# Information Processing

Pessimism of the Intellect, Optimism of the Will     Archive   Favorite posts   Twitter: @steve_hsu

## Tuesday, June 26, 2007

### It's all in the clusters

Nicholas Wade in the Times on genetic clustering, differential selection and race. See here for previous discussion of a metric on the space of genomes and the scientific meaning of race.

I hear the voice of physicist turned evolutionary biologist Greg Cochran in Wade's writing :-)

Wade: Historians [and social scientists!] often assume that they need pay no attention to human evolution because the process ground to a halt in the distant past. That assumption is looking less and less secure in light of new findings based on decoding human DNA.

People have continued to evolve since leaving the ancestral homeland in northeastern Africa some 50,000 years ago, both through the random process known as genetic drift and through natural selection. The genome bears many fingerprints in places where natural selection has recently remolded the human clay, researchers have found, as people in the various continents adapted to new diseases, climates, diets and, perhaps, behavioral demands. ...

Cochran: There is something new under the sun — us.

Thucydides said that human nature was unchanging and thus predictable — but he was probably wrong. If you consider natural selection operating in fast-changing human environments, such stasis is most unlikely. We know of a number of cases in which there has been rapid adaptive change in humans; for example, most of the malaria-defense mutations such as sickle cell are recent, just a few thousand years old. The lactase mutation that lets most adult Europeans digest ice cream is not much older.

There is no magic principle that restricts human evolutionary change to disease defenses and dietary adaptations: everything is up for grabs. Genes affecting personality, reproductive strategies, cognition, are all able to change significantly over few-millennia time scales if the environment favors such change — and this includes the new environments we have made for ourselves, things like new ways of making a living and new social structures. I would be astonished if the mix of personality types favored among hunter-gatherers is "exactly" the same as that favored among peasant farmers ruled by a Pharaoh. In fact they might be fairly different.

(Larger version here.)

From Dienekes:

A new article in BMC Genomics discusses the issue of predicting continental origin using randomly selected markers. The pdf is freely available.

One of the arguments of those who deny the existence of biological races is that their reality is subjective. Some extremists have argued that race is totally socially constructed; this is, however, disproven by the fact that socially constructed race is correlated with physical characteristics. Thus, rather than being separated from biology, the social phenomenon of race is rooted in biology.

A different argument holds that race is correlated with biology, but the differences are "skin-deep", i.e., involve only superficial, visible, (and by some strange logic unimportant) characteristics. According to the proponents of this view, the idea of biological race places an undue emphasis on a set of traits: it is a result of the subjective choice of a set of traits as race-defining. Thus, the commonly recognized races of traditional physical anthropology are discounted as subjective organizations of the biological data: we could just as simply speak of a "lactose-intolerant race" according to this view.

In forensic science and admixture analysis scientists often discover and use polymorphisms which exhibit large inter-population differences. Decoding DNA isn't free, thus, it makes sense to use the most informative, most "biased" markers when one is trying to discover the origin of a biological sample. For example, if Africans have 55% of gene version A and 45% of gene version B, and Europeans have 53% of A and 47% of B, it makes little sense to type this particular gene, since it cannot really tell us whether a sample is European or African. A gene where Africans have 90% of A while Europeans have 5%of A would be much more useful. Race skeptics claim, as with the physical anthropological data, that to privilege such carefully chosen genes is to stress the differences between groups; the implication is that in randomly chosen genes these differences are minor.

The new paper is one of many (you can click on the Clusters label to find more) recent papers that have discovered that no matter what genetic markers you choose: SNPs, STRs, no matter how you choose them: randomly or based on their "informativeness", it is relatively easy to classify DNA into the correct continental origin. Depending on the marker types (e.g., indel vs. microsatellite), and their informativeness (roughly the distribution differences between populations), one may require more or less markers to achieve a high degree of accuracy. But, the conclusion is the same: after a certain number of markers, you always succeed in classifying individuals according to continental origin.

Thus, the emergent pattern of variation is not at all subjectively constructed: it does not deal specifically with visible traits (randomly chosen markers could influence any trait, or none at all), nor does it privilege markers exhibiting large population differences. The structuring of humanity into more or less disjoint groups is not a subjective choice: it emerges naturally from the genomic composition of humans, irrespective of how you study this composition. Rather than proving that race is skin-deep, non-existent, or unimportant, modern genetic science is both proving that it is in fact existent, but also sets the foundation for the study of its true importance, which is probably somewhere in between the indifference of the sociologists and the hyperbole of the racists.

## Monday, June 25, 2007

### Rise of the money machines

As far as I could tell there was only one other physicist at foo camp, and he is CTO and head quant at a big hedge fund. He and Tim O'Reilly ran a discussion called Rise of the Money Machines :-)

From the Economist's review of the new Peter Bernstein book Capital Ideas Evolving (sequel to Capital Ideas, which was quite good).

Economist: ...Indeed, Mr Bernstein seeks to show how financial giants such as Barclays Global Investors and Goldman Sachs Asset Management have built on the insights developed by the academics. If there are ways systematically to beat the markets these days, they probably require men with physics doctorates and massive computer power rather than a smooth manner and the right contact book.

There is the equivalent of a technological arms race as modern fund managers vie to find the best computer models and to trade quickly before their competitors spot the same opportunities. ...

### Curved space and monsters

New paper!

http://arxiv.org/abs/0706.3239

A simple question: how many different black holes can there be with mass M? Conventional wisdom: of order exp(A), where A is the surface area of the hole and scales as M^2.

Using curved space, we construct objects of ADM mass M with far more than exp(A) microstates. These objects have pathological properties, but, as far as we can tell, can be produced via quantum tunneling from ordinary (non-pathological) initial data. Our results suggest that the relation between black hole entropy and the number of microstates of the hole is more subtle than perhaps previously appreciated.

Update! Rafael Sorkin was kind enough to inform us of his earlier related work with Wald and Zhang. We've added the following end-note to the paper.

Note added: After this work was completed we were informed of related results obtained by Sorkin, Wald and Zhang [25]. Those authors investigated monster-like objects as well as local extrema of the entropy S subject to an energy constraint, which correspond to static configurations and obey $A^{3/4}$ scaling. For example, in the case of a photon gas the local extrema satisfy the Tolman--Oppenheimer--Volkoff equation of hydrostatic equilibrium. In considering monster configurations, Sorkin et al. show that requiring a configuration to be no closer than a thermal wavelength $\lambda \sim \rho^{-1/4}$ from its Schwarzschild radius imposes the bound $S < A$. While this may be a reasonable criterion that must be satisfied for the assembly of an initial configuration, it does not seem to apply to states reached by quantum tunneling. From a global perspective configurations with $S > A^{3/4}$ are already black holes in the sense that the future of parts of the object does not include future null infinity.

