Today, I’m talking with Arati Prabhakar, the manager of the White home Office of discipline and Technology Policy. That’s a cabinet-level position, where she works as the chief discipline and tech advisor to president Joe Biden. She’s besides the first female to hold the position, which she took on in 2022.
Arati has a long past of working in government: she was the manager of the National Institute of Standards and Technology, and she headed up the defence Advanced investigation Projects Agency (DARPA) for 5 years during the Obama administration. In between, she spent more than a decade working at respective Silicon Valley companies and as a venture capitalist, so she has extended experience in both the public and private sectors.
Arati and her squad of about 140 people at the OSTP are liable for advising the president on large developments in discipline as well as major innovations in tech, much of which comes from the private sector. That means guiding regulatory efforts, government investment, and setting priorities around big-picture projects like Biden’s cancer moonshot and combating climate change.
Listen to Decoder, a show hosted by The Verge’s Nilay Patel about large ideas — and another problems. Subscribe here!
You’ll hear Arati and me talk about that pendulum swing between public and private sector R&D — how that affects what gets funded and what doesn’t and how she manages the tension between the hyper-capitalist needs of manufacture and the public interest of the national government.
We besides talked a lot about AI, of course. Arati was notably the first individual to show ChatGPT to president Biden; she has a comic communicative about how they had it compose song lyrics in the kind of Bruce Springsteen. But the OSTP is besides now helping guide the White House’s approach to AI safety and regulation, including Biden’s AI executive order last fall. Arati and I talked at dimension about how she personally assesses the risks posed by AI, in peculiar around deepfakes, and what effect large tech’s frequently self-serving relation to regulation might have on the current AI landscape.
Another large interest area for Arati is semiconductors. She got her PhD in applied physics, with a thesis on semiconductor materials, and erstwhile she arrived on the occupation in 2022, Biden had just signed the CHIPS Act. I wanted to know whether the $52 billion in government subsidies to bring chip manufacturing back to America is starting to show results, and Arati had a lot to say on the strength of this kind of legislation.
One note before we start: I sat down with Arati last month, just a couple of days before the first presidential debate and its aftermath, which swallowed the full news cycle. So you’re going to hear us talk a lot about president Biden’s agenda and the White House’s policy evidence on AI, among another topics, but you’re not going to hear anything about the president, his age, or the presidential campaign.
Okay, OSTP manager Arati Prabhakar. Here we go.
This transcript has been lightly edited for dimension and clarity.
Arati Prabhakar. You are the manager of the White House’s Office of discipline and Technology Policy and the discipline and technology advisor to the president. Welcome to Decoder.
It’s large to be with you.
I am truly excited to talk to you. There’s quite a few discipline and technology policy to talk about right now. We’re besides entering what promises to be a very contentious election period where I think any of these ideas are going to be up for grabs, so I want to talk about what is politicized, what is not, and where we might be going. But just let’s start at the start. For the listener, what is the Office of discipline and Technology Policy?
We’re a White home office with 2 roles. 1 is whatever the president needs advice or aid on that relates to discipline and technology, which is in everything. That’s part one. Part 2 is reasoning about working on nurturing the full innovation strategy in the country, especially the national component, which is the R&D that’s done across virtually dozens of national agencies. any of it’s for public missions. quite a few it forms the foundation for everything else in the innovation ecology across this country. That’s a immense part of our regular work. And as we do that, of course what we’re working on is how do we solve the large problems of our time, how do we make certain we’re utilizing technology in ways that build our values.
That’s a large remit. erstwhile people think about policymaking right now, I think there’s quite a few focus on legislature or possibly state-level legislatures. Which part of the policy puzzle do you have? What are you able to most straight affect?
I’ll tell you how I think about it. The reason I was so excited erstwhile the president asked if I would do this occupation a couple of years ago is due to the fact that my individual experience has been working in R&D and in technology and innovation from lots of different vantage points. I ran 2 very different parts of national R&D. In between, I spent 15 years in Silicon Valley at a couple of companies, but most of that was early-stage venture capital. I started a nonprofit.
What I learned from all of that is that we do immense things in this country, but it takes all of us doing them together — the immense advances that we’ve made in the information revolution and in now fighting climate change and advancing American health. We know how amazing R&D was for everything that we did in the last century, but this century’s got any different challenges. Even what national safety looks like is different present due to the fact that the geopolitics is different. What it means to make chance in all part of the country is different today, and we have challenges like climate change that people weren’t focused on last century, even though we now want that they had been.
How do you aim innovation at the large aspirations of today? That’s the organizing principle, and that’s how we set priorities for where we focus our attention and where we work to get innovation aimed in the right direction and then cranking.
Is that the lens: innovation and forward-thinking? That you request to make any discipline and technology policy, and all that policy should be directed at what’s to come? Or do you think about what’s happening right now?
In my view, the intent of R&D is to aid make options so that we can choose the future that we truly want and to make that possible. I think that should be the eventual objective. The work gets done today, and it gets done in the context of what’s happening today. It’s in the context of today’s geopolitics. It’s in the context of today’s powerful technologies, AI among them.
When I think about the national government, it’s this large complicated bureaucracy. What buttons do you get to push? Do you just get to spend money on investigation projects? Do you get to tell people to halt things?
No, I don’t do that. erstwhile I ran DARPA [Defense Advanced investigation Projects Agency] or erstwhile I ran the National Institute of Standards and Technology (NIST) over in the Commerce Department, I ran an agency, and so I had an aligned position, I had a budget, I had a bunch of responsibilities, and I had a blast working with large people and getting large things done. This is simply a different job. This is simply a staff occupation to the president first and foremost, and so this is simply a occupation about looking across the full system.
We actually have a very tiny budget, but we worry about the full picture. So, what does that actually mean? It means, for example, helping the president find large people to lead national R&D organizations across government. It means keeping an eye out on where shifts are happening that request to inform how we do research. investigation safety is simply a challenge present that, due to geopolitics and any of the issues with countries of concern, is going to have an impact on how universities conduct research. That’s something that we will take on working with all the agencies who work with universities.
It’s those kinds of cross-cutting issues. And then erstwhile there are strategical imperatives — whether it’s wrangling AI to make certain we get it right for the American people, whether it’s figuring out if we’re doing the work we request to decarbonize the economy fast adequate to meet the climate crisis, or are we doing the things across everything it takes to cut the cancer death rate in half as fast as the president is pounding the table forward with his cancer moonshot — we sit in a place where we can look at all the puzzle pieces, make certain that they’re working together, and make certain that the gaps are getting addressed, either by the president or by Congress.
I want to draw a line here due to the fact that I think most people think that the people working on tech in the government are actually affecting the functions of the government itself, like how the government might usage technology. Your function seems a small more external. This is actually the policy of how technology will be developed and deployed across private manufacture or government, over time externally.
I would call it integrative due to the fact that we’re very fortunate to have large technologists who are building and utilizing technology inside the government. That’s something we want to support and make certain is happening. Just as an example, 1 of our responsibilities for the AI work has been an AI talent surge to get the right kind of AI talent into government, which is now happening. Super breathtaking to see. But our day occupation is not that. It’s actually making certain that the innovation enterprise is robust and doing what it truly needs to do.
How is your squad structured? You’re not out there spending a bunch of money, but you have different focus areas. How do you think about structuring those focus areas, and what do they deliver?
Policy teams, and they’re organized specifically around these large aspirations that are the intent for R&D and innovation. We have a squad focused on wellness outcomes, among another things, that runs the president’s Cancer Moonshot. We have a squad called Industrial Innovation that is about the fact that we now have, with this president, a very powerful industrial strategy that is revitalizing manufacturing in the United States, building our clean energy technologies and systems, that’s bringing leading-edge semiconductor manufacturing back to the United States. So, that’s an office that focuses on the R&D and all of that large image of industrial revitalization that’s going on.
We have another squad that focuses on climate and the environment, and that 1 is about things like making certain we can measurement greenhouse gases appropriately. How do we usage nature to fight climate change? And then we have a squad that’s focused on national safety just as you would expect, and each of those is simply a policy team. In each 1 of those, the leader of that organization is typically an highly experienced individual who has frequently worked inside and outside of government. They know how the government works, but they besides truly realize what it is the country’s trying to achieve, and they’re knitting together all the pieces. And then again, where there are gaps, where there are fresh policies that request to be advanced, that’s the work that our teams do.
Are you making direct policy recommendations? So, the environment squad is saying, “Alright, all company in the country has promised a million trees. That’s great. We should incentivize any another behaviour as well, and then here’s a plan to do that.” Or is it broader than that?
The way policies get implemented can be everything from agencies taking action within the laws that they live under, within their existing resources. It can be an executive order where a president says, “This is an urgent matter. We request to take action.” Again, it’s under existing law, but it’s the chief executive, the president, saying, “We request to take action.” Policy can be advanced through legislative proposals where we work with legislature to make something decision forward. It’s a substance of what it takes to get what we truly need, and frequently we start with actions within the executive branch, and then it expands from there.
How large is your office right now?
We’re about 140 people. Almost all of our squad is people who are here on item from another parts of government, sometimes from nonprofits outside of government or universities. The organization was designed that way because, again, it’s integrative. You gotta have all of these different perspectives to be able to do this work effectively.
You’ve had quite a few roles. You led DARPA. That’s a very executive function within the government. You get to make decisions. You’ve been a VC. What’s your framework now for making decisions? How do you think about it?
The first question is what does the country request and what does the president care about? Again, quite a few the reason I was so excited to have this opportunity… by the time I came in, president Biden was well underway. I had my interview with him almost precisely 2 years ago — the summertime of 2022. By then, it was already truly clear, number one, that he truly values discipline and technology due to the fact that he’s all about how we build the future of this country. He understands that discipline and technology is simply a key ingredient to doing large things. Number two, he was truly changing infrastructure: clean energy, gathering the climate crisis, dealing with semiconductor manufacturing. That was so breathtaking to see after so many decades. I’ve been waiting to see those things happen. It truly gave me quite a few hope.
Across the line, I just saw his priorities truly reflected what I profoundly and passionately thought was so crucial for our country to meet the future effectively. That’s what drives the prioritization. Within that, I mean it’s like any another occupation where you’re leading people to effort to get large hard things done. Not surprisingly, all year, I make a list of the things we want to get done, and through the year, we work to see what kind of advancement we’re making, and we win wildly on any things, but sometimes we neglect or the planet changes or we gotta take another run at it. But overall, I think we’re making immense progress, and that’s why I’m inactive moving to work.
When you think about places you’ve succeeded wildly, what are the biggest wins you think you’ve had in your tenure?
In this role, I’ll tell you what happened. As I showed up in October of 2022 for this job, ChatGPT showed up in November of 2022. Not surprisingly, I would say mostly my first year got hijacked by AI but in the best possible way. First, due to the fact that I think it’s an crucial minute for society to contend with all the implications of AI, and secondly, because, as I’ve been doing this work, I think quite a few the reason AI is specified an crucial technology in our lives present is due to its breadth. Part of what that means is that it is definitely a disruptor for all another major national ambition that we have. If we get it right, I think it can be a immense accelerator for better wellness outcomes, for gathering the climate crisis, for everything that we truly gotta get done.
In that sense, even though quite a few my individual focus was on AI matters and inactive is, that continues. While that was going on, I think we continued with my large team. We continued to make good advancement on all the another things that we truly care about.
Don’t worry, I’m going to ask quite a few AI questions. They’re coming, but I just want to get a sense of the office due to the fact that you talked about coming in ’22. That office was in a small bit of turmoil, right? Trump had underfunded it. It had gone without any leadership for a minute. The individual who preceded you left due to the fact that they had contributed to a toxic workplace culture. You had a chance to reset it, to reboot it. The way it was was not the way anybody wanted it to be and not for any time. How did you think about making changes to the organization at that minute in time?
Between the time my predecessor left and the time I arrived, many months had passed. What was so fortunate for OSTP and the White home and for me is that Alondra Nelson stepped in during that time, and she just poured love on this organization. By the time I showed up, it had become — again, I would tell you — a very healthy organization. She gave me the large gift of a immense number of truly smart, committed people who were coming to work with real passion about what they were doing. From there, we were able to build. We can talk about technology all day long, but erstwhile I think about the most meaningful work I’ve always done in my professional life, it’s always about doing large things that change the future and improve people’s lives.
The satisfaction comes from working with large people to do that. For me, that is about infusing people with this passion for serving the country. That’s why they’re all here. But there’s a live conversation in our hallways about what we feel erstwhile we walk outside the White home gates, and we see people from around the country and around the planet looking at the White home and the sense that we all share that we’re there to service them. Those things are why people work here, but making that a live part of the culture, I think it’s crucial for making it a rich and meaningful experience for people, and that’s erstwhile they bring their best. I feel like we’ve truly been able to do that here.
You might describe that feeling, and I’ve felt it, too, as patriotism. You look at the monuments in DC, and you feel something. 1 thing that I’ve been paying attention to a lot late is the back-and-forth between the national government spending on research, private companies spending on research. There’s a beautiful tremendous delta between the sums. And then I see the tech companies, peculiarly in AI, holding themselves out as national champions. Or you see a VC firm like Andreessen Horowitz, which did not care about the government at all, saying that its policy is America’s policy.
Is that part of your remit to balance out how much these companies are saying, “Look, we are the national champions of AI or chip manufacturing,” or whatever it might be, “and we can plug into a policy”?
Well, I think you’re talking about something that is very much my day job, which is knowing innovation in America. Of course, the national component of it, which is integral, but we gotta look at the full due to the fact that that’s the ecosystem the country needs to decision forward.
Let’s zoom back for a minute. The pattern that you’re describing is something that has happened in all industrializing economy. If you go back in history, it starts with public investment and R&D. erstwhile a country is wealthy adequate to put any resources into R&D, it starts doing that due to the fact that it knows that’s where its growth and its prosperity can come from. But the point of doing that is actually to seed private activity. In our country, like many another developed economies, the minute came erstwhile public backing of R&D, which continued to grow, was surpassed by private investment in R&D. Then private investment, with the intensification of the innovation economy with the information technology industries, just took off, and it’s been amazing and truly large to see.
The most fresh numbers — I believe these are from 2021 — are something like $800 billion a year that the United States spends on R&D. Overwhelmingly, that is from private industry. The fastest growth has come from manufacture and specifically from the information technology industries. another industries like pharmaceuticals and manufacturing are R&D-intensive, but their pace of growth has been just… the IT industries are wiping out everyone else’s growth [by comparison]. That’s huge. 1 aspect of that is that’s where we’re seeing these large tech companies plowing billions of dollars into AI. If that’s happening in the world, I’m glad it’s happening in America, and I’m glad that they’ve been able to build on what has been decades now of national investigation and improvement that laid the groundwork for it.
Now, it does then make a full fresh set of issues. That really, I think, comes to where you were going due to the fact that let’s back up. What is the function of national R&D? Number one, it is the R&D you request to accomplish public missions. It’s the “R” and the “D,” product development, that you request for national security. It’s the R&D that you request for health, for gathering the climate crisis. It’s all the things that we’ve been talking about. It’s besides that, in the process of doing that work, part of what national R&D does is to lay a very broad foundation of basic investigation due to the fact that that’s crucial not just for public missions, but we know that that’s something that supports economical growth, too. It’s where students get educated. It’s where the fundamental investigation that’s broadly shared through publications, that’s a foundation that manufacture counts on. Economics has told us forever that that’s not returns that can be appropriated by companies, and so it’s so crucial for the public sector to do that.
The question truly becomes then, erstwhile you step back and you say this immense growth in private sector R&D, how do we keep national R&D? It doesn’t should be the biggest for sure, but it surely should be able to proceed to support the growth and the advancement that we want in our economy, but then besides broadly across these public missions. That’s why it was a precedence for the president from the beginning, and he made truly good advancement the first couple of years in his administration on building national R&D. It grew reasonably substantially in the first couple of budget cycles. Then with these Republican budget caps from Capitol Hill in the last cycle, R&D took a hit, and that’s actually been a large problem that we are focused on.
The irony is that we’ve actually cut national R&D in this last cycle in a time in which our primary economical and military emerging competitor is the People’s Republic of China (PRC). They boosted R&D by 10 percent while we were cutting. And it’s a time erstwhile it’s AI jump ball due to the fact that quite a few AI advances came from American companies, but the advantages are not limited to America. It’s a time erstwhile we should be doubling down, and we’re doing the work to get back on track.
That is the national champion’s argument, right? I perceive to OpenAI, Google, or Microsoft, and they say, “We’re American companies. We’re doing this here. Don’t regulate us so much. Don’t make us think about compliance costs or safety or anything else. We’ve got to go win this fight with China, which is unconstrained and spending more money. Let us just do this. Let us get this done.” Does that work with you? Is that argument effective?
First of all, that’s not truly what I would say we’re hearing. We hear quite a few things. I mean, astonishingly, this is an manufacture that spends quite a few time saying, “Please do regulate us.” That’s an interesting situation, and there’s a lot to kind out. But look, I think this is truly the point about all the work we’ve been doing on AI. It truly started with the president and the vice president recognizing it as specified a consequential technology, recognizing promise and peril, and they were very clear from the beginning about what the government’s function is and what governance truly looks like here.
Number 1 is managing its risks. And the reason for that is number two, which is to harness its benefits. The government has, I think, 2 very crucial roles. It was visible and apparent even before generative AI happened, and it’s even more so now that the breadth of applications each come with a bright side and a dark side. So, of course, there are issues of embedded bias and privacy vulnerability and issues of safety and security, issues about the deterioration of our information environment. We know that there are impacts on work that have started and that it will continue.
Those are all issues that require the government to play its role. It requires companies, it requires everyone to step up, and that’s quite a few the work that we have been doing. We can talk more about that, but again, in my mind, and I think for the president as well, the reason to do that work is so that we can usage it to do large things. any of those large things are being done by manufacture and the fresh markets that people are creating and the investment that comes in for that, as long as it’s done responsibly, we want to see that happen. That’s good for the country, and it can be good for the planet as well.
But there are public missions that are not going to be addressed just by this private investment that are yet inactive our responsibility. erstwhile I look at what AI can bring to each of the public missions that we’ve talked about, it’s everything from weather forecasting to [whether] we yet realize the promise of education tech for changing outcomes for our kids. I think there are ways that AI opens paths that weren’t available before, so I think it’s incredibly crucial that we besides do the public sector work. By the way, it’s not all just utilizing an LLM that someone’s been developing commercially. These are a very different array of technologies within AI, but that has to get done as well if we’re truly going to win and thrive in this AI era.
When you say these companies want to be regulated, I’ve definitely heard that before, and 1 of the arguments they make is if you don’t regulate us and we just let marketplace forces push us forward, we might kill everyone, which is simply a truly incredible argument all the way through: “If we’re not regulated, we won’t be able to aid ourselves. Pure capitalism will lead to AI doom.” Do you buy that argument that if they don’t halt it, they’re on a way toward the end of all humanity? As a policymaker, it feels like you request to have a position here.
I’ve got a position on that. First of all, I am struck by the irony of “it’s the end of the world, and so we gotta drive.” I hear that as well. Look, here’s the thing. I think there’s a very garbled conversation about the implications, including safety implications, of AI technology. And, again, I’ll tell you how I see it, and you can tell me if it matches up to what you’re hearing.
Number one, again, I start with the breadth of AI, and part of the cacophony in the AI conversation is that everyone is talking about the part of it that they truly care about, whether it’s bias in algorithms. If that’s what you care about, that’s killing people in your community, then, yes, that’s what you’re going to be talking about. But that’s actually a very different issue than misinformation being propagated more effectively. All of those are different issues than what kinds of fresh weapons can be designed.
I find it truly crucial to be clear about what the circumstantial applications are and the ways that the wheels can come off. I think there’s a tendency in the AI conversation to say that, in any future, there will be these devastating harms that are possible or that will happen. The fact of the substance is that there are devastating harms that are happening today, and I think we shouldn’t pretend that it’s only a future issue. The 1 I will mention that’s happening right now is online degradation, especially of women and girls. The thought of utilizing nonconsensual intimate imagery to truly just ruin people’s lives was around before AI, but erstwhile you have image generators that let you to make deepfake nudes at a tremendous rate, it looks like this is actually the first manifestation of an acceleration in harms as opposed to just risks with generative AI.
The machines don’t gotta make immense advances in capability for that to happen. That’s a present problem, and we request to get after it right now. We’re not philosophers; we’re trying to make policies that get this right for the country. For our work, I think it’s truly crucial to be clear about the circumstantial applications, the risks, the potential, and then take actions now on things that are problems now and then lay the ground so that we can avoid problems to the top degree possible going forward.
I hear that. That makes sense to me. What I hear frequently in opposition to that is, “Well, you could do that in Photoshop before, so the rules should be the same.” And then, to me at least, the difference is, “Well, you couldn’t just open Photoshop and tell it what you wanted and get it back.” You had to know what you’re doing and that there was a rate limiter there or a skill limiter there that prevented these bad things from happening at scale. The problem is I don’t know where you land the policy to prevent it. Do you tell Adobe not to do it? Do you tell Nvidia not to do it? Do you tell Apple not to do it at the operating strategy level? Where do you think, as a policymaker, those restrictions should live?
I’ll tell you how we’re approaching that circumstantial issue. Number one, the president has called on legislature for government on privacy and on protecting our kids most peculiarly as well as broader government on AI risks and harms. And so any of the answer to this question requires government that we request for this problem, but besides for—
Right, but is the government aimed at just the user? Are we just going to punish the people who are utilizing the tools, or are we going to tell the toolmakers they can’t do the thing?
I want to reframe your question into a strategy due to the fact that there’s not 1 place that this problem gets fixed, and it’s all the things that you were talking about. any of the measures — for example, protecting kids and protecting privacy — require legislation, but they would have a broad inhibition of the kind of accelerated spread of these materials. In a very different act that we did late working with the sex policy council here at the White House, we put out a call to action to companies due to the fact that we know the legislation’s not going to happen overnight. We’ve been hoping and wishing that legislature could decision on it, but this is simply a problem that’s right now, and the people who can take action right now are companies.
We put out a call to action that called on payment processors and called on the platform companies and called on the device companies due to the fact that they each have circumstantial things that they can do that don’t magically solve the problem but inhibit this and make it harder and can reduce the spread and the volume. Just as an example, payment processors can have terms of service that say [they] won’t supply payment processing for these kinds of uses. any actually have that in their terms of service. They just request to enforce it, and I’ve been happy to see a consequence from the industry. I think that’s an crucial first step, and we’ll proceed to work on the things that might be longer-term solutions.
I think everyone looks for a silver bullet, and almost all 1 of these real-world issues is something where there is no 1 magic solution, but there are so many things you can do if you realize all the different aspects of it — think of it as a systems problem and then just start shrinking the problem until you can choke it, right?
There’s a part of me that says, in the past of computing, there are very fewer things the government says I cannot do with my MacBook. I buy a MacBook or I buy a Windows laptop and I put Linux on it, and now I’m just beautiful much free to run whatever code I want, and there’s a very, very tiny list of things I’m not allowed to do. I’m not allowed to counterfeit money with my computer. At any layers of the application stack, that is prevented. Printer drivers won’t let you print a dollar bill.
Once you grow that to “there’s a bunch of stuff we won’t let AI do, and there are open-source AI models that you can just go get,” the question of where do you actually halt it, to me, feels like it requires both a cultural change in that we’re going to regulate what I can do with my MacBook in a way that we’ve never done before, and we might gotta regulate it at the hardware level due to the fact that if I can just download any open-source AI model and tell it to make me a bomb, all the remainder of it might not matter.
Hold on that. I want to pull you up out of the place that you went for a minute due to the fact that what you were talking about is regulating AI models at the software level or at the hardware level, but what I’ve been talking about is regulating the usage of AI in systems, the usage by people who are doing things that make harm. Let’s start with that.
If you look at the applications, quite a few the things that we’re worried about with AI are already illegal. By the way, it was illegal for you to counterfeit money even if there wasn’t a hardware protection. That’s illegal, and we go after people for that. Committing fraud is illegal, and so is this kind of online degradation. So, where things are illegal, the issue is 1 of enforcement due to the fact that it’s actually harder to keep up with the scale of acceleration with AI. But there are things that we can do about that, and our enforcement agencies are serious, and there are many examples of actions that they’re taking.
What you’re talking about is simply a different class of questions, and it’s 1 that we have been grappling with, which is what are the ways to slow and possibly control the technology itself? I think, for the reasons you mentioned and many more, that’s a very different kind of challenge because, at the end of the day, models are a collection of weights. It’s a bunch of software, and it may be computationally intensive, but it’s not like controlling atomic materials. It’s a very different kind of situation, so I think that’s why that’s hard.
My individual view is that people would love to find a simple solution where you corral the core technology. I actually think that, in addition to being hard to do for all the reasons you mentioned, 1 of the persistent issues is that there’s a bright and dark side to almost all application. There’s a bright side to these image generators, which is phenomenal creativity. If you want to build biodesign tools, of course a bad actor can usage them to build biological weapons. That’s going to get easier, unfortunately, unless we do the work to lock that down. But that’s actually going to gotta happen if we’re going to solve vexing problems in cancer. So, I think what makes it so complex is recognizing that there’s a bright and a dark side and then uncovering the right way to navigate, and it’s different from 1 application to the next.
You talk about the shift between public and private backing over time, and it moves back and forth. Computing is mostly the same. There are open eras of computing and closed eras of computing. There are more controlled eras of computing. It feels like, with AI, we are headed toward a more controlled era of computing where we do want powerful biodesign tools, but we might only want any people to have them. As opposed to, I would say, up until now, software’s been beautiful widely available, right? fresh software, fresh capabilities hit, and they get beautiful broadly distributed right away. Do you feel that same shift — that we might end up in a more controlled era of computing?
I don’t know due to the fact that it’s a live topic, and we’ve talked about any of the factors. 1 is: can you actually do it, or you’re just trying to hold water in your hand and it’s slipping out? Secondly, if you do it effectively, no action comes without a cost. So, what is the cost? Does it slow down your ability to plan the breakthrough drugs that you need? Cybersecurity is the classical example due to the fact that the exact same advanced capabilities that let you to find vulnerabilities quickly, if you are a bad guy, that’s bad for the world, if you’re uncovering those vulnerabilities and patching them quickly, then it’s good for the world, but it’s the same core capability. Again, I think it’s not yet clear to me how this will play out, but I think it’s a tough road that everyone’s trying to kind out right now.
One of the things about that road that is fascinating to me is there seems to be a core presumption baked into everyone’s intellectual models that the capability of AI, as we know it today, will proceed to increase almost at a linear rate. Like no 1 is predicting a plateau anytime soon. You mentioned that last year, it was beautiful crazy for you. That’s leveled off. I would attribute at least part of that to the capabilities of the AI systems have leveled off. As you’ve had time to look at this and you think about the amount of technology you’ve been active with over your career, do you think we’re overestimating the rate of progression here? Do you think peculiarly the LLM systems can live up to our expectations?
I have a lot to say about this. Number one, this is how we do things, right? We get very excited about any fresh capability, and we just go crazy about it, and people get so jazzed about what could be possible. It’s the classical hype curve, right? It’s the classical thing, so of course that’s going to happen. Of course we’re doing that in AI. erstwhile you peel the onion for truly genuinely powerful technologies, erstwhile you’re through the hype curve, truly large shifts have happened, and I’m rather assured that that’s what’s happening with AI broadly in this device learning generation.
Broadly with device learning or broadly with LLMs and with chatbots?
Machine learning. And that’s precisely where I want to go next due to the fact that I think we are having a somewhat oversimplified conversation about where advances in capability come from, and capability always comes hand in hand with risks. I think about this a lot, both due to the things I want to do for the bright side, but besides due to the fact that it’s going to come with a dark side. The 1 dimension that we talk about a lot for all kinds of reasons is primarily about LLMs, but it’s besides about very large foundation models, and it’s a dimension of expanding capability that’s defined by more data and more flops of computing. That’s what has dominated the conversation. I want to introduce 2 another dimensions. 1 is training on very different kinds of data. We’ve talked about biological data, but there are many another kinds of data: all kinds of technological data, sensor data, administrative data about people. Those each bring different kinds of advances in capability and, with it, risks.
Then, the 3rd dimension I want to offer is the fact that, with AI models, you never interact with an AI model. AI models live inside of a system. Even a chatbot is actually an AI model embedded in a system. But as AI models become embedded in more and more systems, including systems that take action in the online planet or in the physical planet like a self-driving car or a missile, that’s a very different dimension of hazard — what actions ensue from the output of a model? And unless we truly realize and think about all 3 of those dimensions together, I think we’re going to have an oversimplified conversation about capability and risk.
But let me ask the simplest version of that question. Right now, what most Americans perceive as AI is not the cool photograph processing that has been happening on an iPhone for years. They perceive the chatbots — this is the technology that’s going to do the thing. Retrieval, augmented generation inside your workplace is going to displace an full level of analysts who might otherwise have asked the questions for you. This is the—
That’s 1 thing that people are worried about.
This is the pitch that I hear. Do you think that, specifically, LLM technology can live up to the burden of the expectations that the manufacture is putting on it? due to the fact that I feel like that whether or not you think that is actual kind of implicates how you might want to regulate it, and that’s what most people are experiencing now and most people are worried about now.
I talk to a broader group of people who are seeing AI, I think, in different ways. What I’m proceeding from you is, I think, a very good reflection of what I’m proceeding in the business community. But if you talk to the broader investigation and method community, I think you do get a bigger view on it due to the fact that the implications are just so different in different areas, especially erstwhile you decision to different data types. I don’t know if it’s going to live up to it. I mean, I think that’s an unknown question, and I think the answer is going to be both a method answer and a applicable 1 that businesses are sorting out. What are the applications in which the quality of the responses is robust and accurate adequate for the work that needs to get done? I think that’s all got to inactive play out.
I read an interview you did with Steven Levy at Wired, who is wonderful, and you described showing ChatGPT to president Biden, and I believe you generated a Bruce Springsteen soundalike, which is fascinating.
We had to compose a Bruce Springsteen song. It was text, but yeah.
Wild all the way around. Incredible scene just to ponder in general. We’re talking just a couple of days after the music manufacture has sued a bunch of AI companies for training on their work. I’m a erstwhile copyright lawyer. I wasn’t any good at it, but I look at this, and I say, “Okay, there’s a legal home of cards that we’ve all built on, where everyone’s assumed they’re going to win the fair usage argument the way that Google won the fair usage argument 20 years ago, but the manufacture isn’t the same, the money isn’t the same, the politics aren’t the same, the optics aren’t the same.” Is there a chance that it’s actually copyright that ends up regulating this manufacture more than any kind of directed top-down policy from you?
I don’t know the answer to that. I talked about the places where AI accelerates harms or risks or things that we’re worried about, but they’re already illegal. You put your finger on what is my best example of fresh ground due to the fact that this is simply a different usage of intellectual property than we’ve had in the past. I mean, right now what’s happening is the courts are starting to kind it out as people bring lawsuits, and I think there’s quite a few sorting out to be done. I’m very curious in how that turns out from the position of LLMs and image generators, but I think it has immense implications for all the another things I care about utilizing AI for.
I’ll give you an example. If you want to build biodesign tools that actually are large at generating good drug candidates, the most interesting data that you want in addition to everything you presently have is clinical data. What happens inside of human beings? Well, that data, there’s quite a few it, but it’s all locked up in 1 pharmaceutical company after another. Each 1 is truly certain that they’ve got the crown jewels.
We’re starting to envision a way toward a future where you can build an AI model that trains across those data sets, but I don’t think we’re going to get there unless we find a way for all parties to come to an agreement about how they would be compensated for having their data trained on. It’s the same core issue that we’re dealing with LLMs and image generators. I think there’s a lot that the courts are going to gotta kind out and that I think businesses are going to gotta kind out in terms of what they consider to be fair value.
Does the Biden administration have a position on whether training is fair use?
Because this seems like the hard problem. Apple announced Apple Intelligence a fewer weeks ago and then kind of in the mediate of the presentation said, “We trained on the public web, but now you can block it.” And that seems like, “Well, you took it. What do you want us to do now?” If you can build the models by getting a bunch of pharma companies to pool their data and extract value together from training on that, that makes sense. There’s an exchange there that feels healthy or at least negotiated for.
On the another hand, you have OpenAI, which is the darling of the moment, getting in problem over and over again for being like, “Yeah, we just took a bunch of stuff. Sorry, Scarlett Johansson.” Is that part of the policy remit for you, or is that, “We’re definitely going to let the court kind that out”?
For sure, we’re watching to see what happens, but I think that is in the courts right now. There are proposals on Capitol Hill. I know people are looking at it, but it’s not sorted at all right now.
It does feel like quite a few tech policy conversations land on speech issues 1 way or another, or copyright issues in 1 way or another. Is that something that’s on your head that, as you make policy about investment over time or investigation and improvement over time in these areas, there’s this full another set of problems that the national government in peculiar is just not suited to solve around speech and copyright law?
Yeah, I mean freedom of speech is 1 of the most fundamental American values. It’s the foundation of so much that matters for our country, for our democracy, for how it works, and so it’s specified a serious origin in everything. And before we get to the current generation of AI, of course that was a immense origin in how the social media communicative unfolded. We’re talking about quite a few things where I think civilian society has an crucial function to play, but I think these topics, in particular, are ones where I think civilian society… really, it rests on their shoulders due to the fact that there are a set of things that are appropriate for the government to do, and then it truly is up to the citizens.
The reason I ask that is that social media comparison comes up all the time. I spoke to president Obama erstwhile president Biden’s executive order on AI came out, and he made fundamentally the direct, “We cannot screw this up the way we did with social media.”
I put it to him, and I’ll put it to you: The First Amendment is kind of in your way. If you tell a computer there are things you don’t want it to make, you have kind of passed a speech regulation 1 way or the other. You’ve said, “Don’t do deepfakes, but I want to deepfake president Biden or president Trump during the election season.” That’s a hard regulation to write. It’s hard in very real ways to implement that regulation in a way that comports with the First Amendment, but we all know we should halt deepfakes. How do you thread that needle?
Well, I think you should go ask Senator Amy Klobuchar, who wrote the government on precisely that issue, due to the fact that there are people who have thought very profoundly and sincerely about precisely this issue. We’ve always had limits on First Amendment rights due to the harms that can come from the abuse of the First Amendment, and so I think that will be part of the situation here.
With social media, I think there’s quite a few regret about where things ended up. But again, legislature truly does request to act, and there are things that can be done to defend privacy. That’s crucial for straight protecting privacy, but it is besides a way to changing the pace at which bad information travels through our social media environment.
I think there’s been so much focus on generative AI and its possible to make bad or incorrect or misleading information. That’s true. But there wasn’t truly much constraining the spread of bad information. And I’ve been reasoning a lot about the fact that there’s a different AI. It’s the AI that was behind the algorithmic drive of what ads come to you and what’s next in your feed, which is based on learning more and more and more about you and knowing what will drive engagement. That’s not generative AI. It is not LLMs, but it’s a very powerful force that has been a large origin in the information environment that we were in before chatbots hit the scene.
I want to ask just 1 or 2 more questions about AI, and then I want to end on chips, which I think is an equally crucial aspect of this full puzzle. president Biden’s AI executive order came out [last fall]. It prescribed a number of things. The 1 that stood out to me as possibly most interesting in my function as a writer is simply a request that AI companies would gotta share their safety test results and methodologies with the government. Is that happening? Have you seen the results there? Have you seen change? Have you been able to learn anything new?
As I recall, that’s above a peculiar threshold of compute. Again, so much of the executive order was dealing with the applications, the usage of AI. This is the part that was about AI models, the technology itself, and there was quite a few thought about what was appropriate and what made sense and what worked under existing law. The upshot was a request to study erstwhile a company is training above a peculiar compute threshold, and I am not aware that we’ve yet hit that threshold. I think we’re kind of just coming into that moment, but the Department of Commerce executes that, and they’ve been putting all the guidelines in place to implement that policy, but we’re inactive at the beginning of that, as I realize it.
If you were to receive that data, what would you want to learn that would aid you form policy in the future?
The data about who’s training?
Not the data about who’s training. If you were to receive the safety test data from the companies as they train the next generation of models, what information is helpful for you to learn?
Let’s talk about 2 things. Number one, I think just knowing which companies are pursuing this peculiar dimension of advancement and capability, more compute, that’s helpful to understand, just to be aware of the possible for large advances, which might carry fresh risks with them. That’s the function that it plays.
I want to turn to safety due to the fact that I think this is simply a truly crucial subject. Everything that we want from AI hinges on the thought that we can number on it, that it’s effective at what it’s expected to do, that it’s safe, that it’s trustworthy, and that’s very easy to want. It turns out, as you know, to be very hard to actually achieve, but it’s besides hard to measure and measure. And all the benchmarks that be for AI models, it’s interesting to hear how they do on standardized tests, but they just are benchmarks that tell you something. They don’t truly tell you that much about what happens erstwhile humanity interacts with these AI models, right?
One of the limitations in the way we’re talking about this is we talk about the technology. All the interesting things happen erstwhile human beings interact with the technology. If you think models — AI models are complex and opaque — you should effort human beings. I think we gotta realize the scale of the challenge and the work that the AI Safety Institute here is doing. This is simply a NIST organization that was started in the executive order. They’re doing precisely the right first steps, which is working with industry, getting everyone to realize what current best practices are for red teaming. That’s precisely where to start.
But I think we besides just should be clear that our current best practices for red teaming are not very good compared to the scale of the challenge. This is actually an area that’s going to require deep investigation and that’s ongoing in the companies and more and more with national backing in universities, and I think it’s essential.
Let’s spend a fewer minutes talking about chips due to the fact that that is the another part of the puzzle. The full tech manufacture right now is reasoning about chips, peculiarly Nvidia’s chips — where they’re made, where they might be under threat rather virtually due to the fact that they’re made in Taiwan. There’s evidently the geopolitics of China active there.
There’s a lot of investment from the CHIPS Act to decision ship manufacturing back in the United States. quite a few that depends again on the thought that we might have any national champions erstwhile again. I think Intel would love to be the beneficiary of all that CHIPS Act funding. They can’t operate at the same process nodes as TSMC right now. How do you think about that R&D? Is that longer range? Is that, “Well, let’s just get any TSMC fabs in Arizona and any another places and catch up”? What’s the plan?
There’s a comprehensive strategy built around the $52 billion that was funded by legislature with president Biden pushing hard to make certain we get semiconductors back at the leading edge in the United States. But I want to step back from that and tell you that this fall is 40 years since I finished my PhD, which was on semiconductor materials, and [when] I came to Washington, my hair was inactive black. This is truly long ago.
I came to Washington on a congressional fellowship, and what I did was compose a survey on semiconductor R&D for Congress. Back then, the US semiconductor manufacture was highly dominant, and at that time, they were worried that these nipponese companies were starting to gain marketplace share. And then a fewer actions happened. quite a few truly good R&D happened. I got to build the first semiconductor office at DARPA, and all time I look at my cell phone, I think about the 3 or 5 technologies that I got to aid start that are in those chips.
So, quite a few good R&D got done, but over those 40 years, large things happened, but all the manufacturing at the leading edge yet moved out of the United States, putting us in this really, truly bad situation for our supply chains, for jobs all those supply chains support. The president likes to talk about the fact that erstwhile a pandemic shut down a semiconductor fab in Asia, there were car workers in Detroit who were getting laid off. So, these are the implications. Then, from a national safety perspective, the issues are immense and, I think, very, very obvious. What was shocking to me is that after 4 decades of admiring this problem, we yet did something about it, and with the president and the legislature pulling together, a truly large investment is happening. So, how do we get from here to the point where our vulnerability has been importantly reduced?
Again, you don’t get to have a perfect world, but we can get to a far better future. The investments that have been made include Intel, which is fighting to get back in and drive to the leading edge. It’s also, as you noted, TSMC and Samsung and Micron, all at the leading edge. 3 of those are logic. Micron has memory. And Secretary [Gina] Raimondo has just truly driven this hard, and we’re on track to have leading-edge manufacturing. Not all leading-edge manufacturing — we don’t request it all in the United States — but a crucial condition here in America. We’ll inactive be part of global supply chains, but we’re going to reduce that truly critical vulnerability.
Is there a part where you say, “We request to fund more bleeding-edge process technology in our universities so that we don’t miss a turn, like Intel missed a turn with the UV”?
Number one, part of the CHIPS Act is simply a crucial investment, over $10 billion in R&D. Number two, I spent quite a few my career on semiconductor R&D — that’s not where we fell down. It’s about turning that R&D into US manufacturing capability. erstwhile you lose the leading edge, then the next generation and the next generation is going to get driven wherever you’re leading edge is. So, R&D yet moves. I think it was a well-constructed package in CHIPS that said we gotta get manufacturing capacity at the leading edge back, and then we build the R&D to make certain that we besides win in the future and be able to decision out beyond that.
I always think about the fact that the full chips supply chain is utterly dependent on ASML, the Dutch company that makes the lithography machines. Do you have a plan to make that more competitive?
That’s 1 of the hardest challenges, and I think we’re very fortunate that the company is simply a European company and has operations around the world, and that company in the country is simply a good partner in the ecosystem. And I think that that’s a very hard challenge, as you well know, due to the fact that the cost and the complexity of those systems has just… It’s actually mind-boggling erstwhile you see what it takes to make this thing that ends up being a square centimeter, the complexity of what goes behind that is astonishing.
We’ve talked a lot about things that are happening now. That started a long time ago. The R&D investment in AI started a long time ago. The detonation is now. The investment in chips started a long time ago. That’s your career. The detonation and the focus is now. As you think about your office and the policy recommendations you’re making, what are the tiny things that are happening now that might be large in the future?
I think about that all the time. That’s 1 of my favourite questions. 20 and 30 years ago, the answer to that was biology starting to emerge. Now I think that’s a full-blown set of capabilities. Not just cool science, but powerful capabilities, of course for pharmaceuticals, but besides for bioprocessing, biomanufacturing to make sustainable pathways for things that we presently get through petrochemicals. I think that’s a very fertile area. It’s an area that we put quite a few focus on. Now, if you ask me what’s happening in investigation that could have immense implications, I would tell you it’s about what’s changing in the social sciences. We tend to talk about the progression of the information revolution in terms of computing and communications and the technology.
But as that technology has gotten so intimate with us, it is giving us ways to realize individual and societal behaviors and incentives and how people form opinions in ways that we’ve never had before. If you combine the classical insights of social discipline investigation with data and AI, I think it’s starting to be very, very powerful, which, as you know from everything I’ve told you, means it’s going to come with bright and dark sides. I think that’s 1 of the interesting and crucial frontiers.
Well, that’s a large place to end it, manager Prabhakar. Thank you so much for joining Decoder. This was a pleasure.
Great to talk with you. Thanks for having me.
Decoder with Nilay Patel /
A podcast from The Verge about large ideas and another problems.
