On this theme, a recommended textbook path from complete mathematical ignorance to a solid, masters level understanding in applied mathematics.

The same thing for economics, and the same thing for statistics would also be interesting. Best textbook recommendation for each course on that path from the 101 intro course to the master's thesis.

OB economists, your chance to save the world economy - though you have to do it in 1000 words or less:

"VoxEU.org today launches the Global Crisis Debate. The aim is: 1) To broaden the discussion into a truly global debate, and 2) To make the Global Crisis Debate the dominant intellectual forum on the crisis. Thanks to the partnership with the UK government, analysis on the Global Crisis Debate feeds into preparations for the April Summit via the UK Government's web site LondonSummit.gov.uk."

(That's a G-20 summit.)

"We have two main filters to avoid the ‘comment clutter’ typically seen on fully open forums. First, only professional economists can post Commentaries, and they must include their real name and professional affiliation in each Commentary... Second, each Commentary must be substantial, i.e. 200-1000 words."


Retired, the ethics report on military robots was funded by the Navy! The probability that it introduced any set of new or important ideas about robot weapons seems extraordinarily low, as any such idea would necessarily challenge the moral status quo. It will almost certainly be used as a moral rubber-stamp, providing plausible deniability for people in positions of power ("What? our robot weapons accidentally went berserk and killed one hundred thousand people? well, it's not our fault! we had those Calpoly guys study the problem, and they said it was ok!")


Economist John Quiggin thinks your terrorism futures market has been refuted by the financial crisis:

"The strong version, which gained some credence during the financial bubble era says that asset prices represent the best possible estimate taking account of all information, both public and private. It was this claim that lay behind the proposal for ‘terrorism futures’ put forward, and quickly abandoned a couple of years ago. It seems unlikely that strong-form EMH is going to be taken seriously in the foreseeable future, given the magnitude of asset pricing failures revealed by the crisis."

As always, I would enjoy to see this settled in a public debate, perhaps bloggingheads.tv. It seems Quiggin has thrown down the gauntlet...


Mitchell, those rules seem a bit restrictive unless professional economists can be posting surrogates for multiple non-professional economic thinkers.

Question for Robin:

You've talked a bit about gift giving and the incomplete explanation offered by signaling theories. What's the explanation for sponsored activities for charity? It seems that both sponsors and sponsored could signal by simply donating directly to charity, either money or perhaps by volunteering time (a more direct comparison for the sponsored). Perhaps sponsored events allow both parties to more effectively signal to peers than a simple donation since an ordinary donation would have to be advertised to provide an effective signal and this somewhat undermines the effect?

The factors determining the choice of activity in a sponsored event are also interesting. It seems that the activity must be somewhat challenging, even slightly unpleasant (a sponsored spa day seems unlikely to be well received) but extremely unpleasant activities are not common or perceived as being especially valuable. Are sponsors partly motivated by the prospect of seeing the sponsored undergo mild discomfort?

H.A., the answer may be to create a truly open auxiliary forum, like Nick Tarleton's ni.codem.us adjunct to OB (currently offline), and to mention it in the comments at the voxeu.org page above.

Matt: Are sponsors partly motivated by the prospect of seeing the sponsored undergo mild discomfort?

It's possible there's some submerged sadism going on -- I'm sure you've been in a school cafeteria when someone was offered money to drink some ingeniously vile concoction -- but a more likely explanation to me is costly signaling. I think the intended message is something like "If I had your financial means, I'd certainly give a lot of money to this cause. Maybe you should too." This message is easily faked. The greater the discomfort, the harder it is to fake. The discomfort stands in for the financial loss the person is claiming they would be willing to take on themselves.

Are sponsors partly motivated by the prospect of seeing the sponsored undergo mild discomfort?

Doubtful - since there is another explanation: the person who is being sponsored wants to show others that they are doing something for charity. The more effort they expend in the activity, the more they seem to be personally giving - so they pick challenging activities and collect as many sponsors as they can - in order to show how much they care.

A beat poem by Tim Minchin, on quackery, exclusion of supernatural, dishonesty, science, and joy in the merely real.

The way we store knowledge is with concepts. With the exception of proper nouns, concepts do not name single facts, but rather groups of things: "this and not that." So, with the exception of the raw perception, all our knowledge about a particular thing is relative to all the other things we have seen previously. Our knowledge is all intertwined, forming a giant web, with each fact defined in terms of and reenforced by, the others. Knowledge is the whole.