Black hole entropy, curved space and monsters

Stephen D.H. Hsu, David Reeb

(Submitted on 21 Jun 2007)

We investigate the microscopic origin of black hole entropy, in particular the gap between the maximum entropy of ordinary matter and that of black holes. Using curved space, we construct configurations with entropy greater than their area in Planck units. These configurations have pathological properties and we refer to them as monsters. When monsters are excluded we recover the entropy bound on ordinary matter $S < A^{3/4}$. This bound implies that essentially all of the microstates of a semiclassical black hole are associated with the growth of a slightly smaller black hole which absorbs some additional energy. Our results suggest that the area entropy of black holes is the logarithm of the number of distinct ways in which one can form the black hole from ordinary matter and smaller black holes, but only after the exclusion of monster states.

## Sunday, June 24, 2007

### Foo camp impressions

Beautiful weather, but very cold at night.

Too much to absorb!

Larry Page came by helicopter. Jimmy Wales is in a tent like the rest of us. Dirac's grandson is here. Data mining is big. Hedge funds are on the mind, even here. Lots of people working on Web 2.0 and collective intelligence. Facebook's new platform is impressive. MySpace and Facebook in a technology race. Biobricks? Founders of Red Hat, Skype, Wikipedia, Amazon, MoveOn, Digg, Delicious, BitTorrent...

Report and video from WSJ reporter Kara Swisher about the event. I'm at the end of the video, explaining the game "werewolf" to Kara.

Sadly, I was not there to see Google’s Larry Page land a helicopter on the lawn up at Tim O’Reilly’s annual geekfest called Foo Camp, so I could mock him to his face.

(Note to Larry: That better be an awfully big solar footprint you’re building at the Googleplex in Silicon Valley to replace all the carbon emissions your various flying machines spew.)

Since I had to trade kid duty, I could only get up there Friday night for the opening festivities, which are held annually at the Sebastopol headquarters of O’Reilly Media.

Still, it was well worth the trek north of San Francisco to get a short glimpse of some new and sometimes quite wacky ideas about the future of digital development.

The conference is almost entirely user-generated, as people sign up to lead sessions on a variety of sometimes esoteric topics in rooms scattered all through the facility.

While most conferences look at the here and now, Foo Camp is aggressive in its quest to get people to think outside the box. In fact, if there were a box, the brainy denizens of Foo Camp would probably turn it into a time machine/beer dispenser/robot ninja warrior.

It could happen.

Many big wheels and many more big brains were there to figure it all out. Indeed, as you will see from this video, there is still a very pure and very infectious enthusiasm after many years at Foo Camp, even though some have complained about its ever-larger size.

So, for those who wanted to go and could not get in, here is a rather long glimpse at the first night of Foo Camp, including campers trying unsuccessfully to introduce themselves with only three words, a look at the tents, a talk with Tim O’Reilly and some in attendance explaining why they come:

## Thursday, June 21, 2007

### Mark to market

The almost collapse of two Bear Stearns hedge funds investing in mortgage-backed securities is sending a tremor through Wall St. A last minute bailout by creditor Merrill means that $800 million in CDOs is about to be marked to market. In other words, some complex, illiquid securities are about to have a meaningful price, as opposed to the theoretical value on the books. Insert words about Long Term Capital Management and "systemic risks" here. Nobody wants to look at the truth right now because the truth is pretty ugly,'' Castillo said. Where people are willing to bid and where people have them marked are two different places.'' I've written several posts about CDOs here. This and this are getting a lot of traffic right now. Word is that Gaussian copula models are **way** off from real market prices :-) June 21 (Bloomberg) -- Merrill Lynch & Co.'s threat to sell$800 million of mortgage securities seized from Bear Stearns Cos. hedge funds is sending shudders across Wall Street.

A sale would give banks, brokerages and investors the one thing they want to avoid: a real price on the bonds in the fund that could serve as a benchmark. The securities are known as collateralized debt obligations, which exceed $1 trillion and comprise the fastest-growing part of the bond market. Because there is little trading in the securities, prices may not reflect the highest rate of mortgage delinquencies in 13 years. An auction that confirms concerns that CDOs are overvalued may spark a chain reaction of writedowns that causes billions of dollars in losses for everyone from hedge funds to pension funds to foreign banks. Bear Stearns, the second-biggest mortgage bond underwriter, also is the biggest broker to hedge funds. More than a Bear Stearns issue, it's an industry issue,'' said Brad Hintz, an analyst at Sanford C. Bernstein & Co. in New York. Hintz was chief financial officer of Lehman Brothers Holdings Inc., the largest mortgage underwriter, for three years before becoming an analyst in 2001. How many other hedge funds are holding similar, illiquid, esoteric securities? What are their true prices? What will happen if more blow up?'' Shares Fall Shares of Bear Stearns, the fifth-biggest U.S. securities firm by market value, and Merrill, the third-largest, led a decline in financial company stocks yesterday, and the perceived risk of owning their bonds jumped on concerns losses related to subprime home loans may be bigger than initially thought. Both companies are based in New York. The perceived risk of owning corporate bonds jumped to the highest in nine months today. Contracts based on$10 million of debt in the CDX North America Crossover Index rose as much as $10,000 in early trading today to$178,500, according to Deutsche Bank AG. They retraced to $171,500 at 8:28 a.m. in New York. U.S. Securities and Exchange Commission Chairman Christopher Cox said yesterday that the agency's division of market regulation is tracking the turmoil at the Bear Stearns fund. Our concerns are with any potential systemic fallout,'' Cox said in an interview. Bankers and money managers bundle securities into a CDO, dividing it into pieces with credit ratings as high as AAA. The riskiest parts have no rating because they are first in line for any losses. Investors in this so-called equity portion expect to generate returns of more than 10 percent. Fivefold Increase CDOs were created in 1987 by bankers at now-defunct Drexel Burnham Lambert Inc., the home of one-time junk-bond king Michael Milken. Sales reached$503 billion in 2006, a fivefold increase in three years. More than half of those issued last year contained mortgages made to people with poor credit, little loan history, or high debt, according to Moody's Investors Service.

New York-based Cohen & Co. was the biggest issuer of CDOs last year. It has formed 36 CDOs since 2001, including 15 worth a total of $14 billion in 2006, according to newsletter Asset-Backed Alert. Not since 1994 have mortgages with past due payments been so high, according to first-quarter data compiled by the Federal Deposit Insurance Corp., the agency that insures deposits at 8,650 U.S. banks. Lehman analysts estimated in April that the collateral backing CDOs had fallen by$25 billion.

