Overcoming Bias

Lumpaland Parable

Once upon a time ...

Before Willy Wonka came to Lumpaland, hundreds of fierce dragons preyed on ten thousand Oompa Loompas and other wildlife.  A thousand Oopmas tried in their free time to oppose dragons, and a hundred had full time jobs at the Dragon Institute. 

Instituters were impressive - they had charisma, spoke eloquently, made cool devices and mastered hard math.  Others wanted to read about, sleep with, and study under them.  At any one time only a few instituters were out near dragons, usually at a safe distance, focused on a few relatively-safe dragons.  Some complained instituters were too distracted playing institute politics and trying to seem impressive.  But when Oopmas had to choose between an instituter and an amateur, the instituter usually won.

Amateurs were mostly content to read and argue.  And their readings and conversations rarely lingered long on one dragon.  While instituters focused on particular dragons, amateurs prided themselves on having passionate witty opinions on many dragons.  Amateurs were eager to associate with instituters, even as they complained instituters unfairly neglected their writings and favorite dragons.

The few amateurs who focused on particular dragons were considered boring, and amateurs who actually fought dragons were considered dangerous, to be avoided.  If an amateur actually managed to dispatch a dragon, the Oompas nearby trusted to report on the incident were usually instituters, who would if possible take full credit.  If the amateur's role could not be denied, he'd be thanked for his lucky assistance to the institute's grand plan. To get more recognition, he'd have to dispatch several dragons or join the institute.

Question: Biases afflict both amateur and instituter dragon judgments. Perhaps do-nothing amateurs are less biased, but so what?  Isn't it do-somethings' opinions that matter?

Added 16May: This story was inspired seeing amateurs indignant that professionals do not engage their brief writings on difficult complex topics (e.g., many worlds, zombies, Fermi question, nano/robot econ, disagreement, market manipulation).

Science Doesn't Trust Your Rationality

Followup to:  The Dilemma: Science or Bayes?

Scott Aaronson suggests that Many-Worlds and libertarianism are similar in that they are both cases of bullet-swallowing, rather than bullet-dodging:

Libertarianism and MWI are both are grand philosophical theories that start from premises that almost all educated people accept (quantum mechanics in the one case, Econ 101 in the other), and claim to reach conclusions that most educated people reject, or are at least puzzled by (the existence of parallel universes / the desirability of eliminating fire departments).

Now there's an analogy that would never have occurred to me.

I've previously argued that Science rejects Many-Worlds but Bayes accepts it.  (Here, "Science" is capitalized because we are talking about the idealized form of Science, not just the actual social process of science.)

It furthermore seems to me that there is a deep analogy between (small-'l') libertarianism and Science:

1. Both are based on a pragmatic distrust of reasonable-sounding arguments.
2. Both try to build systems that are more trustworthy than the people in them.
3. Both accept that people are flawed, and try to harness their flaws to power the system.

The Failures of Eld Science

Followup to:  Initiation Ceremony, If Many-Worlds Had Come First

This time there were no robes, no hoods, no masks.  Students were expected to become friends, and allies.  And everyone knew why you were in the classroom.  It would have been pointless to pretend you weren't in the Conspiracy.

Their sensei was Jeffreyssai, who might have been the best of his era, in his era.  His students were either the most promising learners, or those whom the beisutsukai saw political advantage in molding.

Brennan fell into the latter category, and knew it.  Nor had he hesitated to use his Mistress's name to open doors.  You used every avenue available to you, in seeking knowledge; that was respected here.

"- for over thirty years," Jeffreyssai said.  "Not one of them saw it; not Einstein, not Schrödinger, not even von Neumann."  He turned away from his sketcher, and toward the classroom.  "I pose to you to the question:  How did they fail?"

The students exchanged quick glances, a calculus of mutual risk between the wary and the merely baffled.  Jeffreyssai was known to play games.

Many Worlds, One Best Guess

Previously in series: Collapse Postulates
Followup to:  Bell's Theorem, Spooky Action at a Distance, Quantum Non-Realism, Decoherence is Simple, Falsifiable and Testable

If you look at many microscopic physical phenomena - a photon, an electron, a hydrogen atom, a laser - and a million other known experimental setups - it is possible to come up with simple laws that seem to govern all small things (so long as you don't ask about gravity).  These laws govern the evolution of a highly abstract and mathematical object that I've been calling the "amplitude distribution", but which is more widely referred to as the "wavefunction".

Now there are gruesome questions about the proper generalization that covers all these tiny cases.  Call an object 'grue' if it appears green before January 1, 2020 and appears blue thereafter.  If all emeralds examined so far have appeared green, is the proper generalization, "Emeralds are green" or "Emeralds are grue"?

The answer is that the proper generalization is "Emeralds are green".  I'm not going to go into the arguments at the moment.  It is not the subject of this post, and the obvious answer in this case happens to be correct.  The true Way is not stupid: however clever you may be with your logic, it should finally arrive at the right answer rather than a wrong one.

In a similar sense, the simplest generalizations that would cover observed microscopic phenomena alone, take the form of "All electrons have spin 1/2" and not "All electrons have spin 1/2 before January 1, 2020" or "All electrons have spin 1/2 unless they are part of an entangled system that weighs more than 1 gram."

Caplan Pulls Along Ropes

Last May I wrote:

The space of all policies ... is huge - with thousands or millions of dimensions. ... The policy world can thought of as consisting of a few Tug-O-War "ropes" set up in this high dimensional policy space.  If you want to find a comfortable place in this world, where the people around you are reassured that you are "one of them," you need to continually and clearly telegraph your loyalty by treating each policy issue as another opportunity to find more supporting arguments for your side of the key dimensions.  That is, pick a rope and pull on it.   If, however, you actually want to improve policy ... then prefer to pull policy ropes sideways.

Bryan Caplan prefers to pull along the ropes:

Endearing Sincerity

Honestly demands I admit that soon after penning that sincerity is overrated, and that fiction typically distracts from reality, I fell in love with a fictional movie celebrating sincerity:  Once, depicting people who really love music, more than money or sex or anything.  It resonated with my cherished memories of being a teen religious cultist (~'74), and a young adult in an idealistic tech community exploring the web, nanotech, and more ('84-93).  People feel tied especially tied to others with whom they share a deep love of an unpopular or unrewarded hobby.  When other rewards loom larger, such as money, fame, sex, etc., we are less sure of our associates' motives.

I'm lucky to be a professor, but alas since this job pays money and prestige, most of the people I deal with seem to primarily seek such rewards.  Gordon Tullock (office next to mine) in 1966:

An investigator wholly motivated by induced curiosity is different in many ways from one motivated by either curiosity or a desire to make practical application of new knowledge.  ... If he could establish and maintain his reputation, and hence his job, by reporting completely fictional discoveries, this would accomplish his end.  ... Those administering a system of induced research ...  must make certain that [such] investigators are induced to pay attention to the real world.  As we have seen, the actual system used by administrators in our present setup is simply to count the number of papers published by a man in journals of various degrees of reputation.  The reputation of the journals, again as we have seen, is determined by their readers.  ... A self-perpetuating process might be set in motion in which a journal read only by people motivated by induced curiosity gradually slipped away from reality in the direction of superficially impressive but actually easy research projects.  In most sciences this does not happen. ... One symptom of the existence of this condition is the development of very complex methods of treating subject which can be readily handled by simple methods (pp56-57).

It is worse that Tullock thought.  Few academic topics are dominated by topic lovers; intellectual progress is largely a side effect of prestige seeking.  And even "sincere" topic love is directed by our ancient evolved coding designed to gain us more basic rewards.  But even so, I miss being part of a community primarily tied by a common love of a topic or activity, vs. wider prestige or money.  Not sure how consistent this is with anything else I've written.

Arbitrary Silliness

When I visited Oxford a few weeks ago I brought up a subject which has been bugging me lately - we don't understand what makes research topics "silly."   For example:

Apparently most people the world over think aliens exist, think searches might find them, think that would be a very important discovery, but think the subject is way too silly to justify government funding. 

Similarly, most people think futarchy (government by betting markets) is silly, even though most think it has a decent chance of performing well, and even though it isn't obviously less likely to happen than a strong world government, which is not nearly as silly.  Or see this giggle-fest on future robot threats. 

At Oxford we listed possible obstacles to dealing better with global catastrophic risks, and we guessed the biggest obstacle is that the topic seems silly.  This puzzle of what makes topics silly seems to have stuck in the mind of Anders Sandberg: 

Regarding some things as silly does not seem to result from an estimation that the probability is extremely low, it seems to be a direct rejection of it as unthinkable and irrelevant - not the same thing, although the rejector will quickly argue that the chances of the things happening are minuscule. The rejection has many similarities to the yuck reaction we see in ethics, where certain possibilities are rapidly rejected as immoral with little reflection (c.f. the work of Haidt). So maybe the best explanation of what makes a paper silly is just that it goes against the social intuitions we have built up about thinkable, serious subjects. Space travel is science fiction and science fiction has low status, so hence papers about the economics of space travel must be silly. Life extension is silly, so papers looking at its consequences must be silly. Framing world government in terms of non-silly globalisation makes it non-silly.

This silliness-taboo has been a thorn in my side all my life, so I'm eager for any insight. 

Initiation Ceremony

    The torches that lit the narrow stairwell burned intensely and in the wrong color, flame like melting gold or shattered suns.
    192... 193...
    Brennan's sandals clicked softly on the stone steps, snicking in sequence, like dominos very slowly falling.
    227... 228...
    Half a circle ahead of him, a trailing fringe of dark cloth whispered down the stairs, the robed figure itself staying just out of sight.
    239... 240...
    Not much longer, Brennan predicted to himself, and his guess was accurate:
    Sixteen times sixteen steps was the number, and they stood before the portal of glass.
    The great curved gate had been wrought with cunning, humor, and close attention to indices of refraction: it warped light, bent it, folded it, and generally abused it, so that there were hints of what was on the other side (stronger light sources, dark walls) but no possible way of seeing through - unless, of course, you had the key: the counter-door, thick for thin and thin for thick, in which case the two would cancel out.
    From the robed figure beside Brennan, two hands emerged, gloved in reflective cloth to conceal skin's color.  Fingers like slim mirrors grasped the handles of the warped gate - handles that Brennan had not guessed; in all that distortion, shapes could only be anticipated, not seen.
    "Do you want to know?" whispered the guide; a whisper nearly as loud as an ordinary voice, but not revealing the slightest hint of gender.
    Brennan paused.  The answer to the question seemed suspiciously, indeed extraordinarily obvious, even for ritual.

Neglecting Conceptual Research

Scott Aaronson and CS theory colleagues complain conceptual insights are slighted relative to technical results:

The trends that worry us are ... Assignment of little weight to "conceptual" considerations, while assigning the dominant weight to technical considerations. ... by "conceptual" we mean the aspects that can be communicated succinctly, with a minimum amount of technical notation, and yet their content reshapes our view/understanding. Conceptual contributions can be thought of as contents of the work that are most likely to be a part of a scientific hallway discussion. ... Once understood, conceptual aspects tend to be viewed as obvious, which actually means that they have become fully incorporated in the worldview of the expert. ... our community should be warned of dismissing such contributions by saying "yes, but that's obvious"; when somebody says such a thing, one should ask "was it obvious to you before reading this article?"

Scott elaborates:

People will often say, "sure, but as soon as you've asked the question / defined the model that way, the answer is obvious." They recognize, but don't sufficiently appreciate, the fact that before the paper in question no one had asked the question or defined the model that way.

Here are a few of the 76 comments.  Travis:

Human Capital "Puzzle" Explained?

Today is Capital Day again, a day to celebrate physical, social, and especially human capital, the added productivity from experience and training. 

Human capital offers an interesting case study in theory versus data.  Just as most people think it obvious that medicine deserves most of the credit for health gains, most people think it obvious that education deserves most of the credit for human capital gains.  Do-gooders the world over have for centuries "known" that what the poor really need is more medicine and education (and religion and art). 

We theorists will tell you that, yes productive people tend to be better educated, but there are many possible explanations for wealth-education correlations.  For example, schooling could be a credible signal of ability, or school could be consumption that the rich can better afford. 

Most who study education are data-crunchers with little patience with such abstract theorizing.  But until recently they were troubled by the fact that data on nations across time seemed to show a negative relation between wealth and education, even after controlling for measures of physical capital!  For example, see this 2001 Pritchett paper.

In the last year, however, education data-crunchers have declared with satisfaction that the problem is now solved, since new data now shows a positive correlation across nations between education and wealth.  See this recent Science paper, which controls (only) for fixed effects in years and nations: