#34 Digital Infrastructure with Nicole Starosielski

When you send a message to a friend—whether by text, email, or social media—you probably don’t think much about where it’s actually going. We’ve grown so accustomed to our free-floating devices and digital clouds and seemingly wireless connectivity that we might not realize just how wired our digital world truly is. The reality is, every message we send travels through a vast physical system: routers, cell towers, data centers, power grids, and miles of undersea and terrestrial fiber networks. And every part of it depends on human decisions. This means that the future of digital infrastructure depends not just on engineers but on policymakers, biologists, economists, designers, even journalists. But how do we bring all these people together?

Nicole Starosielski, professor of Film and Media at UC Berkeley, is working on exactly that. She studies the hidden physical systems that make the internet possible—from undersea cables to global media networks to the environmental costs of connectivity—and is training a new generation of leaders who can navigate this field across all dimensions.

Further reading:

• "Professor Nicole Starosielski (Film & Media) launches world's first certificate in internet infrastructure at Berkeley"
• UC Berkeley's Summer Minor and Certificate in Global Digital Infrastructure (GDI)

This episode was written and hosted by Pat Joseph and Leah Worthington and produced by Coby McDonald. 

Special thanks to Nicole Starosielski, Pat Joseph, and Nat Alcantara. Art by Michiko Toki and original music by Mogli Maureal. Additional music from Blue Dot Sessions.

Transcript

Transcript: 

If you’re like me, when you send a message to a friend—whether by text, email, or social media—you probably don’t think much about where it’s actually going. Physically, I mean.  It feels wireless. Frictionless. Like your message just floats up into some abstract “cloud” and instantly appears on someone else’s phone halfway around the world. Right?

Not quite.

We’ve grown so accustomed to our free-floating devices and instant messages and completely wireless connectivity that we might not realize just how wired our digital world truly is.

The reality is, every message we send travels through a vast physical system: routers, cell towers, data centers, power grids, and miles of undersea and terrestrial fiber networks. And every part of it depends on human decisions.

When a ship drops an anchor and accidentally slices open an undersea cable, someone has to go down and fix it. When a data center gets built, local communities have to make choices about access, energy use, and security. This means that the future of digital infrastructure depends not just on engineers but on policymakers, biologists, economists, designers, even journalists.

The question is—how do we bring all these people together?

[MUSIC IN]

LEAH: This is The Edge, produced by California magazine and the Cal Alumni Association. This interview was conducted by Editor-in-Chief Pat Joseph. I’m your host, Leah Worthington.

Our guest today is Nicole Starosielski, a Professor of Film and Media at UC Berkeley. Nicole studies the hidden physical systems that make the internet possible—from undersea cables to global media networks to the environmental costs of connectivity. Last spring she launched the world’s first certificate program in internet infrastructure at Berkeley, an effort to train a new generation of leaders who can navigate this field across all dimensions—from the technical to the economic to the legal.

This conversation completely changed how I see our supposedly “wireless” world. So if you think digital infrastructure is dry, think again.

[MUSIC OUT]

PATRICK JOSEPH: 2:27

The obvious question, the one you must always get is, you're a professor of Film and Media Studies, what are you doing concentrating on digital infrastructure? 

NICOLE STAROSIELSKI: You know, nobody really starts in digital infrastructure, which is, I guess, part of why I started this program. 

In fact, digital infrastructure wasn't even called that as a kind of coherent entity until, you know, maybe, like, the past 10, 20 years. 

So to back up, I was a doctoral student at UC Santa Barbara and I was in a film and media program. Now, I had done film as my undergrad at USC, so I kept going, and I started thinking about media. But, you know, Santa Barbara was very much like everybody was doing ocean stuff. So I started thinking, Okay, I'm gonna do underwater media, like I was thinking shark documentaries, and, you know, remote operated vehicles, which are very recognizable as media for a general audience. And my advisor said she worked on satellites, satellite TV. She was like, You should look at undersea cables. And I thought, oh my gosh, this was like 2006 I was like, that is so boring. Never gonna get a job. Why would I thought they were old? Yeah, when she said that, what was called to mind undersea cables. Doesn't like telegraph cables. Like, I feel like, if anybody thinks about undersea cables, they're like, Yeah, that was cables were old, like, and that's because you have a phone, and your phone has less and less cables as the years go on, right? You now can charge your phone completely wirelessly, right? You don't need those cables. You don't need to be plugged into a wall. And so I think culturally, we're always moving towards a less cabled future in our imaginary 

PAT: Yeah. I mean, even Wired Magazine, that title seems quaint, right? Because we were very quickly, wireless. 

