AI Superpowers Series: Copying
I’m writing a series in response to AI Superpowers: China, Silicon Valley, and the New World Order by Kai-Fu Lee. For me, this is more like having a dialogue with a book rather than reviewing it. I don’t really want to write a whole review, so let me do a nice overview real fast here.
Lee does a great breakdown of the story of China’s tech sector of the past few decades. He focuses mostly on the development of the Chinese tech sector because that is most likely unknown to his Euro-American audience. All the world already worships Apple and Microsoft, let them hear the praises of somebody else for a change. From copying to smart implementation to modern groundbreaking tech companies, he charts the whole course of the development of China’s tech sector. Then he breaks down what AI is, what we can realistically expect from it, America’s and China’s relative strengths, and how society will have to respond to AI.
In this post, I want to respond to his section about the role of copying in China’s tech development.
Copy not Counterfeit
Lee tells the story of Chinese tech entrepreneur Wang Xing, who copied a number of American social media platforms in the 2000s and ported them to the Chinese internet. Lots of failures but Wang eventually got somewhere, laying the foundations for the company Meituan, currently (2020) a giant in Chinese food services, bike shares, ticket purchasing, and more. I’ve used one or other of the Meituan services more times than I can count, living in China.
Copying an internet platform, while tacky, is not the same thing as counterfeiting. Internet platforms are much more human-driven than handbags. They require communities to build around their brand and the product fails if that community doesn’t work. If I tell my friends to sign up for Facebook and they get stuck on “Fay’s Book” and “Phasebook,” there’s a counterfeiting problem. My American friends won’t get hung up on “Book of Faces in Ruritania,” though it might cut into Facebook’s intended global coverage. A fake Gucci handbag might hurt my trust for that brand. A handbag labeled “Goochee” doesn’t have to deceive anyone to carry some pencils, a jar of olives, and some shag carpet samples. (You ask the owner about why that’s what’s in there. Not my handbag.)
Many knockoffs aren’t even trying to deceive you. I’ve seen a picture of “Crust” toothpaste. I’ve read a story about a guy in northern Iraq who wants to start a McDonald’s franchise but runs a “MaDonal” restaurant until he can get permission to open a proper franchise. In the one case, I guess they can make OK toothpaste. In the other case, the guy wanted to make McDonald’s-style hamburgers, he wanted to start now, and sanctions against Iraq were working against him. Maybe the “close but not quite” is a way of letting you know what quality they aspire to, but they’re not counterfeiting.
The LEGO company has made a huge deal out of crushing counterfeits because the have a specific brand reputation issue: they make their stuff to a certain level of quality. Parents literally trust LEGO products in the mouths of their children (I know, best that it be Duplo bricks). I can attest to the safety of LEGO products. I’ve put them in my own mouth many times. LEGO has a brand to protect, but they are in a niche industry, whereas technology firms own intellectual property (IP) with exponential secondary and tertiary effects for a nation’s economy.
The New Mercantilism
I remember high school history: Britain made all these restrictions on American industry to protect nascent British manufacturing. The idea behind mercantilism was that the world economy was a pie yea big and a bigger piece for you absolutely means a smaller piece for me. Adam Smith comes along with his invisible hand and shows that you can just bake a bigger pie, 1776 and bye bye Britain, it’s America and freedom and pull yourself up by your own bootstraps, go west young entrepreneur.
Before we get back to China and AI, I want to make clear that in American industrial development, we Americans extensively bribed, copied, or stole whatever IP we could get our hands on because we wanted the means to be economically independent of Europe. A lot of IP battles today remind me of mercantilism. While it’s true that brand reputation and safety standards are important, economic self-sufficiency and national security don’t wait for copyright lawyers – but we’ll get to that. So somebody invented a wonderful technology: great! After a certain point, the guy who invented the wheel had to let it go.
Lee talks about the Silicon Valley mindset as “you can change the world!” and so these super smart scientists work to invent technologies to change the world. If you have an ideology, you don’t want false prophets to be able to mess it up. In Asia, he says there is much more of a market orientation. You don’t buy the idea that this street food will change the world, you buy that selling it from a cart by the bus stop saves you some time. Literally. That’s where you buy it. The market in Asia is fast, innovative, and cutthroat. Copying is fast and it gets you right up to where the essential service and delivery questions are. Asian companies have to move faster than American and European companies want to go. The idea of brand being a thing is taking off in China and the rest of Asia, but this doesn’t mean they’re sitting around and waiting for European and North American companies to save them.
Copying is a vital part of the world economy. Facebook has over a billion users, but logging into the internet does not require Facebook to put a product in your hand. Siemens has great train technology, but would a single company like to stake its brand reputation on establishing the entire rail infrastructure of a major Asian continental power? Better to be known as the ones worth stealing from than the ones who couldn’t fulfill promises to do the impossible! Capitalism is all about market forces getting things done at the right price. Copying has a very low price and it lets the world develop, using your ideas, and you don’t have to be the one to go out there to do it.
National Security and Self-Sufficiency
Copying unlocks human potential and jumpstarts development. I never “earned” the right to use the wheel, and programmers in West Africa never “earned” the rank permitting them to use semiconductors. Everybody should get a general education in what technology is all about, but we need an app NOW, so can you code it for us? Countries, like individuals, don’t “earn” the right to have technology. They either own it or they don’t. Having the technology means power; not having it means weakness. In China (and in every country), technological and industrial self-sufficiency is a matter of national welfare.
Every country wants to be able to guarantee its own national security and not rely on the promises of others. If they have to grind their seed corn to pay for permission to use certain IP, they don’t have anything left to continue investing. Technology opens superiority gaps of magnitudes, not mere percentages. They’re not 5% faster than you, they’re 5x faster than you. Market forces don’t just get this done. What’s more, IP is about fairness and just compensation, not about economic development and national security.
A lot of IP protection is by moral suasion rather than legal action. No country wants to have an IP theft reputation, for having shoddy copycat industries. However, morality does not amass interest like money does. Yon peasant there and this here knight in shining armor have the same moral capacities, one just has a bigger budget for specialty varieties of sin. Prestige, however, does amass interest. Yon peasant only has generic “farmers are good, down to earth people” prestige, whereas this here knight can get away with a reputation for murder, theft, and rapine because of his prestige. Nations have prestige as well, and larger ones generate it the fastest.
Copying Gets the Job Done
IP owners really have to figure out how to spin off their stuff so that it can be more widely used. Yes, yes, you get to be repaid for your blood, sweat, and tears for inventing the thing. After a certain point, IP protection becomes a legal weapon rather than a safeguard for inventor compensation. It gets even more ridiculous when companies own IP and not individuals, but then there are reasons for companies to exist and it all gets kinda complicated. Copying gets the job done because it lets people get to work making things rather than paying for the privilege to make things.
The real pivotal point is in implementation vs development. Implementation (making things) is one of the areas, according to Kai-Fu Lee, in which China excels. America has a slight advantage for now in developing technology. One of America’s secret advantages has always been that the brain drain spout for many countries comes out in America. Development is a luxury-expensive necessity. My next post will focus more on implementation vs development.
For now, I will say that copying is really not shameful. Especially when something like software is the closest that we can get to having a physically incarnated idea and computer chips don’t compete on the same visceral level of appeal as a soft suede handbag, copying is closer to chefs sharing recipes than Chef A putting on a moustache and pretending to be the slightly more famous Chef B. I’ll work out some of the copying complications in “implementation vs development.”
