There has been a lot of talk in the press lately about China’s rising technology capabilities. There was a recent New York Times exploration of China’s growing ambitions in AI, as well as countless blogosphere posts on similar topics. The underlying theme of most of this holds that the US risks falling behind China in the AI ‘arms race’.
For those of us who have been studying and visiting China for a long time it all seems very familiar. Over the past forty years the American media has run through regular cycles of China optimism and pessimism. In recent years, this has careened between China as a threat to the US economy and China as an opportunity for US companies. The latest round of China fears reminds us a lot of the rise of Japan in the 1980’s, when that country was seen as a growing threat to the US economy. We are not going to make predictions, but in the case of Japan those fears proved largely misplaced.
We think discussions about China require a much more nuanced approach. Reality sits at neither extreme. In fact the idea of talking about China as some monolithic entity is itself misleading. There is a lot going on in China today, and as with any economy there are many conflicting positives and negatives.
In this post, we want to take a closer look at a few elements of China’s technology industry to drive this point home. A quick warning upfront, due to various NDAs we are going to avoid some details and names, but we do not think this will detract from our central message.
For a long time, in many tech sectors, Chinese competition was viewed with a high degree of skepticism. Chinese hardware is still often dismissed as ‘copycats’ or offering low quality. While there are plenty of examples of IP infringement, in general, we think these criticisms are way off the mark. Chinese hardware companies have immense capabilities, and that has been made painfully clear in markets ranging from mobile phones to core networking equipment.
Another once-frequent criticism holds that Chinese companies cannot ‘do’ software. Our long-time counter to this was that few US companies can ‘do’ software very well either (looking at you Detroit). Now, the press seems to be in the opposite position with China seemingly poised to spring permanently ahead. Let’s dissect that a bit.
First, it is clear that the Chinese government has placed a big emphasis on a national policy for advancing their technology industry. They are spending immense amounts on semiconductors. And there is a clear focus now on Artificial Intelligence (AI). But what exactly is a national policy, and does that help? Old China watchers will say anytime they here about grand, sweeping plans in China, they assume it is some sort of real estate adventure. This may be too cynical, but at the same time it is important to understand that even though the central government in China may launch bold initiatives, these often do not translate into tangible action on the ground.
One of the most frustrating memes to emerge during the 2008 Financial Crisis was the view that China weathered the storm so much better than Western economies because of their centrally planned economy. We heard someone say that on CNBC of all places. We say this was frustrating because it seemed to so casually sweep aside centuries of evidence that free market economies have permanent advantages over centrally planned ones. The same holds true with AI and software. US companies are not ignoring AI, and tech companies here have a pretty good track record of delivering innovation.
That being said, it is clear that some Chinese companies are doing well with AI. They have built large teams, both in China and in the US, and are making real contributions to the state of the art. As we noted at the outset, this is a nuanced process. So while putting a national policy out there may or may not matter in the marketplace, there are some clear areas where Chinese companies are at least competitive with US peers.
One of the big problems with centrally directed economies is that bureaucrats in the Center do not always make the right decisions. The Soviet Union spent a good part of the 20th Century building a world-class 19th Century economy – dams, trains and steel. There is a hint of this today in China as well. If you talk to actual AI practitioners, they tend to drain the hype from AI. The press tends to conflate today’s AI with Hal 2000 thinking robots. AI is important, but we are still far from all-powerful AI systems ruling over us. Benedict Evans had a great tweet suggesting that whenever you see ‘AI’ in the press, substitute the word ‘software’. Seen in this light, China’s AI efforts are part of a natural effort for any advanced economy to develop software.
If we look more closely at China’s software industry we once again have a very nuanced picture. On the one hand, there are some promising companies and initiatives in place. There are many important open source software projects out there which Chinese participants are important, active contributors. On the other hand, if you take a close look at the software stack of many leading Chinese companies, you are likely to find a much cloudier outlook. There pockets of brilliance and there are large swathes of solid competence and there are dark corners of just getting by.
However, one thing is clear to us. For many years we have been telling Western companies to not underestimate the capabilities of Chinese companies. But in the past few years, the conversation has shifted noticeably. Occasionally, we now tell US companies, especially emerging Internet start-ups that they should look to China for inspiration. There is a growing list of companies and products in China that are not available in the US, but should be. Visitors to China often talk about the ease of mobile payment systems as one area where China is clearly ahead. (That example is somewhat marred by the fact that China had no electronic payment systems in place prior to the advent of mobile.) And there are many more examples which we are happy to discuss offline.
It is important for us to point out that we are not criticizing or hyping Chinese tech capabilities. Reality is much more complicated, with many positives and many negatives. There are areas where China lags and areas where China excels. For companies looking to engage with or compete against Chinese companies, you need to move beyond the headlines and the glib analysis of “I went to China for 48 hours and learned everything”. Facts on the ground are incredibly fascinating and not at all obvious from 6,000 miles away.