The Data Disease (and Cure)

The Data Disease (and Cure)

If data is the new oil, free countries can pull ahead by accumulating less of it.

Would you like to live in Saudi Arabia? How about Russia or Venezuela? Countries rich in natural resources tend to suffer from bad government, low productivity, and sluggish growth.

Such economies tend to become over-reliant on a single industry; their institutions fail to develop; and they have low incentive to innovate or invest in education and general welfare. Like a man who won the lottery, countries can be "cursed" with plenty of oil, diamonds, or other minerals and end up profligate, rudderless, and corrupt.

Even developed economies can suffer when they are suddenly blessed with new resources. In 1977, The Economist coined the term "Dutch Disease" to describe a recent gas discovery’s negative effect on the Dutch economy. The term described a more complex form of the resource curse that impacted foreign-exchange rates and made it difficult for the Netherlands to export anything other than gas.

Fossil fuels are no longer in vogue, but something else is. In 2017, The Economist declared that "the world’s most valuable resource is no longer oil, but data." Data is a resource that fuels artificial intelligence systems, which, in turn, are expected to constitute a considerable part of global GDP within a few years.

In 2019, Harvard Business Review started ranking countries based on the amount of internet data they generate — redefining GDP as "gross data product.” China ranked third, trailing the U.S. and United Kingdom. The ranking was partly adjusted on a per-capita basis; in absolute terms, China probably generated and collected more data than any other country.

But China's data trove extends well beyond simple internet traffic. The country has the world's most extensive domestic surveillance network. Per The New York Times:

"authorities can scan your phones, track your face and find out when you leave your home. One of the world’s biggest spying networks is aimed at regular people, and nobody can stop it."

China's surveillance technology is state of the art. It also has enough cheap labor to economically "train" algorithms and augment tasks that machines cannot yet do on their own. At the same time, Chinese citizens are not protected by the privacy laws, constitutional rights, and cultural norms that limit or slow-down the amount of data collected by other countries.

As a result, China is sitting on a massive pile of precious data that is growing at an exponential rate. It "produces" this data faster and at a lower cost than any other major country. This seems like a great advantage, but could it end up eroding its institutions and dampening its economic fortunes?

Perhaps.

Some studies already pointed out that while China collects plenty of data, it is still far behind in terms of actual AI research and even farther behind in terms of its ability to attract the world's best AI scientists. China even struggles to keep its own researchers from leaving the country. Almost two-thirds of the world's leading researchers live and work in the U.S., while only 11% of them are in China.  

The world's best and brightest still want to come to the U.S. (And the U.K., Canada, Australia). Even Chinese researchers end up in the U.S. at a higher rate than anywhere else. China's illiberal regime and immigration policy make it unattractive and difficult to move to. The same system that enables so much data to be collected makes it much harder to attract the researchers who can make the best use of this data.

Perhaps more importantly, the constraints that researchers and companies face in free societies can serve as a source of innovation in themselves. As Caleb Watney points out, the U.S. can offset China's advantage by investing in artificial intelligence techniques that reduce the importance of data:

...things like simulated learning, where basically instead of taking a real-world data set, you're almost just simulating an environment, can end up offsetting the need for real-world data or something like one-shot learning... There are lots of interesting AI techniques that are being developed in that way too. And so the United States, [should be] be thinking about the terrain of AI landscape and thinking about how can we strategically invest in techniques that make it more likely that firms in a US context can be successful without compromising on privacy laws in the same way that China does.

A 2020 report by Tim Hwang from the Center for Security and Emerging Technology at Georgetown University dives more deeply into how democracies can "leverage the malleability of artificial intelligence to offset authoritarians' structural advantages in engineering and deploying AI."

So far, the U.S. seems on track to adopt China's censorship and surveillance model rather than to forge its own path. But we should remember that an alternative path still exists.

Addendum

In March 2020 I spoke to Olivier Knox on SiriusXM radio about China and the West's diverging response to COVID. I was worried about an ultimate convergence:

"because we are not responding [to COVID-19] early enough and with enough seriousness, I'm concerned that we'll have to default to all sorts of more aggressive measures down the line — which might be very soon..."

I mentioned a few things that China was  doing, including:

  • censoring group-chat conversations about the virus,
  • encouraging citizens to snitch on people who aren't wearing masks or seem to be sick, and
  • using facial recognition to identify and shame or punish people who congregate outdoors.

These activities seemed extreme at the time, but they have all come to pass in the U.S. due to America's botched COVID response.

During the same interview, I suggested that instead of mimicking China's response, America should rely on its unique powers — "the free flow of information, the power of American industry develop a vaccine, and [the ability to] allow people to work from home." These, too, seemed far-fetched at the time — only eleven months ago.

My optimism about the vaccine and the viability of remote work was on point. My optimism about privacy and human rights, less so. But there's still time to do better.

Have a great weekend.

P.S.

🎧 Rethinking Real Estate is now available on Audible. Listen, rate, and leave a review.

🇰🇷 Rethinking Real Estate is now available in Korean! Get it here.

🇮🇱  Do you speak Hebrew? I had a blast discussing the future of work, cities, and creativity on the Urbanists Podcast.

Comments