Submind YouTube summaries
Thumbnail for The AI Race Between the US and China (Dealing with Tensions in the West Pacific)

The AI Race Between the US and China (Dealing with Tensions in the West Pacific)

Watch on YouTube

Video summary

The United States and China represent two distinct superpowers pursuing artificial intelligence through fundamentally different governance models and strategic approaches. The US system is characterized by decentralization, relying heavily on private sector innovation, voluntary regulations, and state-level consumer protections rather than a unified federal law. In contrast, China operates under a centralized political structure where the government mandates strict controls over data sets, infrastructure development, and energy resources to drive national objectives quickly without public protest. This centralization allows Beijing to marshal all available resources efficiently but creates significant friction in the US regarding local opposition to large-scale projects like data centers due to concerns about electricity costs, land use, and environmental impact. A critical divergence between the two nations lies in their technological strategies and access to energy, which heavily influences AI development speed and cost-effectiveness. China leverages its centralized control over coal reserves and rapidly expanding renewable energy sources to power massive infrastructure projects that would face intense regulatory hurdles or public backlash in the US. While American technology often focuses on pushing for marginal improvements at high costs with closed-source models, Chinese firms utilize open-source technologies that are cheaper and freely available globally, granting them significant soft power influence. This approach allows China to export AI capabilities widely, though it introduces ideological concerns regarding propaganda embedded within their algorithms, whereas the US maintains a focus on sophisticated but expensive tools primarily for consumer markets and scientific research rather than immediate physical manufacturing applications like robotics or electric vehicles. The geopolitical competition extends beyond mere technological superiority into questions of national security, military application, and global inequality. China is aggressively deploying AI in defense intelligence, weapon systems, and robotic manufacturing to secure its geostrategic interests, while the US lags somewhat behind in integrating these technologies into physical infrastructure despite leading in software innovation. Furthermore, this race risks creating a dangerous divide not only between developed nations but also leaving developing countries like those in Africa or ASEAN at the mercy of whichever superpower dominates AI standards and tools. As both nations tighten export controls on chips and talent to prevent each other from advancing, there is growing concern that these protectionist policies may inadvertently accelerate China's self-sufficiency while destabilizing global trade norms and limiting access for smaller economies. Ultimately, this intense rivalry presents a complex challenge where progress in artificial intelligence cannot be halted without risking significant economic or security consequences for the world. The debate continues over whether export restrictions effectively block Chinese advancement or simply encourage them to develop independent technologies faster, potentially leading to an arms race that escalates tensions rather than resolving them. As both countries push forward with their respective models—one prioritizing open innovation and market freedom, the other emphasizing state-driven efficiency and ideological alignment—the outcome will likely determine not just who leads in AI capabilities, but how this transformative technology is governed globally and whether it serves humanity or exacerbates existing global inequalities.
Read the full video transcript
Welcome to Dealing with Tensions in the West Pacific on Think Tech Hawaii. I'm your host Jay Fidel. This episode is called The AI Race between the US and China competition on AI governance and strategies. Our co-host for the show is Elizabeth Chinhale. Welcome to the show, Elizabeth. >> Thank you, Jay. You know, we need to examine the generally different approaches to AI development and the related challenges we find in the United States and China, the two big players from governance to use to marketing to position strategies. Lots of material here. How are we doing, Elizabeth? How is the US doing? How is China doing? Is the US ahead or behind in chips, in electricity, in AI, AI technology and expertise, in AI distribution, adoption, regulation and governmental policies. Give us a comparison. >> I think these are perceived to be the two super powers in the world in terms of AI development, but they're developing along very two very different lines and the differences are what we have seen elsewhere. It's due to the different systems. So in China because of its political system it tends to be more centralized. The regulations are mandatories and uh but surprisingly their technology is open source but on the other hand US is more decentralized. So except for the executive orders that we see from the presidential office from time to time. Otherwise, I don't really see a federal AI law forming anytime in the near future. What we have are different state consumer protection laws, especially when it comes to sectorized protection such as health data. And um for the most part it's uh voluntary although some states have begin to impose penalties. Uh but our technology is more advanced. It's closed sourced. Um, so I guess it's not surprising given you're we're a capitalist society which tends to leave things to the private sector as opposed to China where everything is state driven. Um, as to who is more advanced, I think each has its own vantage points. Um, so what's really important for AI development? We're talking about power