China Will Overtake the US in AI, said Alphabet's Eric Schmidt • Hacker News Board

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#61

Post by Eridrus » Sun, Feb 11, 2018, 4:17 AM UTC

nopinsight wrote: post #60

seanmcdirmid wrote: post #20

Earlier quoted context omitted.

China has no technological edge in ML yet. Even the most ardent supporter talks about how China will have edges in 2020, but not today. It is still very aspirational. I never said anything about neo tokyo, the fifth generation computing project was never vivid in our imaginations, all most of of us remember about it is the government throwing a bunch of money at AI research and hardware, and that it led directly to t…

What do you think about this? (from part of my other comment) * 9 out of the top 10 algorithms for the Stanford Question Answering Dataset are by Chinese teams, backed by HIT & iFlLYTEK, Alibaba, Microsoft Research Asia (in Beijing), NUDT & Fudan University, and others. They beat out teams by Samsung, Salesforce, CMU, and Facebook, for example. https://rajpurkar.github.io/SQuAD-explorer/ It is not like others do not…

It's basically the same in computer security where you now see most vulnerabilities reported to vendors are from China. I wouldn't necessarily say China has "beaten" the rest of the world, but rather people in the west have basically decided that form of signaling is too expensive.

There's a lot of people publishing ML work on public datasets which barely move the needle. It's usually not too hard to squeeze something extra into a model to do better than the previous work, and if not there are always other datasets, so these raw numbers don't really means lot IMO.

Not to say China won't overtake the rest of the world, eg lots of great work coming out of MSRA, but I think that there is a lot of truth to the idea that once you start measuring something, people start gaming it, and I think the political environment in China encourages this gaming, so I would pay less attention to metrics like this that can be gamed.

Even paper metrics are gameable. Most reviewers acknowledge that once you remove the obvious accepts and obvious rejects you've got a large pile of passable but forgettable work. So if you can just keep cranking out passable work that fits the zeitgeist, you'll get more papers accepted.

I might be wildly off base here, I haven't tried to construct statistics, and academia has a large amount of starfucking going on, so it may be a cultural bias thing, but I don't get the impression that China is anywhere near dominant yet, but yet all these statistics look great!

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#62

Post by nopinsight » Sun, Feb 11, 2018, 4:22 AM UTC

philee wrote: post #59

nopinsight wrote: post #41

To give additional evidence why Eric Schmidt's comment has a basis in reality (and to reply to many sceptical comments here): * 9 out of the top 10 algorithms for the Stanford Question Answering Dataset are by Chinese teams, backed by HIT & iFlLYTEK, Alibaba, Microsoft Asia (in Beijing), NUDT & Fudan University, and others. They beat out teams by Samsung, Salesforce, CMU, and Facebook, for example. https://rajpurkar.…

Counter-example: Geoffrey Hinton (who developed backpropagation), has a self-proclaimed phobia of mathematics. There's plenty more to innovation than solving those IMO problems, Google Code Jams, etc, because they involve very specific practice (and over-fit the needed skillsets). It's instead the encouraging atmosphere and resultant inflows of capital (intellectual and otherwise) that may put China ahead in the futu…

Geoffrey Hinton may not be excellent at math at the professional level but if one watches his talks and lectures, it’s clear that he is more competent than 99% of most populations.

AI requires more than math, but one cannot work competently in AI without being competent at math. Based on teaching experience, I would estimate the lower threshold to be about 2SD above global average (or equivalent to 700 PISA score in math), which might mean less than 2% of US high school students are above that threshold.

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#63

Post by chillee » Sun, Feb 11, 2018, 4:29 AM UTC

Barrin92 wrote: post #52

candiodari wrote: post #46

Earlier quoted context omitted.

