Category Archives: Uncategorized

Science 2.0

By Robin Hanson · June 4, 2020 4:34 pm · Comments (0)

Skepticism … is generally a questioning attitude or doubt towards one or more items of putative knowledge or belief or dogma. It is often directed at domains, such as the supernatural, morality (moral skepticism), theism (skepticism about the existence of God), or knowledge (skepticism about the possibility of knowledge, or of certainty). (More)

Humans have long had many possible sources for our beliefs about the physical world. These include intuitive folk physics, sacred scriptures, inherited traditions, traveler stories, drug-induced experiences, gadget sales pitches, and expert beliefs within various professions. And for a very long time, we paid the most attention to the highest status sources, even if they were less reliable. This encouraged gullibility; we often believed pretty crazy stuff, endorsed by the high status.

One ancient high status group was astronomers, whose status was high because their topic was high – the sky above. It so happened that astronomers naturally focused on a small number of very standard parameters of wide interest: the sky positions of planets and comets (anything that moved relative to the stars). Astronomers often gained status by being better able to predict these positions, and for this purpose they found it useful to: (1) collect and share careful records on past positions, (2) master sufficient math to precisely describe past patterns, and (3) use those patterns to predict future parameter values.

For a long time astronomy seemed quite exceptional. Most other domains of interest seemed to have too much fuzziness, change, and variety to support a similar approach. What can you usefully measure while walking through a jungle? What useful general patterns can simple math describe there? But slowly and painfully, humans learned to identify a few relatively stable focal parameters of wide interest in other domains as well. First in physics: velocity, weight, density, temperature, pressure, toughness, heat of reaction, etc. Then in dozens of practical domains.

With such standard focal parameters in hand, domain experts also gained status by being able to predict future parameter values. As a result, they also learned that it helped to carefully collect shared systematic data, and to master sufficient math to capture their patterns.

And thus begat the scientific revolution, which helped begat the industrial revolution. A measurement revolution starting in astronomy, moving to physics, and then invading dozens of industrial domains. As domains acquired better stable focal parameters to observe, and better predictions, many such domains acquired industrial power. That is, those who had mastered such things could create devices and plans of greater social value. This raised the status of such domain experts, so that eventually this “scientific” process acquired high status: carefully collecting stable focal parameters, systematically collecting and sharing data on them, and making math models to describe their patterns. “Science” became high status.

One way to think about all this is in terms of the rise of skepticism. If you allow yourself to doubt if you can believe what your sources tell you about the physical world, your main doubt will be “who you I trust?” To overcome such doubt, you’ll want to focus on a small number of focal parameters, and for those seek shared data and explicit math models. That is, data where everyone can check how the data is collected, or collect it themselves, with redundant records to protect against tampering, and explicit shared math models describing their patterns. That is, you will turn to the methods to which those astronomers first turned.

Which is all to say that the skeptics turned out to be right. Not the extreme skeptics who doubted their own eyes, but the more moderate ones, who doubted holy scriptures and inherited traditions. Our distant ancestors were wrong to too eagerly trust their high status sources, and skeptics were right to focus on the few sources that they could most trust, when inclined toward great doubt. Slow methodical collection and study of the sort of data of which skeptics could most approve turned out to be a big key to enabling humanity’s current levels of wealth and power.

For a while now, I’ve been exploring the following thesis: this same sort of skepticism, if extended to our social relations, can similarly allow a great extension of the “scientific” and “industrial” revolutions, making our social systems far more effective and efficient. Today, we mainly use prestige markers to select and reward the many agents who serve us, instead of more directly paying for results or following track records. If asked, many say we do this because we can’t measure results well. But as with the first scientific revolution, with work we can find ways to coordinate to measure more stable focal parameters, sufficient to let us pay for results. Let me explain.

In civilization, we don’t do everything for ourselves. We instead rely on a great many expert agents to advise us and act for us. Plumbers, cooks, bankers, fund managers, manufacturers, politicians, contractors, reporters, teachers, researchers, police, politicians, regulators, priests, doctors, lawyers, therapists, and so on. They all claim to work on our behalf. But if you will allow yourself to doubt such claims, you will find plenty of room for skepticism. Instead of being as useful as they can, why don’t they just do what is easy, or what benefits them?

We don’t pay experts like doctors or lawyers directly for results in improving our cases, and we don’t even know their track records in previous cases. But aside from a few “bad apples”, we are told that we can trust them. They are loyal to us, coming from our nation, city, neighborhood, ethnicity, gender, or political faction. Or they follow proper procedures, required by authorities.

Or, most important, they are prestigious. They went to respected schools, are affiliated with respected institutions, and satisfied demanding licensing criteria. Gossip shows us that others choose and respect them. If they misbehave then we can sue them, or regulators may punish them. (Though such events are rare.) What more could we want?

But of course prestige doesn’t obviously induce a lawyer to win our case or promote justice, nor a doctor to make us well. Or a reporter to tell us the truth. Yes, it is logically possible that selecting them on prestige happens to also max gains for us. But we rarely hear any supporting argument for such common but remarkable claims; we are just supposed to accept them because, well, prestigious people say so.

Just as our distant ancestors were too gullible about their sources of knowledge on the physical world around them, we today are too gullible on how much we can trust the many experts on which we rely. Oh we are quite capable of skepticism about our rivals, such as rival governments and their laws and officials. Or rival professions and their experts. Or rival suppliers within our profession. But without such rivalry, we revert to gullibility, at least regarding “our” prestigious experts who follow proper procedures.

Yes, it will take work to develop better ways to measure results, and to collect track records. (And supporting math.) But progress here also requires removing many legal obstacles. For example, trial lawyers all win or lose in public proceedings, records of which are public. Yet it is very hard to actually collect such records into a shared database; many sit in filing cabinets in dusty county courthouse basements.

Contingency fees are a way to pay lawyers for results, but they are illegal in many places. Bounty hunters are paid for results in catching fugitives, but are illegal in many places. Bail bonds give results incentives to those who choose jail versus freedom, but they are being made illegal now. And so on. Similarly, medical records are more often stored electronically, but medical ethics rules make it very hard to aggregate them, and also to use creative ways to pay doctors based on results.

