Computing Cost Floor Soon?

Anders Sandberg has posted a nice paper, Monte Carlo model of brain emulation development, wherein he develops a simple statistical model of when brain emulations [= “WBE”] would be feasible, if they will ever be feasible:

The cumulative probability gives 50% chance for WBE (if it ever arrives) before 2059, with
the 25% percentile in 2047 and the 75% percentile in 2074. WBE before 2030 looks very unlikely and only 10% likely before 2040.

My main complaint is that Sandberg assumes a functional form for the cost of computing vs. time that requires this cost to soon fall to an absolute floor, below which it will never fall, relative to the funding ever available for a brain emulation project. His resulting distribution has costs approaching this floor by about 2040:

SandbergTimingModel

As a result, Sandberg finds a big chance (how big he doesn’t say) that brain emulations will never be possible – for eons to follow it will always be cheaper to compute new mind states via floppy proteins in huge messy bio systems born in wombs, than to compute them via artificial devices made in factories.

That seems crazy implausible to me. I can see physical limits to physical parameters, and I can see the rate at which computing costs fall slowing down. But having the costs of artificial computing soon stop falling forever is much harder to see, especially with such costs remaining far higher than the costs of natural bio devices that seem pretty far from optimized. And having the amount of money available to fund a project never grow seems to say that economic growth will halt as well.

Even so, I applaud Sandberg for his efforts so far, and hope that his or others’ successor models will be more economically plausible. It is an important question, worthy of this and more attention.

GD Star Rating
loading...
Tagged as: ,

Academic Stats Prediction Markets

In a column, Andrew Gelman and Eric Loken note that academia has a problem:

Unfortunately, statistics—and the scientific process more generally—often seems to be used more as a way of laundering uncertainty, processing data until researchers and consumers of research can feel safe acting as if various scientific hypotheses are unquestionably true.

They consider prediction markets as a solution, but largely reject them for reasons both bad and not so bad. I’ll respond here to their article in unusual detail. First the bad:

Would prediction markets (or something like them) help? It’s hard to imagine them working out in practice. Indeed, the housing crisis was magnified by rampant speculation in derivatives that led to a multiplier effect.

Yes, speculative market estimates were mistaken there, as were most other sources, and mistaken estimates caused bad decisions. But speculative markets were the first credible source to correct the mistake, and no other stable source had consistently more accurate estimates. Why should the most accurate source should be blamed for mistakes made by all sources?

Allowing people to bet on the failure of other people’s experiments just invites corruption, and the last thing social psychologists want to worry about is a point-shaving scandal.

What about letting researchers who compete for grants, jobs, and publications write critical referee reports and publish criticism, doesn’t that invite corruption too? If you are going to forbid all conflicts of interest because they invite corruption, you won’t have much left you will allow. Surely you need to argue that bet incentives are more corrupting that other incentives. Continue reading "Academic Stats Prediction Markets" »

GD Star Rating
loading...
Tagged as: ,

Socializers Clump

Imagine that this weekend you and others will volunteer time to help tend the grounds at some large site – you’ll trim bushes, pull weeds, plant bulbs, etc. You might have two reasons for doing this. First, you might care about the cause of the site. The site might hold an orphanage, or a historical building. Second, you might want to socialize with others going to the same event, to reinforce old connections and to make new ones.

Imagine that instead of being assigned to work in particular areas, each person was free to choose where on the site to work. These different motives for being there are likely to reveal themselves in where people spend their time grounds-tending. The more that someone wants to socialize, the more they will work near where others are working, so that they can chat while they work, and while taking breaks from work. Socializing workers will tend to clump together.

On the other hand, the more someone cares about the cause itself, the more they will look for places that others have neglected, so that their efforts can create maximal value. These will tend to be places places away from where socially-motivated workers are clumped. Volunteers who want more to socialize will tend more to clump, while volunteers who want more to help will tend more to spread out.

This same pattern should also apply to conversation topics. If your main reason for talking is to socialize, you’ll want to talk about whatever everyone else is talking about. Like say the missing Malaysia Airlines plane. But if instead your purpose is to gain and spread useful insight, so that we can all understand more about things that matter, you’ll want to look for relatively neglected topics. You’ll seek topics that are important and yet little discussed, where more discussion seems likely to result in progress, and where you and your fellow discussants have a comparative advantage of expertise.