management, talking about big sets of data, we're talking about infrastructure, uh we're talking about talents. So in some ways, uh these two countries are the same. But um China again because it's centralized, it has ability to control the data sets, make them available or not available. It also has uh total control over building of infrastructures uh and data centers. For example, I mean in my hometown of San Jose, our biggest thing is uh not in my own not in my backyard, not exacting in our backyard. Used to talk about ATRs, now it's we don't want a data center near our houses and and I think that's coming up all around the country. The question is where do we then build a data center? You know, I think some of our innovators are beginning to think uh we're going to build in outer space, which is probably a good solution because then we have abundance of energy coming from the sun. We don't need to worry about, you know, polluting or diminishing people's housing values. But then again, China doesn't have that problem. They can build data centers. I just can't believe the US can really build data centers in a different country even if other country is willing to lend us the space, right? And again, we're talking about data sovereignty at some point. You you can't have a data center in another country. What if they cut it off? That means all the AI's capability to be cut off. So that that's a kind of very rough comparison. And also I should add because China is open source, Chinese technology tend to be less expensive. They're making that freely available. So, they're gaining a lot of international influence. >> Let's talk about data centers for a minute. I want to uh you know follow up in some some questions that arose in your in your comments. Um what's the problem with data centers? Why is it nimi? Why are people so opposed to it? It seems like there's a big push back going on. I don't really understand. Um is it the electrical energy issue? Is it the is the the land issue? Um, is it the AI issue? What makes people so concerned about data centers? >> What I'm hearing locally, there may be other reasons, is that whenever you have a data center, these things tend to be monster energy consumers. Then your electricity bills goes up. And I kind of have detected that in RPG and B bill in California where now in winter it used to be, you know, I never get bills above 100 because we tend to be very mild. But excuse me, but lately I'm getting bills which are triple what it used to be. And I wonder is because data centers are competing with our electricity and you you are hearing there are states that mandating that hey if you want to build a data center uh in our state you have to come with your own energy source. Now then we kind of you know what do they go to? they go for these portable nuclear energies which then is you know in the past has not been very welcome right maybe they're smaller managed better and it's a new gen a nuclear technology but still brings out a lot of concerns what are they doing how are they being monitored would they would that produce a natural disaster or this is humanmade disaster that we don't anticipate I think people have fear a lot of concerns and you tend to refuse things that you don't No, >> you know, I'm concerned about the electric electric supply problem. China seems to have more electricity that it can use and it is moving uh faster further on renewables than the US is. Trump opposes renewables. We're doing still doing coal. Uh China's doing coal too, but it it's gaining in renewables. And so the result is that China has more energy. And the old story is that an economy that has adequate cheap energy is going to do better uh than an economy that doesn't have that. We don't have that obviously. And so uh it's kind of it's kind of an inconsistency. Trump on the one hand um doesn't want to uh encourage renewables. He wants to encourage coal and the other hand he wants to build data centers. How do you how do you reconcile those things? That is the policy level things that the White House or our government will have to come to some kind of happy medium. But no doubt I mean again with energy this tend to be centralized and a centralized state like China tends to do better in these areas. I think they still use coal. I hear that they manage their oil crisis quite well probably because they have good relationship with Russia and Iran. they're getting the oil somehow and uh also they started early with electrical energy I mean electrical vehicles building renewable energy sources and again for China this is like a national drive so all the resources are marshal people don't really get to protest and therefore they can implement these things quickly I mean this will be a good topic to kind of dive into about the energy infrastructures in each country and actually United Nations special envoy um praised China in in one of their um conferences in Shanghai recently AI conferences and um but what he's pointed out besides praising China is that this could become another huge divide between developed and developed countries because AI is a fairly it's a wealthy country's game right you're talking about energy you're talking about hot talent. You talk about huge data sets. These are things both China and United States have in abundance. But you think about these smaller countries, you know, in Africa, I mean, even ASEAN, which is kind of considered middle tier country, they're they're kind of concerned too. I mean, how are these countries going to compete? Um, are they be going to relegated to be at the mercy of these big countries have developed AI technology? And we're quickly becoming a world where AI is becoming an essential tool. And uh it's this is going to be a bigger problem than just a digital divide where we're talking about inequality information. Now we're talking about ability to ration to think of new solutions and people are already beginning to think AI is smarter than us in some ways. And I can see that they do reason faster. they have more data at their explosible than my brain can possibly process. So uh I mean I think United Nation is right to be concerned about creating a bigger divide between develop developing countries with this AI movement. >> Yeah. Well, you know, you mentioned uh this whole thing about open source and open source to me means that um it's open that people can get in there and get the code. Um and that's pretty threatening to the United States which is as you said closed and um you know capitalist control uh of of the code and so uh doesn't that give China a tremendous advantage? they can have more people more we know there's a lot of PhDs in computer science in China some huge percentage of college graduates uh go on to PhDs in computer science and they know how to do AI um and so they can use that open source in the US it isn't like that there's a few people wellunded by some of the oligarchs and all um but they keep it to themselves it's a big secret and so that seems to me like the China model is going to work better not only in China but as a something that they can export to the world. Uh we are behind the curve on that aren't we? >> Well that that's precisely the problem but that has traditionally been the case. We always joke about US technology is we spend a lot of money pushing the last 5%. China doesn't do that. You see their technology is generally inferior to US technology but they get to 70 80% and they stop and they say it works. we're just going to settle with that and we run up with more human resources, more repetitions. I think that's also the case with the AI. They use open source, not as sophisticated, probably not as well trained and but they're exporting because they're open source. It's free. So, I think the the less funded countries are going to start using Chinese AI models. Now, now we're not going to talk about robots, which is a different thing, which is machines more costly, but talk about chat bots, right? the the the the the computer the you know what you the chat GPTs the Geminis and we use Chinese are plenty and uh during my recent trip to China these days I I've been you know all we here is deep sea I went to China and I spoke to some of my relatives especially their kids and I realized they have a lot more platforms so I downloaded one it's completely free by Alibaba it's called Chenwin we you know translated Qin in in in English it you know it really is not bad I have to say but then again we're not trying to for for most consumer we're not asking AI to do really I I think the US has been trying to make AI to be really top level thinker maybe being innovators themselves then we can get into the IP issues in another episode but you know for example normal questions you know I I think the Chinese cheaper free model does a pretty decent job you know I mean I asked about traveling about diarrhea, about eating food which you've been sitting out. I it's it's good. What can I say? The only problem that I see with the Chinese model it's um their concern is ideology, right? So they in their models there is a big concern on ideology promoting socialism putting government nothing bad about communism and socialism. So for example, I asked a question about what if I transport an insect in my suitcase and I'm flying across the glo uh the ocean. What the insect die? So they kind of said correctly that no insects will survive. I was asking about cockroaches in particular. So they think tend they're really tough. That's but you know what followed was interesting. He said but Chinese customs are really quite capable. They will remove anything that's movable. So even if they're not caught in flight, they would be removed by Chinese custom. I mean now I think that's just purely propagandistic, you know, trying to promote the Chinese government, their capabilities, putting your trust in the government. But the problem is once they start exporting these models, especially to promote socialism, what do you think is going to happen to these countries that are using these cheap, inexpensive, but okay models? They're they're going to be brainwashed inadvertently, right? you know, so I think that's the threat that US is perceiving. But honestly, I I don't know how we're going to overcome that because our models are just generally more sophisticated, more expensive, and we're probably not going to make it freely available the way Chinese are doing, just gain influence. >> I find that ChibiT and all the others, Perplexity and the like, Gemini, they they're they're tilted. They're tilted. They don't go after Trump in the same way. for example, as they might uh and so you you can feel the the hand of propaganda there in their answers to your questions and all. But I also find that uh in the US we seem to be focused more um on the consumer market uh where you can ask questions and get answers and write your term paper and all that stuff. Uh it's it's more in what do we call it? Information sciences. Um we're not that strong in robotics. we're not that strong in in the in the the physical technology that China is developing, for example, in the cars and in the batteries in um all of that robotic manufacturing that seems to be where their emphasis is. Um and that that really means more because in the in the case of the uh robotic manufacturing, you can tell when it works. You can tell if it's wrong, it won't work. Uh, on the other hand, in the social sciences in in the in the US, you can't tell if they got a slant on it. You can't tell if um, you know, there's something wrong. They try to do propaganda on you. This is the problem. Uh, I I would also say that China must be using it for national security, for intelligence, for geo geopolitical effect, maybe war, certainly weapons. And the US, I don't I don't feel, is in the same place. We're still trying to build uh World War I battleships. Um if we want intelligence as to where deal leader is in tan, we get it from somebody