NICOLE: Yeah, it's, it's, it's wireless, it's unwired. It's the freedom that communication technologies are supposed to give us. I mean, we sign on to our phones, we go on social media in order to communicate with others far away from us or who are sitting next to us. But you know,l to find a way of connecting despite space and time. And we don't usually think about those infrastructures like a railroad. We think about infrastructures that are bound to the ground. So, yeah, so back to 2006 I was like, this is a terrible idea, but, you know, like when one's dissertation advisor is like, you should do this idea. I was like, okay, like, I'll put a chapter on undersea cables into my dissertation. And then I went down a rabbit hole. And it turns out, people go down this rabbit hole all the time. I have so many friends now in the industry who didn't know about undersea cables, and then the moment they found out about them, they were like, oh shit. Like, obsessed with them. And so I found out that almost 100% of trans-oceanic traffic, like intercontinental, transnational traffic, goes through undersea cables. 

PAT: So if I send an email to someone in Switzerland, it's going to go across the ocean in a cable. 

NICOLE: It will go on the very bottom of the ocean. Cables are on the sea floor. So your email, your words, the movies you watch, like everything that everybody does is right now on the sea floor

PAT: And this is fiber optic cable. When we talk about cable?

NICOLE: Yeah,, so it's been it's like glass, and it's pulses of light your words, when somebody is listening to this podcast, if you are far away from San Francisco, let's say you're an ocean away, then these words have been pulses of light on a cable on the very bottom of the ocean with all those crazy sea creatures. 

PAT: It makes me think of Ted Stevens talking about the internet as a series of tubes, and how everyone laughed, but how far off was he really? 

NICOLE: No, it is a series of tubes. It's the plumbing of our infrastructure, and we rarely think about it, because we our interface is wireless. What we're immersed in is a wireless culture. We are increasingly untethered and so it makes sense that we would think about the technologies like these advanced technologies that are making everything possible is also wireless, but it turns out they're really, really fixed. And I think in the last few years, people have started to see this because they've started to make mainstream news. 

PAT: Yeah, how so? How have the—oh, you mean the undersea cables—

NICOLE: Undersea cables, but also data centers, you know, with now, they have aI factories, they're called, so, you know, an old school data center, you know, you'd have a rack, and it would generate a certain amount of energy, and you would, you know, have all these computers into this rack, like you've seen These, you know, server racks probably in, you know, movies, and, you know, there's enough media about them now, so they generate a certain amount of heat, and then you had cooling. Now, AI factories, they generate massive amounts of heat, right? And they have to develop new cooling systems. And these factories can be enormous. And so it's a new model of infrastructure development. I mean, it's in continuity with the existing model of infrastructure, but it's a new scale, it's a new land footprint. It’s a new electrical draw. And so that scale, which is being driven by AI, is now getting a lot of people's attention because it's having an impact and a presence that's much more acutely felt. 

PAT: Yeah, so is it the case we hear a lot about the energy demands that these data centers require, and is it true, then, that the biggest need for that energy is to cool everything? And is it true also of the water that people talk about? 

NICOLE: So it depends. Yes to the energy. So, cooling. I mean, this is true really broadly. If you think about even energy, your energy bill at home, like the usually, the largest portion of that bill is for heating and cooling, and so even if they didn't have ???, the need to provide a stable environment. Now obviously that really depends. In really hot parts of the world you're going to need, you know, an extensive cooling system in order to make sure that you have a stable thermal environment for all of these computers, and for, essentially, AI to run so that it sort of depends on where you are in the world. And then the water usage, it depends on the technical decisions that the those people who own and operate data centers make. 

Now, there are some technologies that do not use water, like, do not draw consistently water, or have a closed loop system where, you know, they draw the water out and then it's done. So I think that that's one of the things that really needs research, right? You need to understand right now we have this kind of flat understanding of data centers and digital infrastructure. People hear about it, they're like, ‘Oh, it's an AI factory in the middle of the country that’s, like, you know, taking up all the farmland.’ That's one model. But there are also, like, small, modular data centers. There are also kind of more community-focused models. There are, you know, people who are looking for data centers in their nations because they want data sovereignty, and they don't want all their data to be in the United States or in the cloud region of a Google or Amazon, right? So I think that we have to start from the position of, who are the people that we're talking about, what is the environment? Are people advocating to have the internet? And then what kind of internet do they want? And then that always requires an infrastructure, so there is no communication or mediation today, like, technically, you know, a facilitated communication or mediation that doesn't have an environmental footprint. You know, paper has an environmental footprint. Recycling itself has an environmental footprint. So once you say, alright, our capacity to communicate with each other has a material impact on the world, and that's not equal, and it has greater impacts on some people and greater benefits for some people. So we need to study that, and we need to develop, like a literacy around that. We need to educate about that. And we need like a, you know, a population that can understand that if you want to make decisions about the system that your life depends on.

PAT: Yeah, yeah. And I think that's, that's the interesting thing is that most of us, including me, are pretty much blind to to, you know, at the point of contact, of, like, typing on my keyboard, I'm not thinking about the data center. I'm not thinking about when I save a short video, what happens with that? Does it have a physical footprint? And I assume it does. But we talk about the cloud, and the cloud makes it sound like it's out there on the ether somewhere, and I can just recall it magically. So, I mean, I do think the cloud more than any other term, seems to me, the most beguiling in terms of how it's made us think about this.

NICOLE: Yeah, and that's why I mean things like, you know, our program, the program that I built here, the global digital infrastructure program, is interdisciplinary. And often we get this question of, okay, but is it just technical? Is it just located in the College of Engineering? Are you just training people to build circuits and AI or whatever? But it's so like all of the industry's issues right now, have a cultural dimension, have an aesthetic dimension. I mean, the issue of the cloud, I mean, that's an imagination of what our infrastructure is, right? We are moving to cloud-based technology. The vision of an AI factory is very different, and so I think our metaphors matter, artistic representations matter. And like we need innovation in artistic representations of digital infrastructure in order to imagine what those futures are. That's one of the things that I always want to tell people, is that, yes, you can learn, you know, the engineering behind data center, you know, development. You can learn the business. And often, I mean, even that's an interdisciplinary process, because a lot of engineers aren't trained in business, and people who are in business aren't trained in engineering, and then you have this level of like policy, which is also politics and regulation. So you’re not going to be able to build digital infrastructure wherever you exist without understanding those elements. And then there's just so many people at the base of the internet. It's all people like you drill down below AI, there are people, and we haven't really paid attention to that human layer. I think the fantasy is both, it's in a cloud, but also, like that cloud is outside of Earth, and nobody lives there. But there are so many people in the cloud.

PAT: Talk a little bit about that. I mean, who are the people that you see at the base of it all? 

NICOLE: So okay, very specifically, on one level, there are people who work in data centers or in cable landing stations to operate those systems so that can range from people who are doing, say, security operations, to sort of being a network technician. I've heard from many people in the industry, the data center digital infrastructure is a living entity. It is an organism. It requires constant engagement. And so there are all these people who are involved. Okay, so that's one level, right? So you think, all right, we got to keep the internet alive. We have a class on that coming out next year or the year after, 

PAT; Called keeping the internet alive?

NICOLE: Called keeping the internet alive. And like the the assignment for that class is developing a game. But you know, based on those processes of keeping the internet alive. Okay, so that's one layer. 

Then you have people who are sort of beyond the facility, who also work to sort of maintain those systems. So for the subsea cable industry, that's a whole maintenance fleet. So I know a lot of people on ships who are like, you know, flying all over the world, moving from ship to ship, like, cables get broken once every three days.

PAT: Really?

NICOLE: Yeah, so you know, they're they're constantly being broken and being repaired. We don't usually see the impacts unless, you know, there's a volcanic eruption in Tonga and it cuts off the country's internet. 

PAT: That actually happened.

NICOLE: That actually happened. But generally, you know, people are maintaining these systems. So that's sort of maintenance beyond and you have beyond the facility, and then you have all of the people who are building these systems.

And that is just, it's like, every single discipline, or, you know, like, we had over 52 majors take our course this past summer, and everyone was like, like, people who are in biology were like, I see, you know, because digital infrastructure is affecting their field, right? So in, let's say it's a biotech company, there's somebody who's sorting out how to use the technology for that company. So the builders are not just the people who are building the data centers. They're the people who are trying to interface with that physical infrastructure and use it, and if they don't understand it, like they can't implement it, right? It's not like AI can just take your company and plug it into a building.

