《聚焦特习会：专家怎么看？》 美中科技竞争，还有合作空间吗？

专家认为，在人工智能安全领域，美中两国可以并且需要进行合作，因为两国在AI发展上处于领先地位。然而，在科技竞争方面，双方的立场和担忧根深蒂固，使得在贸易战等问题上达成协议的可能性不大。尽管存在结构性差异和竞争，但双方仍需努力寻求战略稳定，并在利益一致的领域寻找合作空间，尽管这样的机会日益稀少。

视频链接：

完整文本：

I think the key thing for the U.S. and China to talk about on AI, and we've publicly stated that, we actually had an article in China Daily arguing this, which is the Chinese major state-owned newspaper, that we need to collaborate and work very closely together around AI safety to make sure that the models have certain constraints and don't get out of control.

That's an area where the two of us could work together because nobody else is really working on AI at the level of the US and China. We're the leaders and so we need to be working together on that. But that's very similar to working on infectious diseases, although we saw what happened.

In other words, there are some sort of global challenges. The Ukraine-Russia war, it is important, but I don't think it achieves anything when it comes to the China-US techno-economic trade war. I don't think that's a negotiable issue anymore.

I think the Chinese have always called for the US to ease those export controls or restrictions. I think at this point, there's also a very realistic expectation as for how persuasive Beijing can be or how effective those persuasions will be. Because if you look at the reality, the reality is that the US is extremely concerned with China's technological advancement and becoming a bigger technological powerhouse than the United States.

So essentially, you will be lobbying the United States to take policies or actions that is perceived as countering or undermining.

In fact, I think they won't establish these rules because the competition is too fierce. Both sides feel that artificial intelligence is the most important for the development of high technology, and also has a great impact on military modernization. So perhaps if they agree on some rules, it will be on the most fundamental basis. For example, last time Biden and Xi Jinping met, the two leaders both clearly stated that when it comes to nuclear weapons, it should be a human decision, not an artificial intelligence decision. So if they come up with any rules, I think they will be relatively simple, on a fundamental basis. But the progress and development of artificial intelligence is incredible.

I think if you look at the Commission's annual reports, as Commissioner Slevin mentioned and referenced, we've been very consistent in that view that we are in long-term strategic competition, and that competition now informs virtually every aspect of U.S.-China relations, including related to technology development, innovation.

You know, I think the key is to make sure that the competition remains benign, doesn't spill over into conflict, still leave space and room for cooperation where our interests align. But those seem to be a few and far between these days. And in even areas where you would think there's alignment, oftentimes we find disagreement underlying. So it's very worthy work to try to build a relationship characterized by strategic stability.

But due to a lot of factors, structural differences, differences in governmental systems, etc., well known to your listeners, I'm sure the United States and China will likely remain in competition for as far as at least my eyes can see.