I’ve written many posts on how we could work to pay more for results, and choose more based on track records. And I plan to write more. But in this post I wanted to make the key point that what should drive us in this direction is skepticism about how well we can trust our usual experts, chosen mainly for their prestige (and loyalty and procedures) and using weak payment incentives. You might feel embarrassed by such skepticism, thinking it shows you to be low status and anti-social. After all, don’t all the friendly high status popular people trust their experts?

But the ancient skeptics were right about distrusting their sources on the physical world, and following their inclination helped to create science and industry, and our vast wealth today. Continuing to following skeptical intuitions, this time regarding our expert agents, may allow us to create and maintain far better systems of law, medicine, governance, and much more. Onward, to Science 2.0!

GD Star Rating
loading...
Tagged as: ,

Max Policing Is Disproportionate

By Robin Hanson · June 3, 2020 10:00 am · Comments (0)

NYT: Minneapolis Police Use Force Against Black People at 7 Times the Rate of Whites … most use of force happens in areas where more black people live. Although crime rates are higher in those areas, black people are also subject to police force more often than white people in some mostly white and wealthy neighborhoods, though the total number of episodes in those areas is small. (More)

Star Slate Codex: When restricted to index crimes, dozens of individual-level studies have shown that a simple direct influence of race on pretrial release, plea bargaining, conviction, sentence length, and the death penalty among adults is small to nonexistent once legally relevant variables (e.g. prior record) are controlled.

in 62% of studies, police are not searching blacks disproportionately to the amount of crimes committed or presumed “indicators of suspiciousness”. In 38% of studies, they are. … there are two possible hypotheses here: either police are biased, or black people actually commit these crimes at higher rates than other groups. The second hypothesis has been strongly supported by crime victimization surveys,

Blacks appear to be arrested for drug use at a rate four times that of whites. Adjusting for known confounds reduces their rate to twice that of whites. However, other theorized confounders could mean that the real relative risk is anywhere between two and parity. … the people shot by police are less black than the people shooting police or the violent shooters police are presumably worried about. … 24% of blacks charged with drug dealing are acquitted, compared to only 14% of whites. … There seems to be a strong racial bias in capital punishment and a moderate racial bias in sentence length and decision to jail. (More)

The race and policing literature usually asks if the relative rates of stopping, arrest, conviction, sentences, etc. are disproportionate to race, after controlling for relevant factors like crime rates. I know this not just from the above Star Slate Codex review, but also because I’ve supervised many related research papers while teach grad law & economics.

In that context, the NYT article is crazy biased, as it just reports raw race correlations without controlling for anything. Surely NYT crime reporters know better; they should have at least noted relative black/white crime rates in Minneapolis.

The point of my post here, however, is just to point out that we shouldn’t expect proportionate stops, arrests, etc. in the context of policing designed to max crime reduction.

Imagine you are a police charged with cutting crime in your area. Or more realistically, as you don’t control courts or prosecution, you seek to max the number of crimes you charge that are convicted, weighted by their severity (perhaps proxied by sentence severity). Given this goal, you survey a world of possible and reported crimes, and for each one you estimate the chances that a crime happened there, and if so your chance of finding enough evidence to achieve a conviction. In such a situation, your efforts are often disproportionate.

For example, consider two possible murders, one with 10 and the other with 100 plausible suspects. If you put more effort into investigating that first murder, then each suspect there will get a disproportionate fraction of your attention, relative to their chance of guilt.

As another example, imagine a stream of cars passes you, and you pick which of them to stop. You put each car into a class with which you associate a prior rate of success in finding evidence of crime, weighted by crime importance. As you have limited time, you will limit yourself to stopping the car classes with the highest prior rate of success. This can result in a disproportionate relation between the rate of crime and stopping.

Policing that maxes crime detection and prosecution will not in general effect all people, groups, and situations in linear proportion to their rate or chances of committing crimes, perhaps weighted by severity. It will instead focus on the people, groups, and situations with the very highest rates of crime, and lowest costs of finding and collecting sufficient evidence of those crimes. And often that will produce rates of arrest, etc. that are disproportionate to crime rates.

We do know of one situation where maximizing police effort is proportionate to crime rates: when the chance to find sufficient evidence of crime uncovered goes as the logarithm of police effort re each possible crime. This is when every time you double your effort into a possible crime, say putting in two hours instead of one, you increase your chance of finding sufficient evidence on that crime by the same small amount (say 0.1%).

But while there may be situations where this applies, logarithmic dependence hardly seems a general feature of policing. For example, we we know that it fails at the high end, as the chance of conviction can’t exceed 100%. And it fails at the low end when there are fixed costs to start any investigation whatsoever.

We thus expect disproportionate crime-fighting efforts aimed at groups with higher crime rates, and especially at groups with the highest crime rates for identifiable subgroups. The variance, not just the average, matters a lot; the high tails of these distributions should dominate.

People who are poor, unskilled, and impulsive, and who mix more closely with other such folks, are likely to put in less effort, end less expertise, into hiding their crimes. As a result, police effort is likely to detect such crimes more easily. So police who are tasked with finding crimes are likely to focus their efforts more on these people. If you think that is a problem, an obvious solution would be to prioritize policing of the rich, skilled, and self-controlled. If police seek to max fines or sentence severity, why then increase the fines or sentences of such convictions.

But before you do that, ask yourself honestly: are the crimes of the rich actually more socially harmful? It makes more sense to me to just make punishments proportional to our best estimates of their social harm, regardless of who does the crime. And it makes sense to have punishments compensate for lower probabilities of conviction, to maintain the same expected punishment for someone who commits a crime.

But otherwise, if the poor are actually causing more harm, or making it easier to catch their crimes, well it makes sense to me to pursue those crimes more. Even disproportionately.

GD Star Rating
loading...
Tagged as: ,

The High Cost of Masks

By Robin Hanson · May 29, 2020 1:05 pm · Comments (0)

In my debate last night with Tomas Pueyo, after I emphasized the high costs of lockdowns he said lockdowns are over. From now on we’ll let people go out, as long as they (M) wear masks and keep 6 feet apart when not at home. Which let me to wonder: just how much cheaper is M than lockdowns?

So I tried some Twitter polls. I asked:

  1. [1.14%] If you by law had to M unless you paid, what % of income would you pay?
  2. [4.6%] If you didn’t have to, but were offered % of income to M, what would it take?
  3. [19.6%] Like (2), except it is 2019 and there is no pandemic.
  4. [8.0%] Like (3), except half of the folks in your area are already M-ing
  5. [5.2%] Like (4), except only masks, no 6ft distancing.
  6. [15.8%] Like (4), except both at home and away.

