Rep. Khanna on AI, China, and Industrial Policy
Plus: ChatGPT, TikTok, and NAIRR
What he hopes the China Committee can accomplish
Why ChatGPT let him down
What an effective industrial policy looks like
I’ll be in Phoenix this week! If you’re in town, respond to this email. With a quorum, I’ll try to get a group together.
ChinaTalk is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
The Case for an Economic Development Council
Divyansh Kaushik: The administration put forward what they call the “Modern American Industrial Strategy.” And Congress has put hundreds of billions of dollars toward this whole idea that we have to steer investments. Say, if ten or twenty years from now this were to fail, what would, in your view, be the biggest reason — and what, in your view, could be the biggest reason for its success?
Ro Khanna: The failure would be because we weren’t strategic enough. I’ve called, in a Foreign Affairs article and with Senator Rubio, for the creation of an “Economic Development Council” — like a National Security Council — that would get all of the agencies on the same page. Under the direction of the president, that council would make sure that we had a coordinated response and an expeditious response, that we were targeting places that had faced deindustrialization, and that we had a comprehensive approach — not just the financing of capital for factories, but making sure that we were streamlining permitting without compromising the environment, making sure that labor standards were being upheld, making sure we were investing in the workforce.
So I expect some of the things to succeed — but if there were failures, it would be because of the lack of coordination and planning.
Divyansh Kaushik: And what would be some of your reasons for its success?
Ro Khanna: I think the CHIPS and Science Act was very strong. Of course, I’m biased: I was an original author of it, with Senators Chuck Schumer and Todd Young and Representative Mike Gallagher. But there, you have the government cooperating with the private sector, financing new factories, and effort with the state government also investing in new workforce, and you have investments in new science and technology.
So I think that close collaboration between government, educational institutions, and the private sector — as well as labor — is so critical.
Jordan Schneider: You mentioned being strategic, and you listed a number of potential priorities for American industrial policy — but being strategic ultimately means making trade-offs.
So when you’re talking about regional diversity and looking at industry by industry, what is your preferred framework that you would want to impose on whoever is going to do an analysis that would try to maximize the return that Congress is hoping its money will end up bringing back to the American people and future economic growth?
Ro Khanna: I would look for a moonshot of economic revitalization in factory towns that have been shuttered in communities that have been decimated because of 70,000 factories closing and the offshoring of jobs. I would look at places that have faced that kind of economic devastation and say, “How can we have economic redevelopment there? And what kind of industries and supply chains can thrive in those places?”
The second thing I would look at is, “How do we lower our trade deficit?” — particularly with China, but also globally, and have that as a metric in our consideration.
And the third thing I would look at: “What are some of the key industries where new technology and innovations in manufacturing can be done in the United States?” Semiconductor chips, steel, aluminum, battery appliances all are candidates.
Divyansh Kaushik: Before, you mentioned financing and identifying these industries; you worked with Senator Rubio on this — you introduced the National Development Strategy and Coordination Act recently, and part of that gives $20 billion to treasuries and federal financing investments.
Could you share your vision with our audience? And if you don’t mind, could you also share your thoughts on how the private sector investors should view this opportunity of crowding and capital?
Ro Khanna: This is building on the work of the CHIPS Act. But the idea is that we should have an Economic Development Council that helps finance new industries in places that have been left out of the global economy — particularly in strategic industries — that have all of our federal agencies and government working together to streamline the process of building these new industries, and that is providing federal capital if the private sector is investing in America.
We’re not going to give a tax debt or subsidies for CEOs to pocket that money and then build the factories in Malaysia or Vietnam.
But if they’re building the factories in the United States, and if they can meet the rigors of the free market by getting some seed private capital, then that private capital can be scaled with the federal investment — and the Federal Financing Bank will be looked at as a scaling bank, one that allows market-tested technology that’s being invested in the United States to scale into factories.
Jordan Schneider: Could you talk a little bit about the appetite for bipartisan spending around these sorts of things? A decade ago, I think it was only Senator Rubio on the other side of the aisle — and now we’re having more and more folks of the Republican persuasion be open to this stuff.
Do you see this as something that could last decades or more? Or was there something particularly unique about the post-covid moment that allowed these bills to get across the finish line?
Ro Khanna: I do think there was a uniqueness to the post-covid moment where people said, “We didn’t make masks in America; we didn’t make enough baby formula in America; we don’t make enough Tylenol in America — what happened? How is it that we’re waiting and so dependent and that the trucks are lying vacant because we don’t have a semiconductor chip?” It jolted America into understanding the value of manufacturing and production.
There has been bipartisan activity — but the Republican party honestly is a bit divided, schizophrenic on this. You have Rubio, who takes it seriously on economic development vis-à-vis China. But you have people go on the Freedom Caucus who say, “I don’t want to spend a dollar — I don’t care what it’s for.” It’s a tension within the Republican party — which direction are they going to go: in smart government productive spending, or in vocal opposition to all government spending?
And that’s, I think, why Rubio and I are off by a factor of a couple of zeros. I want $2 trillion of this over ten years to really revitalize industry. He’s for $20 billion. I appreciate, intellectually, him joining this with some courage to get on a bipartisan bill — but to really do this in a way that people are going to feel, “We need to have the investment behind it.”
Jordan Schneider: Let’s talk a little bit about the Democratic caucus for a second. What are the different strains of how people are viewing these sorts of initiatives from your side of the aisle?
Ro Khanna: I think most of the Democratic caucus is fine on the research and technology aspects of it. There’s probably some division on, “How much do we want to support industry in doing it, and how much do we want to support the private sector in doing it? And how much do we just want these to be government jobs?”
I point to FDR’s famous speech where he talked about the right to a job. That was not a right to a job just in government — it was a right to a job in private, in mining, in construction. There were a lot of private industries when FDR talked about it.
Eighty-five percent of jobs in this country are private-sector jobs. We want to make sure that private-sector jobs are part of the development narrative — but they need to be conditioned on good labor standards.
But I think the big debate in our caucus is, “How much do we support a company like Intel in building these factories when it is a private-sector company? And what are the conditions limiting stock buybacks or for-labor pay that are critical?”
Divyansh Kaushik: Talking about supporting Intel on chips and science — last year when Congress was celebrating the CHIPS and Science Act, over four dozen national-security leaders wrote to Congress, including former Secretaries of Defense, Energy, Homeland Security, CIA, NSA directors, even a former Vice Chairman of the Joint Chiefs, who wrote, you know, “We need to pair this with talent reforms.” They specifically called on Congress to look at STEM immigration as a critical component for chips and science — and that clearly did not go anywhere last year.
But you’ve got a great bipartisan working relationship on the hill. Do you think that there is an opportunity to do some of that work in this Congress?
Ro Khanna: I’m going to try to build the Intel factories in Ohio to create the blue-collar jobs. You need to have those PhDs who understand semiconductor production — right now, many of them aren’t in the United States. I want massive investment in engineering programs, PhD programs, and STEM programs in the United States to revitalize our capacity to lead in these areas. But we also need to complement that with immigration. That will allow us to lead in manufacturing in this country.
Of course, it’s personal to me: my father immigrated to America, coming to Michigan, studying chemical engineering in the 1960s when America said we needed immigrants to win the Cold War.
So I will continue to work on immigration reform and link that to our manufacturing agenda, link it to our place-based agenda. We know that, when you add this kind of immigration with incredible talent into a community, it creates economic growth.
Divyansh Kaushik: China is projected to graduate twice as many PhD students by 2025 as compared to the United States. You just talked about the role of immigration reform — do you think there are additional steps Congress should be taking on STEM education for domestic students as well?
Ro Khanna: Yes, we need to make STEM part of national pride, national renewal. I think that starts with as simple things as the President of the United States recognizing — through prizes, through Oscar-like events — the value of STEM in this country.
And then of course, it needs to be financed. I would say we should be financing STEM PhDs: maybe they have a free education; maybe we pay for their living expenses or allow them to earn an income. We need to provide tax incentives for people to go into STEM fields, where they may be saving on their taxes to incentivize people doing these PhDs — instead of just having our best and brightest go to Wall Street or into software in Silicon Valley. We’re going to have to have a whole-of-nation approach in building the PhD and college STEM talent that need.
The House Select Committee on China
Ro Khanna: I voted for the Committee. I have a good relationship with Representative Gallagher: we came into Congress together; he’s a Marine, a PhD. We don’t always agree, but I think he’s going to be serious about the security risk that Taiwan poses, of making sure that there’s deterrence for any military invasion.
But I do want us to focus also on the economic issues, on the immigration issues. We need to make sure, if we are going to compete with China, that we have a rational immigration policy, that we have a rational policy on STEM education that allows for production to return, and that we rebalance our trade deficit. Then if leader Hakeem Jeffries selects me to be on that committee, I will be bringing that perspective to the committee.
Jordan Schneider: Want to talk about export controls for a little bit? Could you take the pulse on Congress — how the sorts of actions that BIS has been pushing have been taken, and what sorts of nudges Congress or the China Committee in particular might be interested in?
Ro Khanna: I think the export controls are important. I think that National Security Advisor Jake Sullivan had a very important contribution here, where he restricted advanced logic chips or the tools that make advanced logic trips from going to China.
What was so impressive about this is that he’s gotten commitments from the Netherlands and Japan. These restrictions have to be with our allies, or you’re just penalizing American companies and not achieving the goal. So I would focus on the export restrictions of the most highly sensitive technology.
On the other hand, we want to be exporting more of our products to China to reduce the trade deficit.
Divyansh Kaushik: When Chairman Michael McCaul was chairing the House GOP China Task Force, they made an offer initially to the Democratic Party as well to join, to put members on the task force; the leadership decided not to, so they went on their own. And Mike Gallagher has also said that he wants to use the work of that task force to guide his work on the Committee.
How do you see that playing into the potential bipartisanship on this committee?
Ro Khanna: A task force, I think, is very different than a committee of Congress. A committee of Congress has a much more formal standing.
And the good news so far is that the ratio, nine-to-seven, is a pretty fair ratio. So it suggests that they want Democratic participation. There are a number of areas where Democrats see eye-to-eye with the Republicans: that China should not in any way militarily invade Taiwan; that there needs to be an economic rebalancing; that there shouldn’t be any intellectual-property theft.
Now, there are places where Democrats want to be sure that we don’t engage in xenophobia, that we don’t engage in anti-China bashing, that we don’t push us toward a cold war, and we will bring that perspective — but I’m hopeful it’ll be a serious committee that looks at what the posture should be toward China in the twenty-first century, and will outlive this Congress and become a permanent select committee, much like the Intelligence Committee.
Divyansh Kaushik: Congress seems to focus a lot on disentangling the US from China. Do you see areas of cooperation where we need to double down instead?
Ro Khanna: There are obviously areas where we do need to cooperate: climate change, where we can; we need to work with China on tackling pandemics where we can; we need to have a dialogue about rules of transparency with the public-health agencies; we need to maintain dialogue so that there isn’t a risk of accidental war.
So I am for engagement, but with an engagement with open eyes and with a concrete goal of lowering our trade deficit to bring jobs home and not have this massive offshoring of our production.
Jordan Schneider: Coming to your Foreign Affairs piece — one of the arguments you make is that there is a universe in which Beijing potentially could get on board with an initiative that the US would push forward to lower the trade deficit.
Talk a little through the logic of what you hope US policymakers, American industry, as well as leadership in Beijing could go down to realize your vision of how this relationship could potentially evolve on a healthier track.
Ro Khanna: First, we need to demand that we’re going to lower the trade deficit — it’s gone from $60, $70 billion to almost $400 billion — and make that very clear to the Chinese government.