You can use this clue to help infer the conversation motives of the people you talk with, and of yourself. I expect you’ll find that almost everyone mainly cares more about talking to socialize, relative to gaining insight.

GD Star Rating
loading...
Tagged as: , ,

Hidden Asset Taxes Must Be Huge

Paul Krugman:

Piketty’s big idea is that we are in the early stages of returning to a society dominated by great dynastic fortunes, by inherited wealth. … Imagine a wealthy family that has managed, somehow or other, to guarantee that a large fraction of its income is used to accumulate more wealth. Can this family thereby acquire a dominant position in society?

The answer depends on the relationship between r, the rate of return on assets, and g, the overall rate of economic growth. If r is less than g, dynasties are doomed to erode: even if all income from a very large fortune is devoted to accumulation, the family’s wealth will grow more slowly than the economy, and it will slowly slide into obscurity. But if r is greater than g, dynastic wealth can indeed grow to gigantic size. …

Piketty tells us something remarkable: historically, r has almost always exceeded g – but there was an exceptional period in the 20th century, a period of rapid labor force growth and technological progress, when r was less than g. And he asserts that the kind of society we consider normal, in which high incomes reflect personal achievement rather than inherited wealth, is in fact an aberration driven by this exceptional period. … A couple of questions:

1. How much of the decline in r relative to g in the 20th century reflected fast growth, and how much reflected policies that either taxed or in effect confiscated inherited wealth? In other words, how much was destiny, how much wars and political upheaval? Piketty stresses both factors, but never gives us a relative quantitative assessment. (more from Piketty here, here)

This rate of return on assets r that Krugman and Piketty discuss is something like the ratio of rental to purchase price of land. I don’t have access to Piketty’s book, but I’ve been pondering this question for a few months, and I’ve concluded that the usual estimates of asset returns r must fail to include many taxes that in practice reduce the actual rate of return r that growing dynasties can achieve. And I think that once we include all hidden taxes, the actual rate of return r that dynasties could achieve in practice must have usually be no more than the economic growth rate g. Let me explain.

Some taxes are explicit, like property taxes. Other taxes are implicit in the property destruction and transfer that result from wars, political upheavals, and legal corruption, and in the costs of reasonable efforts to prevent such losses. Finally, there are implicit taxes resulting from local legal limits on who one may use to manage a dynastic fund. For example, if a dynasty must give its eldest living male wide discretion over spending and investment choices, and if such males often turn out to be spent-thrift fools, this will greatly limit this dynasty’s ability to grow over the long run. An ideal might be to delegate dynasty management to a reputed professional trust that is legally obligated to follow explicit instructions to grow the fund as fast as possible over the long run. But, as I’ve discussed before, most societies have put substantial legal obstacles before solutions like this.

I argue that the net effect of all these hidden taxes on dynastic funds must have been to usually reduce asset returns to below growth rates. My argument is simple: If asset returns had typically been above growth rates, then if any dynastic funds had chosen to grow at the maximum possible rate, then even if those funds had started small they would have come to dominate investments worldwide. And they would have done so on a timescale short compared to the time period over which historical records suggest that asset returns have exceeded growth rates. By competing with each other, such dominating dynastic funds would then have increased the supply of investment so much as to drive down asset returns to or below the sustainable level, which is the economic growth rate.

I conclude that consistently across space and time, the net effects of all forms of taxes on dynastic investment funds, including taxes implicit in limiting who one may trust not to pilfer those funds, has been to reduce real assets returns to below growth rates. Perhaps well below.

Of course, if the main hidden tax in history has been pilfering by dynasty managers, that can result in a world where such pilferers spend a large fraction of world income, without much social value to show for it. One might easily dislike such a scenario, and want to prevent it. But instead of adding more explicit taxes to prevent the growth of dynastic funds, it seems to me better to cut the pilfering tax. Because this should encourage much more investment overall, which seems a good thing. This includes investment in helping and protecting the future, including protection from disasters, including existential risks. Which also seem like good things.

GD Star Rating
loading...
Tagged as: , ,

Why Neglect Social Results?

For many decades I’ve heard people argue about the possibility of ems, i.e., brain emulations (also called “uploads”). Many like to talk about whether ems are possible, when they might happen, and if ems would be conscious, or whether they would “be me.”  People also love to read fiction set in worlds where there are ems. Almost twenty years ago I wrote a short article on the social implications of a world of emulations — what that would actually be like. But that didn’t kick-start much interest in the subject – most discussion is still on possibility, timing, consciousness, identity, and story settings.

Over the years I’ve also heard many people argue about the possibility that we live in a computer simulation. Twelve years ago I wrote a short article “How to live in a simulation,” on how you should live your life differently to take this possibility into account. That article also didn’t kick-start much interest in social implications. Today, most discussion of the simulation possibility continues to focus on using it as a setting for fiction, on the chances that it is true, on clues for inferring if it is true, and on what it implies for identity, consciousness, physics, etc. There remains almost no discussion of life strategies conditional on a simulation.

I just now noticed how similar are these situations, a similarity that cries out for explanation. I see three somewhat related candidate explanations:

  1. The sorts of people who most like these topics are techies, who mostly don’t believe that social and human sciences exist, and thus aren’t interested in hearing about  applications of such sciences.
  2. People are mainly interested in these sorts of topics as ways to stretch and stress-test their basic concepts. So only people with a library of grand social concepts are interested in using these topics to stretch and stress-test such concepts. There aren’t many such people.
  3. I personally did a poor job of introducing these topics. Had someone more prestigious or articulate done the job, there might well be much larger conversations now about these topics.

Whatever the explanation, this bodes poorly for interest in my more elaborated book-length discussion of the social implications of ems. However, I will soldier on nonetheless.

GD Star Rating
loading...
Tagged as: ,

Leadership Fantasies

Predictions about leadership in 2030:

The management consulting firm Hay Group worked with the German futurists at Z-Punkt to identify six mega trends such as globalization, technology convergence and the individualization of careers that will shape the kind of leaders companies will need in the future. I spoke with Georg Vielmetter, Hay Group’s regional director of leadership and talent, about the newly released study “Leadership 2030” that he co-authored. …

I think that positional power and hierarchical power will become smaller. Power will shift to stakeholders, reducing the authority of the people who are supposed to lead the organization. … The time of the alpha male — of the dominant, typically male leader who knows everything, who gives direction to everybody and sets the pace, whom everybody follows because this person is so smart and intelligent and clever — this time is over. We need a new kind of leader who focuses much more on relationships and understands that leadership is not about himself. …

Such a leader doesn’t doesn’t put himself at the very center. He knows he needs to listen to other people. He knows he needs to be intellectually curious and emotionally open. He knows that he needs empathy to do the job, not just in order to be a good person. … We will see a significant decline in physical loyalty between people and organizations. It will be very difficult for leaders to formally bind people to their organizations, so they should not try. This is a battle that leaders can only lose. … What is clear is that leaders in the future need to have a full understanding, and also an emotional understanding, of diversity. That’s for sure. (more)

I call bull. Here’s Jeffrey Pfeffer, in Power:

Most books by well-known executives and most lectures and courses about leadership should be stamped CAUTION: THIS MATERIAL CAN BE HAZARDOUS TO YOUR ORGANIZATIONAL SURVIVAL. That’s because leaders touting their own careers as models to be emulated frequently gloss over the power plays they actually used to get to the top. Meanwhile, the teaching on leadership is filled with prescriptions about following an inner compass, being truthful, letting inner feelings show, being modest and self-effacing, not behaving in a bullying or abusive way— in short, prescriptions about how people wish the world and the powerful behaved. There is no doubt that the world would be a much better, more humane place if people were always authentic, modest, truthful, and consistently concerned for the welfare of others instead of pursuing their own aims. But that world doesn’t exist.

More from Pfeffer last November:

Today’s work world is increasingly populated by millennials with values presumably different from more-senior employees—more egalitarian, less competitive, more meritocratic, less accepting of hierarchy, and more tolerant of all forms of diversity. And if that’s true, surely companies are changing, which means we need new theories about power and influence to reflect these new cultural realities. Strategically expressing anger, building a power base, or eliminating rivals are considered outmoded ways of getting ahead. Certainly, the reasoning goes, in a world where reputations get created and transmitted quickly and anonymously through ubiquitous social networks, people who resort to such bad behavior will suffer swift retribution.

The typical Silicon Valley recruitment pitch, or something to this effect, reinforces this view: “We’re not political here. We’re young, cool, socially networked, hip, high-technology people focused on building and selling great products. We’re family-friendly, have fewer management levels and less hierarchy, and make decisions collegially.”

Unfortunately there’s not much evidence of change but plenty of testimony to the contrary: the power struggles that beset the founding of Twitter (TWTR), the turnover among CEOs at Hewlett-Packard (HPQ), and the experiences of former Stanford MBA students working in the supposedly egalitarian world of high tech who have lost their jobs or been thrown out of companies they founded notwithstanding their intelligence and good job performance. Meanwhile, relationships with bosses still go a long way to predict people’s career success; organizational gossip lives on; and career derailment still awaits those who fail to master political dynamics. (more)

GD Star Rating
loading...
Tagged as: , ,

Prefer Contrarian Questions

Many people are attracted to authority. They are eager to defend what authorities say against heretics who say otherwise. This lets them signal a willingness to conform, and gain status by associating with higher status authorities against lower status heretics.

Other people are tempted to be contrarians. My blog readers tend more this way. Contrarians are eager to find authorities with which they disagree, and to associate with similar others. In this way contrarians can affirm standard forager anti-dominance norms, bond more strongly to a group, and hope for glory later if their contrarian positions becomes standard.

I haven’t posted much on disagreement here lately, but contrarians should be disturbed by the basic result that knowing disagreement is irrational. That is, it is less accurate to knowingly disagree with others unless one has good reasons to think you are more rational than they in the sense of listening more to the info implicit in their opinions.

Today I want to point out a way that contrarians can stay contrarians, taking an authority defying position they can share with like-minded folks and which might later lead to glory, while avoiding most of the accuracy-reducing costs of disagreement: be contrarian on questions, not answers.

Academia has well known biases regarding the topics it studies. Academia is often literature-driven, clumping around a few recently-published topics and neglecting many others. Academia also prefers topics where one can show off careful mastery of difficult and thus impressive methods, and so neglects topics worse suited for such displays.

Of course academia isn’t the only possible audience when selling ideas, but the other possible customers also have known topic biases. For example, popular writings are biased toward topics which are easy to explain to their audience, which flatter that audience, and which pander to their biases.

The existence of these known topic biases suggests how to be a more accurate contrarian: disagree with academia, the popular press, etc. on what questions are worth studying. While individuals may at times disagree with you on the importance of the topics you champion, after some discussion they will usually cave and accept your claim that academia, etc. has these topic biases, and that one should expect your topic to be neglected as a result.

Some academics will argue that only standard difficult academic methods are informative, and all other methods give only random noise. But the whole rest of the world functions pretty well drawing useful conclusions without meeting standard academic standards of method or care. So it must be possible to make progress on topics not best suited for showing off mastery of difficult academic methods.

So if your topic has some initial or surface plausibility as an important topic, and is also plausibly neglected by recent topic fashion and not well suited for showing off difficult academic methods, you have a pretty plausible contrarian case for the importance of your topic. That is, you are less likely to be wrong about this emphasis, even though it is a contrarian emphasis.

Now your being tempted to be contrarian on questions suggests that you are the sort of person who is also tempted to be contrarian on answers. Because of this, for maximum accuracy you should probably bend over backwards to not be contrarian on which answers you favor to your contrarian question. Focus your enjoyment of defying authorities on defying their neglect of your questions, but submit to them on how to answer those questions. Try as far as possible to use very standard assumptions and methods, and be reluctant to disagree with others on answers to your questions. Resist the temptation to too quickly dismiss others who disagree on answers because they have not studied your questions as thoroughly as you. Once you get some others to engage your question in some detail, take what they say very seriously, even if you have studied far more detail than they.

With this approach, the main contrarian answer that you must endorse is a claim about yourself: that you don’t care as much about the rewards that attract others to the usual topics. Most people work on standard topics because those usually give the most reliable paths to academic prestige, popular press popularity, etc. And honestly, most people who think they don’t care much about such things are just wrong. So you’ll need some pretty strong evidence in support of your claim that you actually differ enough in your preferences to act differently. But fortunately, your being deluded about this can’t much infect the accuracy of your conclusions about your contrarian topic. Even if you are mistaken on why you study it, your conclusions are nearly as likely to be right.

This is the approach I’ve tried to use in my recent work on the social implications of brain emulations. This is very contrarian as a topic, in the sense that almost no one else works on it, or seems inclined that way. But it has an initial plausibility as very important, at least if one accepts standard conclusions in some tech and futurist worlds. It is plausibly neglected as having negative associations and being less well suited for impressive methods. And I try to use pretty standard assumptions and methods to infer answers to my contrarian question. Of course none of that protects me from delusions on the rewards I expect to forgo by focusing on this topic.

