China has been progressing quite quickly considering how young their academic/research institutions are.

A very common path for Chinese scientists is to go to a western university for their doctorate and post-doc, develop a research program, and return to a faculty position in China.

Many Chinese grants include salary for the PI, so an investigator getting multiple grants can earn a lot of money. It's an attractive incentive that appears to be working quite well (from my limited exposure).

As for reliability, there are certainly some issues, but I hesitate to paint with a broad brush. Science is often very discipline-specific in terms of reproducibility, openness of collaboration, funding, etc. etc. There absolutely are fields where the Chinese are quite strong.

Chinese AI/CS papers are definitely not going to be ignored. They are really strong on that front, for they have a huge domestic industry to sustain such level of researches and there is economical benefit to going after.

I was just at Ubicomp, and I'd say the level of quality coming out of China was lacking (and a lot of universities outside of China to be fair). But only the papers from China raised red flags if their results were real or not...

Ignoring disreputable journals from China is one thing, but I doubt any serious scientist would claim there are not incredible advances coming from that country. In field after field they've shown they can produce amazing scientific work. There is plenty of low-quality work but let's not forget the amazing work too.

Please look at the number of papers from China in top CS conferences. In general, the reviews are double blind, making it as objective as possible at this moment.

Example: https://nips.cc/Conferences/2018/Schedule

For those wondering, I just clicked through a few randomly and I only saw ~1/10 from China, and it was fairly applied (i.e. not especially novel research). I'd be interested in a more thorough analysis.

Arguably most CS papers / presentations are not replicable because many don't share the data.

Well I guess the first step would be luring Chinese scientists currently in US back to China (many of those have US citizenship and would also be counted as US scientists)

I work with lots of Chinese scientists and they say they are never going back. They don't trust Chinese papers or labs at all.

Survivor Bias at play, the ones do go back, go back. I think the mentality among Chinese research community had changed rapidly, especially in Computer Science field, where China is as practical a choice as US.

The problem with US is for foreign nationals, there is a preset role for us to fit in, while in China, it offers more potentials. I won't say certain it is black and white question at this moment, there is trade-off in between. Living in US is no longer the non-brainer

"report bias" seems a better fit here.

I'm sure there are lots of Chinese scientists that would never go back, but at the same time there're also plenty of Chinese immigrants (including scientists) actually went back or are considering going back when the time is right, including myself. I think overall more people are considering that as a viable option over the years.

Wouldn't there be an implicit bias towards those who don't want to return? The others made their way back to China.

We have only had one leave and go back to China and he was terrible and only here to get the "worked in the USA" seal so he could go back and get a better and higher paying professorship in China.

If you are taking a sample of Chinese scientists 10 years into their career in the US with kids here, then likely true.

My sample based on my circle of Chinese PhD students that have graduated in the last 5 years indicates a more willingness to go back and several have.

You would not be able to work with Chinese scientists that goes back. Sampling issue.

It's not necessarily a sampling issue. Of course they could work with scientists that go back. You work with them, then they go back. The implication is clearly that the parent could have already worked with scientists that have gone back and that isn't occuring.

When I did research and at grad school a large percentage of my peers were either Asian American or Asian, with a large percentage being immigrants from China. As a researcher in the US you spend too much time writing grants and now equity initiatives introduce extra risk in the tenure system for anyone of the wrong identity. I would not underestimate the lure of easy access to grants and a less politicized tenure track system for anyone of Asian or Asian American heritage.

China's current academic system is mostly a copy of that of the United States, and all of the current reforms are pushing further to be more like the United States. Whatever flaws of American grad school are and will be in China's grad school.

It is probably true for the proverbial early birds who went to China, but these days grants won't be that easy unless you are established or well-connected. There are also a lot of complaints about "politicized tenure track system".

> Most of the scientists I know pretty much ignore research out of China ...

Ranking by the 1% most highly cited papers in math and computing for those published between 2013-2016:

6 of the top 15 universities are in mainland China,

6 USA,

2 Singapore, and

1 Hong Kong.

Tsinghua is at number 1. Singapore also seems to employ a number of researchers from China.

The article does state that MIT still leads in the broader STEM fields, but Tsinghua may soon take over.

“The Massachusetts Institute of Technology (mit) still leads in the top 1% of stem papers, but Mr Marginson says Tsinghua is on track to be “number one in five years or less”.”

https://www.economist.com/china/2018/11/17/tsinghua-universi...

In condensed matter physics, there is a lot of good and original work coming out of top Chinese universities and being published in top journals. Also, for those in the 1000 talents programs, the startup grants that they receive are draw dropping.

Me neither. See here:

https://www.wsj.com/articles/american-professor-notes-narrow...

https://www.chinalawblog.com/2016/09/the-china-stock-option-...

You need to read/speak their language to get the good research materials.

A problem not unique to China unfortunately...

This used to be the case but it is changing rapidly. Especially in Math and CS right now.