# Ex-Google CEO WARNS: "You Are Not Prepared For 2027"

*AI Upload* · https://www.youtube.com/watch?v=acUJ_VfT7q8

## Summary

This video presents an interview with former Google CEO Eric Schmidt about the US-China AI race, energy constraints, and AI safety risks. Schmidt argues the US will likely win the AGI race due to hardware restrictions on China, but warns that China dominates robotics and electricity production, and that the US lacks the 92 gigawatts of power it needs by 2030. He identifies misinformation, cyber, and especially biological threats as likely triggers for future AI regulation, and worries that open-source models and distillation make proliferation hard to control. The host adds commentary throughout and disagrees with Schmidt on the open-source question.

## Key points

- Eric Schmidt believes the US looks likely to win the AGI race because hardware restrictions on China limit their access to chips like Google TPUs.
- He says China has effectively won the race in robotics and electric vehicles, citing 172 gigawatts of solar installed last year and the $6,000 Unitree R1 humanoid robot.
- The US needs about 92 gigawatts of new electricity by 2030 for data centers, while almost no nuclear power plants are being started in America.
- Schmidt identifies three main AI threats: misinformation, cyber attacks, and biological threats, with the biological one being the hardest to contain.
- The Biden-era rule requiring registration of training runs over 10^26 flops was scrapped, and the new David Sacks framework focuses on winning the race rather than proliferation.
- Distillation and transfer learning cost about 1% of original training, and moves toward FP8 and FP4 precision make training more efficient, worsening the proliferation problem.
- Schmidt predicts the biggest US models will stay closed while China's biggest models are open source, leading most countries to standardize on free Chinese models.
- The host disagrees with Schmidt on open-source AI, arguing tech CEOs want to criminalize it to protect their investments.

## Notes

### Overview
- This video presents an interview with ex-Google CEO Eric Schmidt about the AI race between the US and China, AI risks, and regulation.
- The host adds commentary between interview segments, including disagreement with Schmidt on open-source AI.

### US vs China in AI
- Schmidt says it "damn looks like" the US will win the AI race toward AGI and superintelligence, because hardware restrictions from the Trump and Biden administrations limit China's access to chips like Google TPUs.
- After visiting Shanghai, he concluded China is playing a different game, applying AI in a more classical way to every product and service.
- He once feared a superintelligence gap could trigger a preemptive attack between nations, but now thinks "we're going to be okay for a few years."

### Robotics and Energy
- China dominates robotics the way it dominated EVs; Unitree launched the R1 humanoid robot for $6,000, and Schmidt ordered one.
- China installed 172 gigawatts of solar last year and has solved its electric power problem, while the US has not.
- Schmidt testified to Congress that the US needs 92 gigawatts of new electricity by 2030 for data centers; a big nuclear plant is 1–1.5 gigawatts, and effectively zero are being started in America.
- The president's deals with Saudi Arabia and the UAE could mean American intelligence is trained in kingdoms.

### AI Threats and Regulation
- Drawing on Henry Kissinger's nuclear arms control analogy, Schmidt says treaties limiting enriched uranium kept us alive.
- Three current threats: misinformation (threatens democracies), cyber (if you can write code you can write attacks), and biological (a pathogen modified to be undetectable, buildable by three people in a basement).
- The Biden rule required reporting training runs over 10^26 flops; Schmidt helped pick that number arbitrarily. It was replaced by the David Sacks document focused on speed and winning.

### Open Source and Distillation
- Schmidt worries more about open-source groups operating in the shadows than big companies, since DeepSeek R1 and R2 reach 80–90% of top closed models.
- Distillation costs about 1% of original training; FP4 gives an 8x boost over FP32.
- He predicts the biggest US models will never be released, while China's biggest models stay open source and free, so most countries will standardize on Chinese models.

## Transcript

[00:00] But, you've been saying recently that, you know, one of the things that's inevitable is some kind of an AI disaster, and we're hoping that it's like 100 people that die, and not 1,000 or 10,000 or a million or or even 100 million. Uh but, you know, suppose that it does play out that there's a catastrophe that's a wake-up call. What do you want to do with that wake-up call? Um the one I'm particularly worried about is biological, and it goes something like this. You take some existing pathogen, and using biological techniques, which I won't discuss, you can modify it enough that it cannot be detected, but it's still quite dangerous. That's an example of a threat. Guys, Eric Schmidt just gave a