Okay. So then beyond that, you have, like, mechanical engineers, electrical engineers, like material science, like trying to figure out the like, how to green this stuff. You've got people in finance trying to figure out, how do you finance, like, a $40 billion dollars, like, the level of like, the deals that are going through, you have people in policy and regulation, like, governments around the world are, like, trying to figure this out right now. They don't even know what's inside the building, right? Like, so, how do you govern something if you don't know it. So all these people are involved in the constitution and then operation of digital infrastructure. Never mind, like all the people after it's already built or needs to be dismantled, there's all these cable recyclers. So there are entire industries built upon this, and they're all these people who are constantly moving around trying to keep this thing that we think is very fixed, alive in operation.

PAT: Yeah, yeah. And again, it's one of those things it's just so easy not to think about, especially the people repairing cables on the sea floor. It's just not something that would ever occur to me. Likewise, I imagine there are places where this cable comes ashore and it's probably visible to a lot of people. You can step over it, but you wouldn't know what you're looking at.

NICOLE: Only a few places in the world. 

PAT: Okay, good to know usually, usually they're buried under the beach, not just because, you know, for security reason, but you don't really want a conduit with a cable going, I mean, so much power goes into this cable. Like, take an ax to a raw cable, like, you would be injured. So it's a safety consideration as well. 

PAT: It sounds like a system that, like you mentioned, Tonga and an earthquake, and that there are a lot of ways that this could be damaged, either on purpose or by accident, and that there would be consequences. 

NICOLE: If you think about infrastructure, digital infrastructure, and the way that you think about houses and roads and transportation networks, there are a lot of similarities, right? So you say, okay, somebody has a house up in the hills on like, a windy road, going to be much more susceptible to a fire. Down by the coast, going to be more susceptible to flooding. Some areas you're going to have a higher security risk.

So similarly, like if you look around the world, there are some systems that are in areas where there's just a high risk so that, you know, the English Channel, yeah, one of these so many ships, I heard someone say, like, you know, look, our cable is just gonna get broken every every year. Like, we're gonna have to pay hundreds of thousands of dollars to repair it. We just build it into our business case. 

PAT: Let me ask, why does the volume of ship traffic have that effect? 

NICOLE: Because the biggest single threat to cables are not, you know, cutting instruments. They're just ships anchors. 

PAT: Anchors? I hadn’t thought about that.

NICOLE: They're just, they just drop anchors on cables and then they break. So if you historically, if you look like you'll see the cable hubs you know, were they tried to locate them away from ship traffic, not successfully in some places like Singapore. But like, you know, there were cable hubs that were sort of dislocated from urban areas because they didn't want all that boat traffic or fishermen to disrupt them. And that's historically been the single largest concern of people who own, you know, internet cables. They're mostly concerned about fishermen and people with their anchors and their boats. 

PAT: How did you educate yourself on all of these disparate topics? Because there's a lot you're telling me a lot of things that sound like geography and engineering and, you know, technology. So was it a slow process of just reading books? Or how did you gather all this? 

NICOLE: I mean, I did read a lot of books, but a lot of this information isn't contained in books. You know, it's contained in people. And you know, I've been studying this for 20 years now, and I have met so many people. So I've gone to industry conferences. I've, you know, reached out to people for conversations. I've talked to people who live around cable landings. And so you start to stitch together through this social network, how the infrastructure works by looking at its interfaces with those many humans around it. And those people can always be found, right? They're always talking to each other in many different places. So oddly enough, like, even though I'm often doing like, sort of humanities kind of study where I'm, like, looking into the history of something, for example, like, a lot of that will come from, like, an oral history from somebody that, like, I was having a one off conversation with at a conference, yeah? And then you just learn so much from those people. 

PAT: Yeah. Well, you mentioned, so we just talk about the certificate program, and you mentioned that you had over, I think, more than 50 majors signing up for it. What do they get out of it? What do they come out of that program with? 

NICOLE: Yeah, so I would say they come out with sort of three different things. The first is, is that they have a general sense of how it all works. So the first course is sort of like the alphabet or the dictionary, where you're like, what is a data center? Where did it come from? What's an undersea cable? What are all these cable repairs? And then you get in the project, how to build a global internet class, you think about almost the grammar, which is to say, Okay, well, how do I build one if my country doesn't have resilient connections to the internet? Like, how do I do this? And then the third class is really about, like, you know, it's called tech wars, and it's about how those systems then come into contact with the rest of the world, right? Like, you know, oh, you're experiencing community contestation. Here are all the different, you know, regulations and policies that don't make sense when put next to each other, right? So you can have like, these inconsistent or contradictory policies. So that's one area where you gain through these lectures, like, a base understanding of how it all works. So if you went through the course, you would then be able to look at your phone, and sometimes we connect it to actual media to say, okay, if I'm playing a game on my phone, it's probably going this way. It's probably depending on these infrastructures. So that way, like it almost lights up an entire imagination of what is the world behind the act that you do on your phone. 

PAT: When you say it's going this way, you mean the data is traveling a certain way. 

NICOLE: Where's the data going? Who has it? Who’s managing it? So that's, that's kind of the first thing. The second is it brings you into that human network. So we have guest speakers. I think this is one of the things that makes our program distinct at Berkeley. We have, it's a summer program. Every week for each course, we have somebody from the industry or a major international organization or nonprofit come in and speak to the students, and we record these but students are also invited to come talk to people, yeah? So they often, like, drop their like LinkedIn handles and connect with students. And from their side, they're like, Wow, this is so cool. And nobody's ever studied us before, yeah. So they're very eager, right? So I would say it's that kind of like industry immersion and like seeing those people behind it. So it's not just like me lecturing, yeah? Which I feel like, you know, I can tell some good stories. You're gonna listen to me a lot, but you're also gonna, like, meet the people behind the scenes, right? And they'll often be able to tell you things that are happening, like, right now.

The third is that, like, you're really able, like all the classes are project based, like, there's some easy quizzes, but like, you are working on pitching your own data center or subsea cable for the build class, it's being taught by Erick Contag, who has built many subsea who has built data centers, right? So you're actually working on projects in relation to people who like understand what those are. And then we have a number of different competitions that take student work and put it out into the industry. So we have, listeners can't see, but we have right here on the table between us, some student articles that were printed in the magazine as, so they actually printed in an industry magazine student research papers. We have this year, a company called Cloud Carib, which is based in the Caribbean, is running a program to give essentially like server space and support for people who want to develop apps who go through the certificate. So the industry has been like, Oh, this is a really great opportunity to support that next generation. So I'd say that's something that people can get out of it. Now, not everybody submits an essay f y that you want to kind of build your career in this space, there's like, room for actual engagement and moving your your work forward. And we have, you know, we have people from major tech companies taking it. So it's a, it can be a standalone certificate for people who are, like, working professionals. That was really important to us, like, we really want to reach Cal students, but there's a massive alumni network. There are people who are out in the world who really want that education.

PAT: Are our students getting hired with this knowledge? I know everybody's anxious about AI sort of chasing away all the entry level coding jobs and computer industry jobs. Is this a leg up for anyone in the job market? 

NICOLE: I mean, we don't do, you know, extensive student tracking afterwards, but like I know of at least five people who have gotten jobs specifically mobilizing what they've learned in the industry. So one person got a job in a cable landing station. Another person got a job for a kind of major organization. Some other people went on to do graduate work. I'd say the infrastructure—so software is a lot easier to sort of manipulate and also replace human labor with machine labor. The physical world, it's much more difficult. Like, you cannot send AI over to, like, talk to a regulator in the Caribbean, like you just can't. You can't send somebody. You have to actually have somebody who can go, who speaks the language, who understands the policy. There's a certain kind of business environment, you need to kind of understand culture like, so there's, like, all these ways that things like area studies are extraordinarily relevant, and people are like, looking to hire somebody who, like, understands. But also it's a major focal point of government policy and regulation right now. And digital sovereignty needs an understanding of the physical layer. 

And then beyond that, like at least in the U.S., in the Bay Area, everything depends on digital infrastructure. So no matter what your company is like, if you work in healthcare, you're getting new digital infrastructure, and who is offering that to you? Where your data is stored? How does that comply with various privacy requirements? That is a physical problem. That's an infrastructure problem, and a lot of times, these companies don't have people who really understand what the cloud is. That can make you susceptible, that can mean that you don't make as much money. So I think it's a set of skills and knowledges that are both useful for, you know, just being a public citizen. But also, if everybody's using these things and relying on them but don't understand them, then, you know, I think that usually operations and you know, the economic model that companies have can benefit from really understanding what they're dependent on, understanding the options, understanding how to manipulate it. Like, do you want these things to be in-house or not? When you pick a cloud provider, an AI provider, what does it really mean? Like, where is that AI factory? You know, I think people are thinking about digital sovereignty because they're like, Well, I don't want data provided by a company that can be that, you know, can be leveraged by a government.