In brackets before each question is the median answer (using lognormal fits to % responses re theses 4 options: <3%, 3-6%, 6-12%, >12%). Note how widely these estimates vary!

The following factors plausibly influence these responses: (A) personal pain and trouble of wearing masks and keeping apart, (B) endowment effect of preferring to stick with what current law seems to endorse, (C) not wanting to look or act too differently from others nearby, (D) wanting to be and seem pro-social and helpful in slowing the pandemic, and (E) wanting to support your side of current culture wars.

The big variation in median answers suggests that non-A effects are big! And the big variation within each poll also suggests that these costs vary greatly across individuals. We might gain lots from policies that let some pay to avoid M-like policies.

Question (4) seems to me to offer the best estimate of the real social cost of doing M. It has the best chance of avoiding effects C & D, and I don’t see a way to avoid E. Regarding B, I think we do want the endowment effect to go this direction, because in fact M was not the legal default a year ago. Yes we enjoy helping our community in a crisis, but we wouldn’t endorse creating crises just to enjoy such effects. So we shouldn’t include then when calculating how much to avoid or reduce crises.

Now this 8.0% median (27% mean) of income cost of masks and public distancing doesn’t include all costs; businesses and other places must also pay to accommodate your distancing. But it is probably a big fraction of costs. And it is quite a bit lower than the 32% of GDP estimate for recent strong lockdowns.

However, my estimate of the total cost of having 50% of the population infected at 0.5% IFR was 3 weeks of income, or 6% of one year’s income. So if we wear masks for 9 months, that single cost equals the entire cost of failing to contain. So as with lockdowns, we should be wary of spending more to prevent infections than the infections would cause. Yes, we should be willing to overspend for very effective preventions, with high elasticity. But masks aren’t plausibly such a very effective method.

Added 30May: I added new polls, included on the above list as #5,6. Seems more than half the cost is from the masks alone. It also seems that masks at home would give similar benefits, and the above suggests that it costs about the same too. So why isn’t there more of push to require that?

GD Star Rating
loading...
Tagged as: ,

Time to Mitigate, Not Contain

By Robin Hanson · May 28, 2020 5:30 pm · Comments (1)

In a few hours, I debate “Covid-19: Contain or Mitigate?” with Tomas “Hammer & Dance” Pueyo, author of many popular pandemic posts (e.g. 1 2 3 4 5 6 7). Let me try to summarize my position.

We have long needed a Plan B for the scenario where a big fraction of everyone gets exposed to Covid19, and for this plan I’ve explored variolation and other forms of deliberate exposure. To be ready, variolation just needs a small (~100) short (1-2mo.) trial to verify the quite likely (>75%) case that it works (cuts harm by 3-30x), but alas, while funding and volunteers can be found, med ethics panels have consistently disapproved. (Months later, they haven’t even allowed the widely praised human challenge trials for vaccines, some of which would include variolation trials.)

One big reason we aren’t preparing enough for Plan B is that many of us are mentally stuck in a Plan A “monkey trap.” Like a monkey who gets caught because it won’t let go of a tasty nut held in its fist within a gourd, we are so focused on containing this pandemic that we won’t explore options premised on failing to contain.

Containment seeks to keep infections permanently low, via keeping most from being exposed, for at least several years until a strong lasting vaccine is widely applied. Mitigation, in contrast, accepts that most will be exposed, and seeks only to limit the rate or exposure to keep medical systems from being overwhelmed, and to maintain enough critical workers in key roles.

Succeeding at containment is of course a much bigger win, which is why containment is the usual focus early in a pandemic. Catch it fast enough, and hit it hard enough with testing, tracing, and isolation, and the world is saved. But eventually, if you fail at that Plan A, and it grows big enough across across a wide enough area, you may need to admit failure and switch to a Plan B.

And, alas, that’s where we seem to be now with Covid-19. Over the last 7 weeks since the cases peak the official worldwide case count has been rising slowly, while official deaths are down ~25%. In the US, deaths are down ~1/2 since he peak 6 weeks ago. You might think, “yay, declines, we are winning!” But no, these declines are just too slow, as well as too uncertain.

Most of the US decline has been in New York, which has just now reached bottom, with no more room to decline. And even if US could maintain the rate of declining by 1/3 every 6 weeks, and repeat that 5 times over 30 weeks (= 7 months), not at all a sure thing, that would only bring daily US cases from 31K at the peak down to 4.2K. Which isn’t clearly low enough for test and trace to keep it under control without lockdown. And, more important, we just can’t afford to lockdown for that long.

You see, lockdown is very expensive. On average, around the world, lockdowns seems to cost about one third of local average income. Yet I estimate the cost of failing to contain, and letting half the population get infected with Covid-19 (at IFR ~0.5%), to be about 3 weeks of income. (Each death loses ~8 QALY.) So 9 weeks of strong lockdown produces about the same total harm as failing to contain! And where I live, we have almost had 10 weeks of lockdown.

If without lockdown the death rate would double due to an overloaded medical system, then paying for less than 9 weeks of added lockdown to prevent that is a good deal. But at that point, paying more than an additional 9 weeks of strong lockdown to prevent all deaths would not be a good deal. So our willingness to pay for lockdowns to cut deaths should really be quite limited. Sure, if we were at a tipping point where spending just a bit more would make all the difference between success and failure, then sure we should spend that bit more. But that’s just not plausibly where we are now.

Yes, sometimes we pay more to prevent harm than we suffer on average from such harms. For example, we pay more for door locks, and security guards, than we lose on average from theft. But those are very effective ways to prevent harm; paying 10% more there cuts harms by much more than 10%. Yet according to my Twitters polls, most see 10% more spent on lockdown as producing much less than 10% fewer deaths. If so, we should spend much less on lockdowns than we suffer from pandemic deaths.

Now if Pueyo is reading this, I suspect he’s screaming “But we’ve been doing it all wrong! Other possible policies exist that are far more effective, and if we use them containment becomes cost-effective. See Taiwan or South Korea.” And yes, other places have achieved better outcomes via better policies. We might not be able to do as well as them now that we’ve lost so much time, but we might well do much better than currently. Pueyo has sketched out plans, and they even seem to be good sketches.

So if we suddenly made Tomas Pueyo into a policy czar tomorrow, with an unlimited budget and able to run roughshod across most related laws or policies, we’d probably get much better Covid-19 outcomes, perhaps even cost-effective containment. But once such a precedent was set, I’d fear for the effectiveness of future czars. Ambitious politicians and rent-seekers would seek to manufacture crises and pseudo-Pueyo-czar candidates, all to get access to those unlimited budgets and legal powers.

Which is a big part of why we have the political systems we do. All around the world, we have created public health agencies tasked with pandemic policy, and political systems that oversee them. These agencies are staffed with experts trained in various schools of thought, who consult with academic experts of varying degrees of prestige. And all are constrained by local legal precedent, and by public perceptions, distrust, and axes of political conflict. These are the people and systems that have produced the varying policies we have now, all around the world.

Yes, those of us in places which have seen worse outcomes should ask ourselves how strong a critique that fact offers of our existing institutions and political cultures, and what we might do to reform them. But there is no easy and fast answer there; good reforms will have to be carefully considered, tested, and debated. We can hope to eventually improve via reforms, but, and this is the key point, we have no good reason to expect much better pandemic policy in the near future than we have seen in the near past. Even when policy makers have access to well-considered policy analyses by folks like Pueyo.

Now it might be possible to get faster political action if Pueyo and many other elites would coordinate and publicly back one specific standard plan, say the “John Hopkins Plan”, that specifies many details on how to do testing, tracing, isolation, etc. Especially if this plan pretty directly copied a particular successful policy package from, say, Taiwan. If enough people yelled in unison “We must do this or millions will die!”, why then politicians might well cave and make it happen.

But that is just not what is happening. Instead, we have dozens of essays and white papers pushing for dozens of related but different proposals. So there’s no clear political threat for politicians to fear defying. Whatever they do, come re-election time politicians can point to some who pushed for some of what they did. So all these divergent essays have mainly limited short term political effects, though they may do much more to raise the status of their authors.

So if political complexity argues against containment now in many places, why doesn’t that same argument apply equally well to mitigation? After all, mitigation can also be done well or badly, and it must be overseen by the same agencies and politicians that would oversee containment. As there is no escaping the fact that many detailed policy choices must be made, why not push for the best detailed packages of choices that we know?

Imagine that you were driving from A to B, and your first instinct was to take a simple route via two interstate freeways, both in big valleys. Your friend instead suggests that you take a backroad mountain short cut, using eight different road segments, many of them with only one lane, and some very wiggly. (Assume no phone or GPS.) That plan might look like it would take less total time, but you should worry about your competence to follow it. If you are very tired, bad at following road directions, or bad at sticking to wiggly roads, you might prefer to instead take the interstates. Especially if it was your tired 16 year old teen son who will do the driving.

Like the wiggly backroad short cut, containment is a more fragile plan, more sensitive to details; it has to be done more exactly right to work. To contain well, we need the right combination of the right rules about who can work and shop, and with what masks, gloves, and distance, tests going to the right people run by the right testing orgs, the right tracing done the right way by the right orgs with the right supporting apps, and the right rules requiring who gets isolated where upon what indications of possible infection. All run by the right sort of people using the right sort of local orgs and legal authority. And coordinated to right degree with neighboring jurisdictions, to avoid “peeing section of the pool” problems.

Yes, we might relax lockdown badly, but we are relaxing toward a known standard policy: no lockdown. So there are fewer ways to go wrong there. In contrast, there are just more ways to go wrong in trying to lockdown even more strictly. And that’s why it can make sense for the public to say to the government, “you guys haven’t been doing so well at containment, so let’s quit that and relax lockdown faster, shooting only for mitigation.” Yes, that might go badly, but it can’t go quite as badly as the worse possible scenario, where we trash the economy with long painful lockdowns, and yet still fail and most everyone gets exposed.

And that’s my argument for mitigation, relative to containment.

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

Pandemic Futarchy Design

By Robin Hanson · May 24, 2020 10:35 am · Comments (0)

Researchers … use … feeds from a global network of students, staff and alumni to construct a “stringency index” that boils down to a single number how strictly governments in 160 countries are locking down their economies and societies to contain the spread of the virus. … plans to include state-by-state measures of stringency in the U.S. … latest version … draws on 17 indicators to determine the stringency of the government response. (More)

Not that I hear anyone eagerly clamoring to try, but let me sketch out how one would use decision markets to set pandemic policy. Just to plant another flag on how widely they could be usefully applied, if only enough folks cared about effective policy.

As you may recall, a decision market is a speculative (i.e., betting) market on a key outcome of a decision, conditional on which discrete decision is made. To apply these to the current pandemic, we need to pick

  1. key ex-post-measurable outcome(s) of interest,
  2. likely-enough key decisions which could substantially influence those outcomes,
  3. participants capable of learning a bit about how decisions related to outcomes,
  4. sponsors who care enough about informing these decisions, and
  5. legal jurisdictions that may allow such markets.

Regarding participants, sponsors, and permission, it makes sense to be opportunistic. Seek any sponsors interested in relevant questions, any participants you can get to trade on them, and any jurisdiction that let you want to do. Alas I have no sponsor leads.

For key decisions, we could consider using bills before legislatures, administrative rulings, or election results. But there are a great many of these, we don’t get much warnings about many, and most have little overall impact. So I’d prefer to aggregate decisions, and summarize policy via three key choice metrics per region:
Lockdown Strictness. As described in the quote above, some have created metrics on lockdown strictness across jurisdictions. Such metrics could be supplemented by cell-phone based data on trips outside the home.
Testing Volume. The number of tests per unit time, perhaps separated into the main test types, and perhaps also into accuracy classes.
Tracing Volume. The number of full-time equivalent tracers working to trace who infected whom. Perhaps supplemented by the % of local folks use apps that report their travels to tracing authorities.

Yes, worse pandemic outcomes will likely cause more lockdown, tests, and tracing. But one could look at outcomes that happen after decisions. Such as how average future outcomes depend on the decisions made this month or quarter.

For key outcomes, the obvious options are deaths and economic growth.

For deaths, we can avoid testing problems by looking at total deaths, or equivalently “excess” deaths relative to prior years. It helps to note the ages of deaths, which can be combined with local mortality tables to estimate life-years lost. Even better, if possible, note the co-morbidities of those who died, to better estimate life-years lost. And even more better, have estimates of the relative quality of those life-years.

For economic growth, just take standard measures of regional income or GDP, and integrate them many years into the future, using an appropriate discount factor. Assuming that the temporary disruption from a pandemic is over within say 10 years, one could end the bets after say ten years, projecting the last few years of regional income out into the indefinite future.

As usual, there will be a tradeoff here re how far to go in accounting for these many complexities. I’d be happy to just see measures of life years lost related to lockdown strictness, perhaps broken into three discrete categories of strictness. But I’d of course be even happier to include economic growth as an outcome, and tests and tracing as decisions. Either aggregate all outcomes into one overall measure (using values of life years), or have different markets estimate different outcomes. For decisions, either separate markets for each type of decision. Or, ideally, combinatorial markets looking at all possible combinations of outcomes, decisions, and regions.

GD Star Rating
loading...
Tagged as: ,

Five Ways to Rate

By Robin Hanson · May 19, 2020 9:05 pm · Comments (2)

When we have people and orgs do things for us, we need ways to rate them. So we can pick who to have do what, and how much to have them do. And how we evaluate suppliers matters a lot, as they put great effort into looking good according to the metrics we use.

Results – When we buy a pound of roast chicken, or have someone to mow our lawn, we can pretty directly pay for the things we want. The more aspects of what we want that we can articulate and verifiably measure, the more of them we can specify in a contract. When action is risky, not always reliably producing desired results, paying for results means those who do things face payment risk, which they don’t like. And competitors might find a way to produce better results and displace them, a risk they also don’t like.

Record – If we get similar things over and over in a relatively stable context, and also stick with one provider for a while, then we can see a track record of how well they do for us. So we can pay them more of a steady fee not as closely tied to what we get, and drop them when their record seems unsatisfactory. We might switch between providers to sample quality, or hear gossip from associates about their experiences. Groups like Consumer Reports can collect stats on overall customer results. Suppliers face lower payment risks here, though still substantial competition risks.

Prestige – When data on customer results isn’t available, we may rely on a general opinion based on many weak clues about the quality of relevant people and orgs. For people, such clues include wealth, attractiveness, intelligence, social savvy, well-connectedness, etc. For orgs, there is also name-recognition, sponsorships, prestigious projects, and many other elements. Early education and training varies in prestige and adds to individual prestige. When individuals affiliate with orgs, the prestige of each adds to the prestige of the other. I count network effects under prestige; you use the system everyone else respects, As prestige is usually pretty stable over time, suppliers chosen by prestige tend to have a secure position.

Loyalty – While we less often admit it, we often choose suppliers to show our loyalty to “our sides”. Those we choose make sure to signal which sides they are on, and we help to ensure that our associates see those signals, so they can credit us for loyalty. It matters less to us that these signals actually correlate with the things we claim that our side seeks to achieve, as long as they are widely seen as clearly marking folks as on our side relative to other sides. The more stable are sides and signals, the more security a supplier can gain by clearly picking a side.

Procedure – Often specialists create official procedures re how to do something, and rules saying what not to do along the way. Then suppliers can brag to customers that they follow good procedures, they may be required by regulators to do so, and may be punished by courts as negligent if they deviate. Civil servants, for example, are typically paid and promoted based on following official procedures, and on internal politics, not on rates or results. Divisions within private orgs also try to become silos evaluated via rules and procedures, not results. Suppliers tend to like being evaluated by stable rules and procedures that can be achieved with limited effort, as this ensures high job/supplier stability.

While all these methods have their place, the first one, results, seems the most solid and hardest to corrupt from the customers’ point of view. Yes there are obstacles to applying it widely, but such problems are often exaggerated to excuse the other methods. Suppliers would generally rather be evaluated via prestige, loyalty, and procedure, as once these are established they can usually look forward to long stable lucrative relationships, unthreatened by upstart competitors.

I want to find ways to tell people that we could pay for results far more than we now do. Yes, suppliers would resist such a switch, but we customers would get more of what we wanted.

GD Star Rating
loading...
Tagged as:

Farrell On Who Pushes To Open

By Robin Hanson · May 18, 2020 7:04 pm · Comments (0)

Recently I wrote:

While the public will uniformly push for more opening, elites and experts push in a dozen different directions. If elites would all back the same story and solution, as they did before, they would probably get it. … So elites and experts don’t speak with a unified voice, while the public does. And that’s why the public will win. While the public tends to defer to elites and experts, and even now still defers a lot, this deference is gradually weakening. We are starting to open, and will continue to open, as long as opening is the main well-supported alternative to the closed status quo. … public pushes will will tend to be correlated in a particular direction, in contrast with the elite pushes which are much more diverse.

Also, my poll on “If in your region, pandemic continues to grow, or decline very slowly, for N more months, you will support a weaker lockdown,” got a median answer N = 1.2.

Henry Farrell responded:

People … are in favor of stay at home orders, … It could be that people are lying …but it would be unprecedented for so many of them to be lying. … So why then, may the equilibrium break down? It’s clearly not because of express demand from the public. Nor because of cheating … The plausible answer is that private power asymmetries are playing a crucial role in undermining the equilibrium. Some people – employees with poor bargaining power and no savings – may find themselves effectively coerced into a return to work as normal. … this is certainly a situation where the state and private actors are looking to get into cahoots, but not in the ways that public choice economists have devoted significant analytic energy to. And if you want an illustration of Marxist arguments about the “structural power of capital” to threaten politicians … Where public choice people seem to perceive a “public” that collectively wants to return to work, I see something different – a set of asymmetric power relations that public choice scholars are systematically blind to …

So here’s my bet. If the public choice analysis is right, and this is about some kind of broad and diffuse “public” pushing back against impossible regulations, then we will see a return to the economy sooner rather than later. But we can reasonably presume that this return will be roughly symmetric. … In contrast, if I’m right, we will see a very different return to “normality.” The return to the economy will be sharply asymmetric. Those who are on the wrong end of private power relations – whether they are undocumented immigrants, or just the working poor – will return early and en masse. Those who have the choice and the bargaining power will tend instead to pick safety. …[This] highlights the frequently brutal power relations that public choice scholars shove under the carpet when they talk about the “public” wanting an end to lockdown and a return to past economic relations.

So, Farrell says that it might look like people want to work, but this is only because “power asymmetries” let bad firms force them. If those firms would just be good enough to keep paying full wages, why then workers would happily stay home forever. Thus proving its not ordinary people who push to open, it is really the evil capitalists pushing. As also proven by those polls. Economic theory (what “public choice” scholars use to study politics) just can’t explain why poor people with less savings might be more eager to work; for that you need “private power relations” theory.

Wow. As you can see, I am not making this up.

Look, even if we kept printing plenty of money to hand out to workers to all stay home, eventually there wouldn’t be anything to buy, because no one was making and distributing products and services. And then no one would be happy to stay home. The fact that someone needs to work in order to make our world function isn’t some conspiracy foisted on us by the asymmetrically powerful, it is just a basic fact about our world that would also apply in Marxist heaven, or any other social system.

There are many channels by which such fundamental pressures will be communicated to the political equilibrium, but they must all lead to the same place. We are paying a real price for not working, we will only continue if we see sufficient value in it, and even then for only a limited time. Every non-crazy theory of politics must predict this. And economists, including those who study politics, do in fact predict that the poor will push first and harder, as they run out of resources first. Maybe you blame the very existence of the poor on evil capitalists, but this outcome happens regardless of why there are poor, it only requires that they exist.

Who pushes the least? Rich elites who can work from home selling abstract arguments that blame all problems, even pandemics, on (other) evil rich.

Note that we also have many non-work social needs that push us to open. And employers are similarly willing to stay home if we keep sending them checks.

Added: Farrell quickly responded:

GD Star Rating
loading...
Tagged as: ,

Lazar’s Dreamland

By Robin Hanson · May 16, 2020 1:50 pm · Comments (0)

I’ve always enjoyed science fiction, in part because such big things tend to be at stake there. But over the decades as I’ve learned more about the world, the less sense most of it makes. And I enjoy it less. Authors work hard to have their stories make sufficient sense to their median reader or reviewer, but not much beyond that.

Biographies are more realistic. They may not be exactly accurate, but they try harder to seem so. Not as many big things happen there, though I recently enjoyed Chernow’s Titan, a biography of John D. Rockefeller, wherein events are plenty big. I found I liked and admired him; he deserves a better reputation than he has.

Recently, I discovered that stories of probable hoaxes can offer a great compromise, as they try both to have big things happen, and also to seem realistic even to knowledgeable but skeptical investigators. In that spirit, I very much enjoyed physicist Bob Lazar’s Dreamland, the story of his working briefly for the US government in 1989 near Area 51 on alien UFO tech, and then publicizing that fact.

I was born seven months after Lazar, and like him studied physics, worked at secretive west coast US government labs, hung out with relatively colorful characters, and was prone to take more chances than the people around me. I was at NASA ’89-93 and Lockheed ’84-89, were I once had a top secret clearance. Lazar is a type of person I knew, describes a world I knew well, and does so believably.

Someone somewhere complained that Lazar isn’t very deep, which is true, but also realistic. Lazar is a much more hands-on intuitive guy, while I’m more of a theorist. He put a jet engine on his bike and car, and he throws around physics theory concepts in ways that I find sloppy. But that seems realistic for a person like him, and it makes sense that someone might think it would make sense to hire a person of his style to do the task he claims to have been assigned. His sort of person might even be tempted to embellish a few not-central-to-story details when telling his story.

I’ve also watched his documentary and Rogan interview, where Lazar comes across as more trustworthy than the people around him. So I’m inclined to believe him – except for that one fact: his key claims sound batshit crazy. Sorry, this isn’t the sort of thing I can believe on the testimony of one person, no matter how credible.

Reading Lazar’s Dreamland makes me a bit more eager to see a good overall stat analysis of a large dataset of UFO reports, where ideally his case is one datapoint. And more eager to read other probable-hoax biographies; what else ya got?

GD Star Rating
loading...
Tagged as: ,

Modeling the ‘Unknown’ Label

By Robin Hanson · May 12, 2020 10:45 am · Comments (1)

Recently I’ve browsed some specific UFO encounter reports, and I must admit they can feel quite compelling. But then I remember the huge selection effect. We all go about our lives looking at things, and only rarely do any of us officially report anything as so strange that authorities should know about it. And then when experts do look into such reports, they usually assign them to one of a few mundane explanation categories, such as “Venus” or “helicopter.”  For only a small fraction do they label it “unidentified”. And from thousands of these cases, the strangest and most compelling few become the most widely reported. So of course the UFO reports I see are compelling!

However, that doesn’t mean that such reports aren’t telling us about new kinds of things. After all, noticing weird deviations from existing categories is how we always learn about new kinds of things. So we should study this data carefully to see if random variation around our existing categories seems sufficient to explain it, or if we need to expand our list of categories and the theories on which they are based. Alas, while enormous time and effort has been spent collecting all these reports, it seems that far less effort has been spent to formally analyze them. So that’s what I propose.

Specifically, I suggest that we more formally model the choice to label something “unknown”. That is, model all this data as a finite mixture of classes, and then explicitly model the process by which items are assigned to a known class, versus labeled as “unknown.” Let me explain.

Imagine that we had a data set of images of characters from the alphabet, A to Z, and perhaps a few more weird characters like წ. Nice clean images. Then we add a lot of noise and mess them up in many ways and to varying degrees. Then we show people these images and ask them to label them as characters A to Z, or as “unknown”. I can see three main processes that would lead people to choose this “unknown” label for a case:

  1. Image is just weird, sitting very far from prototype of any character A to Z.
  2. Image sits midway between prototypes of two particular characters in A to Z.
  3. Image closely matches prototype of one of the weird added characters, not in A to Z

If we use a stat analysis that formally models this process, we might be able to take enough of this labeling data and then figure out whether in fact weird characters have been added to the data set of images, and to roughly describe their features.

You’d want to test this method, and see how well it could pick out weird characters and their features. But once it work at least minimally for character images, or some other simple problem, we could then try to do the same for UFO reports. That is, we could model the “unidentified” cases in that data as a combination of weird cases, midway cases, and cases that cluster around new prototypes, which we could then roughly describe. We could then compare the rough descriptions of these new classes to popular but radical UFO explanations, such as aliens or secret military projects.