The big question is whether these forced liquidations represent a tipping point in the market,'' said Carl Bell, who helps manage $63 billion in fixed-income assets as head of the structured-credit team at Boston-based Putnam Investments. It may put pressure on other hedge funds pursuing similar strategies'' as the Bear Stearns funds, he said. Biggest Names The Bear Stearns funds are run by senior managing director Ralph Cioffi. One of the funds, the 10-month old High-Grade Structured Credit Strategies Enhanced Leverage Fund, lost 20 percent this year, the New York Post reported. Officials at Bear Stearns and Merrill declined to disclose the losses. The funds had borrowed at least$6 billion from the biggest names on Wall Street. Aside from Merrill, other creditors included Goldman Sachs Group Inc., Citigroup Inc., JPMorgan Chase
& Co. and Bank of America Corp. All of the firms are based in New York, except Bank of America, which is based in Charlotte, North Carolina.

As the funds faltered, Merrill sought to protect itself by seizing the assets that were used as collateral for its loans. JPMorgan planned to sell assets linked to its credit lines before
reaching agreement with Bear Stearns to unwind the loan, people with knowledge of the negotiations said yesterday.

Bear Stearns was still in talks late yesterday with creditors to the funds to rescue the funds, said the people, who declined to be identified because the negotiations are private.

Russell Sherman, a Bear Stearns spokesman, and Jessica Oppenheim, a spokeswoman for Merrill, declined to comment.

Pretty Ugly'

Merrill's decision yesterday to accept bids on $800 million of bonds it took as collateral for its loans further stifled trading in CDO securities, said David Castillo, who trades asset-backed, commercial-mortgage and CDO bonds in San Francisco at Further Lane Securities. Nobody wants to look at the truth right now because the truth is pretty ugly,'' Castillo said. Where people are willing to bid and where people have them marked are two different places.'' The perceived risk of holding Bear Stearns bonds jumped to a three-month high, according to traders betting on the creditworthiness of companies in the credit-default swaps market. Contracts based on$10 million of its bonds rose $5,800 to$45,500, according to composite prices from London-based CMA Datavision. An increase in the five-year contracts suggests
deterioration in the perception of credit quality. Contracts on Merrill jumped $4,700 to$33,000, CMA prices show.

Long-Term Capital

Shares of Bear Stearns fell for a fourth day, declining 19 cents to $143.01 at 9:32 a.m. in New York Stock Exchange composite trading. The stock was down 12 percent this year before today, compared with the 0.4 percent advance of the Standard & Poor's 500 Financials Index. Merrill dropped 20 cents to$87.48 and Citigroup fell 13 cents to $53.31. The reaction to the Bear Stearns situation is reminiscent of Long-Term Capital Management LP, which lost$4.6 billion in 1998.

Lenders including Merrill and Bear Stearns met and agreed to take a stake in the Greenwich, Connecticut-based fund and slowly sold the assets to limit the impact of its collapse.

We're not surprised to find the principal circle of players is pretty interconnected,'' said Roy Smith, professor of finance at New York University Stern School of Business and
former head of Goldman's London office. `What we're looking for is whether the interconnection creates a negative domino effect: Whether Hedge Fund A creates a problem for other hedge funds,
which in turn creates a problem for the prime brokers that are lending to them.''

## Wednesday, June 20, 2007

### Fallows on Shenzhen

Speaking of Shenzhen, here is the recent Fallows article in the Atlantic (subscription required): China makes, the world takes. Narrated slideshow (free) here.

I first visited Shenzhen on a day trip from Hong Kong in the early 90's. I was excited to see it, since it was one of the first Special Economic Zones (SEZ) set up by the communist government. My friends in HK couldn't understand why I would want to visit -- they suggested scenic Guilin instead. But I was already fascinated by the rapid economic changes taking place; I wanted to see China's future, not its past. At the time, everything was rough: roads were terrible, new buildings had no sidewalks or landscaping, our bus was caught in some of the worst traffic jams I have ever seen. Now you can take the subway/train there directly from HK, there are 8 million residents, and (I'm told) the city is full of parks and green spaces.

From the article (excerpts found here):

...Has the move to China been good for American companies? The answer would seemingly have to be yes—otherwise, why would they go there? It is conceivable that bad partnerships, stolen intellectual property, dilution of brand name, logistics nightmares, or other difficulties have given many companies a sour view of outsourcing; I have heard examples in each category from foreign executives. But the more interesting theme I have heard from them, which explains why they are willing to surmount the inconveniences, involves something called the “smiley curve.”

The curve is named for the U-shaped arc of the 1970s-era smiley-face icon, and it runs from the beginning to the end of a product’s creation and sale. At the beginning is the company’s brand: HP, Siemens, Dell, Nokia, Apple. Next comes the idea for the product: an iPod, a new computer, a camera phone. After that is high-level industrial design—the conceiving of how the product will look and work. Then the detailed engineering design for how it will be made. Then the necessary components. Then the actual manufacture and assembly. Then the shipping and distribution. Then retail sales. And, finally, service contracts and sales of parts and accessories.

The significance is that China’s activity is in the middle stages—manufacturing, plus some component supply and engineering design—but America’s is at the two ends, and those are where the money is. The smiley curve, which shows the profitability or value added at each stage, starts high for branding and product concept, swoops down for manufacturing, and rises again in the retail and servicing stages. The simple way to put this—that the real money is in brand name, plus retail—may sound obvious, but its implications are illuminating.

At each factory I visited, I asked managers to estimate how much of a product’s sales price ended up in whose hands. The strength of the brand name was the most important variable. If a product is unusual enough and its brand name attractive enough, it could command so high a price that the retailer might keep half the revenue. (Think: an Armani suit, a Starbucks latte.) Most electronics products are now subject to much fiercer price competition, since it is so easy for shoppers to find bargains on the Internet. Therefore the generic Windows-style laptops I saw in one modern factory might go for around $1,000 in the United States, with the retailer keeping less than$50.

Where does the rest of the money go? The manager of that factory guessed that Intel and Microsoft together would collect about $300, and that the makers of the display screen, the disk-storage devices, and other electronic components might get$150 or so apiece. The keyboard makers would get $15 or$20; FedEx or UPS would get slightly less. When all other costs were accounted for, perhaps $30 to$40—3 to 4 percent of the total—would stay in China with the factory owners and the young women on the assembly lines.

Fallows' guide is an Irish entrepreneur who couldn't get a US green card (found here):

...Mr Fallows centers his story on an accidental entrepreneur, Irishman Liam Casey. Mr Casey lives in Shenzhen because that's where the action is. But he'd be living in Laguna or Newport Beach if he could. He loved it there. But he couldn't get a green card or a long-term work permit, so he moved on. Mr Fallows notes parenthetically,

I might as well say this in every article I write from overseas: the easier America makes it for talented foreigners to work and study there, the richer, more powerful, and more respected America will be. America's ability to absorb the world's talent is the crucial advantage no other culture can match - as long as America doesn't forfeit this advantage with visa rules written mainly out of fear.

The supply chain in action (here):

Here are a few examples, all based on real-world cases: You have announced a major new product, which has gotten great buzz in the press. But close to release time, you discover a design problem that must be fixed—and no U.S. factory can adjust its production process in time.

The Chinese factories can respond more quickly, and not simply because of 12-hour workdays. “Anyplace else, you’d have to import different raw materials and components,” Casey told me. “Here, you’ve got nine different suppliers within a mile, and they can bring a sample over that afternoon. People think China is cheap, but really, it’s fast.” Moreover, the Chinese factories use more human labor, and fewer expensive robots or assembly machines, than their counterparts in rich countries. “People are the most adaptable machines,” an American industrial designer who works in China told me. “Machines need to be reprogrammed. You can have people doing something entirely different next week.”

Or: You are an American inventor with a product you think has “green” potential for household energy savings. But you need to get it to market fast, because you think big companies may be trying the same thing, and you need to meet a target retail price of $100. “No place but China to do this,” Mr. China said, as he showed me the finished product. For the world’s most efficient supply chain: Or: You are a very famous American company, and you worry that you’ve tied up too much capital keeping inventory for retail stores at several supply depots in America. With Mr. China’s help, you start emphasizing direct retail sales on your Web site—and do all the shipping and fulfillment from one supply depot, run by young Chinese women in Shenzhen, who can ship directly to specific retail stores. Over the course of repeated visits to Shenzhen—the breakfasts!—and visits to other manufacturing regions, I heard about many similar cases and saw some of the tools that have made it possible for Western countries to view China as their manufacturing heartland. Some involve computerized knowledge. Casey’s PCH has a Google Earth–like system that incorporates what he has learned in 10 years of dealing with Chinese subcontractors. You name a product you want to make—say, a new case or headset for a mobile phone. Casey clicks on the map and shows the companies that can produce the necessary components—and exactly how far they are from each other in travel time. This is hard-won knowledge in an area where city maps are out of date as soon as they are published and addresses are approximate. (Casey’s are keyed in with GPS coordinates, discreetly read from his GPS-equipped mobile phone when he visits each factory.) If a factory looks promising, you click again and get interior and exterior photos, a rundown on the management, in some cases videos of the assembly line in action, plus spec sheets and engineering drawings for orders they have already filled. Similar programs allow Casey and his clients to see which ship, plane, or truck their products are on anywhere in the world, and the amount of stock on hand in any warehouse or depot. (How do they know? Each finished piece and almost every component has an individual bar code that is scanned practically every time it is touched.) The idea that your supply chain is critical intellectual property is obvious. That is why we know so little of the origins of our shiny new MacBook Pro or HP widescreen display. It also highlights why how a company differentiates through innovation and design [e.g., Apple] because it’s products are likely made on the same assembly line as competitors. I could describe many installations, but I was fascinated by two. The first represents one extreme in automation. It is owned and operated by Inventec, one of five companies based in Taiwan that together produce the vast majority of laptop and notebook computers sold under any brand anywhere in the world. Everyone in America has heard of Dell, Sony, Compaq, HP, Lenovo-IBM ThinkPad, Apple, NEC, Gateway, Toshiba. Almost no one has heard of Quanta, Compal, Inventec, Wistron, Asustek. Yet nearly 90 percent of laptops and notebooks sold under the famous brand names are actually made by one of these five companies in their factories in mainland China. I have seen a factory with three “competing” brand names coming off the same line. The Inventec installation I saw was in an export-processing zone in Shanghai specially created for the company, in which imported components for manufacturing and finished products for export were free of the usual duties or taxes. It turns out more than 30,000 notebook computers per day, under one of the brand names listed above. Each day, an Inventec plant on the same campus produces hundreds of large, famous-brand-name server computers to run Internet traffic. This is today’s rough counterpart to the Ford Motor Company’s old River Rouge works. In the heyday of The Rouge, rubber, steel, and other raw materials would come into the plant, and finished autos would come out. Here, naked green circuit boards, capacitors, chip sets, and other components come in each day, and notebook computers come out. Some advanced components arrive already assembled: disk drives from Taiwan or Singapore, LCD screens from Korea or Japan, keyboards and power supplies from other plants in China. The overall process looks the way you would expect a high-tech assembly line to. Conveyers and robots take the evolving computer from station to station; each unit arrives in front of a worker a split second after she has finished with the previous one. Before a component goes into a machine, its bar code is scanned to be sure it is the right part; after it is added, the machine is “check-weighed” to see that its new weight is correct. Hundreds of tiny transistors, chips, and other electronic parts are attached to each circuit board by “pick and place” robots, whose multiple arms move almost too fast to follow. The welds on the board are scanned with lasers for defects. Any with problems are set aside for women specialists, looking through huge magnifying glasses, to reweld. Why did this factory invest so much in robots and machine tools? I asked a supervisor from Taiwan. “People can’t do it precisely enough,” was his answer. These factories automate not what’s too expensive but what’s too delicate for human beings to perform. Did you know that new USB headset you just bought was packed and shipped from China? The other facility that intrigued me, one of Liam Casey’s in Shenzhen, handled online orders for a different well-known American company. I was there around dawn, which was crunch time. Because of the 12-hour time difference from the U.S. East Coast, orders Americans place in the late afternoon arrive in China in the dead of night. As I watched, a customer in Palatine, Illinois, perhaps shopping from his office, clicked on the American company’s Web site to order two$25 accessories. A few seconds later, the order appeared on the screen 7,800 miles away in Shenzhen. It automatically generated a packing and address slip and several bar-code labels. One young woman put the address label on a brown cardboard shipping box and the packing slip inside. The box moved down a conveyer belt to another woman working a “pick to light” system: She stood in front of a kind of cupboard with a separate open-fronted bin for each item customers might order from the Web site; a light turned on over each bin holding a part specified in the latest order. She picked the item out of that bin, ran it past a scanner that checked its number (and signaled the light to go off), and put it in the box. More check- weighing and rescanning followed, and when the box was sealed, young men added it to a shipping pallet.

