Weaknesses in Kai-fu Lee’s Argument

One thing that I want to mention about Kai-fu Lee’s AI Superpowers is that the examples he gives of the supposedly amazing AI breakthroughs in China are surprisingly few and unimpressive. He keeps repeating the same three things because that’s all there is.

This isn’t a criticism of China. For now, at least, AI brings no discernible benefit or “convenience,” as the evangelists call it, to the users. All it does is facilitate spying on them. Kai-fu Lees of this world benefit from it but the users don’t. “Smart homes” aren’t more convenient, recommendation algorithms aren’t more precise, and the destruction of computers in favor of apps is a huge inconvenience.

Here are the 3 “big AI achievements” of China that Kai-fu Lee keeps listing:

1. The use of QR codes to pay for everything. He gives an example of paying street vendors with your phone. There aren’t crowds of street vendors in the West, so who cares. But the idea of using micropayments to show appreciation for favorite online authors is great. I’d definitely pay 10-15 cents for online articles I like if it’s done on a voluntary basis.

2. Another favorite example of Lee’s is the enormous number of bicycles available for rent through your phone all over Beijing. Again, this is something that exists in the West but isn’t massively used not because the West doesn’t have the technology – it’s really not that sophisticated of a thing – but because everybody drives, and nobody is fit enough to do a lot of biking around big cities. How people manage to bike through Beijing, with its horrid air and unbearable stench, is a whole other discussion.

3. And then there’s the issue of mega apps that let you “do everything without ever leaving the app.” I already wrote about this, so I won’t repeat myself. Really not that impressive.

The huge and impressive things that AI does can’t be revealed to consumers because consumers won’t like them. And the whole narrative of “yes, you get spied on but look at the benefits!” falls flat because there are no benefits.

But even if we take Kai-fu Lee’s argument completely uncritically, it still doesn’t work. He says that China’s great advantage is that it has a larger population than the US and that population uses data-gathering apps a lot more. So more information is gathered. But it’s information on Chinese users. Which, as Kai-fu Lee explained at length, cannot be extrapolated onto anybody else because of cultural differences. And unlike the US, China hasn’t made its culture globally attractive. If anything, it’s done the opposite.

5 thoughts on “Weaknesses in Kai-fu Lee’s Argument

  1. Reply to: Weaknesses in Kai-fu Lee’s Argument

    I don’t think it is true that there are no advantages to AI for the user. The first one that occurred to me is that self driving vehicles will allow blind people to travel alone, by car. There are others.


    1. By the way, where are those self-driving cars? Or are they going the way of the drone delivery programs, which have already been phased out?


      1. How are these AI, rather than just ubiquitous computing and internet? UPC codes have allowed for easy scanning on the vendor side since the 1980s, and smart phones with QR scanners now put that on the customer side. Why is AI needed for that? An app that lets you do everything from your phone is, IMO, just a web browser. Again, how is AI needed? I can see the use of AI to predict where to put the bikes for biggest revenue or most customers, but I don’t see how AI is necessary or even useful to the customer for the other two to function.


        1. The AI part is not on the consumer side. It’s on the side of the seller of the service. Of course, it’s not useful to the customer! It’s very detrimental to the customer. And that’s the whole point.

          Once you connect your bank account to the Chinese version of FB, for instance, it accesses your entirely banking history and sells that information. That’s the AI. Collecting gigantic amounts of data and using them to influence human behavior.


  2. I think there are advantages in medicine. AI is beating humans in detecting pathologies from images. But this is a benefit that will be experienced by comparatively few people in rich countries, whereas everyone but the richest will experience being surveilled.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.