I disagree. Time and time again data advantages and computing power advantages are clobbered by algorithm advances. For instance, there is no amount of cluster power that'll let you best resnet with a non-residual network. Well perhaps there is, but even Google's mighty clusters can't make up the difference. In audio, end-to-end models that take 2 weeks to train on a dual GTX 1080 machine, and take > 1 week to train…

>The same goes for data. We can outperform VGGNet trained on all of imagenet (~ 15 million images) using resnet trained on a 1% random sampling of imagenet. We have papers putting forth statistics on specific model architectures outperforming with okay granted, but isn't this largely part of the academic field, openly accessible and transferable? People at Baidu and Tencent are likely up to date on the state of the a…

You see the overall big techniques being published, but there's still a huge gap between papers being published and actual implementation in products. Implementing papers into problem-specific domains always introduced all sorts of challenges as well as new structure to take advantage of. For example, none of the openly available machine translation implementations are anywhere near the proprietary ones.

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#64

Post by nopinsight » Sun, Feb 11, 2018, 4:41 AM UTC

Eridrus wrote: post #61

nopinsight wrote: post #60

Earlier quoted context omitted.

What do you think about this? (from part of my other comment) * 9 out of the top 10 algorithms for the Stanford Question Answering Dataset are by Chinese teams, backed by HIT & iFlLYTEK, Alibaba, Microsoft Research Asia (in Beijing), NUDT & Fudan University, and others. They beat out teams by Samsung, Salesforce, CMU, and Facebook, for example. https://rajpurkar.github.io/SQuAD-explorer/ It is not like others do not…

It's basically the same in computer security where you now see most vulnerabilities reported to vendors are from China. I wouldn't necessarily say China has "beaten" the rest of the world, but rather people in the west have basically decided that form of signaling is too expensive. There's a lot of people publishing ML work on public datasets which barely move the needle. It's usually not too hard to squeeze somethin…

Thanks for your thoughtful reply! I think everyone, including the Chinese, agrees that they are not (yet) dominant in AI, just that they are catching up very fast to the best in the world. They might even beat all others in optimizing in existing paradigms.

Part of their 2030 plan to become preeminent in AI is to develop indigenous breakthroughs by 2025 and be equal to the leading nation by 2020. They are very much on track to achieve the closest goal given evidence to date.

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#65

Post by chriskanan » Sun, Feb 11, 2018, 4:51 AM UTC

olympus wrote: post #28

He claims the US government isn't investing in AI. That's false. They're investing in AI, they just aren't giving any money to Google. What would the government do with an AI tuned for advertising? This is likely just Schmidt complaining about not getting free money from the government for once. Besides, it seems like Google is investing plenty of their own money into AI research without the government needing to con…

The US government has been investing in AI, but that spending has been relatively flat compared to China. China is increasing its spending in AI each year and has been advertising that they aim to be number one in AI. They are even adding AI to pre-college education.

https://www.ft.com/content/856753d6-8d31-11e7-a352-e46f43c58...

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#66

Post by seanmcdirmid » Sun, Feb 11, 2018, 5:03 AM UTC

nopinsight wrote: post #60

seanmcdirmid wrote: post #20

Earlier quoted context omitted.

China has no technological edge in ML yet. Even the most ardent supporter talks about how China will have edges in 2020, but not today. It is still very aspirational. I never said anything about neo tokyo, the fifth generation computing project was never vivid in our imaginations, all most of of us remember about it is the government throwing a bunch of money at AI research and hardware, and that it led directly to t…

What do you think about this? (from part of my other comment) * 9 out of the top 10 algorithms for the Stanford Question Answering Dataset are by Chinese teams, backed by HIT & iFlLYTEK, Alibaba, Microsoft Research Asia (in Beijing), NUDT & Fudan University, and others. They beat out teams by Samsung, Salesforce, CMU, and Facebook, for example. https://rajpurkar.github.io/SQuAD-explorer/ It is not like others do not…

Just to be clear, not all researchers in Microsoft Asia are Chinese (I wasn't, anyways). One of the best ML guys I knew at MSRA was actually European (he started working on GPU-based deep learning way early in 2010 when we were still at Sigma, he works in Redmond now but was a long timer like I was but had to leave because of the pollution, which is why most laowai left). If China could attract more of that world talent, they might be able to do it, but otherwise, its like a talent pool of 1.4 billion competing with a talent pool of 7.4 billion (many of the ML researchers in the states aren't American either).