More formally, assume we have a space of class models, parameterized by A, models that predict the likelihood P(X|A) that a data case X would arise from that class. Then given a set of classes C, each with parameters Ac and a class weight wc, we could for any case X produce a vector of likelihoods pc = wc*P(X|Ac), one for each class c in C. A person might tend more to assign the known label L when the value of pL was high, relative to the other pc. And if a subset U of classes C were unknown, people might tend more to assign assign the label “unknown” when either:

  1. even the highest pc was relatively low,
  2. the top two pc had nearly equal values, or
  3. the highest pc belonged to an unknown class, with c in U.

Using this model of how the label “unknown” is chosen, then given a data set of labeled cases X, including the unknown label, we could find the best parameters wc and Ac (and any in the labeling process) to fit this dataset. When fitting such a model to data, one could try adding new unknown classes, not included in the initial set of labels L. And in this way find out if this data supports the idea of new unknown classes U, and with what parameters.

For UFO reports, the first question is whether the first two processes for producing “unknown” labels seems sufficient to explain the data, or if we need to add a process associated with new classes. And if we need new classes, I’d be interested to see if there is a class fitting the “military prototype” theory, where events happened more near to military bases, more at days and times when those folks tend to work, with more intelligent response, more noise and less making nearby equipment malfunction, and impressive but not crazy extreme speeds and accelerations that increase over time with research abilities. And I’d be especially interested to see if there is a class fitting the “alien” theory, with more crazy extreme speeds and accelerations, enormous sizes, nearby malfunctions, total silence, apparent remarkable knowledge, etc.

Added 9a: Of course the quality of such a stat analysis will depend greatly on the quality of the representations of data X. Poor low-level representations of characters, or of UFO reports, aren’t likely to reveal much interesting or deep. So it is worth trying hard to process UFO reports to create good high level representations of their features.

Added 28May: If there is a variable of importance or visibility of an event, one might also want to model censoring of unimportant hard-to-see events. Perhaps also include censoring near events that authorities want to keep hidden.

GD Star Rating
loading...
Tagged as: ,

Non-UFO Local Alien Clues

By Robin Hanson · May 9, 2020 1:30 pm · Comments (0)

[US] Department of Defense formally released three Navy videos that contain ‘unidentified aerial phenomena.’ … When the videos were published in 2017 and 2018 by The New York Times …, they gave new hope to those looking for signs of extraterrestrial life. … ‘it’s time that we make progress to understand the extraordinary technology being observed during these events.’ (More)

Still, when you run all the arguments through your mind, is it not possible to come away with an estimate of at least a one-in-a-thousand chance that alien visitations are a real thing? Even such a small chance would be worthy of more discussion. (More)

Alexander Wendt, a professor of international relations at Ohio State University. Wendt is a giant in his field of IR theory, but in the past 15 years or so, he’s become an amateur ufologist. … ‘It’s possible they’ve been here all along. … They could just be intergalactic tourists. Maybe they’re looking for certain minerals. It could just be scientific curiosity. It could be that they’re extracting our DNA. I mean, who knows? I have no idea. All I know is that if they are here, they seem to be peaceful.’ (More)

In the above, two social scientists, economist Tyler Cowen and political scientist Alexander Wendt, say to take UFOs-as-aliens more seriously. But in a quick search I can’t find any serious social analysis of this hypothesis. I see studies of why humans might want to believe in aliens, or why they might have a taboo against considering aliens. But not an analysis of alien social behavior, to help us evaluate the UFOs-as-aliens hypothesis. What I find is mostly like the above, “who knows?” So let me try.

To do Bayesian inference well, we need a set of not-crazy scenarios describing what might really be going on, we need a prior describing our beliefs about which of these scenarios seems how likely using our background knowledge, we need some more specific data to consider, and we need likelihood functions that say how likely each piece of specific data would be given each scenario.

Note: to study priors and likliehoods, I’ll need to make some assumptions, and see where they lead. That doesn’t mean I actually believe them.

While many UFO reports can be easily dismissed, a remnant of reports seems harder to dismiss, apparently showing artificial physical objects in the sky with amazing velocities and accelerations, but without the usual physical effects on nearby things.

Regarding these puzzling UFOs, I see three key explanation categories:

  • Measurement Error – What look like artificial objects with crazy extreme abilities are actually natural stuff looked at wrong. Perhaps due to intentional fakery. This is widely and probably correctly judged to be the most likely scenario. Nevertheless, we can’t be very confident of that without considering its alternatives in more detail.
  • Secret Societies – There really are artificial objects with amazing abilities, though perhaps somewhat overestimated via partially misleading observations. These are created and managed by hidden groups in our world, substantially tied to us. Secret local military research groups, distant secret militaries, non-state Bond-villains, time-travelers from our near future, dinosaur civilizations deep in the Earth’s crust, etc.
  • Aliens – Again these objects really do have amazing abilities, and are created by hidden groups. But in this case the relevant groups are much less integrated with and correlated with our societies and history. Little green men, their super-robot descendants, universe-sim admins, gods, etc. If these groups had a common origin with, competed with, or were much influenced by the groups that we know of, such things mostly happened long ago, and probably far away.

These three alternatives don’t obviously exhaust all options, but then again I can’t really think of much else.

Assuming that third scenario, hidden groups whose history and features are not much integrated with ours, we can confidently conclude that they most likely arose long ago and far away. Otherwise their space-time correlation with us would be an unlikely coincidence. Perhaps we and they arose from stars in the same stellar nursery, or Earth life was seeded by them, but that still leaves huge relevant durations and distances. And these pretty strongly support their having spectacular technologies and capacities. They have progressed and innovated for many millions and perhaps billions of years more than we. So they can travel very long distances, and survive very long durations.

Now it isn’t at all crazy to expect that many alien powers might arise over the scope and history of the universe. Our prior there has to start out high. But it is a bit more surprising that over billions of years this hasn’t resulted in visible changes to the universe we see. Somehow, all these advanced aliens have not widely rearranged galaxies, deconstructed stars, and so on. Once we condition on the “great filter” fact that we don’t see aliens out there, it become much less clear how likely we should consider aliens to be, especially aliens capable of and inclined to come near us. But that scenario also isn’t obviously impossible, so let us continue.

To consider UFOs-as-aliens, we must consider ancient aliens who were once very far away long ago, had spectacular tech and capacities, did not visibly change the universe, eventually traveled to here now, and are doing stuff around here now. The most likely scenarios consistent with that description tend to have those aliens be clearly visible around here. They’d be living near, building things, using local resources, dumping trash, fighting with each other, etc. But they are not clearly visible. So we must downgrade our prior again, perhaps a lot, to consider scenarios where active local aliens are clearly visible neither on cosmic scales nor on local scales.

