The New Tech Era — 2025 College and University Educators Workshop
About College and University Educators Workshops
In this session of “The New Tech Era,” Adam Segal addresses U.S.-China technological competition and implications for national security and Nicol Turner Lee speaks on the societal implications of AI. The conversation is moderated by Calvin Sims.
The goal of the workshop is to find new ways for college and university educators to encourage their students to learn about international relations and the role of the United States in the world. It provides an opportunity for educators to explore the wide array of CFR and Foreign Affairs teaching and research resources available to the academic community, participate in substantive briefings with subject experts as well as in group discussions, and share best practices and educational tools for bringing global issues into the classroom.
The New Tech Era
Transcript
FASKIANOS: If everybody could take their seats, please. If everybody could take their seats, please. Thank you very much. Oh, you’re—I love it, everybody coming to attention. Excited to have our next panel, The New Tech Era. And I’ll invite our panelists on the stage, and this session will be moderated by Calvin Sims.
(Pause.)
SIMS: Well, good morning, everyone.
I’m Calvin Sims, and I have been a reporter, a foreign correspondent, a number of things that journalists do, and I have just come in from the cold for a little while and it’s great to have you all here for this session.
So I wanted to start by introducing our two panelists: Adam Segal, who’s the Ira Lipman chair in emerging technologies and national security, and director of the Digital and Cyber Policy program at CFR; and we’re also very privileged to have Nicol Turner Lee, who’s the senior fellow in governance studies and director of the Center for Technology Innovation at the Brookings Institution.
So I think we’re going to have a really good back and forth today and this really is always focused on new journeys for this institution, and so now we will be looking at places we haven’t before. That is the new tech era, and for this I’d like both Adam and Nicol to start us off by talking a little bit about how things have transitioned and transitioned so widely across not only the United States but the world.
So I’m going to start with Adam and you can kick this off. Thank you very much.
LEE: I’m so glad you got that first. (Laughter.)
SEGAL: Yeah. Yeah.
So I guess I would point to at least two things, two that I thought of already. So one, of course, is just the dramatic pace of change, right? I think it is a cliché but it is true that the pace of technological change has increased and we keep getting caught unaware by developments.
ChatGPT is the clearest example of that but then, you know, DeepSeek just a couple of months ago, and so I think we are struggling both intellectually and from a policy perspective on how do we keep up and how do we think about those technologies.
The second that I’ll just note is the convergence of all these technologies and the digital and the physical. So, again, AI is a great example of that—you know, the application of AI to biology and synthetic biology and all those places where there’s going to be huge breakthroughs.
So intellectually, you know, having a cyber and digital program over time makes less and less sense because it becomes part of every issue that we deal with and so that, I think, is a second.
I’ll just add the third, which is that, of course, as we’ve been talking about probably in all the other panels beforehand it is now a major focus of geopolitical competition, right? It is not the second or third or fourth issue that gets brought up but almost every way we think about competing, particularly with China but just across the board economically, militarily, through systems of governance, all express themselves through technology.
SIMS: Nicol?
LEE: Yes. So great to be here as well.
I would add off of the three points—I’ll stick to my three points, too, Calvin, right, as opposed to giving a fourth.
SIMS: OK.
LEE: I would first say that what we’re seeing in the tech era is this lack of agency as systems become much more autonomous, and we’re seeing more and more humans sort of displaced and moved out of the loop. I wrote a book about this. It’s called Digitally Invisible, right? (Laughs.) He’s like, that was really good, wasn’t it? (Laughter.)
SEGAL: That’s really fast, too.
LEE: It really was fast. (Laughter.) I said I was going to show it. There’s collateral out there. But I wrote this book right before the pandemic, Digitally Invisible: How the Internet Is Creating the New Underclass, through the Brookings Press and Rowman & Littlefield, which is available wherever books are sold.
But the book’s focus was really on people, right, who have been affected by this shift from very transactional, tactile machines and computers to what we see now, which is much more autonomous systems and I’ll talk a little bit about it—it’s my third point.
The digitally invisible they live in farming communities. They live here in New York City. They are people who are connected to the digital ecosystem but they do not have the wherewithal to be able to engage in equitable matters like everybody else, and you’d be surprised, despite the pandemic moving a majority of people online, there were already a lot of people who were offline.
I call them digitally invisible populations that we did not know about as researchers and we really did not understand just how disconnected they were from this shift of analog to more digital to now more autonomous systems. So that’ll be my first point.
My second point is we do have a concern as we look at artificial intelligence with regards to the lack of human agency that has evolved with these technologies also has consequential outputs on where they show up. And so at the Brookings Institution two years ago I started what we call the AI Equity Lab where workshopping issues like artificial intelligence and its impact on education, health care, criminal justice, financial services, but we’re doing it with the experts like many of you in this room sitting at the table because more and more as these autonomous systems develop, which is sort of the—sort of pivot point in this new tech era we’re seeing less humans in the loop or outside of the loop sort of guiding and advising and interrogating these systems.
And then I just want to pick up on that third point in terms of the geopolitical concern. We have traditionally seen tech sort of posed against our relationship or competition with China. I think now that has changed, particularly in the last couple of weeks and hours—I’m sorry, looking at some of the trade wars—but in particular we’re seeing technology much more collateralized around assets and the value chain.
So I do a lot of work in 5G technology and how that actually interacted with China and Africa. I’ve written on that as well. But my thing is what I’m seeing differently is that we’re seeing a lot more collateralization of Global South countries, African Union, others, and not sort of investments in this geopolitical conversation outside of Europe and, you know, other places which—and China, which tend to be predominant verticals when we talk about technology.
The space is so much bigger. In my book and in my writings the expansive use, for example, of technology in the African Union is beyond our wildest imagination, but yet we don’t see them as a competitive vendor or partner or ally in these conversations when we start to talk about the evolution of technology.
So a lot of this is the work that I do at Brookings but definitely part of the new tech era.
SIMS: For both of you, can you talk a little bit about the ethics of this new renaissance, if you were to call it that, and how do we start to engage this and in what way? It would be interesting to hear both of you, because this is mainly what you’re looking and reporting and thinking about, right?
LEE: So it’s Renaissance AI?
SIMS: Yes.
LEE: (Laughs.) Well, I’ll start. I mean artificial intelligence for me—I’m a PhD in sociology, not a computer scientist, so all of my work pretty much undergirds the same methodological approach, which is to look at the equitable distribution and access to these technologies.
So AI is different because, one, it doesn’t necessarily require that the people who are the subjects of the technology sit at the table and design and develop those technologies. So from an ethics standpoint it already starts with the absence of what I call many unhidden figures like educators and health practitioners and others who need to be sitting at the design phase of the modeling versus coming in and becoming more of the subjects of the technology.
So that’s one of the issues that we face. The other thing is artificial intelligence—and people tend to love this. I’m sure Adam has this little pet peeve as well. There are so many forms of it, right?
There is predictive AI, which tends to be based on what we think in terms of repetitive behavior of the machine learning algorithms, what’s going to happen as a result. There’s now, you know, generative AI, which is basically for a layperson. I call it eyes in the back of the head because it’s definitely sort of scraping what is already available text-wise, image-wise, voice-wise.
And now there’s agentic AI, right, which is pretty much autonomous in making decisions not just based on human behaviors but an agentic AI for people who are, like, where is this lady going. You know, it’s like your autonomous nurse or the person who might be a bill collector who is trying to actually walk you through the next stage of a payment plan.
So, for me, these technologies have advanced so quickly that we all have been sort of—not—I don’t want to say coerced but we have been in many ways using these technologies because of the convenience that they offer us without knowing what’s behind the black box, and as a result of not knowing what’s behind the Renaissance the work that I do is there are going to be decisions that are pretty innocuous.
Like, I love black jackets. I want the recommendations for that type of commerce purchase. But I also don’t want my child to be profiled or stereotyped in the classroom simply because AI told an educator to maximize their learning you need to put them in the front or the back or you need to use this stair stage to be able to do that.
So, for me, when we start getting into the renaissance, into consequential areas that improve quality of life or create wealth or have eligibility decisions then for me it becomes problematic and that’s where I find myself in this work because we just don’t do enough on the ethics of design, deployment, and interrogation.
We talk about ethics as this big umbrella but for me it’s what’s under the umbrella and what are we digging into, and how do we dig into those things in areas where there are truly some red boxes that we need to pay attention to.
SIMS: Adam?
SEGAL: Yeah. So I’m a political scientist so I’m focused on what the state is going to do both at the domestic and the international level. We are in a massive period of transition, right? The Biden administration had an approach that was essentially focused on trust and accountable systems. It tried to balance voluntary commitments from the firms, right?
So thirteen of the largest AI and tech firms agreed to voluntary commitments that the Biden administration—about safety and transparency and accountability, plus the role of the federal government.
So the executive order saying we’re going to use these systems and we’re going to make sure, again, that they have certain levels of accountability and transparency there, and then at the international level let’s build these discussions around safety and accountability so through what became kind of the Hiroshima process that then went to London and then to Seoul and a little bit through the UN and the OECD and some other international organizations.
Now, the Trump administration essentially said this was the wrong balance that focused too much on the costs and the risks and that they were slowing the U.S. down and in particular, you know, we’re going to lose the race to China so overturned the executive order. Basically, told Kratsios, who’s the new head of the science and technology policy in the White House, go as fast as you can.
Vice President Vance went to the AI summit in Paris and basically is like, I’m not here to talk about the risks; I’m here to talk about the opportunities. Oh, and by the way, America is going to dominate this space. If you want to, you know, come along with us that’s OK but we’re going to dominate.
So the messaging is no longer really about how do we regulate or control. It’s about how do we race faster. You know, the Europeans, I think, are still focused on regulation—the European EU AI Act—but there’s been some hesitation, right?
They slowed down the rolling out of part of the accountability act around AI because they are also worried about getting caught behind and the firms, clearly, are embracing race faster.
There was an article in the FT today that said OpenAI has cut the amount of time that they give outside analysts to do safety testing of new models. So they’re feeling the pressure and they’re moving there.
We’re seeing a lot at the state level. So California, you know, got pretty close to an AI regulation act. Connecticut and others—Maryland—are thinking about it. Firms, clearly, don’t want that, right? They don’t want fifty separate AI legislation but that’s where we are there, I think, on the control.
SIMS: Talk a little bit about access and who gets the access and under what, you know, plateau or whatever that we want to talk about it. How does that actually play out over a period of time?
LEE: Mmm hmm. Well, I’d love to pick up on what Adam also talked about, too. I mean, the right space framework also assumed that we were also going to democratize technology in many respects and we were going to find ways to sort of invest in, you know, more R&D, more conversations on accountability, more red teaming, and to test out and weed out the vulnerabilities.
With the new change in administration, obviously, that’s not the focus. It’s not necessarily on the socio technical side or the safety side or the human side. So what does that mean in terms of access?
If you go back to my original dialog, we are all subjects of AI, right? Is there anybody in this room that owns a copyright or trademark on any AI product?
Well, guess what? Like my mother used to tell me, if you’re not in the kitchen you’re on the menu. So we’re all on the menu. Just letting you know, right? (Laughter.)
We are part of the menu of services that AI will actually bring. So in my opinion—and I’ve been struggling with this, right? It’s not necessarily an either/or. It’s a both/and, right? And so you can have accelerated artificial intelligence tools integrated into health care.
I’ve seen some tremendous applications of AI in certain spaces where it’s helping us with diagnosing disease quicker. It’s helping us with tracking and monitoring. And you also have to be able to understand the limitations of AI. So in my work, for example, Black women are underrepresented in clinical trials.
So there may be great AI when it comes to, you know, looking at breast cancer but if people aren’t fully represented it’s inadequate. So for me that’s an accessibility question, right? What does the data look like?
How is it represented among the various people that we want to be in this sample so that we create better AI or AI at least that is much more sufficient towards people? That’s one.
Two, we know that with this new, burgeoning AI, right, Adam, that it’s going to take a lot more energy, a lot more data centers, and that seems to be the particular conversation. Just recently on Capitol Hill there were two hearings in the matter of two weeks. One was on AI data centers and energy and the one that just happened this week, I believe it was, was with Eric Schmidt and others on AI and how it’s actually going to be part of our international competitiveness.
Again, based on where you frame it accessibility can be defined by a larger global norm where if we do the right things we’re going to be able to empower not just AI but the future of quantum, in all honesty. So we’ll have the system—the systemic—the system’s infrastructure in place.
But if you are a kid that lives in the Bronx and you are a teacher who wants to use AI, accessibility will look very different. So when I speak with educators, for example, who are so excited about using AI-enabled education products I say, well, where are you from?
And if you say you live in a rural community I say, well, you’re digitally invisible. You probably don’t even have basic internet access, right, if you’re a local college, et cetera. Or if you’re in pockets within big urban areas you may not also have the infrastructure to empower your ability to do AI-enabled products and services.
So we have somewhat of a mismatch, I think, when it comes to trying to solve the supply and demand, and we’re behind. Under the Biden administration—I’ll just say this last—we did put—and I reference this in my book.
My book is about the digital divide but more so about digital competitiveness. But I go through the six administrations starting with Clinton and Gore, who actually defined the digital divide. But what I say is we’ve never solved it and what we’ve done is we’ve thrown a lot of money at it.
So under the Biden-Harris administration we put in the $65 billion to start—went up to trillions—in terms of infrastructure. Those projects are not done and they were only to provide basic connectivity in communities via our states that didn’t have it.
So when you start talking about AI systems, Calvin, it sort of exponentially changes the nature of the game and how advanced we’re going to be if we’re going to actually run many of the products and services.
And I think you’re so right. That’s why we’re hearing more of the tech companies say, give us more data centers. Give us more energy. You know, give us the ability to move fast in communities because we need to actually be able to empower these systems like ChatGPT-4 to do this stuff.
Although I did hear a professor once say the energy of ChatGPT-4 you actually use more energy eating a hamburger. It was some kind of—I had to find this professor. I had to go up to him and say, OK, are you sure this is right?
But it was the way that he explained it. It may be an exaggerated conversation in terms of the amount of energy that these tech companies are demanding. (Laughs.)
SEGAL: So I’ll shift it up again. So to the international level, you know, Nicol mentioned 5G. So I spent fourteen months inside the State Department working at the Cyber and Digital Bureau and our big focus was we were not going to get Huawei-ed again, right?
So Huawei means, you know, here we have a Chinese telecommunications manufacturer that is producing and deploying 5G in most of the developing world and the main reason is because the U.S. has no product, right?
We stopped producing telecommunications equipment for a lot of reasons but we stopped doing it. We tried to convince those countries that, you know, the security was bad and they were going to get spied on and they—you know, they shouldn’t use a Chinese product. But that was not a particularly powerful argument, right?
Most of those countries are, as Nicol said, worried about the digital divide and getting their population online and their assumption is, I’m going to get spied on by the Chinese and the U.S. no matter what I do so I might as well at least put in cheap product and get it there as quickly as possible.
So while I was in the department we were trying to always think of, well, how can we provide access to technology that is secure but also addresses their sustainable development goals and economic development.
We came up with a strategy that was built around this idea of digital solidarity so the argument being that we were actually made stronger by other countries having their own tech capabilities, right?
If we’re not beating up the Europeans because they’re beating us up because they have to rely on us for cloud, if we can kind of get that out of the way then we can concentrate on getting shared values and other things and competing with the Chinese and other places there.
Heavily reliant on USAID and other tools to build out that capacity in those places. You know, right now the Chinese are doing a pretty good job at being able to paint the United States as saying because of the use of export controls and other tools to say, we’re trying to control that technology, right?
We’re using it for ourselves. Everybody else is going to be cut off from it. China can use its usual rhetoric of win-win and all boats are going to be lifted. I don’t think that’s necessarily true, but it is pretty attractive to all of those developing economies and how they can think about access.
Now, of course, digital solidarity is not going to survive. The Trump administration is going to have a different focus on that. But what we’re seeing is lots of countries talking about AI sovereignty, right? They want to have their own models and their own languages, and so there’s going to be a massive kind of differentiation based on capacity.
You know, Singapore is going to probably do a pretty good job of developing ASEAN language models. India will probably have some pockets of success. Brazil might do OK. But given all the resources and energy and everything else you’re going to do we’re, clearly, going to see a divide on AI capabilities as well.
LEE: I was going to say we just recently published a writing series. One of the fellows that we have in our department is focused on the Global South so she’s been trying to figure out ways to bring in voices, and we’ve leveraged the Paris AI summit and the conversation around safety and we just actually published people from Oceana region, African Union, Southeast Asia, India. They all published indigenously where they thought, you know, they could be part of this conversation.
You’re so right. I think some of the inadequacies that we see in many of the large language models, for example, is the insistence to train on big language versus micro language and so you lose some of the diversity that comes with, you know, linguistic adaptations or challenges.
That’s actually also the case in the United States, right? It tends to train on traditional language where you lose some of the other nuances and cultural capacities of other languages. So to your point, we’re seeing a lot of that as well and I think our writing series, which is also available at Brookings on our Center for Technology webpage, but it gives you a nice look at what you’re talking about, that there is going to be this upshoot of people wanting to define how they want to exist within this space as well.
SIMS: One more question before we turn it over to the audience. The ethics of this and who determines what is right, what is wrong, who saves people who need to actually have access to it—are we that far away from it or, you know, how long will it take us to actually get there?
LEE: You start. (Laughter.)
SEGAL: We’re pretty far away, I think it’s safe to say. Again, so I think—let me just start on the international level. Like, there—it’s clear that technology is racing faster than any international agreement is going to get there.
You know, we had all of these models that were floated around briefly. Should we have an IAEA for AI? Should we have Atoms for Peace for AI? None of those are going to happen, right?
And as, you know, Nicol pointed out, AI is not one thing. It’s numerous, numerous systems, and so the idea that we’re going to get one international agreement strikes me as being very, very unlikely, just having done in the last fifteen years in cyber where we really have—we have agreed upon state norms that were—that everyone for the most part ignores, right?
So we don’t have a lot of agreement on that space. I think we will see kind of regional agreements, you know, the EU, clearly, being one. China has a very well-developed AI system. It may not be the one that we like but it’s, you know, extremely regulatory, sophisticated, and they keep rolling out things there.
So I think we’ll see regional kind of discussions about that emerge and we may—you know, we’ll see some broad agreements, for example, around AI and military use. That has happened at the UN I think we’re up to over sixty countries have signed that agreement.
But it’s pretty broad, right? It’s about, you know, some very broad guidelines about how we might do that.
At the national level—I think we’re going to see it at the states. I don’t actually think we’re ever going to get anything in Congress just given that, you know, the difference between the Republicans and the Democrats on big tech and anti-monopoly versus—how content moderation has become an issue for the right more than it has been for the left. And so I would be very surprised if we get any agreement there. There may be a default to basically saying, look, we have regulations already in place; just they’re sectoral and let’s just apply them as they already exist. But I would expect to see something, you know, from California or Maryland or some other of those leaders first.
LEE: Yeah, and I would just sort of tag onto this.
So I think at the international level I do believe we’re going to continue to have conversations with EU and the U.S. will continue to dominate what does safety look like, et cetera, in the competition with China and Russia, and I do believe the military will probably be the only area where we can actually find some common ground in terms of where we want to deploy AI.
I do, again, want to echo that I’m seeing in my research just a groundswell of participation of countries. The recent summit was held in Rwanda, right, and we’re just seeing a lot more Global South countries come to the table to sort of assert, I think, their value principles in this which I think is going to actually continue, particularly since the 5G debate where we’re now seeing more of those countries figure out ways to have some independence on the—of the supply chain.
I think on the federal level we are going to see some type of AI movement. I mean, let’s not be confused. If we are seeing a lot of the tech companies sort of at the White House more, obviously, there’s going to be some self-interest in trying to ensure that they can have the investments reliability from the U.S. to be able to engineer new R&D.
I think in those spaces we’ll see that, and I think we’ll have some normalization, and something just came out, for example, on AI use by federal agencies, which was actually started during the first Trump presidency. So I think that conversation is there.
The thing to actually throw into the mix, however, is with the federal layoffs, for example, in the obvious movement of government towards privatization we’re going to see more AI use in government, which I think will be quite interesting. I’ve written on this.
Because what that means is that if you have a digital divide there will be constituents that will not get access to that AI-enabled process when it comes to government, but more importantly it may also bypass some of the traditional procurement and compliance, and safety and security measures that we’ve had with government when it comes to technology, which is an area that I’m particularly interested in and the degree to which we see a lot more embedded AI without even knowing that it’s being embedded into government systems.
And then I think on the state level I would actually offer that I think the states will come out with their patchwork of enforcement activities. Texas—I was sitting in Texas and I heard a legislator say something to the effect that he is pushing for more agency among people, more disclosures, more appeal capacity.
I tell you, I was in Texas. I had to go find his staffer and say, am I in Texas in a red state? (Laughter.) I mean, is he really saying that on stage? He said, yes, and I think we’re going to see more of that, right?
I think we’re going to see state legislators because of their ability to interact closely with their attorney generals sort of think about ways in which they can enforce some of the sectoral issues when it comes to compliance with fair housing and fair credit and other areas—disclosure of AI-generated decisions, and they’re going to be watching this, right, because they’re going to be accountable to their constituents at the state and local level.
So I think in the absence of any type of federal movement on rights we’ll probably see a lot of that actually shift to the states and New York is another example. And it’s going to be by sector. Here in New York they are shifting away from a hiring-based AI.
So we’re going to see more of that come out, which is going to create some challenges for those of us who cross borders because that means that there will be different requirements for each.
I want to just put this out there as well. I’d be remiss in terms of where we are as well. You know, AI is also being used in many respects, regardless of how you voted, to sort of normalize and homogenize our communities here and our democracy, and so that’s another area that I think as a sociologist I’m particularly interested in how this looks because we’re using AI to weed out words and weed out processes and policies.
I mean, there’s a whole list of triggers that AI has that is now defunding colleges and universities and, you know, making it very difficult for people to engage in free speech. So I think we’re going to see AI in this type of weaponization, potentially, where we won’t feel the effects of those until, like, the next five to ten years. We’ll actually know what really happened with many of those data sets that are no longer publicly available or constrained or have been stripped down.
And so I want to put that in there as well because I think that’s another area of the democratic conversation that we need to have as well in terms of artificial intelligence.
SIMS: So we would like to have the audience engage as well, and so I’m going to bounce around but I’d like to start with you, and someone will bring you a mic.
Yeah. Please identify yourself. Yeah.
Q: Hello, all. So I’m Kelley Littlepage and I’m from the University of Houston, and what I wanted to ask is—I love that you said we’re all on the menu and this is what provoked this thought.
But what is the future with AI and this new emergent technology in concerns for fundamental rights like privacy, which is already under attack from many other fronts than just technology? But I wanted to know what you thought the future with this new technology and privacy would be both domestically and internationally, please.
LEE: I mean, the problem we’ve had and the reason we continue to have these conversations is because the U.S. has no national federal privacy legislation. We’ve been talking about federal privacy legislation.
Cam Kerry, who is the brother of John Kerry, I’ve known him since he was on the Obama administration. He has been talking about privacy for at least twenty years since he was at Commerce and we still don’t have it, right? He works on our team at Brookings.
Part of the reason why is because we can’t agree on just some fundamental principles on what we should be protecting and how we should be protecting it and how we enforce those protections, and it always goes back to—two things of disagreement is, you know, private right of action—if people have the collective right to sue—and most importantly, like, how are you protected, and I think Adam suggested this innovation phase of technology while still allowing for some protections by consumers.
That is probably our biggest issue. I’ll let Adam talk about all the privacy legislation that’s happening outside of the United States. But had we solved that issue I think the conversations around responsible AI would have been different, and because it’s not been solved, again, more data is available. And you’d be surprised. It’s not just, you know, data that you expect.
I mean, AI is very unique from the standpoint that it does not need particular data to actually empower it. It can do proxies, inferential data. It can figure out—you know, there’s a study that we actually have completed that aligns with Latanya Sweeney at Harvard University that in the employment space it is the sounding of your name that triggers whether or not you’re going to get a pre-hire screening or not, and there’s some AI that you give a video capture and the AI checks to see if you’re looking in the camera, if you’re looking away, how many times you smile, which then determines whether or not you’re a suitable candidate for a job.
So it’s so embedded without privacy protection we basically have no agency over these technologies.
SEGAL: Yeah. So I—look, it’s around privacy and who controls and has access to the data. You know, we’ve been jerry rigging this approach with the Europeans for the last fifteen years where we say, OK, we promise we’re not going to do this stuff with the data and then European courts say OK and then Max Schrems sues and then we go back to square one.
So Max Schrems, for those who don’t know, is a German privacy activist and he has won at least three cases in the European Court of Justice that basically says U.S. standards are not appropriate for European data control and until that’s in place you can’t send data over to the—from Europe to the United States because of GDPR and privacy controls there.
We have an agreement in place now, but one of the things that President Trump has done is not appoint replacements to the privacy—PCLOB?
LEE: C. He took them all off. (Laughs.)
SEGAL: He took them all—right. So there’s a—there is a bureau that’s supposed to inspect—kind of account for those agreements because Europeans are allowed to basically say, I think you’ve collected data on me that’s inappropriate and I want to push back against that.
Without this thing having anybody on staff then there’s no way for that to happen. So it’s very likely that that agreement will also collapse. We’ll be back to square one. Usually it’s OK for the big firms. It’s not great for the little firms. Big firms manage to figure it out through other types of contracting but small startups have a much more difficult time.
So, you know, it’s still going to be just a huge issue that is going to really get—once we get—once—if we get through the tariffs problem, if we get through antitrust, if we get through content regulation with the Europeans then the privacy thing is also still hanging out there and I don’t see any easy resolutions.
LEE: It gets kicked to the bottom every time. (Laughter.) Starts here at the beginning of a session and then it lands down here.
SIMS: We’re going to try to get a couple more questions before we—let’s go back here. You have in green a question. Is that—
Q: Good morning. My name is Sandeep Mertia. I’m a professor of technology and society at Stevens across the river here. Thank you so much for these insights.
As an anthropologist of computing I’m wondering how much good faith are we giving—showing the tech industry by continuing to use their own language? Because a lot of powerful criticism of these technologies from the likes of Dr. Alondra Nelson, Dr. Timnit Gebrum, and various other historians of computing has been that what is this word AI that we use and why are we buying into their hype cycle discourse, which loves projecting these superhuman technologies, which are basically, if I may use a research term by Dr. Ilana Gershon, who’s a linguistic anthropologist, that these are bullshit generators. (Laughter.)
LEE: You said that, not me.
Q: Oh, it’s a published article. (Laughter.) I’m just quoting. So while I understand that there are very important privacy concerns, geopolitical concerns, harms to address, by buying into their language are we not helping them project their power? Thank you.
LEE: So, I mean, I think that—and the people you mentioned we just actually acknowledged them as part of “Unhidden Figures: Women Leaders in AI” as part of Women History Month.
So I think you’re right, right? So let’s just start with the internet. OK. So the internet, essentially, in the United States started as something that grew out of deregulation and, you know, this shift from analog to digitization was, obviously, part of the Clinton era. It was actually under two bipartisan FCC—Federal Communications Commission—chairs, Michael Powell and Bill Kennard.
But the goal is that the internet was this thing that we didn’t know about. It was very far beyond the rotary telephone and the telegraph, and it actually did a really great thing. It actually created the smart phone, right? And out of the smart phone came ridesharing services and out of ridesharing services came much more interest in 5G.
I’m an internet historian so, like, when I think about where we are with AI I’m a little bit baffled, right, because out of the deregulation of the internet basically has brought us to where we are today, and we were sort of anticipating a lot of this with the net neutrality debate in terms of who owns the internet. And power has always been coalesced around the internet because the information of this now fabricated privately-run town square is essentially determining how we actually govern ourselves.
I mean, I’m, like, I have to be on X to find out what happened in Capitol Hill, right, you know, where decisions are being made. It’s a whole different—I mean, Reed Hundt, who was a former FCC chair, used to say the internet is the new town square. It’s, like, way beyond that because now we have these advanced capabilities that create these algorithmic filters and amplification and all this other stuff.
So to your point, it’s very much determined by power and what we have found is that whereas most people who are internet users—and I can give this great example for my book. When I went around and talked to people from farming communities to cities in the book I would say, what is your opinion of the internet? And many people would say, oh, I don’t trust the internet. I hate the internet. I don’t use the internet.
And then I would say, how do you get money from your customers? And they say, Cash App—(laughs)—or they say the internet was the Facebook.
So we as a country, as a world, really don’t know what the internet has become and we’ve allowed power to sort of shape what that narrative is for us and I think at this point, as the technology has become much more advanced, much more integrated as the business case, not just by tech companies but by banks and school, ed tech—the big area as well—we’re in this state where, to Adam’s point, we either assert more regulation and prescription or we allow it to continue to grow.
The difference is whereas other internet technologies we had a one-to-one relationship with—we kind of have a relationship with our phone. We had a relationship with what’s on there. This is one in which we’re sort of passive adapters of this technology in ways I think that is sort of outside of our control and it’s making decisions based on our ability to live.
And that’s the conversation I would urge all of you as educators who are working in spaces with the inquiry of your students to have them have a conversation around the history of the internet and to understand where this fits in this wide spance of how technology has shifted from consumer base to something that is way beyond—you know, consumer enterprise based to something that is much more democratically rooted.
That’s my long story answer. That is, like, my second book I want to write. You just gave me—I’m not going to name it that expletive, though. (Laughs.) But it’s an interesting question I have found myself in a lot lately in terms of bringing it back to the internet.
SIMS: Yeah. I want to get a couple more—
LEE: I’m sorry. I’ll get off my soapbox on that one, Calvin.
SIMS: Right here.
Q: My name is John Scianimanico. I’m with New York University here locally.
My question is actually about AI in education that you brought up, Nicol. I’m sure you saw that China starting in the fall is making AI education compulsory for its primary and secondary students—
LEE: Yes. Yes.
Q: —and I was wondering if you could comment on how the United States should be thinking about AI literacy in our schools. Of course, we all know that our students in higher education use it but I sometimes worry that it’s too late by the time that they come to us for us to start teaching them about how to use it responsibly.
Do you feel like this should be something that should be taught in schools, and with a shuttered U.S. Department of Education are states prepared today with the resources and expertise and energy to be able to inform schools and district leaders on how to do this well?
SIMS: Do you want to start, Adam?
LEE: You want to start on that one, Adam?
SEGAL: No, I think it’s for you. (Laughter.)
LEE: I’ll take that one, right?
So I do think AI literacy should be embedded in a K-20 pipeline, right, because I think you’re correct, the students that go into college they’re basically familiar with it but the difference is they’re familiar with it based on the school district they came out of.
So I’m sure many of us in this room hear the same thing I hear often which is we can’t let certain students use AI because they’re cheating, and so this debate actually resonates through a lot of conversations because we look at where AI starts normally for higher income school districts on the K-12 side.
Those kids are learning how to use it for prompt building, critical inquiry skills, research, so that when they get to college they’re ready, right? They know how to—my daughter is in a Fairfax County school. She runs all her stuff through an AI authenticator to make sure it does not replicate that. She has those skills.
Some kids who live, you know, in the Bronx or other parts of Washington Square Park they may not have those same skills introduced to them in K-12. So I do think that we need to have a K-20 but here’s the problem. We tried to integrate media literacy in 2010 when the Knight Foundation came out with a commission on what an information democracy should look like and we were unsuccessful in getting digital citizenship embedded into our schools.
And so I do agree now we need to have that conversation. There’s so many groups—and I would love to talk to you offline—that are actually thinking about ways to embolden AI literacy. But what I would say to the—(inaudible)—and to colleges is this conversation I’ve been having with all of you, is your school should be thinking about your guidance principles and your practices, and you should start creating some parameters for how you want AI integrated into your content.
So I’ll give you a great example I just talked about with some college professors. Your syllabus—how many hours of this should be AI related? How many of the hours should be based on non-AI-related research, and sort of indicating that in your syllabus that this is an AI activity that’s acceptable, this is not.
So we have to change the way we as educators also sort of present this to students to give them the same framework of literacy that they can also take back. So we got to start somewhere and we can start with us.
Q: Hello. Catherine Odari. I’m a professor of history at Spelman College in Atlanta, Georgia.
Mine is more of an international. So I recently watched a documentary. It was a BBC documentary about how tech companies have outsourced the job of moderating the content of AI to African countries. So in this particular one it was Kenya, and the people who were tasked with the job of moderating this content they worked long hours and not paid, and if paid at all it was very little. But the impact of what they had to watch for those long hours was tremendous in terms of their mental health. And so my question is, how do we as a society—a global society—ensure that it is not these countries who, again, are bearing the brunt of something that is benefiting the developed world at the detriment of these societies?
SEGAL: Yeah. I mean, we saw this with social media content moderation—the first wave of it, right? A lot of it was offshored to people that then had to watch terrible, terrible things for hours and hours again.
I think what only had an effect was—quite honestly, was transparency so the companies were embarrassed and were forced to spend more money and pay more and do some of it. But, you know, clearly, other firms are going to rise and take advantage of that as well.
So there’s not a lot of regulatory tools I see in place and I would suspect it’s mostly going to come from embarrassing the firms and pulling them out. I think we’re also going to have the issue about language expertise and people on the ground that are—you know, the most well-known example was in Myanmar—in Burma, right.
Facebook had one or two people that could actually speak the local language and so could not track the genocide as it was developing. I’m sure we’re going to see something similar to that again, given that the firms are going to find the cheapest people to do it.
Q: Hello. I’m Kue at Idaho State University.
So I know internationally there’s a lot of push for digital government and also digital inclusion, but also I think other countries also kind of emphasize the interoperability of different ministry works. And I know in Idaho one of the things that we see is that there’s kind of a disconnect of what’s going on in the world versus what’s happening in the local government. Systems are not integrated and much more difficult to find relevant information. So I’m wondering how do you guys see this divide, how do you see movements from international spheres as they struggle or kind of push for interoperability for efficiency and effectiveness purposes but also at the same time struggling with the fact that iCloud data servers are probably not in their country, right, and so in other countries that’s oftentimes outside their jurisdiction. So, thank you.
LEE: So your question is actually pretty relevant, right. We see—I go to the Mobile World Congress every year and, you know, countries, like—other countries are doing digital twins. They’re figuring out how to do smarter cities and to rely more on compute power to be able to integrate government services, consumer services, et cetera. That conference is amazing, by the way, because you get to see how cities are thinking about public safety or, you know, online enrollment.
Idaho, you got a little bit of work to do, right, you know, partly because of the broadband concern. On the infrastructure side we know that rural communities, for example, just have a lack of facilities. I call it more cows than people because there are more cows than people.
But we can’t discount that because there are farmers that want to be productive that have the ability—we have—I visited a five-room school with twenty-five students who were all Amish during my travels and they were the top robotics team in the state of Maryland even though they couldn’t actually play with robotics at home.
But the idea was that they would be more productive if they knew these skills for their families, who were actually entrepreneurs and independent factory makers.
My point is—you know, the best way I can explain it is we talk about China. When I go to China and I go to some of the smaller markets I pull out money. They pull out a QR code. And until we have, I think, the political will to sort of see how we use integrated technology to do some really amazing things that make us much more smarter, much more mobile, much more efficient, for me it’s the lack of political will to do it in a bipartisan fashion that keeps us in the same space of being much more innovative.
SEGAL: Yeah. I always like to point out the Chinese complain about this too, though, right. There’s plenty of writing in the Chinese press—
LEE: It’s true. It’s true.
SEGAL: —about, like, data islands and this agency won’t share data with that agency and everything else. So it is a kind of structural problem that anybody that has data generally doesn’t like to share it.
LEE: Right.
SEGAL: But, you know, look, we’re Americans. We’re not going to learn from the rest of the world. (Laughter.)
LEE: You think? (Laughter.)
SEGAL: But there are great examples, right? I mean, Estonia and Singapore and other places have extremely well-developed systems that protect user privacy, that integrate—you know, you have your ID and you can vote online, you can open your business in twelve hours, and you can do all these other things.
Now, we don’t have the structures in place to do that but that I—you know, there’s lots of great models out there that one could think about building and, again, at the local level there are probably, you know, mayors and governors who are thinking, you know, how do we build on that.
LEE: Yeah. But you also have to be careful of greater surveillance, right? So when I do go to those conferences, Estonia being an example I think it was—maybe it wasn’t Estonia. It was another conference. It was one of the years I was there.
They said, we have these smart bridges and then we can also see when people are walking on the bridge and they’re standing there for a long time, and then we send the police out. (Laughter.) And I said, well, why do you send the police out? Well, we don’t want them to hurt themselves.
They also have to be careful of surveillance, right, because those added layers of surveillance when people get this bug can also be detrimental to our individual freedoms here of what we expect.
SIMS: Go back—farther in the back. Yeah.
Q: Thank you. Ginta Palubinskas, West Virginia State University.
One of the things that we don’t hear much about is the impact of AI on the environment or environmental concerns. And I wonder if you could address the compatibility of the rising use of AI and its impact on the environment in terms of energy use, water, and global warming. Thank you.
SEGAL: Yeah. So, I mean, the energy issue is going to be huge—
LEE: It’s real.
SEGAL: —even if it’s more than eating a hamburger or less than eating a hamburger.
LEE: (Laughs.) Yeah, it was—I got to figure out the analogy. (Inaudible.)
SEGAL: So it’s going to be huge. I mean, that’s why, you know, people like Altman and Schmidt are all investing in modular—small modular reactors. Water is a huge issue, especially on—for the chips and production of the chips. You know, I was out at TSMC in Arizona three weeks ago—last month and, I mean, just always remember—they’re, like, water is not going to be a problem. I was just reminded it’s a desert but they insist that water is not going to be a problem, that they’ve figured it all out.
So I think people are paying attention of it. But, of course, you know, the tech companies say that, you know, we’re going to solve the problem with nukes or other spaces there. There were some—there was an interesting study out of Duke two months ago or three months ago that basically said actually energy is not going to be as big a demand as we thought it was going to be, that you can meet a lot of the demands with the grid we have now which was, I think, pretty surprising but about rebalancing and other things there.
But I would say that, yes, there is a fairly large focus on it, at least in how do we provide it. Now, again, the response on the climate change and the pollution and environmental and urban degradation I suspect will be local as opposed to at the national level, considering that the administration has essentially signaled that, you know, the most important thing is to go as fast as possible.
LEE: And I would just say too, I mean, I think that was—and this is an area that I know you’ve been working in, Adam—that’s where it was interesting when China came out with the DeepSeek model and sort of suggested that they could do the same type of high-capacity processing with less energy and less infrastructure.
It sort of challenged many of the investment models, and it goes back to this gentleman’s question in the back about power that many of the companies had suggested it was going to take to be able to run their models.
I’ve had the opportunity to see one of the chips. It’s about my height, and the key thing is the cooling of those spaces. And I live in northern Virginia where we’re opening a lot of, for example, data centers where we need these to be running a long time.
I think the environmental question is one but we should also be looking at the workforce question, too. You know, as we’re building these data centers in actuality it takes less workers to actually man them because they’re pretty much self-sufficient. And we also have to think about—and this is an area I have to admit that we’ve not gotten a lot of climate change experts to sort of chime in yet because of the larger conversations of recommissioning nuclear sites.
We don’t know where some of these places are going to go, right, at this point and we’re not sure, again, how they’re going to interact with the local community. We’re hearing bits and pieces. But I think, again, in the next two to three years we’ll probably hear more of the climate implications or the community implications on water systems and others later.
SIMS: Yes?
Q: My question is, all the rest equal, what will be the effect of AI on inequality both internationally and also within United States from state to state, and as educator even more so in light of the shrinkage of the Department of Education and the trend to leave more and more responsibilities for education to the states?
LEE: States, yes.
Q: Full disclosure, I just asked AI the same question. (Laughter.)
SEGAL: What did it say?
LEE: And what did it say?
Q: It will exacerbate.
LEE: Inequality, yeah. Well, they got that right. (Laughter.) It depends on which one you’re using.
So, you know, this is my space, right? I mean, again, I try not to sit in the Debbie Downer space so I try to think of the and—the both/and. I’m very conscious of doing that because I still think, again, AI can be very helpful in stair-stepping some more personalized learning.
Invisible assessments is an area I’ve been looking into recently where a kid doesn’t have to be put in the back of the classroom to do an assessment. They can actually do it through AI where it’s actually scaffolding their learning in ways that they are prompting them for different research questions.
So it can actually present itself as inequality. Again, going back to my first point, educators are not sitting at the table when these technologies are designed and they don’t necessarily have a vested interest in these technologies helping us to advance educational achievement.
Think about it, folks. I mean, our procurement in the education space of a lot of shiny objects over the years tend to be on what is sold as something that’s going to just make our jobs easier.
Now, if you talked to the Department of Education maybe three years ago this is what they would have said because I was on a panel with them and they said, listen, we’re working on ways in which we do responsible, equitable AI and that’s what they did, right?
They wrote many manuals over the course of the Department of Education and the Biden-Harris administration on how to use it responsibly within school districts, how to think about the integration into the classroom.
With all that being gone, I am afraid that we’re going to lose some of that repository of learning and best practice because, to your point, states are going to decide where they want to put more or less AI and how they want to use it, and it’s not necessarily going to be presumptive of me to say that in areas where there are probably low-performing schools that they may move away from AI and go back to some basic cognitive abilities.
But you ought to be honest, too. Across the country we have a resistance to technology. Shut off your phone. Don’t use it for this and that, which is also potentially a mistake because everything that kids are learning today is technologically based.
Our challenge when it comes to inequality—and I’ll say this to all of you as educators and I will sound like Dr. Martin Luther King—it is in your hands as educators to come up with good guidance, professional development practices on how it’s going to work within your university, and to develop cross collaborations on what that looks like.
I just spoke to the California Community Colleges, their association of 750 people. They’re starting to have these conversations among their association—how do we want to use it so we do not replicate inequality and what does that look like.
Arizona State University is another really great example. They’ve developed a student and faculty advisory to work with the tech companies to ensure that the tech companies are not just bringing their products and then creating different disparities within their student population, which tend to be commuters or tend to be bilingual.
So you have that agency to really have these conversations in a very deliberate and intentional way, knowing that if we don’t do this at these sectoral levels the AI will run wild and will enhance the—I don’t think it’s going to create. It’ll just enhance the inequalities that we have.
SEGAL: I’m sorry. I’ll answer that question.
SIMS: No, please. Go ahead.
SEGAL: I thought we had to end at 11:00.
So I don’t think we know yet what the diffusion of AI is going to look like, right? If it’s a general purpose technology that looks eventually like electricity then its impact on kind of imbalances of power, I think, will be less than if it is something that’s tightly controlled and only used by some states.
And I don’t think we know yet, right? I think it’s very—we’re kind of—there’s a strong argument to be made that it doesn’t matter as much as if you innovate in the space as it is if you diffuse it, right. So that means lots of countries could be fast followers and apply it in interesting, creative ways and so that doesn’t matter so much that they don’t have the one or two companies that do the breakthroughs.
That said, short term I think it totally exacerbates inequality.
Q: Thank you so much. My name is Mindy Haas. I’m from the University of Pittsburgh. I study intelligence and international security.
My question is about what implications do you see for AI in international security technology, right? Nicol, you had mentioned, you know. I suspect that states will be able to find some military common ground with regard to AI. Is that going to look like autonomous weapons? Is that going to look like targeting systems? Is that going to look like Cambridge Analytica intervention, right? Sort of what do you see on the horizon for AI and security?
SEGAL: Yeah. So I think there will be some broad—nothing on social media or information warfare or intelligence. All of that will be, you know, every state for themselves. On military applications there will be a kind of continued discussion about application of international law and how that should guide it.
You know, I think we’re already past the man or human-in-the-loop conversations. Like, most people now believe that that will be not possible because the decision making is just going to happen so fast.
So then you have to have something that’s human-on-the-loop kind of, but then really it starts—it puts the responsibility on designing the system itself to make the decisions based on international law and other provisions there.
So we will probably get some broad agreement among likeminded about that, but the others will probably refuse to go along in those discussions and not have a lot of agreement on this space.
So, look, in cyber all we have are these state norms which, again, are not really followed and I suspect AI will be the same thing. Will not be binding treaties. There’ll be a difference about countries arguing about the application of international law.
So we’ll still say, you know, no, we expect proportionality. We expect distinction. We expect neutrality, and other countries will resist and say, no, no, we need a whole new international treaty and that means going to the UN which means, really, no progress.
So that’s where I would expect. So kind of like minded, a similar view about that, and then the others.
SIMS: In the back, yes.
Q: Good morning. Laine Munir, Arizona State University.
LEE: Yes. So Lev Gonick. Yeah. Yeah.
Q: Yes, thank you.
So I have a very praxis-oriented question, if we could look at AI use in universities from the ground up.
At ASU, we have been given four verbatim paragraphs that we can choose from to put into our syllabus. The first is no AI use permitted, period. The next one is limited AI use under these conditions with permission. The third is use AI as you wish and the fourth is you must use AI in order to complete this course.
For example, with the idea at our law school that a graduate who uses AI will always outperform a graduate who doesn’t use AI, so when we’re talking about social inequality and diversification and democracy within AI use, what strategies might we as individual educators implement in our daily work to help ensure that we are teaching our students to use AI as allies of democratizing ChatGPT and making it more inclusive.
So one thing that we are always worried about is at this point I’m wondering what can I do in my daily work to help us get to a place where AI is inclusive and democratized?
LEE: You know, I would say on that, I mean, obviously, there’s an institutional piece. So Arizona I know that university. I know Lev really well so I’ve been part of this, like, advisory that they’ve put together.
But, again, on the institutional level your institutions should be deciding how you interact with the AI companies and what products actually make it onto your campus, and how it’s actually sensitive to the linguistic diversity of the students—the learning diversity of the students.
And then on top of that—and some universities have already done that—you have some type of infrastructure that allows students to use the—access the resource whether on campus or off campus to address digital divide concerns.
When I was at Spelman, you know, there is actually an AI-based course that is actually introduced to the students at Spelman College where they go through, you know, this very—it’s almost like a competency for students, and the reason I know that they know that stuff is because they asked really hard questions when it was related to AI and these were some of the computer science students, the math students, and the social science students.
So I think as an individual educator, first and foremost, I love the four bullets. We have to come to some agreement at the university level where are we allowing and permitting the technology to be used.
We have the same type of policies at Brookings. It should be the same thing across universities, just a really outright transparent disclosure of where AI is assessed in the students’ use for you as well as the educator to be able to know here’s where it’s an acceptable norm for my classroom.
But then, two, I think it’s really important for you as individual educators to check your values, norms, and assumptions at the door, OK, because there is the assumption that everybody who uses AI that may be a low performer in the classroom may be using it to cheat. I want to keep going back to this.
Listen, you know when a student is cheating if the topic was on, you know, the Alaskan wildlife and they’re talking about Greenland—(laughs)—because they didn’t check the AI that they actually pulled it from. I’ve seen those. I, you know, teach. I’ve seen them in college essays. I’ve seen it in scholarship applications where the student has not checked the fact that they pulled a prompt and it has nothing to do with the question. So that’s obvious.
But I think we need to also figure out what type of grace we give it. This goes back to the question of AI literacy. What kind of grace do we give students as educators to say, where’d you get that prompt from? Is that something that you’re interested in? Is that comfortable? It’s a different way of teaching.
We did an education in AI workshop at Brookings over the summer. We had educators at the college level. We had Khan Academy there. We had college professors. We had Google and technology companies, and what we realized is that the educators are the most disserviced because they don’t know what their job description is.
So starting there, to your point, really helps us to frame how we teach, how we interpret, how we interrogate, and how we actually get our students to learn from the technology because it is going to be part of the workforce, going forward.
SEGAL: So I teach adjunct at SIPA once—like, one class a year, and so we had the bullet points but we had three bullet points, but the second one was if you use it tell me how you used it, what the prompts were, why you used it.
And so out of a class of twenty-five I only had two students that chose two, which I was—that prompt, the—which I was surprised by. But then I was confronted with—so I, you know, read the essay and I was, like, oh, this is a pretty good essay. And then I got to the back and I saw all the prompts and I was, like, oh. (Laughter.)
And then I realized that they had just clipped and pasted the responses and I was, like, that’s not what I—that couldn’t be what I meant, right, just clipping and pasting the responses.
But I hadn’t explicitly said that. So I was kind of stuck. And then I realized, you know, at SIPA the default thing you do when you have a problem is you give the student an A-minus. (Laughter.)
That was a graduate seminar, and I realized that the student had wrote the prompt that said answer the question as if I’m a 90 percent student, and I was, like, why did they choose that? Like, why did they go for 90 percent?
LEE: Because that’s the inequality.
SEGAL: So, you know, I’m a political scientist so plural of—you know, data is a plural anecdote. So I guess my only response is that it’s a learning iterative process so I’m not going to change the policy next year but I’m certainly going to be more explicit. You can’t just clip and paste the answers.
Like, you at least have to reshape and rewrite the wording so I saw that you engaged with it. But you still have to show me all the work. I’m still not completely comfortable with it, quite honestly.
LEE: And can I—I know we’re short on time but it goes back to, like, the question of the internet, right? So I’m going to date myself. I know I look a lot younger.
But, like, I remember I used to hand write all my research papers, right, and then a typewriter came out with the white-out function, and then I used to type my papers, right, and I used to feel so accomplished when I handed it to my teacher and said, here’s my typed paper.
And I realized some of my peers in my—I went to school in New Rochelle, New York, not too far from here—they didn’t have a typewriter, right? And so I felt that I would get more credit if I typed it, and I spent a lot of time ripping the paper out and putting it back in because I thought that was going to allow me to be on par with students who had that resource.
My point is we’re too premature in our settings in education to make that a level-up tool for our kids, right—for our students. So we have to actually explain the culture in which we want this tool to be used, and when you explain the culture, over time we’ll get more young people to see that this is a transferable skill that has its benefits. It has its drawbacks.
But, most importantly, it will help us move a little bit beyond some of the conversations that I tend to be in the middle on where we’ve got to get rid of all the technology because it’s just blasting the critical inquiry skills of our students.
It’s actually very helpful, you know, at times and we as adults use it all the time with that presumption that we’re writing—I mean, employers are writing performance reviews using ChatGPT, believe it or not.
People are using it for research and talking points. We didn’t do our talking points. Maybe Calvin’s questions were AI-generated. We don’t know, right? (Laughs.) We need some journalists here. But we got to figure out how to use it in a way that—
SIMS: Journalists being—(inaudible). There you go. Yes. (Laughter.)
LEE: I had to put that out there. I was on a panel where somebody said, I did all my responses from ChatGPT. I was, like, well, you cheated here, right? (Laughs.) All of us. (Laughs.)
SIMS: Well, I want to thank both Adam and Nicol and everyone here today, and we hope to see this again, you know, in another, say, months or a year from now to see where we are.
But thank you all for coming and participating. (Applause.)
(END)