“Parasitic Technology”: Could Jensen Huang/Nvidia Scandal Prompt Rethink About AI?

AMY GOODMAN: We begin today’s show looking at the fallout from Jensen Huang’s resignation as the CEO of Nvidia, one of the world’s most powerful companies. Nvidia’s advanced AI computer chips have powered much of the artificial intelligence boom. Huang cultivated close ties to President Trump and is credited with helping to shape U.S. policies around artificial intelligence and rules around the export of advanced AI chips.

Huang’s resignation came days after reports emerged he had an affair with a Chinese businesswoman named Alice Huang — no relation — whom U.S. intelligence officials reportedly suspected of spying for Beijing and helping China obtain advanced Nvidia chips, which are banned by U.S. export controls. The Information news outlet reports Alice Huang helped route at least $7.5 billion worth of Nvidia chips to China via a Singapore-based cloud computing company called Megaspeed International. Jensen Huang confirmed he had an affair with the woman. Later, published reports revealed he also had affairs with two employees at Nvidia, including with a senior executive who reported to him.

Nvidia has announced its longtime chief financial officer, Colette Kress, would serve as interim CEO as the board searches for a permanent replacement.

On Thursday, U.S. Attorney General Todd Blanche took questions about Nvidia.

REPORTER: When did the Justice Department learn, with regard to Nvidia, that Alice Huang might be a suspected Chinese asset? And there’s been concerns raised that the Commerce Department was giving favoritism to Nvidia and Megaspeed, which is Alice Huang’s company. Do you share the concerns that maybe the Commerce Department was impeding the ability to uncover the smuggling network that Alice Huang was involved in?

ATTORNEY GENERAL TODD BLANCHE: Well, your first question assumes something that I have never said and that I don’t know to be true. So, you assume that DOJ learns that she is, in fact, a spy. I’m not — nobody said that. I certainly haven’t said that. And I don’t — all I said a minute ago, and all that I think I will say, is that, of course, DOJ is investigating or looking at it. So, there’s no — there’s no conclusions reached by the Department of Justice or by me or by Director Patel around who she is or is not.

I don’t have concerns with the Commerce Department. I don’t. We have a good relationship with the Commerce Department. I’ve read some of the stories, which, you know, are sourced or not sourced, and so I don’t have any — but, no, I don’t have — the answer to your question is, no, I don’t have any — I don’t have any concerns.

AMY GOODMAN: On Capitol Hill, Republican and Democratic lawmakers are calling for probes into Nvidia’s actions and the Trump administration’s policies regarding chip exports to China. Republican Senator Tim Scott, the chair of the Senate Banking Committee, has requested Jensen Huang and Commerce Secretary Howard Lutnick testify. Democratic lawmakers Ron Wyden of Oregon and Richard Neal of Massachusetts are investigating the Trump administration’s decision to allow Nvidia to sell advanced chips to China.

For more on Nvidia and questions about the concentration of power in the AI industry, we’re joined now by two guests. In Toronto, Canada, we’re joined by Paris Marx, a tech critic and host of the award-winning Tech Won’t Save Us podcast. He writes the Disconnect newsletter, author of the book Hyperscale: The Ambition and Excess of Big Tech’s Data Empires. And in Germany, we’re joined by Ramesh Srinivasan, professor of information studies, director of the AI Futures Lab at UCLA and host of the podcast Utopias.

Paris Marx and Ramesh Srinivasan, welcome back to Democracy Now! Professor Srinivasan, let’s begin with you. What should we understand about what has just taken place? And talk about the centrality of Nvidia to the whole AI bubble, to the whole AI industry.

RAMESH SRINIVASAN: Right. Thank you so much for having me, Amy.

AI is not magic. It comes from the planet, and the technologies that produce it often return to the planet as electronic waste. It’s powered by water, energy sources, labor and our data, and those come from us, from the planet itself. Right? So, what has occurred is this astonishing technology that reduces life to data and numbers, and reduces intelligence to pattern recognition, has been — has now been governed and run by just a few people on the planet, mainly men, and men that you’ve discussed, you know, commonly on Democracy Now!, who were there at the inauguration and so on.

So, now when we have soap opera sorts of moments, like Mr. Jensen Huang being connected to, you know, this Chinese — potential Chinese spy, various affairs and so on, you can see how precarious the governance of this AI truly is, because the administration has completely handed the keys to this incredibly powerful technology, that feeds off of all of us and the planet, to basically a few people, and they can do what they want — their vagaries, their mishaps, their corruption in certain cases, their gluttony. That is what is being accumulated more and more.

So, the story of AI is a story that’s tied to data centers — right? — which you cover all the time, which are, you know, computing and so on, much like server farms in the past. They’re tied to chips, and that’s very key to this story, the Nvidia chips, or superchips, that were originally used in the context of gaming and now were applied to artificial intelligence and large language models. But it also comes from all of our data, and then the screens and systems that determine and shape our lives.

So, this is a technology that is harvested, extracted from our planet and all of us to create potentially trillions of dollars of valuation, which might be a massive bubble — and we can talk about that — with returning almost next to nothing for the rest of us, other than anxiety, insecurity and deep concerns. The technology doesn’t have to go in that direction, but that’s the direction the Trump administration has taken it.

And I think it’s incredibly notable that it’s the China — it’s the China scare yet again, just like we saw with Huawei, just like we saw with COVID. It’s the China scare that’s forcing all of this action and this spectacle and this controversy, rather than the real conditions of AI, which are disenfranchising and robbing people of many aspects of their futures.

AMY GOODMAN: Paris Marx, if you could continue, talking about what is the role of Nvidia when it comes to, overall, fueling so much? And explain the issue of the ban on these — on these superchips, why they’re so significant, what kind of monopoly Nvidia has on this, and why you believe he is so close to President Trump, who didn’t want him to resign.

PARIS MARX: Absolutely. Great to join you this morning, Amy.

You know, Nvidia is key to this AI market. You know, it controls about 90% or so of the chips that go into, you know, all these data centers that are powering the AI boom that we have seen occurring now for the past few years, right? And as a result of that, Nvidia has become one of the most valued companies, and for a while has been, you know, the most valuable publicly traded company on the planet, which is just staggering to imagine, right? You know, it went from kind of just, you know, a regular old tech company, you know, not particularly notable — sure, it made some graphics cards for gamers — to being now one of the key companies that, as you say, you know, is so important for what so many other companies are doing.

And it’s not just the fact that it creates these chips — right? — these chips that are just essential to, you know, the generative AI workloads behind ChatGPT and the image generators and those sorts of things, but because Nvidia has made so much money for the past number of years, it has also become a major investor, plowing more than $120 billion into other AI companies across the economy, across the United States, around the world, which has meant that now there are all these companies that not just depend on Nvidia’s chips, but also depend on Nvidia’s financing. You know, I’m sure you’ve been talking on the show about the upcoming Anthropic IPO, you know, Anthropic looking to go public. It is expecting about $10 billion from Nvidia as part of its public offering. And so, Nvidia has been really key here.

And because of that, you know, because Nvidia was valued at something like more than $5 trillion until recently, it’s no surprise, then, that Jensen Huang has become very close with Donald Trump, right? It’s a very valuable company. It is a company that makes the technology that is very important to the broader AI boom that has been kind of lifting the U.S. economy and the U.S. stock market — right? — because of the investment, not just in all these AI products, but in all these data centers that, you know, have been being built across the country, as well. And so, Jensen Huang has become a major adviser to Donald Trump on AI policy and has certainly been pushing him to weaken the restrictions on, you know, basically, his ability to sell these chips to China.

And I think now in light of these revelations, there are questions as to whether, you know, that is kind of a firmly held belief that Jensen Huang had because there was obviously, you know, as someone who runs a major company, a big market opportunity in selling those chips to China, or whether there was influence on his thoughts, you know, from Alice Huang. And I think those are questions that we don’t have answered yet, but certainly is probably, you know, what some of these lawmakers are hoping to try to get more insight into by having Jensen Huang come speak to congresspeople.

AMY GOODMAN: Paris Marx, in your book, you quote one investor who describes Nvidia as “the central bank of AI.” And there are many descriptions of how Huang is really the conduit from the other tech billionaire bros to reaching President Trump and making the demands that they want.

PARIS MARX: Absolutely. You know, as I was mentioning, when we look at Nvidia, it is not just a company that is selling all these chips. It is a company that has gone to a lot of these other AI and data center companies and made sure that they had investment, sure, so that they could, you know, grow, expand their businesses, what have you, but also that they — so that they could then buy more Nvidia chips from Nvidia, right? So, it’s kind of like a circular financing scheme, where Nvidia gives it the money to operate, and then they buy the chips from Nvidia, and Nvidia makes more money in that way, right? So the money comes back to Nvidia all the time.

And then, of course, linked with that is it has also used that money to try to support companies that were in difficult situations. So, for example, a company called CoreWeave, that creates data centers to power AI and kind of sells that computing power to these AI companies, has been in trouble, on and off, for a number of years, and Nvidia has often been there to support that company when it has been in dire straits. And it’s not the only one like that.

And so, of course, you know, linked to that, Jensen Huang has developed this really close relationship with Donald Trump. You know, we saw this just the other day, just yesterday, after Jensen Huang resigned, and, you know, Donald Trump came out and said that he was talking to Jensen and, you know, that things are difficult, and, you know, basically, had this really kind of personal relationship to him, seems to be someone who he really trusts. And as a result of that, I’m sure that Donald Trump has been open when Jensen Huang has been discussing AI and tech policy with him.

And, of course, Jensen has been pushing to make sure that there is not regulation of AI products, in the way that your other guest has been talking about, you know, is really necessary, but then, on top of that, again, trying to weaken these restrictions on the ability to sell chips to China. And, of course, you know, the U.S. government’s concern and many lawmakers’ concern there is that by selling those chips, because they are some of the — because they are the most advanced chips that are out there, that if they get sold to China, you know, China will not just be able — Chinese companies will not just be able to use them to power their own AI products, but that there’s a concern that, you know, Chinese companies will be able to kind of reverse-engineer the technology and then have a big kind of leap in their — in their kind of chip-making capabilities, essentially.

AMY GOODMAN: In October of last year, Jensen Huang spoke to the media about whether China should have access to Nvidia’s most advanced chips.

JENSEN HUANG: The China market is large, and it’s a vibrant, vibrant market with a lot of developers. And those developers create AI technology that’s exported from China all over the world. And so, to the extent that American tech stack can run and operate those AI models, it’s good for the United States and around the world.

REPORTER: So, would you want the Blackwell to be sold in China?

JENSEN HUANG: I hope so. I hope so. But that’s a decision for President Trump to make.

AMY GOODMAN: So, Professor Ramesh Srinivasan, if you can explain what he is saying there? And also, this whole issue of the so-called Chinese spy, what evidence do we have of this? Is it other competing companies saying this? Is there a big division within the U.S. government in the intelligence community who had informed Trump of this months and months ago?

RAMESH SRINIVASAN: Right. It’s not clear what the evidence is as of this point. You could even see in the story, in the introduction to this story, that even the Attorney General Todd Blanche is, you know, very elusive and vague in his comments about, you know, what is — what exactly occurred here.

But I do think it’s very important to note a couple things that were already mentioned here. First of all, the AI industry, there’s incredible amount of concern that this is an overspeculated bubble. That’s because many of the revenues are occurring in the circular way that your other guest just described. So the revenues seem to be inflated here.

And talking about China, as we actually discussed before, Amy, back, you know, about almost two years ago, Chinese models around AI are performing nearly as well as the top American so-called frontier models. I mean, the question of what we’re building all these models for is a worthwhile question, I think a big elephant in the room. But they’re doing it at a fraction of the cost. So, we have to really consider what’s occurring here and what’s really occurring here, right?

So, the president’s response around all these concerns is often to say that the economy is super healthy. And when he says that, what he’s talking about — Trump, I’m saying — is he’s talking about the S&P 500. But the vast majority of people in America and around the world do not have large amounts of investment capital. Investment capital is how people make money, but there’s a deep amount of concern that these — this is not going to actually produce the type of revenues that even legitimate it for investors.

But what about the rest of us, as workers, as consumers? AI is being pushed out as a technology of replacement, not a technology to augment or expand or help us, you know, create together. It’s not a telescope. It’s not, you know, a microscope. It’s not what Steve Jobs once famously called a “bicycle for the mind.” It’s being — it’s using us. It’s surveilling us. It’s turning us into data. It’s turning life and intelligence into patterns to replace us, to read, write, speak for us. In fact, a CEO in Big Tech told me recently, because I’m from Silicon Valley and I have friends in that industry, “Hey, Ramesh, you don’t even have to do this interview with Democracy Now! Very soon, an AI can do that for you, and better than you.” What does that even mean? So, this is what’s at play here.

And so, this has been cast, as I alluded to earlier, as some sort of flat, you know, geopolitical battle between the U.S. and China, when, in reality, what we need is cooperation between the two nations, and, most importantly, outside of the elites in those two nations, governance by stakeholders, not just in the Global South and around the planet, where I’m part of a Global South AI alliance, but also from workers, consumers and so on, to ensure that a technology that uses patterns to predict the present and construct the future is a technology that’s applied to the right domains. And no one, at the end of the day, has been paid for their data. None of us have equity in any of these companies, but these companies are feeding off us to potentially replace us. And that’s what I would call a parasitic technology. But it doesn’t have to go in that direction.

AMY GOODMAN: Ramesh, can you show us your hands? AI models are notoriously bad at generating the correct number of fingers. I’m only kidding. OK, Paris Marx, Paris, this was from Bloomberg. You have Bloomberg in — let’s see, says, “Jefferies Financial Group Inc.’s Chris Wood said the most likely long-term outcome of the AI boom will be massive capital destruction in the U.S.” Do you see that coming now?

PARIS MARX: It’s hard to say when exactly that would come. You know, I think that there are multiple moments over the past few years where it looked like, you know, maybe — maybe the bubble is going to start bursting, and then, you know, it has continued. I think that it is facing a serious threat in this moment, and not just because of the resignation of Jensen Huang from Nvidia.

And to be clear, you know, Nvidia was a company that was really like Jensen Huang’s fiefdom. You know, it was not set up to have a lot of kind of people who could easily take over from him if he were to leave the company. You know, a number of his family members are also executives within the company. And so, you know, I think that there’s real concern about what this is going to mean for Nvidia’s future — right? — as such a key player in the AI market, you know, again, one of the most valuable companies in the world, supplying the key technology that these companies need, but also providing a lot of financing to other companies.

But we’ve also been seeing, through this year, a lot of major companies kind of pulling back on their generative AI spending, because they’re finding they’re not getting the return on investment from what they’re doing there. There was just a report at the end of October from an analyst at Apollo called Torsten Slok, and he basically looked at the S&P 500 to see, you know, whether companies had really been benefiting from this rollout of generative AI, and found that with the big seven tech companies, you could clearly see an increase in profit margins and in revenue per employee, you know, from rolling out generative AI products, but for those other 493 companies within the S&P 500, he did not see, you know, kind of like a notable difference or that generative AI had really benefited them.

So, there are growing questions there about the economic benefits of actually using this technology, not to mention, you know, kind of the many social harms that your other guest was talking about that need to be addressed and reined in. And so, you know, there are already — there’s already kind of skepticism and questions about all the money that has been put into these AI models, into these data centers in order to power them, whether they’re actually going to be as useful or whether the market for them is actually going to be as big as the companies are promising. And so, on top of those existing concerns, now these issues at Nvidia are kind of adding on top of that, which means that there is a potential kind of greater risk there, you know, of a collapse of some sort coming in the near future. But I certainly wouldn’t want to go as far as to call that this is the moment that it’s all going to happen.

AMY GOODMAN: And, Paris Marx, as you look down on us in the United States from Toronto, and across, of course, at Alaska, I wanted to get your comment on the midterms. You’ve written extensively about the anti-data center movement in the United States. It certainly fueled a lot of people, might have driven a lot of people to the polls. If you can also talk about the possibility of Democrats investigating tech CEOs? And, of course, right now we have the demand of Democrats and Republicans to investigate Jensen Huang.

PARIS MARX: Yeah, I would love to see it. Let’s bring on more investigations of the tech CEOs. Start with Jensen, and we can go from there.

But I do think that the midterms were not just a repudiation of the Republican Party and Donald Trump, but also of these tech CEOs that have been wielding so much power over, you know, our societies collectively for such a long period of time. And especially it feels like they really took this to a new level with what they’ve been doing the past few years — right? — not just the rollout of this generative AI technology that, as I was saying, is not as beneficial as they promised, but also had a lot of negative social and environmental consequences that came along with it, but then, paired with that, trying to push these massive hyperscale data centers into so many communities across the United States and around the world.

And so, what we saw is voters really push back against that. You know, so many races in the United States had a data center component to them. And if we look at key states like, say, Texas or Georgia or Ohio, we can see that data centers were really consequential in the battles that played out there, and that Democratic lawmakers recognized that, you know, going against data centers, promising moratoriums, was going to win the votes.

I think now the question is where this goes from here, right? Are we actually going to see those Democrats follow through on those promises? Is the pressure going to keep up? Or is the tech industry, which I’m sure it is, you know, about to pivot at this moment, hiring a ton of lobbyists to try to find a way to shape some of what Democrats are going to try to do, try to shape that kind of tech regulation, data center regulation agenda to make sure it’s not going to be as biting as it might otherwise appear? And so, I think that’s what I’ll be watching, you know, going forward, how this is going to play out.

AMY GOODMAN: Well, Paris Marks, I want to thank you for being with us, author of Hyperscale: The Ambition and Excess of Big Tech’s Data Empires, speaking to us from Toronto, and Ramesh Srinivasan, director of the AI Futures Lab at UCLA, today speaking to us from Germany. Thanks so much, both, for being with us.

Coming up, an Israeli strike on Nasser Hospital in Gaza has killed at least six people, including a surgeon. We’ll speak to Dr. Feroze Sidhwa, who volunteered at Nasser and worked with Dr. Abu Shaira. Stay with us.

[break]

AMY GOODMAN: “Garden” by Nicolás Jaar.