First things first
1. The newsletter will be published sporadically over the next few weeks because of travel. I'll try to do at least once a week, though.
2. I'm breaking my own rule by writing something strictly for this newsletter, mostly because I'm trying to tackle this subject in a longer post and am struggling to capture its immensity without doing 9,000 words. So you all get some half-baked thoughts on it ;-)
The subject is China, and the misrepresented/misunderstood narrative over the country's artificial intelligence prowess and strategy. However, I think it's actually a bigger story than just AI, and has to do technology and open source software on a broader scale. And all this without getting into tariffs and other trade-war issues.
I've linked to a handful of stories on this earlier this week (and, below, from the past year or so), and here are four more that got me thinking and that highlight just how complex this story really is:
- Intel, Alibaba launch a joint edge computing platform (ZDNet)
- Alibaba’s “honey badger” AI chip company hopes to take on bigger beasts (MIT Tech Review)
- Former head of Google China foresees an AI crisis—and proposes a solution (IEEE Spectrum)
- Former Google CEO predicts the internet will split in two — and one part will be led by China (CNBC)
It's not a matter of any country winning or losing on technology, because open source software and the international nature of most large technology companies has left national borders more or less obsolete. When everyone has access to the same products, techniques and code bases, it really becomes a matter of what data governments and companies have, and how they're able/willing to use it. Those applications will be some combination of innovative, unethical, scary, beneficial, etc., etc.
Neither China nor the United States (their governments and companies) can afford to work in a vacuum, because then they're missing out on all great work going on elsewhere, particularly in open source communities. Some outlets presented China's push for a global AI community as a softening of its stance on building out its overall AI-domination strategy, but that ignores the nature of how technology is built today. Especially at government and web-giant scale, if you're not building off the work of others in areas like distributed systems, big data, AI, etc., then you're actually putting yourself at a disadvantage.
One might argue that it's a one-way flow of benefit because so many open source projects are first developed in the United States, but the massive scale of Chinese web companies makes them large adopters of many technologies and often some of the largest users. When they push technologies to their scalability limits and contribute back, everyone wins.
Getting back to the competition part: If, as Alphabet's Eric Schmidt suggests, the internet bifurcates into two -- one led by China and one led by the U.S. -- is that inherently bad? Or if Chinese society adopts AI more readily into their everyday lives than U.S. society does -- in part because of the differences in Chinese culture and urban life -- has it won anything? Is it losing if AWS doesn't dominate every country, but must cede portions of Asia and Africa to Alibaba Cloud?
I certainly don't know the answers; I'm just trying to wrap my head around the enormity of it all. We can draw some clear lines around surveillance, cybercrime, actual warfare, etc., but beyond that it starts getting foggy. For better or worse, technology is open. How anyone decides to use it seems less like a competition and more like a natural evolution influenced by who, what and where it's happening.
And if it must be a competition, the only way to victory is by having better ideas and execution. Trying to cut the other guy off at the knees seems like a fruitless effort.
Some more thoughts on this and interviews with people doing AI in China:
- China vs. America is an AI red herring
- The convergence of cloud computing, Alibaba, Google and China
- Malong Technologies CTO on commercializing computer vision
- Skymind CTO on the realities of selling AI to enterprises and operating in Asia
- Baidu’s Andrew Ng on the economics of AI and what tech companies owe the labor force