Added 7Mar: People are already in the habit of pleasantly tolerating a wider range of opinion on which questions are important, both because differing values contribute, and because people tend to strongly overestimate the importance of the questions they work on personally.

GD Star Rating
loading...
Tagged as:

Moral Legacy Myths

Imagine that you decide that this week you’ll go to a different doctor from your usual one. Or that you’ll get a haircut from a different hairdresser. Ask yourself: by how much do you expect such actions to influence the distant future of all our descendants? Probably not much. As I argued recently, we should expect most random actions to have very little long term influence.

Now imagine that you visibly take a stand on a big moral question involving a recognizable large group. Like arguing against race-based slavery. Or defending the Muslim concept of marriage. Or refusing to eat animals. Imagine yourself taking a personal action to demonstrate your commitment to this moral stand. Now ask yourself: by how much do you expect these actions to influence distant descendants?

I’d guess that even if you think such moral actions will have only a small fractional influence on the future world, you expect them to have a much larger long term influence than doctor or haircut actions. Furthermore, I’d guess that you are much more willing to credit the big-group moral actions of folks centuries ago for influencing our world today, than you are willing to credit people who made different choices of doctors or hairdressers centuries ago.

But is this correct? When I put my social-science thinking cap on, I can’t find good reasons to expect big-group moral actions to have much stronger long term influence. For example, you might posit that moral opinions are more stable than other opinions and hence last longer. But more stable things should be harder to change by any one action, leaving the average influence about the same.

I can, however, think of a good reason to expect people to expect this difference: near-far (a.k.a construal level) theory. Acts based on basic principles seem more far than acts based on practical considerations. Acts identified with big groups seem more far than acts identified with small groups. And longer-term influence is also more strongly associated with a far view.

So I tentatively lean toward concluding that this expectation of long term influence from big-group moral actions is mostly wishful thinking. Today’s distribution of moral actions and the relations between large groups mostly result from a complex equilibrium of people today, where random disturbances away from that equilibrium are usually quickly washed away. Yes, sometimes they’ll be tipping points, but those should be rare, as usual, and each of us can only expect to have a small fraction influence on such things.

GD Star Rating
loading...
Tagged as: ,

Open Thread

This is our monthly place to discuss related topics that have not appeared in recent posts.

GD Star Rating
loading...
Tagged as:

Fail Faster

It looks bad for a manager to have one of his projects fail. So to “cover his ass”, such a manager often tries to prevent any records showing that people saw failure coming. After a failure, he wants to say “this was just random bad luck; no one could have foreseen seen it.” His bosses up the chain of command tend to allow this, because they also want to avoid being held responsible for failures during their watch. So they also prefer the random back luck story.

Unfortunately, this approach tends to prevent organizations from getting signals that would let them mitigate failures, such as by quitting projects earlier. For example, most startup firms don’t fail until they have spent nearly all of the cash they were given. It is rare for a startup to admit it isn’t going to work out, and give some cash back to investors. Similarly, government agencies created to achieve some purpose rarely recommend to legislatures that they be eliminated when their find that they aren’t achieving their intended purposes.

Of course bosses don’t want to be too obvious about silencing possible signals of failure. They find it hard to silence what have become standard signals, like cost accounting measures.

A great application of prediction markets is to give better and clearer warnings of upcoming failure, to enable better mitigation, such as quitting. Of course project bosses anticipate this, and oppose prediction markets on their projects, for exactly this reason. But we can still hope that prediction market warnings may someday become a standard signal, and thus hard to silence:

I hope prediction markets within firms may someday gain a status like cost accounting today. In a world were no one else did cost accounting, proposing that your firm do it would basically suggest that someone was stealing there. Which would look bad. But in a world where everyone else does cost accounting, suggesting that your firm not do it would suggest that you want to steal from it. Which also looks bad.

Similarly, in a world where few other firms use prediction markets, suggesting that your firm use them on your project suggests that your project has an unusual problem in getting people to tell the truth about it via the usual channels. Which looks bad. But in a world where most firms use prediction markets on most projects, suggesting that your project not use prediction markets would suggest you want to hide something. (more)

GD Star Rating
loading...
Tagged as: ,