With that out of the way, all of those are not very relevant. Its like judging the quality of programmers by ACM programming contest wins, or by judging hardware advancement on super computing benchmarks that no one really cares about anymore.

China has declared that they will be the winner in AI in the future. But they definitely aren't today, and that assumes that everyone else stays absolutely where they are already.

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#67

Post by nopinsight » Sun, Feb 11, 2018, 6:06 AM UTC

seanmcdirmid wrote: post #66

nopinsight wrote: post #60

Earlier quoted context omitted.

What do you think about this? (from part of my other comment) * 9 out of the top 10 algorithms for the Stanford Question Answering Dataset are by Chinese teams, backed by HIT & iFlLYTEK, Alibaba, Microsoft Research Asia (in Beijing), NUDT & Fudan University, and others. They beat out teams by Samsung, Salesforce, CMU, and Facebook, for example. https://rajpurkar.github.io/SQuAD-explorer/ It is not like others do not…

Just to be clear, not all researchers in Microsoft Asia are Chinese (I wasn't, anyways). One of the best ML guys I knew at MSRA was actually European (he started working on GPU-based deep learning way early in 2010 when we were still at Sigma, he works in Redmond now but was a long timer like I was but had to leave because of the pollution, which is why most laowai left). If China could attract more of that world tal…

I agree China does not yet surpass all other nations regarding creating innovations in AI, although they have come close to the top of the league.

As shown by PISA and other international tests, the percentage of AI-capable people varies wildly among the populations in that 7.4 billion global pool. China performs very well, esp relative to their development level, and above most OECD countries.

http://ncee.org/2016/12/pisa-2015-results/

If China can hold onto most of its talent, it would be a relatively large percentage (say 4-5%) of 1.4 billion vs a smaller percentage (2.0-2.5%) of 1.3 billion (OECD country populations) and a very small percentage (say 0.2-0.5%) from 4.9 billion (the rest of the world) who can potentially work productively in AI. This assumes the Gaussian distribution of education quality, skills, and perserverance and a fairly high bar (2+ SD above OECD average) for a lower threshold to work competently in AI.

It is obvious that not all talent in the latter two would like to immigrate to the US, so it is unclear if the US will continue to have a quantitative advantage despite a more global talent pool. Chinese government's deliberate policy in nurturing and developing AI talent will also add to the number of local practitioners. There is also clear evidence that more Chinese nationals are moving back home after graduating from abroad.

The rough estimates above are relevant for AI optimizations and applications in many real-world settings since a sufficient quantity of R&D practitioners is necessary for that.

For breakthroughs, other factors might come into play and it is not clear if cultural factors and upbringing would affect the potentials of people to create paradigm-shifting ideas, which means quantitative analysis in isolation is not nearly as predictive.

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#68

Post by pedrosorio » Sun, Feb 11, 2018, 6:28 AM UTC

fooker wrote: post #39

pedrosorio wrote: post #5

"Schmidt provided examples to support his claim that China is poised to dominate the field of AI. While many assume that the Chinese educational system is inferior to the U.S., he pointed out that Chinese people "tend to win many of the top spots" in Google's coding competitions." In the last 5 years, participants from China reached the Google Code Jam finals (top 25 contestants) 12 times. Russian participants did th…

This is evidence that the Chinese education system is not inferior. Nothing more. The argument about AI lead is based on other facts, not particularly this one.

I don't think performance of the best on these contests says much about the education system (other than possibly how much exposure does the education system provide to this kind of contest from a young age).

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#69

Post by seanmcdirmid » Sun, Feb 11, 2018, 7:01 AM UTC

nopinsight wrote: post #67

seanmcdirmid wrote: post #66

Earlier quoted context omitted.

Just to be clear, not all researchers in Microsoft Asia are Chinese (I wasn't, anyways). One of the best ML guys I knew at MSRA was actually European (he started working on GPU-based deep learning way early in 2010 when we were still at Sigma, he works in Redmond now but was a long timer like I was but had to leave because of the pollution, which is why most laowai left). If China could attract more of that world tal…

I agree China does not yet surpass all other nations regarding creating innovations in AI, although they have come close to the top of the league. As shown by PISA and other international tests, the percentage of AI-capable people varies wildly among the populations in that 7.4 billion global pool. China performs very well, esp relative to their development level, and above most OECD countries. http://ncee.org/2016/1…

Except those PISA tests only compare urban shanghai to larger more diverse samples in the western world, this is not a very good metric, while education in china is hard from uniform! But sure, not all is going to be even in 7.4 billion people, like not everyone in china is a smart scientist rather than a taxi driver or fuwuyuan. 2% is way too high, especially considering that education on average in china is still much worse than the west.

The number of Chinese returning after studying abroad is increasing because the numbers studying abroad are increasing, these are mostly rich fuerdai kids not studying STEM who have no chance getting jobs in the west anyways; the amount of STEM talent retained, especially at the important post graduate level, has remained fairly constant.

Overall, China cannot achieve dominance in this area with its current policy of nationalistic tech isolationism (e.g. blocking off internet sites, not allowing immigration). Even if we ignore the west, there is still India and other countries that have large roles to play; that have their own promising characteristics.

Re: China Will Overtake the US in AI, said Alphabet's Eric Schmidt

#70

Post by nopinsight » Sun, Feb 11, 2018, 7:46 AM UTC

seanmcdirmid wrote: post #69

nopinsight wrote: post #67

Earlier quoted context omitted.

I agree China does not yet surpass all other nations regarding creating innovations in AI, although they have come close to the top of the league. As shown by PISA and other international tests, the percentage of AI-capable people varies wildly among the populations in that 7.4 billion global pool. China performs very well, esp relative to their development level, and above most OECD countries. http://ncee.org/2016/1…

Except those PISA tests only compare urban shanghai to larger more diverse samples in the western world, this is not a very good metric, while education in china is hard from uniform! But sure, not all is going to be even in 7.4 billion people, like not everyone in china is a smart scientist rather than a taxi driver or fuwuyuan. 2% is way too high, especially considering that education on average in china is still m…

I referred to the 2015 PISA test when making those estimates. In 2015, four Chinese provinces took the test including Guangdong, which student's native language is often not mandarin. Their total population is 230 million people.

Granted, they are likely above China's average in terms of skills but I read elsewhere that someone involved in PISA said regarding about unofficial PISA 2009 research in other Chinese provinces that even some very poor regions scored close to OECD average.

We will likely see China's whole country results in 2019 when PISA 2018 official results come out and I believe it will still be better in math than many OECD countries.

Your point regarding lower average quality of education in China could be true but that means in 10 years they could be improved with economic development and their skills would be even better than now. Developed countries in East Asia and Singapore where populations share the same cultural traits with China have for some time dominate PISA rankings in all subjects.

India has many talented people but because of many factors, as a percentage, the figure is probably many times lower than China. Check out their former PISA results before they pulled out. They will improve but it is a long way to go and we know from history that network effects and other factors tend to make tech dominance into natural monopoly, at least until a major paradigm shift happens.

No other country has both the size of native population and the distribution of population's skill levels taken together to compete on equal footing with China in the next 1-2 decades.

The US can hold its own and perhaps keep the lead by being a desirable place for global top talent but it remains to be seen if the government's policy and other factors will erode that advantage away.