Sat - June 25, 2005

The Animation of Old Age - Howl's Moving Castle



After much anticipation, I saw Howl's Moving Castle last weekend. It was kind of disappointing - the storyline was convulated, and the characters not all that convincing. But it was still nice to see the master filmmakter at work.

Posted at 11:03 PM    

What Level Playing Field? On CNOOC's Unocal Bid



CNOOC, a Chinese state-owned oil company, has made an unsolicted all-cash bid for Unocal, a California-based oil and gas concern. Now that some in Congress has asked the Bush administration to block the deal, Corporate America has come to China's vociferous defense. The Wall Street Journal reports that the CEO of Bank of America said the US should not interfere with CNOOC's bid, because investment can't be a "one-way street," and his company desperately wants to invest in China. The same sentiment is echoed by American corproate types in other news stories.

The "level playing field" argument is disingenous. Its proponents say that the U.S. must permit Chinese investment so there can be a level playing field in foreign direct investment in both countries. It's true that there is currently no level playing field, but the problem is Chinese, not American, protectionism. Suppose Unocal tries to bid for CNOOC. Or IBM for Lenovo (Lenovo is the Chinese firm that just bought IBM's PC division). The Chinese will laugh in their faces. Large Chinese companies are nearly all state-owned, and absolutely not open to foreign control. In addition, the Chinese government does not permit majority foreign control, period, in many sectors, even if the investment does not involve the takeover of a local company. Foreign investors in China, whether in automobiles or financial services, must accept only minority stakes in partnership with a majority Chinese-state-owned firm. Take Bank of America - it recently acquired a non-controlling 9% stake in a Chinese state bank. It probably would have liked majority control, but the Chinese would not consider it. Earlier this year, Goldman Sachs finally got a foothold in investment banking in China, but only as a minority partner in a joint venture with a state bank. In contrast, the US has no legal barriers to foreign control in most sectors; foreign investment usually goes though unless the government intervenes, whereas in China foreign investment always requires government permission. American CEOs are perfectly aware of this. Yet they make these silly "level playing field" arguments to support Chinese investment access to the US, essentially asking the US to turn the other cheek. What gives?

The reaction to CNOOC illustrates the extent to which China has succesfully held Corporate America - and foreign investors in general - hostage by controlling access to its market. Everyone wants access to China, and because China is an authoritarian state unconstrained by law, it wields an axe over foreign investors, turning them into its spokesmen and political footsoldiers. The most dramatic example of this is Hong Kong, where the business class has become the most important defenders of Chinese authoritarianismin the former British colony. But the same logic applies to big Western and Japanese corporations. Bank of America is positively head-over-heels giddy over its pathetic 9% stake in a Chinese venture, which it obtained only after protracted negotiations and relationship building with the Chinese government. For American corporations like BofA to maintain and broaden its access to the Chinese market it must voice support for Chinese political and strategic interests in the American political arena. China, it is sometimes easy to forget in all the euphoric press coverage over its fast growing economy, is not a democracy. George Soros can campaign against the Bush Administration and still do perfectly good business in America. No such freedom exists in China. To survive there, foreign investors must tow the government's line. This is the downside of growing economic integration with China. America's largest trading partners have traditionally been democracies (Europe, Canada, Japan), so these types of problems have not been at issue. But now policymakers must see that excessive economic entaglement with a Communist dictatorship could produce political and business dynamics that threaten liberty at home.

Posted at 10:47 PM    

Sun - June 12, 2005

Release Notes



Okay, I am restarting the blog for real now. I promise. Really.

Posted at 07:36 PM    

MacIntel: The Empire Strikes Back



I think the Intel move is a smart move for Apple. The risks are enormous - if developers do not get fully on board, the future of the entire platform will be in doubt. And Mac sales could dry up over the next year as consumers hold off purchases in anticipation of the new Intel machines. Nonetheless, if executed cleanly, the move could significantly increase Apple's market share in the long run.

The G5 is a great chip, but it it not acceptable if IBM cannot deliever a laptop edition that can run without liquid cooling or a heat sink the size of a lunch box. The PowerBook line is becoming a joke; a $600 Mac Mini outperforms a $2000 PowerBook in many tasks. When Intel delivers its dual-core mobile chip (codenamed "Yonah") next year, Wintel laptops will simply blow away G4 PowerBooks in hardware performance. The G5, derived from server chips, was not designed with mobile computing in mind. If Apple wants to hold onto the laptop market, which it must, PowerPC has to go.

Weening the Mac platform off PowerPC also permits Apple greater product flexibility at much lower technological risk in the long run. IBM has few incentives to invest heavily in cpu development for personal computers. Server and game console cpus, which persent very different engineering needs from that of personal computers, account for most of IBM's microprocessor revenue. This means that Apple will have to bear nearly all the cost for cpu development for personal computing, and also bear all the risk of technological failure. Since Apple accounts for less than 3% of IBM's cpu business, Big Blue will suffer little harm from any problems with Apple microprocessors. For Apple, however, the fate of the whole company rides on the cpu roadmap. Dependence on IBM is, in short, dangerous for the Mac community.

In contrast, Intel lives and dies with personal computing. It will invest heavily in pc cpu development on its own, without requiring investments and subsidies from customers. Moreover, if Intel falters, Apple will have the option of turning to AMD, which also produces for the x86 platform. Life on x86 will be less risky, and technologically more promising, than holding out on PowerPC.

Many in the Mac community appear to be shell-shocked by the once-unthinkable embrace of Intel. I was myself deeply apprehensive in the days leading up to the June 6 announcement. The move to x86 makes the Mac less different and less special. As John Siracusa pointed out in his elegant and heart-felt elegy of Mac-on-PowerPC, this ensures that Macs will always be no worse than PCs in hardware, but at the cost of killing any hope that Macs can be better. The G5, developed from some of the world's most powerful server cpus, had given Mac users hope that they might at last attain hardware superiority over pcs. This hope was further fueled by the high-profile deployment of Apple G5 servers in supercomputing clusters. But all that power does most Mac users little good, if the G5 can't be deployed in notebooks or small desktops like the Mac Mini, which uses many laptop components.

Standardizing on x86 ensures that Mac users will have access to the best hardware the pc world has to offer. Given the size of the pc world, this is not something to sneer at. I will eagerly await the arrival of the first Intel Macs.

Posted at 07:19 PM    

©