PAT: Yeah, like Saudi Arabia could buy up all the data centers or build them themselves, and then we're suddenly dependent on those. Not to pick on Saudi Arabia, but—

NICOLE: Location matters, right? Countries matter border. I mean, I think at least people are recognizing that now, like in our current geopolitical moment, borders are are incredibly important for so many different social formations. There they play real world and the digital has so many borders everywhere, and so many conduits and they’re physical. 

PAT: Yeah, we talked a little bit about the the sheer amount of financial leveraging that's going on to ramp up the data centers and the infrastructure, and the question soon arises, like, who's gonna Where's, how does this pay for itself? You know, first, how do we afford it, and who's, who's paying for it? And then, yeah, at the end of the day, where's the, where's the model, you know, to extract the money from us, presumably, to make that investment worthwhile?

NICOLE: Exactly. Our program, we have, like, full time researchers just dedicated to developing research that feeds into the curriculum. And our curriculum goes by experts in data centers, and they offer their input on this. Sometimes they argue about what should be included, about where the industry is going, and all of that students get exposure to.

PAT: How big is the enrollment? 

NICOLE: So right now, I think we have across like four or five classes, we have like 400 students in various classes. Last year we had about 30 students who completed the full certificate.

PAT: Is this one of a kind? Are there other programs like it? 

NICOLE: I'm proud to say it's one of a kind. It's one of a kind. I've done, I've done my research, and, yeah, there's nothing out there like this. I would say, for those of you that want to be data center engineers, there's some great programs out there that are just within engineering, that just focus on the technical dimension. I know that there are schools in policy where people are starting to think about the policy. You know, business schools are starting to think about like project finance and these other areas for digital infrastructure, but there is no program that integrates the various elements, and the industry has been like trying to figure this out. So I sort of stumbled into the industry, and I realized that my desire to develop education and research really actually matched their desire to generate sort of like more informed citizens that could better understand what was going on behind the scenes.

PAT: I mean, sounds like the one commonality amongst all those students, because they, I mean, you name so many motivations for why they might be interested in the course, but it's just the broad curiosity about the big picture.

NICOLE: A hundred percent, because at the end of the day, student is going to get hired because they're good at what they do, right? They're good, they're a good mechanical engineer, or, you know, they've done a degree and a language, and they really understand that language and that culture. So the students, but what they lack is an understanding of the phenomenon itself, right? So often companies will hire, they'll be like, I hired the best accounting person, and then I had to teach them everything that's unique about digital infrastructure, because it's global, right? It's at scales or costs that you just don't have elsewhere. So they have to do a lot of the training internally, and then they have to repeat it, yeah, and this is the idea is that if you could pair your existing major and your area of expertise, then those you know that people who are hiring are going to want to hire the best journalist, the best computer scientist, but they're going to want to hire somebody who understands the object of study. So then you would want to apply your disciplinary training and your domain expertise to this, like rapidly evolving, you know, infrastructure for the Internet.

PAT: So it’s a certificate program now, is there any chance, or can you imagine down the line it becoming a major? 

NICOLE: I mean, it's actually fingers crossed, will become a minor in the next few weeks. It's going through the minor submission proposal process right now, so just getting sign off from every school that they'll allow their students to take this minor, and I think it could easily become a major. But one of the powerful things about this program, and then about the about digital infrastructure itself is that it brings so many people together, and that really it's in that exposure to people in different parts of the world, and like, you know, in different disciplines, that like, you have to know how to collaborate to work in digital infrastructure. Like you have to, like an architect has to, and a civil engineer need to be able to talk to, like, you know, somebody who's doing, like, you know, electrical systems or developing generators. And then there's this human element to all of it, so you know, and how do you like connect to communities, right? Like, there hasn't been any work really on community engagement, and like, how to apply frameworks of community engagement. So, so it would be hard to have a major that was just digital infrastructure, but I think I can imagine, like, either a minor or a sort of major structure where you still get your disciplinary expertise that’s going to give you, you know, everything you need to know to be an electrical engineer, but is going to give you, like, the understanding of, like, what is going on right now, and give you the network of people to work with.

PAT: Yeah. I mean, from where I stand as a journalist nearing the end of my career, I think it would be a great, you know, add on to a journalism degree or even experience at the Daily Cal, you know, the student newspaper. But knowing, you know, journalists don't generally specialize, but this is a beat that is so large that it really does seem like it would, it would pay off. 

NICOLE: I mean, we would welcome journalists. I mean, there are so many different I mean, this story, the number of people around the world who call me all the time for news like and and they just need, sometimes, the basic facts, or they need to understand the geopolitical context for, like, a specific set of cables or data centers. And that, I think, if you look at, like, you know, other areas of tech journalism, you could say, Okay, well, people who are doing, who are reporting on Apple, you know the history of Apple, you know, like, you're like, Oh, I know who the leadership was. I know all the technologies. Is that true of people who are reporting on data centers today? And that's because there isn't a lot of information out there. So, you know, one of the really successful students from this year was somebody who wrote these really beautiful articles about that made, I mean, they were able to make digital infrastructure sound so interesting and really bring us into the world. And I was like, this skill, like, nobody's doing storytelling, yeah, in this domain

PAT: And that's what it takes, because most of us don't want to dive into a dry, analytical data-filled—

NICOLE: Ugh, terrible.

PAT: Yeah, it does sound like stories about people unspooling cable across the ocean—that alone would be a great story. 

It's Berkeley, and I'm guessing there are a lot of students who are very socially aware and activist-oriented, who worry about the haves and have nots, the people who are left out of this whole digital future that we're creating and whether they even want to be in that, you know, swept up into it. 

NICOLE: I mean, that's woven throughout the entire curriculum. We have examples from around the world, and the way that we think about it is that there are a lot of areas that are sort of historically marginalized in digital infrastructure development. So this is to say, these places and people don't have a digital infrastructure. They might have, like phones, but they don't necessarily have adequate infrastructure. And then beyond that, they're, you know, who owns that infrastructure? Who owns that data? Is that extractive model? Is it beneficial? And that is where I think some of the most interesting activity is happening today. So there's really great you know, nonprofit, the global digital inclusion partnership, who we've worked with, and they have, like, developed these frameworks of meaningful connectivity, right? So it's not just like you have a phone, it's what does it mean that you have the infrastructure to support things that you want to do? Do you want that infrastructure? And I think the thing that I hear, I can't say universally, because there are no universals, I think, when you look at a kind of like globally distributed system, but I hear people say we want the ability to make our own decisions about the infrastructure that we have. We want to be able to decide if we want to own it or not, because there are some cases where, if you own it, you have to repair it, right, and it can be real expensive. So like, you know, owning your own infrastructure is not always the choice that people would make. So I think that that question of, how do you get to an equitable infrastructural substrate for digital activity, that's something that needs decision-making around the world, on a local, a regional, national level. Lof people are involved. But if you don't understand the object, like, say, You don't understand like, well, what does it mean that I have an AI factory? Is that factory processing data for me? Is it benefiting me and muse of AI? Is it benefiting somebody else? Where is that stored? To whose laws is that subject? Right? Will it all disappear? Will it suddenly increase in price?

PAT: It's mind boggling, really, when you think about how fast this is moving, and then consider how many questions have gone unasked.

NICOLE: I think that when you dive into something like digital infrastructure, I have had so many questions that I didn't even know that I needed to ask because I was having conversations with people in so many different places and so many different disciplines with so many different backgrounds. 

PAT: Can you give an example of one?

NICOLE: Yeah, so I have a really like, Okay, so there's a cable landing station I went to visit, and I was talking to the cable landing station owner—the manager, and he was like, Look, I have been trying to do sustainability at this cable landing station, trying to make things greener. Worked with a you know, company to try to, like, bring in, you know, green technologies. So you could see from the outside, this is not supported by his government or anything. It was like somebody who was really trying to implement this, and he was like, Okay, well, I discovered that you could, like, set your equipment up in your cable landing station, so that way air would flow, like hot air would come out, and then cool air would go in and and that would save a lot of—would mitigate the energy. So he was like, he told his team, put the cabinets there, on the side walls, put them in this configuration. He told them for months, they didn't do it. They didn't want to do it. They're like gloves, like all their stuff was in the cabinet. They wanted it right near the door, and so there's like, kind of, like social inhabitation of the space. Finally, the manager went there. He moved all like the cabinet, like he and like that energy use dropped. 

PAT: Wow. 

