Remember a while back there was a lot of chatters about how "China is leading in AI innovations"? Whenever I hear that I always roll my eyes. I don't know where it originated from but it certainly isn't from computer science space. My first response is: Really? What innovation?

Has China written a better SciPy? Contributing to Apache Hadoop and Spark? What new AI alogrithms have they developed? Which Chinese university is leading such research? Are there more Chinese contributors to AI/ML/Big Data projects? Obviously none of that is true.

When I asked friends who repeated this narrative uncritically, some said it's because China is a surveillance state and they could provide a much larger dataset for analysis to tune their algorithms. I was like, where's the 'innovation" there? That's application not innovation.

Big Data is made to scale. A much larger corpus produces better ML results. If all China is doing is using existing ML technologies and feed them 1.4b profiles instead of 5m, that's literally zero innovation involved. Not to mention the privacy implication.

Yes Chinese companies are becoming more active in open source. Even Taobao and Alibaba have projects on GitHub now. But most of them are focused on cloud computing infrastructure and orchestration. There just aren't any significant Chinese AI/ML OSS projects I know of.

And I don't say this just to diss China's tech space. AI/ML is an extremely research-intensive field. For it to thrive in China would require a lot of academic freedom and research focus.

I don't think it's controversial to say China's tech "research and innovations" are mostly focused on the practical rather than the theoretical. So more often than not they'll be chasing iterative improvements rather than technological breakthroughs.

Btw a pretty good barometer of where China's AI/ML is at is comparing Chinese language search results from Google and Baidu. ¯\\_(ツ)_/¯