[00:40] pretty shocking interview about AI. In this interview, he talked about where AI is headed right now, and how dangerous it's it's going to get in the next 2 years. He warned that if governments of the US and China don't put any safeguards around AI, it's going to take over literally everything. Eric also talked about AI disasters, which could destroy our world if our governments don't regulate this industry right now. Guys, the audio of this interview is a little bit choppy in the first 2 minutes, but I guarantee that this interview is a real mind-opener. Let's watch the interview, and I'll explain everything as we go. >> So, are we going to win the AI race, and you know, what are the scenarios where

[01:17] we win and lose? It damn looks like we will. And uh let me define the way And so, I think the the separate that's so consensus, which is I call it, is what people in separate that's so believe, which charitably was you. Uh Kids, that I you only just be built from current magnetic computing to various forms of recursive self-improvement to eventual AGI and superintelligence. Um in order to do that, it requires that you use the enormous amount of hardware. Uh Google TPUs, the biggest chips, and so when everybody in the audience knows us that it sure looks like the hardware restrictions that the Trump and Biden administrations have put on China are going to prevent them from competing at that space. I was recently in Shanghai

[02:09] for a few days. I have good relationships with the Chinese. And my conclusion is they're they're they're fighting a different game. They're regular about AI in every product, every service, everything, but in a more classical way. Whereas America's been a secret for AGI. And I was quite worried that we would end up in a superintelligence phrase where you would end up with such enormous gains that one side would have to actually pre- would have to actually attack the other one. Uh and you know, there's a preemptive attack. It looks like that fear from me was not well grounded in fact. I think we're going to be okay for a few years. Really? What what about robotics? We seem to be pretty far behind in that

[02:51] front. Yeah, if if if it's okay for me to be completely blunt, um the Chinese are doing the same thing in robotics that they have done in electronic vehicles. Um in case you're confused while the current government in the US gets rid of solar and wind subsidies and promos, China put in 172 gigawatts of solar last year. I think it's the number. It's remarkable. So, the the Chinese race around solar and EVs, electronic vehicles of one kind or another, looks like they've won. They're using all of those technologies, in particular new ways of building stepper motors and other very inexpensive, very powerful, you know, physical thing. Good example is Unitree just launched the R1 uh for $6,000 uh available in December. I'm I've

[03:45] ordered one. We'll see how good it is. But the arrival of humanoid robots um is likely to be dominated by China. I'm not suggesting there won't be areas where the US will play. The US will still have spaces of very high-end, very sophisticated stuff. Our software is so much better than the Chinese software, but at the hardware level, I think you should assume that the world be awash in inexpensive Chinese robots. In the same sense that they'll be awash in inexpensive Chinese electric vehicles. Who knew? And so this is going to be my last question, but let me bump it to the front since you just beautifully segued into it. In the race for AGI, but then also in parallel robotics, China's going to be miles ahead in

[04:29] electricity production. And in the very short term, we're all going to be chip constrained, TPU constrained, but if you look 3-4 years out, they know that the fabs are running at full throttle, the chips are coming out by the millions, then you're suddenly electricity constrained. Is there a vulnerability for America there? It's a huge issue. So, again, let's use China versus the US as a metaphor. What are China's strengths? And I'm not praising China. I'm just be trying to report it. They have solved their electric power problem. They also have full control over social media, so they don't have the kind of problems that people here complain about. They have enormously talented software people, and they don't have enough

[05:12] hardware. I think that's roughly where they are. They're also incredible capitalists. They they call it Chinese Chinese socialism with socialism with Chinese characteristics, but trust me, it's pure raw capitalism. Let's call it what it is. In the US, we have the many benefits everybody understands. We do not have enough electricity, and our hard and our our at least in our consumer stuff, the Chinese are likely to beat us. Our hardware architectures are fantastic, and I'm including the Amazon chip, obviously the Nvidia chips, TPU V5, which I'm happy to say I was part of TPU V version 1. All of this stuff is incredible. In the following part, Eric talks about the lack of energy production in the United

[06:00] States. He says that the US is going to fall behind China if we don't generate enough electricity to pursue our AI goals. Guys, in my opinion, the way China is moving right now, I feel like we've already lost the AI race. I mean, the amount of electricity we produce is nowhere close to China. China has been upgrading its grids massively just to meet the energy demands of its data centers. So, I mean, the real question is, can the US build power plants fast enough to compete in the AI race? The answer, in my opinion, is no. So, if you think about it, and I testified in Congress a month or two ago on this, we looked at the amount of electricity required in the United States to power the expected demand

[06:45] of data centers, and we needed, in our calculation, by 2030, 92 gigawatts to be built. And for reference, a big nuclear power plant is somewhere between 1 and 1 and 1/2 gigawatts. To give you a sense of how many nuclear power plants are getting started in America, effectively zero. So, we had hoped, and I've hoped in my lobbying and test testifying that the government would fast track availability of all kinds of electricity. And indeed, they have promoted oil and gas, but they've also hobbled solar and wind to a terrible degree, which is an error. The country needs more energy. And if we don't get more energy, we're not going to be able to fully exploit the lead we have in AI and AGI. It's very clear.