So when I met with the Chinese ambassador, I said that. I also said, to the extent we’re going to affirm our One-China Policy, that becomes easier to do if the trade deficit is lowered — where people in this country don’t think that their parents’ jobs are were shipped to China and their jobs are continuing to be shipped to China. One of the interests that China has is to have a more constructive relationship with the United States.
Second interest: they have a huge market in the United States. If we started imposing tariffs, that’s going to hurt them a lot more than it will hurt us, just because of the nature of the trade deficit.
The third incentive they have is the long-term. Xi Jinping has spoken about increasing consumer demand. They have an economy that is way too dependent on a few Communist-Party apparatchiks and factories, and not on a broad-based, consumer-demand-driven economy. Now they have a need for jobs, and that’s why they’ve been prioritizing it. But if they don’t increase consumer demand, that’s not a long-term sustainable strategy. So rebalancing, to some extent, is in their long-term interest.
Jordan Schneider: Do you think China could do anything over the near- to medium-term to change the temperature of how Congress sees China or the future of the relationship?
Ro Khanna: Yes: stop the artificial depreciation of their currency by buying up American reserves — and work with us to lower the trade deficit, to identify places where they would buy from American industry and where they would stop subsidizing cheap imports that are dumping into the US markets.
Jordan Schneider: If China’s growth slows into the medium-term — and we’re talking one- to three-percent growth over the next three to five years — does that make them a less scary competitor going forward?
Ro Khanna: Well, they’ve slugged about three percent for the first time. I don’t wish any country a lower growth that hurts their own people, and I don’t think that we should be counting on the other side failing — that would be like going into a football game saying, “We’re going to score only seven points because the other team is not going to score.” I think we need to make sure that we have a strategy that assumes that China will rebound and be.
Jordan Schneider: Any thoughts on TikTok? Is this something Congress should take on? Okay to leave to the executive branch?
Ro Khanna: I don’t think Congress should get involved. My view of the ideal solution is the forced sale of TikTok so that people can still use it, but their data isn’t going to the Chinese Communist Party.
Congress Meets ChatGPT
Jordan Schneider: Congressman, have you used ChatGPT?
Ro Khanna: I have.
Jordan Schneider: What’d you ask it to do?
Ro Khanna: I was candidly very unimpressed. I was on the floor of the House of Representatives, and Jonathan Jackson (brilliant guy, son of Jesse Jackson, a civil rights hero) told me that Thoreau — who I knew had influenced Gandhi’s civil disobedience — was actually influenced by the Bhagavad Gita. And I didn’t know that Jonathan Jackson knew that. And he said, “You should really read Thoreau’s writing — he was influenced by the Gita, and that informed his views on civil disobedience.”
So I was feeling like, “Well, what is this ChatGPT thing? Can it give me as much insight as Jonathan Jackson?” Jackson is not a philosopher, but he certainly understands the history of the civil rights movement.
And I asked ChatGPT, “In what way was Thoreau’s writing on civil disobedience influenced by the Gita?” I got an essay that I wouldn’t have given a passing grade if they had written that at Stanford. It was basically saying, “Thoreau looked to the Gita; Gandhi looked to the Gita” — dressed up with fancy words and paragraphs, but with no depth of analysis.
Then I realized, “If coders are writing how to do the searches, are the coders philosophers? Are the philosophers at the level that Jonathan Jackson is thinking about this?”
So I think it’s going to help people better than CliffNotes and probably write fine papers and probably do a lot of tasks — but it’s far away from deep thinking.
Jordan Schneider: In six months, we’ll ask it the same question and see if it does any better.
Why should people read the Gita?
Ro Khanna: It is one of the most profound books. It’s like the Bible. It’s full of lessons about life. The central lesson, of course, is to do your duty without concern for the external rewards, and that ultimately if you do that, the rewards won’t follow.
Next, we get into Rep. Khanna’s views on the National AI Research Resource Task Force as well as his most pressing analytical questions.