By the time the night shift was ready to leave—8 a.m. China time, 7 p.m. in Palatine, 8 p.m. on the U.S. East Coast—the volume of orders from America was tapering off. More important, the FedEx pickup time was drawing near. At 9 a.m. couriers would arrive and rush the pallets to the Hong Kong airport. The FedEx flight to Anchorage would leave by 6 p.m., and when it got there, the goods on this company’s pallets would be combined with other Chinese exports and re-sorted for destinations in America. Forty-eight hours after the man in Palatine clicked “Buy it now!” on his computer, the item showed up at his door. Its return address was a company warehouse in the United States; a small Made in China label was on the bottom of the box.

## Monday, June 18, 2007

### A modern Borges?

As you can probably guess, I read a ton of science fiction when I was a kid. There was a time when I could walk into a book shop and recognize every title in the sci fi section. But I stopped reading it about the time I reached high school, and, except for the occasional guilty pleasure (e.g., stuff by Vernor Vinge), haven't kept up since. Being a physicist makes it tough to read science fiction -- I particularly hate it when the sci fi assumptions or their implications aren't self-consistent or at least treated in an sophisticated way. It's also true that there is very little new under the sun (I suppose this applies to all literature). Every time I pick up a new title I can more or less recall several stories of which it is derivative. Sometimes I wonder if the author knows this.

Recently I came across the short fiction of Ted Chiang. It's the first really good science fiction I've read since I was a kid. I guess others agree, because Chiang has already won 3 Nebulas and a Hugo, despite having only published a dozen or so stories. (Read the reviews at Amazon for his one published collection here.)

This is probably over the top, but he reminds me of Philip K. Dick and even Borges. Some will violently disagree with me, but I think Chiang really understands the scientific or metaphysical concepts that appear in his stories, while at times I feel Borges is just alluding without deep understanding.

Here are some of his stories, available online.

Division by Zero. A mathematician has a mental breakdown when she constructs formal machinery showing the internal inconsistency of mathematics. The theme is eerily reminiscent of these comments by Greg Chaitin!

Understand. A shorter, updated version of Flowers for Algernon?

What's Expected of Us. A brief meditation on time travel and free will :-)

### The bunnie and the chumby

Hardware hacker Andrew "Bunnie" Huang has a blog. Huang grew up in Michigan and earned SB and PhD degrees from MIT. Well known for his reverse-engineering exploits (e.g., Xbox), he co-founded hardware startup Chumby (their product is a low-cost wifi data device) and is working in Shenzhen to set up production.

I highly recommend the following posts from his blog.

On the US technology education deficit.

...I was chatting with UCSD high speed integrated circuits professor Jim Buckwalter about the nature of the graduate student applications he has received.

The statistics were astonishing. Of the thousands of applicants, only 80 were from the US. To put this in perspective, he had more applicants with the surname “Lee” alone than he had domestic applicants. And UCSD engineering is no slouch; according to the rankings they are #11 in engineering overall.

...The enormous disparity in domestic applicants to higher education in crucial fields such as high speed circuit design is a bit disturbing. With numbers like these, it is inevitable that the US will lose its edge in technology. I guess it wouldn’t be as bad if these foreign students actually stayed in the US and started companies, but my experience in China has shown that just about every company I talked to had US-educated management from schools like Berkeley and Stanford.

At the SEG electronics market in Shenzhen (what Akihabara used to be).

Ten years ago, Akihabara was the place to be for the latest electronics and knick knacks and components. I’m convinced the new place to be is the SEG Electronics Market in Shenzhen (although to be fair I heard there is a competing market in Korea that’s supposedly even better–the Japanese test-market their stuff there even before they try it in Akihabara!).

...Chips that I couldn’t dream of buying in the US, reels of rare ceramic capacitors that I only dream about at night. My senses tingle, my head spins. I can’t supress a smirk of anticipation as I walk around the next corner, to see shops stacked floor to ceiling with probably a hundred million resistors and capacitors.

...Oh my god! Sony CCD and CMOS camera elements, I couldn’t buy those in the US if I pulled teeth out of the sales reps–and behind the counter, the guy sometimes has a datasheet–ask for it. A stack of Micrel regulator chips–over there, a Blackfin DSP chip for sale. The smell, the bustle, the hustle. It’s the ultimate electronic component flea market. Over here, a lady counting 256 Mbit DRAM chips…trays of 108 components, stacked twenty high, a row of perhaps 10 of them–she has the equivalent of Digikey’s entire stock of DRAM chips sitting right in front of me.

Reflections on capitalism in China: electronics manufacturing, basic labor economics, efficiency, corruption, etc. (See also Shenzhen diary.)

Minimum wage In Shenzhen, the minimum wage is about $0.60/hour. However, there is a very competitive labor market in China–there is a shortage of workers and mobility between factories is unimpaired by employment agreements. Therefore, employers must provide a very competitive benefits package for their employees, which typically includes dormitory housing, food, medical care, schooling, and day care; there are no retirement or unemployment benefits. ...workers have an 8-hour day, 5 days a week, and employers are required to pay 1.5x overtime and 2x on weekends. As far as I can tell, employers honor this. So in the end, these laborers earn a discretionary income of at least$100 per month, or $1200 per year. This is surprisingly comparable to the$2,075/yr discretionary income of US households that earn under $50,000 (link), which is probably the correct reference point for comparing minimum wage workers in both countries. I haven’t even adjusted for the cost of living difference between China and the US–but let’s just say 100 RMB goes a loooong way if you are just buying food... ...The fully-burdened rate of a worker in China is around$1.80 it seems–this is the rate that the employer pays once all the benefits (free food, housing, medical care, day care, etc.) are factored in. At these wages, laborers are cheaper than pick-and-place machines. In the US, you typically pay between $0.05-$0.25 per component placed on a PCB with a pick and place machine in low volume (100’s to 1000’s).

...In the end, I guess the trillion-dollar question is: will the Chinese economy surpass the US? I think, after being on the ground there and seeing where things are going, the answer is an unequivocal yes. While their current position is beneath the US, the first derivative is positive, the second derivative is also positive. Even if the economy were to start cooling down today (second derivative goes negative), I think they have enough inertia to soundly position themselves above the US for total GDP in about a decade or two. [OK, that's off a bit, but Bunnie is an engineer, not an economist. See here for what I think is a more realistic estimate.]

## Saturday, June 16, 2007

### Games gone wild

A great article in the Times Magazine on the current state of online gaming. Millions of people now spend a significant fraction of their waking hours immersed in a virtual gameworld. The numbers will only continue to increase, along with the richness and complexity of the worlds. Just as gaming drives the development of graphics cards and even CPU chips (see Sony's Cell processor), it may eventually come to drive AI development, as designers seek more realistic and complex interactions between game characters and players. (Other drivers of AI development: search, human-machine collaboration, cat and mouse security competition between malware/spam automatons and security systems, ...)

Why anthropic "typicality" arguments (all the rage in the current nutty world of theoretical physics) are meaningless without an assumption about the possibility of virtual worlds.

A slideshow of game players and their avators, like the one below.

NAME Jean-François de la Fage BORN 1979 OCCUPATION Journalist LOCATION Paris AVATAR NAME Dark Freeman AVATAR CREATED 2005 GAME PLAYED City of Heroes HOURS PER WEEK IN-GAME 21 CHARACTER TYPE Natural hero SPECIAL ABILITIES Invincibility

NYTimes Magazine: ...More than eight million people around the world play World of Warcraft — approximately one in every thousand on the planet — and whenever Li is logged on, thousands of other players are, too. They share the game’s vast, virtual world with him, converging in its towns to trade their loot or turning up from time to time in Li’s own wooded corner of it, looking for enemies to kill and coins to gather. Every World of Warcraft player needs those coins, and mostly for one reason: to pay for the virtual gear to fight the monsters to earn the points to reach the next level. And there are only two ways players can get as much of this virtual money as the game requires: they can spend hours collecting it or they can pay someone real money to do it for them.

...In 2001, Edward Castronova, an economist at the University of Indiana and at the time an EverQuest player, published a paper in which he documented the rate at which his fellow players accumulated virtual goods, then used the current R.M.T. prices [R.M.T. = real money trading] of those goods to calculate the total annual wealth generated by all that in-game activity. The figure he arrived at, $135 million, was roughly 25 times the size of EverQuest’s R.M.T. market at the time. Updated and more broadly applied, Castronova’s results suggest an aggregate gross domestic product for today’s virtual economies of anywhere from$7 billion to $12 billion, a range that puts the economic output of the online gamer population in the company of Bolivia’s, Albania’s and Nepal’s. Not quite the big time, no, but the implications are bigger, perhaps, than the numbers themselves. Castronova’s estimate of EverQuest’s G.D.P. showed that online games — even when there is no exchange of actual money — can produce actual wealth. And in doing so Castronova also showed that something curious has happened to the classic economic distinction between play and production: in certain corners of the world, it has melted away. Play has begun to do real work. ## Friday, June 15, 2007 ### Why is it dark at night? Many years ago someone made a short film, asking Harvard graduates at commencement to explain why we have seasons. It's Harvard, so each answer is eloquent and delivered with perfect confidence -- and every one is badly wrong. I wish I could find a copy of that film on YouTube! Motivated by the film, I've been asking the following question on exams given in courses on "conceptual physics" or physics for non-scientists. I've found that at both Berkeley and U Oregon, about 25% of students answer the question incorrectly (note this is at the end of the term, not the beginning). Remember, in our political system the votes of the people who get this question wrong count as much as yours or mine! Question: It is dark at night in Eugene/Berkeley because a) the night sky absorbs UV light b) the moon blocks out the sun c) the earth blocks out the sun d) none of the above Hint: draw a picture! Further comments: Surprisingly the success rate for Berkeley students was about the same as for Oregon students, despite the difference in admissions selectivity. Nearly all of the students in these classes are humanities or social science majors. Many are upperclassmen, even seniors, trying to fulfill a minimal science requirement in order to graduate. These days the main claim of humanities programs is that they "teach critical thinking" :-) See here for related discussion. ## Monday, June 11, 2007 ### It's all in da gene: muscles II "Horses ain't like people, man, they can't make themselves better than they're born. See, with a horse, it's all in the gene. It's the fucking gene that does the running. The horse has got absolutely nothing to do with it." --- Paulie (Eric Roberts) in The Pope of Greenwich Village. The Times has a nice article on the myostatin mutation in dogs, which leads to increased speed (one copy of mutant gene) and muscle mass ("bully whippet", with two copies). In the picture below, the bully whippet is on the right. We discussed this in an earlier post, including data on the frequency of the human myostatin mutation in different population groups. Two copies of a myostatin mutation in humans can lead to super babies. (40% more muscle mass than an ordinary kid -- one baby could do the iron cross at 5 months old!) ...Free of most of the ethical concerns — and practical difficulties — associated with the practice of eugenics in humans, dog breeders are seizing on new genetic research to exert dominion over the canine gene pool. Companies with names like Vetgen and Healthgene have begun offering dozens of DNA tests to tailor the way dogs look, improve their health and, perhaps soon, enhance their athletic performance. ...“We’re on the verge of a real radical shift in the way we apply genetics in our society,” said Mark Neff, associate director of the veterinary genetics laboratory at the University of California, Davis. “It’s better to be first confronted with some of these issues when they concern our pets than when they concern us.” ...A mutation similar to the one that makes some whippets faster also exists in humans: a sliver of genetic code that regulates muscle development, is missing. “It would be extremely interesting to do tests on the track finalists at the Olympics,” said Elaine Ostrander, the scientist at the National Institutes of Health who discovered that the fastest whippets had a single defective copy of the myostatin gene, while “bullies” had two. “But we wouldn’t know what to do with the information,” Ms. Ostrander said. “Are we going to segregate the athletes who have the mutation to run separately?” For the moment, it is whippet owners who find themselves on the edge of that particular bioethical frontier. It was not exactly news to breeders that speed is an inherited trait: whippets were developed in the late 1800s specifically for racing. But knowing that one of her dogs was sired by a carrier of the gene, said Jen Jensen, a whippet owner in Fair Oaks, Calif., makes its championships seem “less earned.” Ms. Jensen’s suggestion that a DNA test be required for all dogs and that the fastest ones without the mutation be judged and raced separately, however, has not gone over well. At a recent race here in southern New Jersey, some whippet owners wanted the mutation eliminated altogether, even if that meant fewer fast dogs. But as the dogs pounded after a lure at 35 miles per hour, several owners allowed that they would prefer a whippet with the gene for speed. “It’s more fun having fast dogs than slow dogs,” said Libby Kirchner, of Glassboro, N.J. ### Foo camp! I'm off to Foo Camp in a couple of weeks. Got my tent and sleeping bag all ready :-) You can see the social networking site set up for the 200-odd invitees here. (So far it's viewable without a password.) Most of the participants have a blog or personal web site, so if you're looking for some interesting thoughts on the future of technology, software, the internet, AI, media, venture capital, etc., you might want to have a look. (Note that true big shots like Larry Page of Google haven't filled in their profiles yet...) I'm one of many entrepreneurs but few professors and even fewer physicists invited. That's me -- ever the maverick outsider! ;-) ### Call your buddy New research shows that mutual fund managers get a higher return from investments in companies where one of the top execs is someone they went to school with. The effect is most pronounced with Harvard Business School alumni. Having worked with a lot of HBS guys, I am not surprised :-) This kind of study can only be done with mutual funds and public companies, where all the records are available. Whatever is going on here is probably going on even more in private equity and hedge funds, but isn't easy to quantify. Clearly, social networks like school ties give access to information that average investors don't have. As they say in poker, if you don't know who the sucker is at the table, it's probably you... PS I see a trading strategy here. Set up your own real-time database of fund pm and C-exec educational bios, and mirror all trades where school ties are indicated. Perhaps you can capture the 20% return indicated in the study! Mutual fund managers invest more money in companies that are run by people with whom they went to college or graduate school than in companies where they have no such connections, the study found. The investments involving school ties, on average, also do significantly better than other investments. ...“Something about these social networks is allowing portfolio managers to better predict the future returns of companies within the network,” said Lauren Cohen of Yale, another author. The study looked only at mutual funds, which are required to report their holdings and performance regularly. It did not examine hedge funds, which are investment pools for wealthy individuals and institutions; hedge funds do not have to disclose their holdings publicly. ...Their study, titled “The Small World of Investing,” examined 85 percent of the total assets under management from 1990 to 2006 and looked at different levels of university connections. In the weakest kind of connection, a fund manager and one of a company’s top three executives shared nothing more than an alma mater. They could have attended different schools within the university and have been on the campus decades apart. In the strongest connection, a fund manager and one of the top three executives attended the same school at the same university, and their time on campus overlapped. The most common shared school in the study, by far, was Harvard Business School. On average, investments in companies where there was no connection returned 11.7 percent a year before fees, according to the economists’ estimates. Investments in companies with the closest level of connection — when a fund manager attended school with an executive — returned 20.1 percent a year. As might be expected, investments with weaker connections had returns that fell somewhere in between, with returns of more than 11.7 percent and less than 20.1 percent. The most benign explanation for the pattern is simply that fund managers who attended school with executives have an easier time learning about the companies where those executives work. They are more likely to travel in similar social circles and may even remember their old classmate’s strengths and weaknesses. “The results are much more consistent with the story that you went to college with Person X and know they’re really smart,” said Steven N. Kaplan, a finance professor at the University of Chicago. “My guess is that the whispering is going on, too, but the question is the relative amounts.” Supporting Mr. Kaplan’s view, the paper does not offer clear evidence that the investments by the fund managers did unusually well in the weeks and months immediately after a stock was purchased. On the other hand, investing based on school ties seems to have become less popular recently, which could suggest that financial regulations passed after the demise of Enron cut down on the exchange of inside information. Last year, 7.1 percent of fund managers invested in at least one company that had a top executive with whom they had gone to school, down from 15 percent in 2002. The average during the 1990s was 11 percent. ## Saturday, June 09, 2007 ### All about the benjamins Sunday's Times Magazine is all about money and income inequality. You can track your income status over the years in inflation-adjusted 2006 dollars here. Makers of real money won't really fit on the chart (which only indicates 10th, 90th percentile and median figures), although most academics will :-( Even the figure below doesn't do justice to the money men, although it does indicate what's happening in our society. Statistical comment: by looking at wealth data you can see that most of the people in the top 1% of income in a given year are not there each year -- typically it's an upward fluctuation (like selling a business) that brings someone into the top 1%, although there are others like movie and sports stars, surgeons, top lawyers and financiers whose income is more consistent. This interview with the creator of Entourage is telling -- note what his next show is about. The Harvard-educated doctors he refers to are described here. Q: As the creator of the hit show “Entourage,” which follows four buddies from Queens on their adventures in Hollywood, you’ve become a leading chronicler of the American dream at its most nakedly materialistic. So perhaps it’s not surprising that you’re planning a new HBO series devoted to a group of hedge-fund guys in New York. Why do you find the subject of money so interesting? Sadly, right now, that’s the world people are aspiring to. You have Harvard-educated medical doctors who would rather work at an investment bank than try to cure cancer, and that’s the wish-fulfillment lifestyle that I play into. Q: The desire to make money is nothing new. No, but now people are looking for the home runs constantly. That’s the difference. People are seeing shortcuts and easy routes. Q: “Entourage” is itself a study in “shortcut culture,” to coin a phrase. The male characters don’t work very hard, except at playing video games and having casual sex and wanting to be famous. Is the show intended as a sendup of Hollywood excess? No. In fact when we first started, HBO would say satire, and I said this is not satire. This is scripted reality. I consider the show 100 percent realistic. I have friends who wake up in the morning and want a Bentley and they go get it. I find it funny. Meanwhile, in Nepal: ... Outside Africa, no country is poorer than Nepal. Its per capita income looks like a misprint:$270 a year. Sudan’s is more than twice as high. Nearly two-thirds of Nepalis lack electricity. Half the preschoolers are malnourished. To the list of recent woes add regicide — 10 royals slaughtered in 2001 by a suicidal prince — and a Maoist insurgency.

