Monthly Archives: August 2008

Beauty Bias

In a TV game show, pretty contestants were not better or more cooperative players, but other contestants seemed to act as if they were:

It’s an uncomfortable truth that beautiful people make more money. … Now a study of a TV game show supports the prejudice hypothesis. … V. Bhaskar … analysed 69 episodes of Shafted … At the end of a round, the highest-scoring player picks a contestant to eliminate. Although the least attractive players scored no worse in the show than others, they were twice as likely to be eliminated in the first round. The contestants did not seem to base their decision on other factors such as age or sex. …

Contestants also confused attractiveness with cooperativeness. In the final round of Shafted, the last two players vie for an accumulated pot of money. Each player must opt to share the prize or attempt to grab it all for themselves. If one player opts to grab while one opts to share, the grabber takes the lot. If both try to grab, they both leave empty-handed, so game theory dictates that the leading contestant should pick a fellow finalist who is likely to cooperate. 

Even though attractiveness was found to have no bearing on cooperativeness, the leader often elected to play the final round with the most attractive of their remaining rivals. In 13 shows, these looks-based decisions even overrode a simple imperative to choose their highest-scoring rival, which would have led to an increases in the ultimate prize fund. In these cases, the prize was E350 lower than it could have been, on average.

It also seems we think everyone is prettier when we are tipsy:

Researchers … randomly assigned 84 heterosexal students to consume either a non-alcoholic lime-flavoured drink or an alcoholic beverage with a similar flavour. … After 15 minutes, the students were shown pictures of people their own age, from both sexes.  Both men and women who had consumed alcohol rated the faces as being more attractive than did the controls … The effect was not limited to the opposite sex – volunteers who had drunk alcohol also rated people from their own sex as more attractive.

So, do we think everyone is better and more cooperative when we are tipsy?

GD Star Rating
Tagged as:

Invisible Frameworks

Followup toPassing the Recursive Buck, No License To Be Human

Roko has mentioned his "Universal Instrumental Values" several times in his comments.  Roughly, Roko proposes that we ought to adopt as terminal values those things that a supermajority of agents would do instrumentally.  On Roko’s blog he writes:

I’m suggesting that UIV provides the cornerstone for a rather new approach to goal system design. Instead of having a fixed utility function/supergoal, you periodically promote certain instrumental values to terminal values i.e. you promote the UIVs.

Roko thinks his morality is more objective than mine:

It also worries me quite a lot that eliezer’s post is entirely symmetric under the action of replacing his chosen notions with the pebble-sorter’s notions. This property qualifies as "moral relativism" in my book, though there is no point in arguing about the meanings of words.

My posts on universal instrumental values are not symmetric under replacing UIVs with some other set of goals that an agent might have. UIVs are the unique set of values X such that in order to achieve any other value Y, you first have to do X.

Continue reading "Invisible Frameworks" »

GD Star Rating

Dark Dreams

Here’s another reason to prefer reality over dreams; dreams are darker:

We collected dream reports (N=419) and daily event logs (N=490) from 39 university students during a two-week period, and interviewed them about real threat experiences retrievable from autobiographical memory (N=714). Threat experiences proved to be much more frequent and severe in dreams than in real life, and Current Dream Threats more closely resembled Past than Current Real Threats.

If someday we have tech to suppress dreams (or at least memories of them), will it be considered cruel to allow your kids to dream?  HT to Tyler.

GD Star Rating
Tagged as:

No License To Be Human

Followup toYou Provably Can’t Trust Yourself

Yesterday I discussed the difference between:

  • A system that believes – is moved by – any specific chain of deductions from the axioms of Peano Arithmetic.  (PA, Type 1 calculator)
  • A system that believes PA, plus explicitly asserts the general proposition that PA is sound.  (PA+1, meta-1-calculator that calculates the output of Type 1 calculator)
  • A system that believes PA, plus explicitly asserts its own soundness.  (Self-PA, Type 2 calculator)

These systems are formally distinct.  PA+1 can prove things that PA cannot.  Self-PA is inconsistent, and can prove anything via Löb’s Theorem.

With these distinctions in mind, I hope my intent will be clearer, when I say that although I am human and have a human-ish moral framework, I do not think that the fact of acting in a human-ish way licenses anything.

I am a self-renormalizing moral system, but I do not think there is any general license to be a self-renormalizing moral system.

And while we’re on the subject, I am an epistemologically incoherent creature, trying to modify his ways of thinking in accordance with his current conclusions; but I do not think that reflective coherence implies correctness.

Continue reading "No License To Be Human" »

GD Star Rating

Are UFOs Aliens?

I’m a social scientist with a high estimate of the power of social science (especially economics and sociobiology) to trace the outlines of a wide variety of social behavior.  I even use social science to estimate our distant descendants’ future, and the astronomical signatures that aliens might leave.

Some complain that such efforts reflect an overconfidence in social science, that academic insights today have little power to generalize to such distant situations.  Some even say social science does not exist, i.e., that it fails to offer much insight even into human society today.  But most of these social science skeptics also want to say we are pretty sure UFOs are not aliens – that aliens are not regularly visiting Earth today.  And during a long drive from DC to Indianapolis and back with Bill and Chris Dickens (to attend GenCon), I realized this is contradictory: social science is our main theoretical basis for thinking no UFOs are aliens!

Humans have long reported various rare and odd phenomena, from angels to faeries to bigfoot to sea monsters to ghosts.  We are pretty sure most such "weird" reports are errors, i.e., mistakes, frauds, or misunderstandings.  We also reasonably believe most weird report categories are entirely errors  — for example, we reasonably believe all faery reports are errors. 

Nevertheless, we can’t be especially confident that a category of weird reports is all error, merely because it is weird.  After all, some weird categories have been vindicated, i.e., we now think many reports really were as weird as claimed.  Meteorite reports, for example, were once thought to be crazy – why the heck would rocks fall from the sky?  I’d love for someone to survey categories of weird reports made a few centuries ago, identifying the categories most clearly settled by now, and seeing what fraction of settled categories have been vindicated.  My guess would be roughly 5%.

Continue reading "Are UFOs Aliens?" »

GD Star Rating
Tagged as:

You Provably Can’t Trust Yourself

Followup toWhere Recursive Justification Hits Bottom, Löb’s Theorem

Peano Arithmetic seems pretty trustworthy.  We’ve never found a case where Peano Arithmetic proves a theorem T, and yet T is false in the natural numbers.  That is, we know of no case where []T ("T is provable in PA") and yet ~T ("not T").

We also know of no case where first order logic is invalid:  We know of no case where first-order logic produces false conclusions from true premises. (Whenever first-order statements H are true of a model, and we can syntactically deduce C from H, checking C against the model shows that C is also true.)

Combining these two observations, it seems like we should be able to get away with adding a rule to Peano Arithmetic that says:

All T:  ([]T -> T)

But Löb’s Theorem seems to show that as soon as we do that, everything becomes provable.  What went wrong?  How can we do worse by adding a true premise to a trustworthy theory?  Is the premise not true – does PA prove some theorems that are false?  Is first-order logic not valid – does it sometimes prove false conclusions from true premises?

Continue reading "You Provably Can’t Trust Yourself" »

GD Star Rating

Caplan Gums Bullet

Bryan Caplan:

I often disagree with people who know more about a given topic than I do. … I reason, if I did immerse myself in the modern literature, it’s a lot more likely that I would arrive at a sophisticated version of my current view than that I would radically change my mind. … When I argue with people who are better-informed than I am … [I ask] "If I saw and read everything that you’ve seen and read, what would I conclude?" … Even though the disputants are not on a level playing field, that isn’t the real reason why they hold different views. …

I suspect that Robin Hanson will be disturbed by my heuristic.  After all, its lets every person retain his view that his prior is "special."  You could even call my method the Anti-Hansonian Heuristic, because it deliberately ignores the fact that lots of smart people persistently disagree with you.  In response to Robin, though, I’d say that (a) it’s almost impossible to convince anyone that his prior isn’t special – and my heuristic improves the quality of beliefs despite this impasse; and (b) since my prior is special (laugh if you must!), this is a great heuristic for me to live by.

I’d like to say Bryan bites a bullet here, but alas he just gums it, as he doesn’t engage the hard questions:  what exactly is his better-origin scenario/story, and what evidence supports that story over less-flattering stories?  That is, how could Caplan tell the difference between a situation where his prior was good and mine bad, vs. a situation where his prior was bad and mine good?  If he grants that a reasonable person, long before our births, would have thought these two situations equally likely, what later evidence could have convinced this reasonable person that Bryan’s prior turned out better? 

GD Star Rating
Tagged as:

Dumb Deplaning

So I just traveled to Portsmouth, VA for an experimental conference – in the sense that I don’t expect conferences of this type to prove productive, but maybe I should try at least once – in the unlikely event that there are any local Overcoming Bias readers who want to drive out to Portsmouth for a meeting on say the evening of the 20th, email me – anyway, I am struck, for the Nth time, how uncooperative people are in getting off planes.

Most people, as soon as they have a chance to make for the exit, do so – even if they need to take down luggage first.  At any given time after the initial rush to the aisles, usually a single person is taking down luggage, while the whole line behind them waits.  Then the line moves forward a little and the next person starts taking down their luggage.

In programming we call this a "greedy local algorithm".  But since everyone does it, no one seems to feel "greedy".

How would I do it?  Off the top of my head:

"Left aisle seats, please rise and move to your luggage.  (Pause.)  Left aisle seats, please retrieve your luggage.  (Pause.)  Left aisle seats, please deplane.  (Pause.)  Right aisle seats, please rise and move to your luggage…"

Continue reading "Dumb Deplaning" »

GD Star Rating

Mundane Dishonesty

It bothers me that my commitment to overcoming bias is contracted daily by these dishonesties:

  1. I am tragically uncertain about how I should divide my time between work and play, among various possible work projects, and with who I should ally and spend my time.  These are terribly important decisions about which I have only very weak clues.  But I usually act as if I know what I’m doing.
  2. I actually care a lot what other people think of me, and in most conversations the major topic for most everyone is who praises or blames who else how much.  But this strong subtext is usually not acknowledged in our explicit text.  Like most people, I act as if we were talking about other topics, and only indirectly make points of praise or blame.
  3. I think about sex an awful lot – it is not far from my awareness anytime I am in the presence of, or thinking about, most any healthy female.  But I almost never acknowledge that fact directly via my actions or words.

Now I don’t think I’m very different from most people on these points.  And there are obviously very good reasons why we are dishonest in these ways.  Telling associates explicitly how uncertain we are about associating with them would seem like threatening to "break up" with them.  Talking explicitly about who we like how much would sound like bragging and insecure requests for praise.  And talking explicitly about sexual undercurrents would usually be seen as sexual harassment. 

So like most people I am stuck in a signaling equilibrium where my best strategy is to act in a way that seems to me dishonest.  Oh you might say that everyone knows about all this so I’m not really fooling  anyone.  But while we "know" at some level, to function effectively it seems we must self-deceive enough to often take appearances at face value.   (See a nice related quote by Nagel, courtesy of Richard.)

GD Star Rating
Tagged as:

The Cartoon Guide to Löb’s Theorem

Lo!  A cartoon proof of Löb’s Theorem!

Löb’s Theorem shows that a mathematical system cannot assert its own soundness without becoming inconsistent.  Marcello and I wanted to be able to see the truth of Löb’s Theorem at a glance, so we doodled it out in the form of a cartoon.  (An inability to trust assertions made by a proof system isomorphic to yourself, may be an issue for self-modifying AIs.)

It was while learning mathematical logic that I first learned to rigorously distinguish between X, the truth of X, the quotation of X, a proof of X, and a proof that X’s quotation was provable.

The cartoon guide follows as an embedded Scribd document after the jump, or you can download from as a PDF file.  Afterward I offer a medium-hard puzzle to test your skill at drawing logical distinctions.

Continue reading "The Cartoon Guide to Löb’s Theorem" »

GD Star Rating