I can think of a few reasons off the top of my head, all of which are at least slightly interwoven:
Chinese companies and universities are cranking out a lot of AI research and products—sometimes beating U.S. institutions to the punch in certain areas—and hiring from these universities will help improve Google’s internal research efforts. Just like Chinese internet giants are setting up U.S. offices to hire talent here, Google needs to hire in mainland China if it wants to acquire talent that might otherwise end up at Baidu, Tencent or Alibaba.
- Hiring in the backyard of Baidu, Tencent and Alibaba suggests Google is not yet willing to cede the AI-powered economy in China to those companies. It probably has a lot of catching up to do in terms of market penetration and actually understanding Chinese consumers, but hiring locally will help with the latter. Google’s sheer size, combined with some smart acquisitions, could help with the former.
- International law is becoming murkier with respect to immigration and data sovereignty. Should U.S. immigration law become more restrictive, an established AI team in Beijing will serve as a natural home base for Chinese workers who might have otherwise ended up in the U.S. Further, new Chinese laws about data protection went into effect on June 1, and it’s possible Google has deemed the best/only path to compliance with regard to AI—which requires lots of data—is to carry out China-centric research and build China-centric products in China.
I would also point out that a dedicated Google AI outpost in Beijing is further proof of my previous argument that claims about the U.S. “falling behind” in AI miss the point
, at least with regard to commercial AI. Advances in AI today are coming from universities across the globe and from companies with offices across the globe. While data is still a precious, guarded commodity
, research techniques are shared fairly openly. Multi-national companies invest in AI where there’s adequate talent and market demand, but advances made in one geography aren’t somehow stuck there.
There’s a lot of other good content in this issue that I didn’t have time to comment on or write clearer headlines for (in part because I spent a good chunk of my day recording podcast interviews / battling Skype), but definitely scroll through and take a look. Here are a handful that are definitely worth reading: