Not Now, But Soon: A Hurricane of Data

By Steven Gonzalez, Malka Older

Our new miniseries, Not Now, But Soon, challenges the stories we often tell about disasters and explores how we can use speculative fiction to create better futures and policies. On our first episode, host Malka Older talks to Steven Gonzalez, an anthropologist of technology who researches the human labor behind data centers. 

Gonzalez is also a speculative fiction writer under the byline E. G. Condé. He is one of the creators of Taínofuturism, which incorporates Indigenous Caribbean traditions to imagine futures from a radically different perspective. His novella, Sordidez, explores how survivors’ efforts to rebuild after a hurricane intersect with oppression and conflict in a future Caribbean. 

In this episode, Gonzalez compares these different types of disasters: the dramatic and immediate impacts of the hurricane, and the slower, steadier, and often overlooked disasters of environmental destruction, resource depletion, and exploitation of human labor associated with our internet infrastructure.

Resources

• Visit Steven Gonzalez’s website to learn more about the impacts of cloud computing. 
• Find Gonzalez’s fiction under the byline E. G. Condé. Check out his novella, Sordidez, winner of the Indie Ink Award for “Writing the Future We Need: Latinx/Latine Representation by a Latinx/Latine author.”
• This podcast is part of the Future Tense Fiction project. Read all Future Tense Fiction stories at issues.org/futuretensefiction.

Transcript

Malka Older: What do you think of when you hear the word “disaster”? Disasters are often seen as periodic eruptions of chaos, random strikes of misfortune that are corrected with the return to normality. If you look at media coverage, the disaster story goes like this. The disaster happens, the community is devastated. Aid is sent, things get fixed, we move on. Even if the affected community can’t.

But disasters are deeply rooted in the way our societies are constructed, and studying them can tell us a lot about our assumptions and blind spots. Shifting the stories we tell about disasters can help us understand them—not as random events, but as the consequences of a series of decisions, decisions that could be made differently in the future.

I’m Malka Older. In this podcast miniseries Not Now, But Soon, I’ll be talking with friends who, like me, think about and work on disasters. Together, we’ll expand the idea of what a disaster is and what to do about it, and imagine how we can build a more robust, resilient future for all of us.

In the first episode, I’m joined by Steven Gonzalez. Like me, Steven wears multiple hats. He is one of the creators of Taínofuturismo, which builds on often ignored Indigenous Caribbean traditions and belief systems to imagine futures from a radically different perspective.

Hey, Steven. It’s great to see you. How you doing?

Steven Gonzalez: Pretty good, pretty good. I’m excited to be here with you.

Older: Can you tell us about what you do? I know this is more than one thing.

Gonzalez: I’m an anthropologist of technology. I’m really interested in how people interact with and making meaning out of technology. In particular, my research is about data centers and computing infrastructures. But this very high-tech world also really syncs up with my other life, which is as a speculative fiction writer under the byline E.G. Condé. I draw a lot from my scientific experiences to create science fiction that I think explores a lot of the themes of my research, but also connects a lot with my heritage and background and roots in Puerto Rico.

Older: When you tell people that you’re a science fiction writer, or if you talk about, I know you’ve referred to your work as Taínofuturismo or Taínofuturism. How do people see that? What are the stories they tell about science fiction writers?

Gonzalez: I think a lot of people struggle with what I am, what I’m interested in. There’s an artifact of that in my research I think. When I’m interviewing data center technicians or people, they’re telling me about the technology and they’re resisting my attempts to find out about who they are and what they think about their world, and so on.

There is a story of survival, this story of refusing extinction that I wanted to tell.

Taínofuturism, it started really as a dream. I had been reading a lot about Taíno history and archeology. I do have some connection to this part of the story of Antillean heritage in my own family, as well as African elements. I had this really bizarre dream that was almost like an animated painting. It was a combination of all these elements of indigeneity, of animals, and birds, and plants, and cybernetic machines that were all intertwined in really strange ways. I wanted to really lean into it. This journey of Taínofuturism for me began with a story called Somnambulist.

From there, I started to think of it as a way to recuperate indigeneity in the Caribbean—in the Antilles in particular—and to resist a narrative that I learned when I was growing up in American schools that the Taíno went extinct, that they were exterminated in their entirety. When in fact, many Puerto Ricans—but also Dominicans, and Jamaicans, and Haitians, and others in the Caribbean—have maintained connections and ties, and even bits of language. There is a story of survival, this story of refusing extinction that I wanted to tell.

Older: I was excited to read Steven’s work because we need more stories about the future written from Indigenous perspective. His stories have a particular resonance for me because my mother’s from Cuba, where Taíno people lived before the Spanish Conquest. Indigenous cultures and peoples are often devalued and even erased in popular narratives. Steven’s novella Sordidez brings reimagined Taíno governance structures and values to a future Puerto Rico in crisis.

As you can probably guess from my fiction, I always want new stories about governance and policy. We need to experiment more with our democracies and science fiction can be an effective thought experiment to kick that off.

Gonzalez: I certainly think that the science fiction life and the academic life, the feedback goes both ways. Specifically, with the discipline and the practice of anthropology, it is a practice of I would say maybe scientific documentation of observation. But unlike other sciences, this is a science that is about building relationships with people, about cultivating a sense of empathy to understand communities in a more intimate way, and to help tell a part of their story.

I see it as a very collaborative process, but this notion of storytelling is incredibly important for the discipline of anthropology. I do see that these experiences of writing science fiction, of writing stories, of writing novellas and exploring all these genres has really helped me also in my scientific work or in my research life, and also has helped me to write and publish about my research in an accessible way that reaches an audience outside of academia that’s not so laden with jargon and that speaks to the more human dimensions of something like running a data center. People nowadays maybe know about data centers, these warehouses filled with computer servers that are powering AI and everything. I don’t think as many people really wonder about the story of the people who take care of those facilities, who work in them, what their lives are like. That has been my obsession for almost 10 years now.

Older: When I think about it, I imagine something Mission Impossible-y, but from the first one, where there’s a building with all these machines that are faceless and just humming, but there’s no people in this image. Yet clearly, what you’re doing in your work is all about the people who are there. The conceptions that people have about this, how does that affect the way they interact with you or the way they understand your work?

There are technicians working on either end of the internet to make sure that this is a relatively smooth and seamless experience. I think we take that for granted.

Gonzalez: The reason I’m doing this research is to tell a story of these technicians who are mostly men, actually. It’s a very masculine-ized profession and there’s a lot of issues with that. There is rampant misogyny and homophobia, and other dimensions that I have observed in some cases. But it’s also a story of men who are mostly very overworked and very stressed out, and working in these conditions of precarity and constant vigilance takes a toll.

If you can imagine, this very Zoom call is possible because there are technicians working on either end of the internet to make sure that this is a relatively smooth and seamless experience. I think we take that for granted because we have all these digital applications and they’re always… okay, occasionally they fail, you’ll get the “server unavailable” message, but it’s actually quite rare, especially now. But there is a human cost to that. I wanted to tell that story by spending time in these data centers. And also, to dispel this notion that all the data centers look like this aesthetic of the spy film, the espionage thriller with the dark cave filled with computers, and it’s kind of beautiful in a way.

I’ve seen a much different version of that story and it’s been a privilege to get to know this world in a more intimate way. And to also participate, doing some of the less glorious tasks like untangling cables and cleaning filter media that are used to cool servers that get all this sediments and gunk on them from water supplies that have high mineral content, for example. But this is the story that I thought it was important to tell.

Older: Okay. Now that you said they don’t look like that, can you give us an idea? I’m sure there’s a wide variation.

Gonzalez: Yeah, absolutely. When I first started this research back in 2015, I think a lot of people still didn’t really know about data centers that much. There was some media coverage that was starting, people were starting to think that Netflix has an environmental toll. But before that, there wasn’t a lot of knowledge. Then I think even today, the hyperscalers like Google and Amazon, and the big tech players, their images of data centers is like the most technologically sophisticated, the most aesthetically interesting, capture the imagination of I think this espionage aesthetic that a lot of people know.

The reality is there are a lot of data centers out there that are not so pretty. These places, there are no windows for security reasons. It’s almost like a casino, there’s a sense of timelessness, of eternity. These rows upon rows of servers, but also there’s a lot of mess, too. Of floor tiles, there’s these underfloor plenums. If you can imagine, the cold air is pumped into this pressurized plenum, and then these little floor tiles have these little perforations. Then that pressure enables the cooling to come up. You can see that there’s this hidden underworld with all of these cables. Sometimes there’s mice in there and mouse traps.

Older: You mentioned the environmental cost. Is this a disaster? Is it an environmental disaster? