Overcoming Bias

Immodest Caplan

Bryan Caplan is concerned about this objection to his book:

Caplan says that people tend to be irrational on questions where there are no direct material costs of being wrong.  But there are no direct material costs to Caplan of being wrong on most if not all of the questions he addresses in this book.

Bryan responds:

I take this possibility seriously.  In all honesty, my situation is precisely the kind in which I claim that people's attachment to rationality is weakest. ... Rather than making a mostly futile effort to convince you of my own cognitive virtues, I prefer to direct your attention inwards.  Do your best to put your feelings and ideological commitments aside, and judge my claims on their own merits. ... When dealing with abstract, "impractical" areas like politics and economics, it is unfair to infer that a viewpoint is wrong just because it is unpopular.  The best way to evaluate contrarian arguments, once again, is to put your feelings and ideological commitments aside and actually listen. ...

A final interesting possibility is that I am basically right but fail to temper my judgments with due modesty. ... Pleas for greater modesty should be viewed with suspicion on both strategic and intrinsic grounds.  Strategically, the problem with modesty is that we live in a culture of energetic self-promotion.  In this environment, humility is the equivalent of unilateral disarmament. ...  More fundamentally, though, neither I nor any other economist I know claims to be infallible or anything close.  All I claim is that on average, economists' judgments about economic policy are a lot more trustworthy than the public's. ... Remember that limited expertise is better than none at all.

Alas Bryan stumbles twice here.  First, if we are, as he says, much more irrational on no-direct-material-cost questions, the reasonable response is to have much lower confidence in our opinions on those topics, relative to other topics.  Bryan says he is not claiming to be infallible and that experts are better than amateurs, but those are beside the point.  The issue is simple: does Bryan in fact have much lower confidence in his political opinions, vs. his other opinions?  If so, he is consistent; if not, not. 

Second, Bryan seems to mistakenly conclude that one is more justified in relying on one's own direct evaluation of arguments, relative to the evaluations of others, on topics where people tend to be more irrational.  But that irrational tendency will afflict both his evaluation and others' evaluations; it doesn't obviously afflict others' more than his.  The reason to weigh the evaluations of others, and not just focus entirely on your own, is that you might make mistakes.  That reason surely remains as relevant on topics where most everyone makes big mistakes. 

'Anyone who thinks the Large Hadron Collider will destroy the world is a t**t.'

This week is Big Bang Week at the BBC, with various programmes devoted to the switch-on of CERN’s Large Hadron Collider (LHC) on Wednesday morning.  Many of these programmes are covered in this week’s issue of the Radio Times—the BBC’s listings magazine—which also features a short interview with Professor Brian Cox, chair of particle physics at the University of Manchester. Asked about concerns that the LHC could destroy the earth, he replies:

‘The nonsense you find on the web about “doomsday scenarios” is conspiracy theory rubbish generated by a small group of nutters, primarily on the other side of the Atlantic.  These people also think that the Theory of Relativity is a Jewish conspiracy and that America didn’t land on the Moon.  Both are more likely, by the way, than the LHC destroying the world.  I’m slightly irritated, because this non-story is symptomatic of a larger mistrust in science, particularly in the US, which includes things like intelligent design. [… A]nyone who thinks the LHC will destroy the world is a t**t.’ (Final word censored by Radio Times.) [1]

Who counts as a nutter and a t**t on this reckoning?  It is true that anyone who thinks there is a 100% chance that the LHC will definitely destroy the world is confused—but it’s probably also true that not many people really think this.  On the other hand, if anyone who thinks that it is worth taking seriously the (admittedly very slim) possibility that the LHC will destroy the world is a t**t, then there are many apparently very clever t**ts knocking about in our universities.  Among these are several of my colleagues: Nick Shackel has previously blogged about the risks of turning on the LHC, as has Toby Ord; and Rafaela Hillerbrand, Toby Ord, and Anders Sandberg recently presented on this topic at the recent Future of Humanity Institute-hosted conference on Global Catastrophic Risks. And, despite having chatted to each of these people about the LHC at some point or another, I’ve never heard any of them express sympathy for the view that the Theory or Relativity is a Jewish conspiracy or that nobody landed on the Moon.  So, are they t**ts or not?

Disagreement is Disrespect

Consider these dueling bumper stickers:

Here religious conservatives do seem unfairly maligned: seeing a behavior as immoral is not at all the same as "hating."  These folks also rightly seethe at how they are usually portrayed in popular film and TV, and at seeing their democratic ideals violated when even local voting majorities can't prevent their kids from being taught evolution in public schools.  You can feel this resentment in the enthusiasm for Palin.  (Of course since I'm not religious about God, sexual preference, or democracy, this all bothers me lots less.)

But this does seem a handy opportunity to repeat that while disagreement isn't hate, it is disrespect.  When you knowingly disagree with someone you are judging them to be less rational than you, at least on that topic.  (Judging them less informed or experienced by itself can't create disagreement.)  It might be only a minor disrespect, if you think this disagreement suggests little about whether you'd disagree with them elsewhere.  But disagreement is disrespect, nonetheless.

Added: Wikipedia says hate speech is:

Speech intended to degrade, intimidate, or incite violence or prejudicial action against a person or group of people based on their race, gender, age, ethnicity, nationality, religion, sexual orientation, gender identity, disability, language ability, moral or political views, socioeconomic class, occupation or appearance (such as height, weight, and hair color), mental capacity and any other distinction-liability. [emphasis added]

How exactly do you disagree with someone's moral views without degrading them?  Can you really say pedophelia is disgusting without degrading pedophiles?

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? 

Hot Air Doesn't Disagree

Followup to:  The Bedrock of Morality, Abstracted Idealized Dynamics

Tim Tyler comments:

Do the fox and the rabbit disagree? It seems reasonable so say that they do if they meet: the rabbit thinks it should be eating grass, and the fox thinks the rabbit should be in the fox's stomach. They may argue passionately about the rabbit's fate - and even stoop to violence.

Boy, you know, when you think about it, Nature turns out to be just full of disagreement.

Rocks, for example, fall down - so they agree with us, who also fall when pushed off a cliff - whereas hot air rises into the air, unlike humans.

I wonder why hot air disagrees with us so dramatically.  I wonder what sort of moral justifications it might have for behaving as it does; and how long it will take to argue this out.  So far, hot air has not been forthcoming in terms of moral justifications.

Physical systems that behave differently from you usually do not have factual or moral disagreements with you.  Only a highly specialized subset of systems, when they do something different from you, should lead you to infer their explicit internal representation of moral arguments that could potentially lead you to change your mind about what you should do.

The Bedrock of Morality: Arbitrary?

Followup to:  Is Fairness Arbitrary?, Joy in the Merely Good,  Sorting Pebbles Into Correct Heaps

Yesterday, I presented the idea that when only five people are present, having just stumbled across a pie in the woods (a naturally growing pie, that just popped out of the ground) then it is fair to give Dennis only 1/5th of this pie, even if Dennis persistently claims that it is fair for him to get the whole thing.  Furthermore, it is meta-fair to follow such a symmetrical division procedure, even if Dennis insists that he ought to dictate the division procedure.

Fair, meta-fair, or meta-meta-fair, there is no level of fairness where you're obliged to concede everything to Dennis, without reciprocation or compensation, just because he demands it.

Which goes to say that fairness has a meaning beyond which "that which everyone can be convinced is 'fair'".  This is an empty proposition, isomorphic to "Xyblz is that which everyone can be convinced is 'xyblz'".  There must be some specific thing of which people are being convinced; and once you identify that thing, it has a meaning beyond agreements and convincing.

You're not introducing something arbitrary, something un-fair, in refusing to concede everything to Dennis.  You are being fair, and meta-fair and meta-meta-fair.  As far up as you go, there's no level that calls for unconditional surrender.  The stars do not judge between you and Dennis - but it is baked into the very question that is asked, when you ask, "What is fair?" as opposed to "What is xyblz?"

Ah, but why should you be fair, rather than xyblz?  Let us concede that Dennis cannot validly persuade us, on any level, that it is fair for him to dictate terms and give himself the whole pie; but perhaps he could argue whether we should be fair?

The hidden agenda of the whole discussion of fairness, of course, is that good-ness and right-ness and should-ness, ground out similarly to fairness.

Abstracted Idealized Dynamics

Followup to:  Morality as Fixed Computation

I keep trying to describe morality as a "computation", but people don't stand up and say "Aha!"

Pondering the surprising inferential distances that seem to be at work here, it occurs to me that when I say "computation", some of my listeners may not hear the Word of Power that I thought I was emitting; but, rather, may think of some complicated boring unimportant thing like Microsoft Word.

Maybe I should have said that morality is an abstracted idealized dynamic.  This might not have meant anything to start with, but at least it wouldn't sound like I was describing Microsoft Word.

How, oh how, am I to describe the awesome import of this concept, "computation"?

Perhaps I can display the inner nature of computation, in its most general form, by showing how that inner nature manifests in something that seems very unlike Microsoft Word - namely, morality.

Consider certain features we might wish to ascribe to that-which-we-call "morality", or "should" or "right" or "good":

• It seems that we sometimes think about morality in our armchairs, without further peeking at the state of the outside world, and arrive at some previously unknown conclusion.

Someone sees a slave being whipped, and it doesn't occur to them right away that slavery is wrong.  But they go home and think about it, and imagine themselves in the slave's place, and finally think, "No."

Can you think of anywhere else that something like this happens?

Moral Error and Moral Disagreement

Followup to:  Inseparably Right, Sorting Pebbles Into Correct Heaps

Richard Chappell, a pro, writes:

"When Bob says "Abortion is wrong", and Sally says, "No it isn't", they are disagreeing with each other.

I don't see how Eliezer can accommodate this. On his account, what Bob asserted is true iff abortion is prohibited by the morality_Bob norms. How can Sally disagree? There's no disputing (we may suppose) that abortion is indeed prohibited by morality_Bob...

Since there is moral disagreement, whatever Eliezer purports to be analysing here, it is not morality."

The phenomena of moral disagreement, moral error, and moral progress, on terminal values, are the primary drivers behind my metaethics.  Think of how simple Friendly AI would be if there were no moral disagreements, moral errors, or moral progress!

Richard claims, "There's no disputing (we may suppose) that abortion is indeed prohibited by morality_Bob."

We may not suppose, and there is disputing.  Bob does not have direct, unmediated, veridical access to the output of his own morality.

Incomplete Analysis

In reading the comments on my variance-induced test bias post, I was reminded of a big bias loophole in social science: judging when an analysis is complete "enough."  We usually have some status quo policies, and some analyses relevant to those policies.  Each analysis tends to favor some possible policies relative to others, but alas most every analysis is incomplete, leaving out relevant considerations. 

Now we do need to assess which analyzes are most relevant to any given policy question, but at least here experts can, when analyses are similar enough, usually bring to bear some relatively "objective" criteria.  When we ask if the relevant analyses are good "enough" to justify action, however, we can usually appeal only to much weaker standards of evaluation. 

Why Argue Values?

We know truth-seekers should not knowingly disagree about facts.  Many are eager to justify their disagreements, however, by noting that values infuse most common disputed topics, such as politics, morality, music, and so on.  And yes, there may be nothing irrational about you preferring chocolate while another prefers strawberry.  But while it may not be problematic to have differing opinions on values, it is problematic to argue about opinions due to values. 

In arguments, people commonly offer reasons, such as evidence, analysis, and compelling examples, to support their opinions.  Furthermore, people think these reasons should have a good chance to be persuasive, to induce reasonable listeners to change their mind in the suggested direction.  We are often surprised and indignant to see others unpersuaded by what we consider strong arguments.  But if it were just a matter of each person having different values, why should arguments change our minds? 

If when you argue, you try to show people the bad consequences of the policies they endorse, then your dispute is on facts of those consequences.  If you try to show people that their values are at odds with other commonly accepted values, then your dispute is on facts of which values are at odds, or facts of whether these people embrace those other values.  More generally, if you think others are mistaken on their values, and that your reasons should help them see their true values, then your dispute is on facts of what are their actual values.