NICOLE: Now this is not to say that one person was right or wrong, yeah, but it was to say that sometimes, when we sit in academia and we have opinions about things, the question is, why aren't, you know, X, Y and Z places, adopting sustainable technologies? And then from the other side, you might be like, why is this company, which is a historical legacy of, like a, you know, maybe something that we wouldn't agree with from a sort of humanities perspective, why is it? Why is it that they're telling the workers to disrupt their everyday routines? And then you get into it, and I'm just like, well, I like, and that was when I discovered, I was like, oh, there's this problem of, like, implementation in translation, like, literally, translation of language, translation of culture. 

PAT: Like, yeah, the the human friction.

NICOLE: The human friction. And I, like, I had been working for years to develop this cable landing station, like, set of best practices. And in that moment, the person I was talking to, I was like, Oh, my God, I can't just develop a set of best practices. I have been asking over and over, oh, what are your best practices? How do we do this? Like, but wait, this has to be translated into the vernacular world and languages and embodied routines, given a value, made understandable. Because you actually need people to kind of come together to achieve something, and they also have to want to achieve it. And if I come in and say, Okay, well, you all need to, like, use less energy, like it, I can't occupy that position. 

PAT: Right.

NICOLE: I think I realized it's all much more complicated. And so that's another reason I started this program, because I think I was just like, Okay, there's so much sort of like, like, if, if we want to be an activist, we have to listen, we have to understand the place that we're going to if you do this in a global context, like, can't just sit back in Berkeley and be like, Oh, I think that people should do this right? Like, I have to go talk to them. I have to understand, I have to, like, work to extend to a place that is sort of in between. And I think that's something that I got from the cable industry. 

They in the like, mid 20th century, they had something called a half circuit, which was like, okay, the US and Japan are building a cable together. And they're collaborating on every piece, yeah. And then, like, they developed this concept of, like, the U.S. owns to midpoint in Japan owns to midpoint. But, like, they built the whole thing together. But there was, like this kind of, like negotiation of like, of of shared interest and benefit, which was connection, and also sort of, like trying to figure out where that like middle point is, yeah, that made sense so everybody could sort of sustain the system.

PAT: I just keep thinking of two children having the same sort of discussion, like, we've put the Legos together. Here's my end of the Legos, and there's your end of the Legos. And do you think, well, it's all ours and we'll share it. We now we've built it together. Or, I think more likely, you say, No, I built this half. You built that half. That's yours. That's this is mine.

NICOLE: Yeah. There's no individual digital infrastructure, like, it is all part of a network. So while we might want to parse it up and say and draw these borders around it, like, fundamentally, you're talking about a system of interconnection. 

PAT: When you first started imagining this course that you built, or the certificate program that you built, were we in the era of AI already, or we were still in the big data era, which was, you know, not so long ago, or the web 2.0?

NICOLE: Yeah. I mean, I have always sort of been teaching this, but there hasn't been a widespread public interest, nor has it been as powerfully impactful as it is now. Now, I mean, yes, Credit Card Systems always depended on digital infrastructure. You know, banks depend on digital infrastructure. But now everyday users, like, if digital infrastructure goes down, like your phone isn't going to work as your credit card, like, if you didn't bring cash somewhere, like you're not going to be able to pay for things. So there's been this way that, like our ability to sustain our lives, at least here in the Bay Area, in different ways, this operates in different parts of the world, but like I can say, in the Bay Area, most of my students, their ability to sustain their everyday life depends on these infrastructures working in a very specific way. And if those are disrupted, the impact is huge. I mean, we're just talking about millions and billions of dollars of impact when these infrastructures don't work. So there's a significance now to how people eat and survive and live and pay their rent and have jobs that we see. You know that really came to the forefront during covid, and really now is accelerated with AI. So it's not that it wasn't there before. It's not that I wasn't thinking about it before. It's that it didn't feel like such a pressing need that I had to develop this. You know, I could go, like, write scholarly books and do, like, much more long form things, and now I, you know, I feel like this is knowledge and information that people can use to actually alter their conditions of existence, and that feels really meaningful to me. 

PAT: Excellent. That was fun. 

NICOLE: Yeah, it was great.

[MUSIC IN]

LEAH: This is The Edge, brought to you by California magazine and the Cal Alumni Association. I’m Leah Worthington. This episode was produced by Coby McDonald, with support from Pat Joseph and Nat Alcantara. Special thanks to Nicole Starosielski. Original music by Mogli Maureal.

[MUSIC OUT]