A few hours east of the city, a gravel road juts across a talc quarry, where the work would be disturbing enough even if the workers were not under five feet tall. Scores of young teenagers, barefoot and stunted, lug rocks from a lunar pit. The journey continues through a district capital flying Communist flags and ends, 12 hours after it began, above a forlorn canyon. Halfway down the cactus-lined slope, a destitute farmer named Gure Sarki recently bought four goats.

The story of Gure Sarki’s goats involves decades of thinking about foreign aid and the type of program often seen as modern practice at its best. Two years ago, an organizer appeared in the canyon to say that the Nepal government (with money from the World Bank) was making local grants for projects of poor villagers’ choosing. First villagers had to catalog their problems. With Sarki as chairman, Chaurmuni village made its list:

“Not able to eat for the whole year.”

“Not able to send children to school.”

“Lack of proper feed and fodder for the livestock.”

“Landslide and flood.”

“Not able to get the trust of the moneylender.”

“Insecurity and danger.”

How, again, does one justify keeping economic migrants out of rich countries? Oh yes, by weighting the well being of our fellow citizens an order of magnitude more than that of others born elsewhere.

## Tuesday, June 05, 2007

### China's Silicon Valley

Creative destruction at work in Beijing. I was in Z-Park (short for Zhongguancun Science Park) briefly a few years ago, but didn't have time to look around in any systematic way.

Businessweek: ... Z-Park, China's homegrown Silicon Valley, is the jewel in his innovation crown. The biggest and oldest of the 53 national high-tech zones in China, Z-Park has become an important port of call for global corporations looking both to develop R&D for all sectors and to get a foothold in the Chinese market. For China watchers, it also serves as a window into how the evolving giant interacts with global market and social forces—and how it is turning to private resources to build its vaunted innovation economy.

Z-Park is made up of a group of seven parks, covering an area of about 100 square kilometers at the northwest edge of Beijing, close to the city's internationally esteemed educational institutions such as Tsinghua University, the University of Beijing, and the Chinese Academy of Science (CAS). It was founded in 1980, when Chen Chuxian, a researcher at CAS, returned from a trip to Silicon Valley. He opened the Advanced Technology Service Assn., the first privately funded, civilian-run, scientific and technological consulting firm in China. Soon after, other scientists came to the district, attracted by the support afforded by both CAS and the central government. A virtuous circle began, with new ventures spinning off from Chinese universities. Foreign companies also set up shop.

... by 1996, global corporations such as IBM (IBM), Sun (SUNW), Nokia (NOK), and Microsoft (MSFT) had all established R&D centers there. Non-tech companies such as P&G were also lured, while homegrown (often university-grown) companies such as Lenovo (originally known as Legend), Founder, and UFSoft all had their start in Z-Park.

Today, some 18,000 companies operate in Z-Park, including more than 1,500 foreign firms. In 2006, Z-Park generated $85.75 billion in revenues and$12.6 billion in exports. From January to November of last year, the IT industry within Z-Park generated $45 billion in revenues, including$5.8 billion in technology income, $16.8 billion from new products sales, and$7.29 billion in exports.

Companies such as Vimicro (VIMC) and ARCA are making real strides in mobile technology and integrated chip design. Vimicro's intellectual property, for instance, entered the world market in 2003, when the company's Starlight 4 mobile-phone color message processing chip was adopted by Sprint. While these signs are rarely visible to the world at large, partly because many of the R&D centers' efforts are focused on designing products for the domestic market, the result is there are currently 73 Z-Park companies listed on the Shenzhen, Shanghai, Hong Kong, and NASDAQ stock exchanges.

Former Expats Are Today's Recruits

Z-Park has aggressive growth plans. Its representatives are constantly visiting high-tech areas such as Silicon Valley and Research Triangle Park in North Carolina in an effort to poach companies, experts, and investors. Offering benefits such as reasonable rent, travel perks, high wages, and easier startup conditions, they have lured tens of thousands of Chinese expats away from foreign high-tech centers.

Jennifer Pan, for instance, graduated in computer science from the University of Beijing, obtained her EMBA, and worked in Houston with Compaq and BMC Software (BMC) for eight years before returning to China to work with Z-Park. In 2005, she founded ChinaSense, a company that provides business services to faculties and students of foreign universities. Pan saw China as a better place to pursue her "American dream" and Z-Park as the gateway to success. As she puts it, "There are many issues in China, but there is also an innovation imperative here. China is like a first-time parent with a teenager that is full of energy, eager to learn, and full of hope. It is doing the best it can, while the U.S. is not. It is an exciting place to be."

In nurturing Z-Park, the government has played a key role in designing an environment in which high-tech ventures and high-growth enterprises can flourish (or die). Entry is regulated, driven by a set of conditions stipulating that 50% of revenues must be from high-tech projects, while R&D expenditure cannot be less than 3% of total revenue. In addition, at least 20% of employees must have a college degree.

A Laboratory for Policy

With its focus on high-growth software and integrated circuit design companies, Z-Park provides appealing financing arrangements through cooperation with commercial Chinese lending institutions. Small and midsize companies such as Analogix, a venture capital-funded Silicon Valley semiconductor company, arrived in 2006, attracted by financial sources such as the Z-Park Development Fund or the Beifang Microelectronic Industry Development Fund.

Tax incentives for those companies headquartered in the park have also been highly effective: The income tax rate is 15% for foreign-funded high-tech enterprises (10% if the export output exceeds 40% of the gross output). "High-tech" enterprises are exempt for the first three years, and the rate is reduced by 50% for the following three years.

A set of very open regulations—translated by law firm Perkins Coie as "Anything not prohibited is allowed"—allow for a fuzzy definition of the business scope of an enterprise within Z-Park, and these encourage startup and emerging venture capital, in contrast with general business regulations in China, which are markedly more prohibitive.

...

Plan is "Not a Reality Yet"

The park and its ilk are part of a systematic national effort to bring an educated, creative workforce to China, while waiting for the new generation of creative thinkers to come of age. It's a tough balancing act, and as professor Henry Rowen, co-director of the Stanford Project on Regions of Innovation and Entrepreneurship, and co-author of Making IT: The Rise of Asia in High Tech, points out, "The nation's self-reliance plan is not a reality yet." According to him, despite the impressive figures, "several thousands of Z-Park's companies are not real," meaning that it includes many micro- and small-business enterprises with doubtful futures.

He's not wrong: A small or medium enterprise crashes and burns within Z-Park every nine minutes. He also points out that "all technologies have come from the outside." But Rowen also argues that the government is moving in the right direction, toward a role of indirect support, and that there are signs of innovation in R&D, business models, and services. The U.S. and other nations, he believes, will face new opportunities for collaboration—as well as significant challenges in competing with the rise of an innovative China.

For its part, China remains focused on building an innovation infrastructure to encourage the development of formidable companies like global telecommunications giant Huawei, which generated \$8.5 billion in sales in 2006 with a 10% spend on R&D. As Jan Gronski, general manager of Cisco's (CSCO) China R&D Center, puts it, Huawei "did not get there just by copying." Others would be ill-advised to underestimate his assessment.

## Monday, June 04, 2007

### Print on demand

Soon we'll have every book in the world stored digitally -- probably taking up as little as a few cubic centimeters, if current storage trends continue. The technology for print on demand is also coming along, as are business models.

NYTimes on this year's BookExpo America: ...In a pavilion outside the main exhibit hall Jason Epstein, the former editorial director of Random House and the creator of the Anchor Books paperback imprint, and Dane Neller, founders of OnDemandBooks.com, demonstrated their Espresso Book Machine, which can print a small paperback book on site in less than five minutes. “This could replace the entire supply chain that has been in existence since Gutenberg,” Mr. Epstein said.

Chris Morrow, whose parents founded Northshire Bookstore in Manchester Center, Vt., three decades ago, said he would be installing one of the machines. He said he planned to print local histories and Northshire-brand titles from the public domain, like “Middlemarch” or “Moby-Dick.”

“There are lots of challenges in bricks-and-mortar book selling, and I see this as a way of expanding our business,” Mr. Morrow said.