Hacker News Reader

Hacker News

Comment by melling | Hacker News Reader

Full thread

melling

By some measures China has already passed the United States:

https://www.bloomberg.com/view/articles/2017-10-18/who-has-t...

China, and eventually India, matching the United States will hopefully help increase R&D.

For example:

- Solar energy prices have dropped significantly because of China's investment.

- Nuclear energy is dead in the US but lives in China.

- AI - China will match the US, accelerating development

- Mass Transportation - High-speed rail, maglevs

- Electric vehicles - China has large EV quotas and they're the largest car market.

View on HN

Only in PPP, which doesn't mean much because many things are more expensive in China than the USA (housing, cars, high-end goods, schooling, ...). For an equivalent lifestyle, you often need to earn more in China than the USA, so PPP seems to be a weird metric.

- Solar is suspected to be heavily subsidized by the government. That isn't so bad for us as long as we don't produce panels, it has led to a solar revolution in the USA as well.

- China still hasn't done its own production nuclear plant yet. Citizens are extremely wary of the government to do it right.

- The point here is "will" rather than "does". China keeps touting how they will beat the USA in AI by 2020 or 2030 or whatever, but it is still all aspirational. China has a lot of talent and a lot of data work with (privacy isn't much of a concern), but claiming you are going to win a race is much different from winning it.

"Citizens are extremely wary of the government to do it right." - I don't think citizens have much say in the matter...

- it's worth noting that along with data and talent, there is money - China's dedicated 150 billion USD to AI as part of it's 5 year plan. Meanwhile the US is considering cutting scientific research funding by as much as 15% as part of 2018 budget

> "Citizens are extremely wary of the government to do it right." - I don't think citizens have much say in the matter...

They actually have more say than you think. And also, the government also has a lot of its own introspection about nuclear, especially after Fukushima. Couple with the huge capital costs, and it isn't a sure thing for China.

> it's worth noting that along with data and talent, there is money - China's dedicated 150 billion USD to AI as part of it's 5 year plan.

Much of the current advances in AI are not coming from federal funding (rather, private companies who are using it). Nor can you expect Chinese academia to soak up that much money and use it efficiently. Again, China has a chance at it, but it is far from inevitable. It is annoying to me when they talk about how their success is a sure thing, the Chinese government used to talk about how they would clean up their pollution problem by 2015 in 2010 because they were going to spend a lot of money on it....

"They actually have more say than you think" - I disagree. China forcibly relocated over 1 million to complete the 3 gorges project. It doesn't have the best track record of reacting to citizen concerns. I think the other points you mention are more material in it's nuclear considerations.

Fair about the money and pollution. I would assume those commercial advances you mention however are mostly non-military, nor will players like Google or Apple likely soon contribute to US government/military AI research initiatives. I guess in the end the "AI race" is more like "AI olympics" in that no country will win all the events

Reply on news.ycombinator.com