[07:39] And by the way, the obvious next question is, what do we do? Well, there are scenarios. For example, the president went to Saudi and the UAE and did huge deals for multiple gigawatts. And so we might find ourselves in a situation where our training for our most important thing, the thing which are the the essence of America, which is American intelligence, is actually being developed in kingdoms. And that may be the only fallback we have. That is really weird, and that begs a very difficult question. Do you mind if I ask you a a tough one? Go ahead. Uh I've been, you know, uh kind of uh inspired by you for most of my adult life. And uh I I really have have been watching your uh podcast where you're

[08:22] talking about, look, when you you remember you when you guest lectured Erik Brynjolfsson's class over at Stanford, actually you said when you were running Google you felt like you made decisions three times faster than any company on the planet. But then when you got into the federal government, you felt like the decision making was 1/3 as fast as even a slow company. So from your point of view, it's like, you know, a tiny fraction of the pace that we need to move. >> Okay, next, Erik talks about some pretty serious stuff. He says we're going to see a lot of AI disasters if the governments around the world don't regulate AI companies. He warned that the pace at which AI is advancing, it has the potential to topple democracies.

[08:58] Guys, in my opinion, I believe that every major government around the world knows that AI is both a risk and a competitive advantage. If one country slows down, the other won't. That's why major players like the US and China are not putting any guardrails around their AI companies. They want that AI edge. >> But you've been saying recently that, you know, one of the things that's inevitable is some kind of an AI disaster, and we're hoping that it's like 100 people that die, and not 1,000 or 10,000 or a million or or even 100 million. Uh but, you know, suppose that it does play out that there's a catastrophe that's a wake-up call, what do you want to do with that wake-up call? What what's the next move for Eric

[09:42] Schmidt after the wake-up call? Well, so so the background here is that uh Dr. Kissinger, Henry, and I spent an awful lot of time talking about the period in the 1950s where he was a key component of all of these things. And the So so what he did was he he used the fact that we had used the nuclear bomb to negotiate over about a 15-year period a set of treaties that restricted nuclear proliferation. Those treaties when they were um uh when they were negotiated have allowed us to be alive today. So these were centrally important. Without but without controlling the spread of enriched uranium and the other secrets, we would all be toast, literally, because of crazy people and so forth. Is there an

[10:31] analogous set of things that we can do? The problem here is uh I guess the good news is we're not in a war, we haven't had a nuclear bomb, we don't have that thing to discuss. So we can talk about it, but governments tend to act reactively. So let me talk There are three obvious threats right now, which I think are fairly well understood. The first is misinformation and the software that we're all collectively giving people um allows for all sorts of misinformation, fake videos, fake news, what have you. We all understand this. It's all open source, that's done. That's a threat to democracies and maybe to dictatorships, certainly to democracies. The second one is cyber. And I think one way to understand cyber

[11:13] is that if you can write code, you can also write cyber cyber attacks. It's the same logic. And you have these incredible gains in software. It's frightening how good these soft Remember, my career is as a programmer. These things program better than I ever did. It's like shocking, right? And then the third one is bio. And I think most people believe that one of those three will create some kind of mini crisis that will then cause the governments to say, "Hang on. Let's have a conversation about how to really deal with the downsize." The upsides are incredible. Right? And I want America to win, and I want us to run as fast as we can, and we're indeed doing that with the Trump administration, which is great.

[11:56] Right? But we have to be aware that these possi- these things are possible. Um the one I'm particularly worried about is biological, and it goes something like this. You take some existing pathogen, and using biological techniques, which I won't discuss, you can modify it enough that it cannot be detected, but it's still quite dangerous. That's an example of a threat. There are many others. Yeah, I that's also the easiest, I think, which is scary. Um of the of the immediate, you know, CBRN threats, threats, so cyber, biological, radiological, nuclear. Uh biological is the one you can kind of do in a basement with three people, but it's also the one that's hardest to contain. So, anyway, uh I hope the theory is right.

[12:35] So, so then uh how do you deal with proliferation? You know, the Biden approach was, "Okay, let's contain AGI to these five companies, you know, Google being one of them, and then let's say any model with over 1E26 training flops has to register with the federal government, and then we'll keep it all contained." Uh so, that all got scrapped immediately, you know, after the election, and got replaced by the new David Sacks document, which the David Sacks document is much more about how do we move as quickly as possible and and win the race. Um but it doesn't really address proliferation, and obviously in America, we want startups uh and researchers to have incredible access to technology and compute. On the other hand, you know,

[13:19] the three people in a basement making a biological weapon is is a real scary thing. So, how do you balance those? Look, it turns out if you read the David Sacks President Trump announcement, they're very clear that they want to continue to study the security aspects of AI, especially in a geopolitical way. And that's code for China versus the US. And so they continue in their proposal to fund things involving nation-state attacks and so forth and so on. And I fully support that kind of stuff. Um there's probably a difference on the misinformation social media stuff between the two. But for example, the Biden rule was simply that you had to report if you're doing a a training run 10 to the 26 or

[14:02] greater. I was part of the group that made that number up. Uh we made it up cuz we had no better number. Uh I'm not suggesting it's the right number. And the the conclusion you come to is it's probably the case that that we know our government, I don't know this, but I'm guessing, knows where the training runs are going on in in China and because of espionage. And it's probably the case that the Chinese have espionage on the US knowing where our training runs are. So I'm not sure the nature of the training run runs is a secret. And frankly, everybody knows where the data centers are cuz they're immense. So uh we will see. Another attack on the 10 26 is that the training is getting more um more cost-efficient. If you look at the

[14:50] moving they move from something called FP16 to FP8, that means eight precision floating point. People are now moving to four four bit floating point, which is bizarre. It turns out these trainings algorithms seem to be quite tolerant for floating point in in precision, which is a shock to me. So again, we're getting more efficient in training. This creates more of a proliferation problem. Okay, next Eric talks about the threat from open-source AI. Guys, I don't agree with Eric here. I believe these tech CEOs are going after open-source models because they just want to protect their investments. Their goal here is to criminalize open-source AI. I don't think open-source is the real threat. The real

[15:32] threat is that one of the big corporations will eventually train a super-intelligent AI with the wrong terminal goal. If I can If I can say in general the proliferation issue, I'm not worried about big companies and big countries because I can count them. You know, there'll be 10 huge data centers, 10 huge training runs around the world. I'm much more worried about the open-source groups which can operate in the shadows. They don't have to solve every problem. They just have to do one thing well. They can patch together the open-source which is generally available and is good enough. If you look at the quality of uh deep seek R R2 R1 and now R2 coming, sure looks like it's in 80% or 90% of these

[16:14] top closed models. The closed models people, which obviously includes Google, get very defensive over this because they say, "Look, those are kind of synthetic. They have diffused the models. They've trained our best information." They're not All of that is true. They're correct. But they are nevertheless useful for specific things and they could be used in a proliferation issue to do various forms of cyber and biological attacks. Well, yeah, the current rule of thumb is that distillation and transfer learning is about 1% of the cost of the original training but gets you to the same destination. So, if you take, you know, FP4 which you mentioned, you know, you get an 8x performance boost over

[16:52] FP32. So, you get an 8x on the on the, you know, the the weights and then you've got another 100x on the transfer learning and it's really hard to draw an analogy to nuclear or chemical weapons in the past cuz you couldn't kind of shrink them a thousand x under the covers and get the same result. But AI is really weird that way. It's It's very compressible, very fluid. Uh well, again, we we we will It does not look like we have a very good solution for distillation. It looks like opponent of a company can mask their queries to look like normal sets of users and then distill the models. One of my friends has thought about this a lot and he thinks that the eventual state in the United States is that the biggest models will

[17:43] never be released. And that the companies will distill their own models down for that reason. That's his opinion. We'll see if that's true. And this produces a bizarre outcome where the biggest models in the United States are closed source and the biggest models in China are open source. And the geopolitical issue there, of course, is that open source is free and the closed source models are not free. And so the vast majority of governments and countries who don't have the kind of money that the West does and so forth will end up standardizing on Chinese models, not because they're better, but because they're free. And so we'll see if that's true or not, but I do worry about that. All right, guys. That's it

[18:25] for today's video. I'll see you guys again next week with another one.