The Pebble and the Avalanche

Moshe Thumbnail
Current Revolutions in Business and Technology

by Dr. Moshe Yudkowsky,

author of The Pebble and The Avalanche: How Taking Things Apart Creates Revolutions

 

Tue, 2010-Mar-09, 07:56

Story Marker
Proper Answers to Census Questions

Yesterday I received a letter from the Census Bureau that informed me of next week's arrival of the census form. I always fill it out the same way: I give the government all the information to which the government is entitled, namely a count of the individuals living in my home. The rest of the information, including their names and ages, is none of the government's business.

This year the census form includes questions about race, as it always does — there's quite a few this time around, and I'm always bemused by the arcane process that selects the ethnicity to measure. But I do wonder: how does government-mandated disaggregation into racial groups affect governance and society? While it's very easy indeed to come up with negative consequences, I draw a complete blank when I try to think of an instance when officially-mandated racial disaggregation resulted a positive innovation.

Mon, 2010-Mar-01, 09:18

Story Marker
Assassination In Dubai

My friend Bruce Schneier asked me to chime in on the question of the assassination of Mahmoud al-Mabhouh, a Hamas terrorist, in Dubai. I started writing this on that day; put it aside for long while; and while the information is just as relevant, I think that I'll just give up and post what I've written so far.

One of the most important things to do when you read a police report or press release is to keep an open mind, and this adage applies double in the politically-charged atmosphere of the Middle East. I have a long list of basic questions about the assassination; I'll post a few here as samples.

Was there an assassination?

The Dubai police claim that Mahmoud al-Mabhouh died of asphyxiation, and the coroner noted that this determination was the "hardest of his career" and took ten days to discover. Asphyxiation generally leaves quite noticeable forensic evidence, which make me wonder not only why it took so long to discover that this was an assassination and not natural causes, but also whether or not it was an assassination at all.

Some people may recall that when Yassir Arafat finally died, of a disease that seems to have been HIV/AIDS, a good portion of the Arab press claimed that the Israelis had poisoned him. The Dubai press doesn't seem to suffer from this sort of problem, but it's a question worth keep in mind.

I'll proceed through the rest of this post on the assumption that Mahmoud al-Mabhouh actually did die of asphyxiation by person or persons unknown. Furthermore, since Mahmoud al-Mabhouh was a terrorist, I will use the words "assassination" or "execution" to describe this presumed extra-legal death.

Did a Team Do This?

A quick glance at the video released by the Dubai police shows a group of travelers; I haven't made the time to look at the entire thing. Without the narrative the Dubai Media Office supplied ("teams," "sophisticated communication devices"),for the most part we have nothing but boring video of ordinary travelers, to the point that I don't have the patience to sit through all the video that's been released.

For the remainder of this essay I'll assume, arguendo, that the Dubai police have constructed but not released to the public more convincing evidence.

How Much Did This Cost?

The Dubai police orginally claimed that 17 persons participated in this operation and they keep adding more people to the roster of suspects every day. Round-trip airfares are about $600 per person between Austria and Dubai; figure that each person took another 2 round-trip tickets to cover their costs. The hotel costs are about $150 per night, per diem costs to maintain a cover as a tourist are perhaps up to $200 per day.

The "retail" cost of a blank, stolen British passport is about $3,500 (American passports are worth less, Canadian passports more).

Total out of pocket costs for the "observable" part of the operation are therefore in the range of $300,000. Of course the actual costs (especially if it was a government operation) were many times that amount, but the scope of the operation is well within the reach of anyone with a lot of money and a grudge.

What Kind of People Did This?

According to Dubai police, a team of seventeen thirty-odd people carried out the assassination. That's a lot of people, which tells us something about who they are: they are either government employees, a terrorist organization, or members of a religious group.

They are not mercenaries. Perhaps I'm naive, but I don't believe you can hire such large gangs of well-trained operatives; there simply isn't enough call for such work to keep a gang this size in business. (Of course you can hire gangs this size for small-scale warfare, but not for assassination.) Of course you can hire freelance mercenaries and assemble them into your own team, but that runs into issues of trust.

In an operation this size, any participant faces several security risks from the police. He may be caught, of course; one of his colleagues may be caught and induced to talk; the people running the operation may captured, sloppy, or otherwise compromised and again induced to talk. In turn, this implies that the operatives must have a high degree of trust in each other, which argues against a team this size as a group of mercenaries hired at random. Worse, random mercenaries or even members of a mercenary group will likely find themselves, one day after capture on a different operation, confronting a police officer and in a position to bargain for their freedom by discussing an old operation...

The size of this team argues that the team consisted of government agents, who have patriotic cohesion and the substantial resources of a government to help them; or they are terrorists, who share an ideology and have resources to enforce discipline and perhaps free captured operatives; or they are coreligionists who share a devotion to a cause and some sort of structure to intercede to support captured operatives.

Who Ordered This Operation?

At first the Dubai government refrained from speculating who commited this assassination, which made me respect their professionalism, but apparently that's gone by the wayside. Be that as it may, who would commit this assassination?

As a Middle East terrorist, al-Mabhouh had many enemies:

  • Israel. al-Mabhouh not only committed terrorist attacks in the past; he actively supported the Hamas terrorist organization's infrastructure in Gaza. I have to admit that his public profile does not seem to support a high-cost Israeli attack against al-Mabhouh now — why would the Israelis assassinate someone like al-Mabhouh now? If he is just a middleman for Iranian missle supplies to Gaza, that means he's just another middle manager who can be replaced.

    Aruging against the Israelis is that six of the forged passports used the names of Israeli citizens. There's no reason for Israel to do this, and implicating your own citizens (however briefly) smacks of desparation. Was the operation really so slapdash? Was the selection of Israeli citizens a message? See the next paragraph for further discussion.

    Some of the alleged assassins exited Dubai via Iran, but that does not indicate these were not Israelis. It just indicates that, if these were Israelis, they were confident of their ability to transit Iran on their passports.

  • Fatah. In case you've forgotten, Hamas and Fatah fought a pitched battle for control of Gaza, and Fatah lost. Fatah controls the West Bank, Hamaz controls Gaza, and the only thing preventing a full-scale civil war between the two is the physical separation proivded by Israel and Israeli interdiction of heavy weaponry. I also suspect that Fatah's long-term civic corruption has reached the level that interferes with their abiilty to mount attacks.

    Again, it isn't clear why Fatah would move at this time against al-Mabhouh.

    The use of Israeli identies in the forged passports argues in favor of a Fatah operation. Fatah and criminal gangs operate in Israel, carrying off anything isn't nailed down, including cars. I find it easy enough to believe that Fatah could "steal" identities of Israeli citizens.

  • Anti-Iranian Regimes and Groups. Hamas is Iran's proxy in Gaza, and Iran continues to pressure the Arab world to accept Iran's hegemony. Lebanon is of course the most extreme case, with Iranian troops supporting the Hezbollah terrorist organization. Any Middle East regime, from Saudi Arabia to Pakistan, has a vested interest in the destruction of Iran's proxies.
  • France has a continuing interest in Lebanon, and may have found a reason to assassinate a Hamas operative with ties to Iran, given Iran's malignant presence in Lebanon.
  • If al-Mabhouh was actively involved in some more high-profile but not publically known activity, related to (for example) smuggling weapons of mass destruction to use against European targets, any European country would be a possible sponsor. This might be far-fetched, given the wretched diplomacy of most European countries regarding Iran and Hamas, but I would not be surprised to see a competent French inteligence unit deciding that al-Mabhouh posed a threat to French interests.

The best bet is still the Israelis, who will be more than happy to quietly take credit for this operation.

And so on and so forth. The main point I'm trying to get across is simple: we don't know who did this. We don't know why. The Dubai police have a narrative, and the Israelis certainly have the motive, means, opportunity, and above all the resolution to see this sort of operation through. But so do others.

Tue, 2010-Feb-16, 10:13

Story Marker
Europe: United When It Should Not Be

The captain of a US Navy ship enjoys tremendous freedom to innovate — but is absolutely accountable for not only his actions, but the actions of the people whom he appoints to positions of authority. If a young officer makes a mistake in the middle of the night when the captain is asleep, the captain is accountable.

I have many reasons to be leery of the European Union, not the least because the EU comprises yet another layer of legislative, judicial, and regulatory government on top of the current national governments. But from where I sit, the EU enjoys wide power but with little or any accountability, which inevitably leads to both abuse and failure.

The current situation in Greece would be a disaster for that country, but because Greece shares a common currency with other members of the EU, the Greek problems can easily spread. EU member countries routinely ignore the budget and debt rules — the basic rules that are supposed to support the Euro currency. Unlike the military, where accountability is a matter of law, in the EU adherence to the rules is governed by politics, which means that popularity and congeniality trumps accountability.

The EU has made the choice to handle Greece's problems by a political process, namely a bailout package. We here in the US are familiar with this theory; two of the three largest US-owned automobile manufacturers have been kept on government life support for years, with no hope in sight for recovery. If the EU is unwilling to disaggregate their monetary system, then they must either allow Greece to fail or use a unified process to make certain that Greece's budget conforms to EU rules. While there may be a clever idea out there, just about any other option I can think of leads directly to failure.

Tue, 2009-Nov-17, 09:25

Story Marker
Breast Cancer Screening Guidelines

The latest breast-cancer screening guidelines from the US's "Agency for Healthcare Research and Quality" may or may not be scientifically sound; I'm not qualified to judge. Many independent societies have gone ballistic over the new guidelines; but there's a deep fallacy that's gone unnoticed.

I disagree with the entire exercise of publically-defined health care benefits, which reverses disaggregation of care and service. This study takes a collectivist approach: so many women will live, so many will die, assign cost/benefit analysis to the results and decide if breast-cancer screening makes economic sense to "society as a whole." This is, of course, how state-run medicine/national health care works in Europe; but here in the US we are not quite yet cattle that can be culled from the herd when we become too expensive to heal. As Mark Steyn noted in this same context, when the government health care system meets its budget and the people who receive no treatment don't whimper too loudly, that's defined as success; but in a society of free individuals, Mark Steyn runs his own "health care system," and in the Mark Steyn health care system going without treatment — as per President Obama's suggestion to the elderly to forgo corrective surgery, take a pill instead and live with the pain — is the very definition of failure. We've applied disaggergation to our ordinary lives, and we can customize just about anything we wish from cars to shoes to ordinary house paint; if we standardize and federalize health care we will take a giant leap backwards.

Mon, 2009-Nov-16, 08:55

Story Marker
GE in China

According to today's paper, GE just signed a deal with a Chinese company. Mainland China still suffers from a lack of disaggregation between the government and business: a favored business can expect special treatment when it comes time to enforce or interpret contract law. Partners from the US tend to suffer. This raises a question: Has any US business had a decent, ten-year relationship with a Chinese company in China? Or have all large US businesses suffered in the end?

Wed, 2009-Nov-11, 09:52

Story Marker
Bear Convictions Bear Market

The US government just lost its case against two managers at Bear Stearn. Prosecutors tried out a new theory of criminal wrongdoing, which the jury correctly rejected.

This case originally received a great deal of attention when the prosecutors announced their indictments and released emails that purported to show that the defendants had committed deliberate fraud — a clear case of prosecutorial misconduct that has become all too prevalent. At trial, the prosecutors released the entire contents of the email messages and not just short excerpts, at which point the jury rejected the prosecutors' case. Disaggregation is all fine and good, but chopping up documents to mislead the public and pressure defendants into agreeing to a plea bargain ought to be, but unfortunately is not, grounds for throwing the prosecutors themselves into jail.

Tue, 2009-Jul-21, 08:49

Story Marker
Wisconsin's Practical Joke

Wisconsin's Department of Transportation played a wonderful practical joke on me Sunday afternoon. (Of course, it might just be a mistake brought about by poor design; we'll discuss that below.)

When I headed south on 94, after being on the road for 10 hours, I exited at Highway 50 for a short break and to fill up with gas. The traffic on 94 was quit heavy due to road construction and I decided to take a chance on the "Alternate 94" route. That's when the fun and games began.

The "Alt 94" route brought me a bit south to Highway D; the signs sent me west on Highway D. After a long, long ride the signs sent me north through the town of Bristol, and then eventually east on Highway 50. Yes, indeed, after a good long tour through the countryside, the "Alternate 94" signs brought me back to the same place I'd started.

After examining the map I picked my own alternate road to avoid 94 southbound, and I was not too surprised to see "Alternate 94" signs along the route once I actually reached the road itself (not along the way to the road, of course).

I believe this incident illustrates a fundamental truth: disaggregation requires communication. I don't expect the construction workers on the road to set up the signs for the alternate routes; these are separate jobs. But unless the two separate agencies communicate with each other — unless the sign department tracks the progress of open and closed roads and the current state of the signs — then this sort of silly and maddening behavior becomes almost inevitable. Wisconsin DOT can get away with this sort of miscommunication (are you going to drive on a non-DOT road?), but private companies with non-captive customers don't have that luxury.

Then again, maybe it was a practical joke, and there's a betting pool somewhere in the Wisconsin Dept. of Transportation. If so: good one, guys. I'll remember it when you visit Chicago...

Thu, 2009-Jun-18, 08:49

Story Marker
When You Lose Sight of Your Goal

One of the most important things to know about disaggregation is that you must still maintain some sort of feedback loop if you want the disaggregated parts to work together. For example, if a factory that makes nuts and bolts splits into a factory that makes nuts and another factory that makes bolts, the two new factories must work together to make certain the sizes, shapes, and material work together.

We see this problem in the US drug industry. The drug companies create new products, but the government decides if the prodcuts can be sold. While in theory this might enhance drug safety, in practice the government agency has entirely different goals than the drug company and the people who consume drugs: the bureaucrats' main goal is the preservation of their careers. Their performance is not measured in terms of profit earned or lives saved; their performance is measured by how well they follow the rules and whether or not a product they approved harms anyone. As a matter of course this means that the government demands draconian, hideously expensive testing and then refuses to let even the safest drugs be released for sale unless they're "better" than existing drugs; this makes no sense if the goal is safety, but the real goal is to avoid possible bad publicity. Complete disaggregation, in this case, leads to fewer lifesaving drugs.

Bear this in mind as the Obama administration attempts to create new regulatory agencies. One proposal would create a consumer oversight agency with broad powers to regulate any new financial products intended for consumers — an agency that would be completely disaggregated from the people who provide the products and the people who consume them. Without any doubt, this new agency will stop financial innovation dead in it tracks.

Mon, 2009-Jun-01, 08:51

Story Marker
The GM Disaster

While I don't usually discuss politics, I can't turn aside from the Obama Administration's seizure of General Motors. President Obama and his advisers have committed a mistake of historic proportions, the kind that future generations look back on in amazement.

The fundamental rules of capitalism dictate separation between government and business. Unless these two systems — with their very different goals and their very different rules — remain disaggregated one from the other, the inevitable result is disastrous. At this point, the only question is how bad the situation will become.

The past several months have show just how deadly government interference can be. Over $30 billion of wealth has been destroyed as government subsidies disappeared to no effect: can you imagine any private institution that would dole out money with such abandon and with so little prospect of success? As bankruptcy approached, President Obama used his official and unofficial powers to browbeat bondholders.

As for the prospects for the future, one need only look to previous horrible examples, from UK (Jaguar) to France (Citroen) to Romania (Dacia). Governments run car companies not based on how well the cars will run, how much profit will be made, or return on equity; they run car companies as they run any other enterprise: politicians maximize votes, not profits. I expect a series of ever-more desparate political interventions in the marketplace over the next several years to keep GM and Chrysler afloat regardless of the economic sense or economic consequences.

Wed, 2009-Apr-29, 09:18

Story Marker
Chrysler and GM In a Single Pool

I'm afraid there's no gentle way to say this: the Obama administration is about to commit yet another major blunder. As part of the administration's interference in the US automobile industry, the administration proposes to merge the lending arms of the Chrysler and GM companies into a single unit. This course is, of course, precisely backwards: the key to innovation, recovery, and viability is disaggregation.

Instead we'll see one huge behemoth. While easier for the US to influence and administer, especially given the Obama administration's problems with staffing, the combined organization will have exposure to political and financial risk from both Chrysler and GM. A failure at one of these companies — or, more likely, a political crisis in advance of failure — will lead to risky lending at the new combined organization. Because of its size, the organization will no doubt be deemed too risky to fail, and so the merry-go-round of increased subsidies followed by increased government interference will continue to spin.