else like the Israelis. Uh you know, we we are not using it in our in our military adventures in the same way that China is. So I think you know how you deploy AI is a big question and it's a national security issue. uh certainly an issue as to the the productivity the usefulness of AI as to how you develop it and deploy it. Do you agree? I think so and I think our entrepreneurs some of them are more idealistic than others right because recently there is a question about who's going to be used in a department defense and I think I and honestly I had to look up again it's between open AI and anthropic I can't remember I I think it's probably more anthropic that it kind of denied us DoD's uh demands and comes to certain kind of informations and our our technology companies have the ability and have done that in the past. They they're willing to stand up and say no to the government although I think they back down later because I guess the the the consequences fin consequences are so huge. Uh and but then again the the Chinese technology companies probably don't have that luxury of saying no to the government. So if government wants to summon their technology and they not going to ask the question as to where it's going to be deployed but although I I think maybe we're not deploying in the military area but I heard our pharmaceuticals are you using AI very extensively and that helps a lot in new drugs discovery so it's being put to good commercial use for for for humans right and when yeah counter your points about robots Yeah, those you know 10 20 years of manufacturing China has done for the world is doing it some good right I mean they have been I've seen their advertisement for robotic dogs for robots I have no doubt that they're going to roll models uh faster certainly cheaper than the US just like your electrical vehicles I mean think about it Tesla really couldn't meet its production demand until they opened the Shanghai factory right so even we have the technology, we're not going to have the capabilities to build as fast and as cheaply as China when we move to the next phase where we're talk about physical AI such as robots. So what about export controls? You know, I find that uh in in China, you know, which is totalitarianism, um they want to control it, they can control it, no problem. Um and they can deal with export controls or importation or sale overseas, what have you. Uh or or giving it away as you say, that's really important for what do you call it? Soft power in various places in the world. And good for them for that. But we have all these export controls and our policies seems to be in chaos. Congress doesn't Congress doesn't really understand AI. I'm sorry to tell you as they never understood technology anyway. Um so the problem is do we have good policies? Do we have good export controls? I don't I don't think that anything Trump does is a good policy on export. Um but there's a big difference between the way China handles the engagement with the outside world uh and with the US and the way the US handles export controls and policies and selling and giving not only AI um but chips and all those other accessories uh to allow China to develop AI. Uh what are your thoughts about that? Are we making a mistake here? Is the right thing? Well, we're we're becoming kind of unamerican in the process in this competition with China. That's the unfortunate thing, right? Because we tend to be the open one. We believe in open economy, free trade. We rarely put restrictions on exports, but now we're beginning to. China has always had very strict restrictions on export technology in my field. I mean if we're buying and selling patent to get a patent from China into a US company's hand it was a very difficult process you had to jump to hurdles um but now we're becoming the same you know the question is is this effective I mean I'm hearing two lines of thinking some people are saying including Jensen Huang and the video saying hey you can't really block things anyway you're just encouraging China to develop it own technology ology. If we don't have such controls, they will continue to depend on technology. Wouldn't that be better? I also hear the other line say that's precisely what China wants you to believe. The truth is China are years behind and they can never develop the video level uh chips anyway. So, we should continue to blockade uh the export technology. But this can escalate. You know what I'm hearing lately is uh China's beginning to put restrictions on exports of human talent outside the country because remember there's a period where their top best and brightest used to come to us and receive education in working US companies. Apparently their policy is switching and maybe that's not going to be as open as before. Certainly not as encouraged but at the same time Trump is doesn't want foreigners to come to our universities anymore either. So maybe it just says well I don't know. >> Elizabeth, we're out of time. Uh what are your thoughts about this in general? Give us one minute uh for a takeaway. >> I think this is the AI is one of the most interesting human developments uh in the world. We're developing perhaps carbonformed intelligence that will rival sorry we're carbonformed. We're developing a silicon based intelligence which may compete against us. We're all full in interpretation as to what its potentials are. Is it going to do good for us? Is it going to do bad for us? And we have two superpowers really competing from along very different ways. This is a race to be watched. Uh but progress cannot be stopped. There's no turning back. >> Okay. Thank you to our co-host Elizabeth Shanhal. I'm Jay Fidel, host of Dealing with Tensions, and these certainly relate to Tensions in the West Pacific. The title of the episode has been the AI race between the US and China. Thanks to our viewers for watching. We'll see you again soon, very soon for the next one. Aloha. Aloha, Elizabeth. >> Aloha. Thank you, Jay.