This is why human progress must be gradual, step-by-step, instead of great leaps. Progress requires knowledge, but to be knowledge a fact must be integrated in to the whole, and that can only be done when things are discovered sufficiently close to it's edges. Even if someone did make a giant leap, and discovered something way outside of what we already have, they would not be able to recognize it as such. Only years later, as their conceptual vocabulary grew towards it from the edge of the whole, would they say "Aha!"

So where does that leave the singularity? Great leaps are epistemologically impossible. It's not that we're dumb, and that genius AI could do it, it's that knowledge must remain a whole. There is an upper limit of the size of jump that is possible. Once you have enough intelligence to make that size jump, more is just wasted. At that point, more speed beats more brains.


"New Singularity University may overpromise the infinite"

- http://arstechnica.com/web/news/2009/02/singularity-u-launches-vanishes-after-exceeding-web-quota.ars

"NASA, Google team to form Singularity University"

- http://www.bizjournals.com/sanjose/stories/2009/02/02/daily32.html

The Science article "The Psychology of Transcending the Here and Now" has been discussed in a bunch of posts lately (it's one with the whole near-far thing). I thought I'd mention that for those without a Science subscription or university access, a non-gated copy is available from the other author's website.

Oh, and there's also a bunch of other articles discussing the same topic available from there.

My father participates in sponsored charity events. He doesn't care about the charity, he just likes to go on long bicycle trips with his friends.

I have a question concerning mind, thought and the functioning of the brain.

Is it possible for the mind, to have a moment in time (t=0)where there is no thought processed simultaneously to the sensory experience. If this is possible for the mind, is the mind still creating the concept of the ego in this state of mind? Is this state of mind in a moment in time (t=0) the meaning of enlightenment in buddhist philosophy?

Why is Zen boring?

I also have a question concerning the individual perception of the moon (object in space). I was looking at the moon (full moon) a couple of years ago, and I noticed, that I was kind of seeing the moon as a 2 dimensional object attached to the "socket" of the sky. I realized, that this was a false perception or a cognitive bias because the moon obviously is a 3 dimensional object (like the earth), that is located in the 3 dimensional space of the universe. From that time on, I am able to choose between seeing the moon 2 dimensional and 3 dimensional, whenever I look at the moon. I asked my friends about their view of the moon and most of them said, that whenever they look at the moon, they rather see it 2 dimensional then 3 dimensional.

I want to share this question with the overcomingbias community. How do the community members perceive the moon (works best with full moon)?

2 dimensional or 3 dimensional?

Why do some people perceive the moon as a 2 dimensional "painting" when it is proven to be a 3 dimensional object in the universe (p < 0.05)?

I am sorry, this could be a stupid question.

I want to add that I have been looking at moon for many years, never questioning my 2d vision of the moon. It is just an honest, naive question :)

Tim Tyler wrote:

"The more effort they expend in the activity, the more they seem to be personally giving - so they pick challenging activities and collect as many sponsors as they can - in order to show how much they care."

But if they want to demonstrate effort why not volunteer to work for a charity rather than to perform pointless labour as an indirect means of persuading others to donate to the charity? There are plenty of challenging activities that could directly assist many charities. What is the logic behind persuading someone else to pay a third party for your pointless efforts, where the more effortful the pointless effort the greater the expected payment to the third party?

I think Doug S.'s explanation makes a bit more sense, that participants actually enjoy the sponsored activity and that is a large part of why they do it. If that is the case though, why do the sponsors feel more compelled to make a donation to charity because someone else is expending their leisure time in that particular way?

But if they want to demonstrate effort why not volunteer to work for a charity rather than to perform pointless labour as an indirect means of persuading others to donate to the charity?

Why do more students not do volunteer to work for charities - instead of the apparently-ridiculous sponsoring business? A few do - but mostly they have other commitments, couldn't help much anyway - and actual work doesn't perform any signalling role: they still have to actually go and tell their friends about how caring the job they have is.

Blog bug? Where did the "Share likelihood ratios" posting go?

Psy-Kosh, I tried to change a typo in Steve's and my "Share likelihood ratios" post, and it turned the post back into "draft" status. I didn't realize that was how permissions worked. I called Eliezer, who will repost it when he gets home at 3:30pm PST or so.

Anna: ah, okay. That post was particularly good, so I'd hate to see it vanish permanently.

(I say that with embarrassment in that I already knew that for independent evidences, one simply multiplies the likelihoods. I understood that. And yet it hadn't really explicitly occurred to me that sharing the likelihoods (and enough info to determine which are likely to be independent) is the obvious and natural simple way for rationalists to share information and come to agreement. DOH! :))

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