For example, perhaps these aliens have found other attractive resources somewhere else hard-to-see nearby, perhaps dark matter or another dimension, resources so much more attractive than ours that they see no point in using the stuff we see. (But then why do their UFOs come here and interact with our matter?) Or perhaps they’ve coordinated to make our region into a nature preserve, not to be used much. Or perhaps they want to observe Earth and human evolution untouched and uninfluenced. Not crazy scenarios, but also not obviously the most likely ones consistent with our prior knowledge.

We have so far had to cut down our aliens prior to account for the lack of clear alien visibility at both cosmic and local scales. But we still have at least one more puzzling data point to integrate into our analysis: these aliens are sometimes somewhat visible as UFOs. Surely such advanced aliens are well aware of our existence, and can figure out roughly what we can see and infer about them. So either they are purposely allowing us to see glimpses of them in this way, or they are failing to prevent such glimpses.

So far, everything I’ve said I’ve heard before from others. Now come my original points, which I haven’t heard from elsewhere, though I wouldn’t be surprised if others have said them. (Far more is written on this than I have time to survey, as I lack good quality filters in this area.) Under either of these scenarios, purposeful or accidental revelation, it isn’t obvious that UFOs would be the only or main channel of such revelations to us about aliens.

If UFOs are shown to us on purpose, to influence our society in some way via a weak suspicion of local aliens, surely such capable aliens would also have a great many other way to influence us. And it is hard to imagine a purpose, or ability package, which would limit their influence to letting us see UFOs. They could edit our DNA, start pandemics or earthquakes, whisper hints to key leaders or innovators, kill off opponents, etc. And while they might be able to do all these things in ways that remain quite hidden, they could also work less hard to hide, and let some of us sometimes get glimpses of their influence.

Perhaps the fact that we see strange UFO behavior is due to accidental failures to sufficiently monitor or incentivize local alien actors who would otherwise want to influence us. Their abilities to prevent such failures would be quite good, but not quite perfect. But if so, similar failures could also allow local aliens to influence us in other ways. On our end, perhaps editing DNA, whispering hits, etc. On their end, they must at some points collect materials and energy sources, stay at home locations, and discard trash.

So under both types of scenarios, if UFOs are due to aliens we should also expect to sometimes see rare but striking alien influences in many other domains. Thus we should be able to get data to confirm or refute this UFOs-as-aliens theory by looking at many other areas of life, not just at strange objects in the sky. (Or in sea, caves, forests, and other sparse places.)

Sure, it is logically possible that aliens intend for us to see them only via strange sky objects. But our prior doesn’t at all favor that, even after modification to condition on low visibility at cosmic and local scales. So a lack of apparent alien influence in many other areas of life must count as evidence against the UFOs-as-aliens scenarios, both the purposeful-but-weak and the barely-accidental versions. Conversely, UFOs-as-aliens would be confirmed by a consistent pattern of rare but striking hard-to-explain influences in other areas of life, influences that aliens might plausibly want to cause.

I am somewhat of a polymath, pursuing a wider range of areas and topics than do most intellectuals or social scientists. So I consider myself to be more qualified than most to consider the possibility of strange influences on human behavior. And while I have in fact seen many strange things, for almost none does alien influence seem especially helpful in explaining what we see.

Now you might argue that aliens want to limit their purposeful revelations to one main most-effective area, or that due to varying costs of coordination and enforcement, one main area will end up being the hardest to control, and thus the area where the most accidental revelations occur. So why couldn’t strange stuff in the sky be that main area in either case?

Yes, that’s not crazy. But assume then that aliens are trying hard to just barely weakly reveal themselves in only one area, or that they are trying hard to prevent us from seeing them but just failing a bit in one worse area. Neither of these scenarios offer much encouragement for more careful analysis of this UFOs-as-aliens theory. In both cases, aliens are controlling how much we see, and so can plausibly quickly adjust their efforts to hide better if we surprise them with being more perceptive than expected. And if we are less perceptive than expected, they can relax their efforts a bit, to make it easier for us to see.

This is somewhat like the problem of inferring that we live in a sim via errors in the sim. If we lived in a sim, and the people running it could see us noticing errors, then they cold stop the sim at that point, back it up, and restart after putting more effort into cutting errors. So we’d only remember errors if they wanted us to remember them. In this scenario, if we just barely sometimes notice errors that we are not very sure are errors, our putting more effort into studying possible sim errors would only be rewarded by stronger efforts on their parts to hide their errors.

That is, if there really are gods around who don’t want us to easily see them, but who sometimes reveal themselves to some of us, we can’t gain much by trying to together better analyze our shared data, to see if they exist. They can control what we see, and control us more directly, in enough ways that we will only know and see what they want us to know and see. Yes, okay, maybe they intend to reward us by revealing themselves to us, but only after we do a good enough collective analysis of our data. But really, given all the other plausible motives and priorities that they might have, our prior has to count that as a quite unlikely scenario. Most likely, when they want us to see them, we’ll see them, but not before.

Yes, aliens might just happen to be at the edge of detectability to us, but not due to efforts on their part to prevent or encourage detection. Yet if that edge region in detectability space is narrow, then it seems that our relative prior on that scenario should be low. The fact that UFOs have remained near our edge of detectability for 80 years of improving sensor tech and increasing sensor density also weakly suggests that more than coincidence is at work here. However, an increasing taboo against UFO-as-aliens may be an adequate explanation for this, and the edge region of detectability space may not in fact be narrow.

Of course even more likely, perhaps, no nearby aliens cause UFOs. But if they do, the best hypothesis, for its combo of likelihood and productivity, seems to be aliens who can travel very far in space and time, who sometimes travel near us, but who care little about us or the types of resources that we can see or use. They do visit here sometimes, where we sometimes meet them accidentally. The rest of us only hear of such meetings when our taboo against reporting them happens to be especially weak. Weirder than I expected, but then the universe has been weirder than we’ve expected before.

Added 5a: A creative scenario is humans finding & using ancient alien tech. Alas, the prior chances seem quite low that alien tech would be abandoned near us, found, still functional after this long, functional outside supporting resources of their civilization, useable by us, and still kept hidden.

GD Star Rating
loading...
Tagged as: ,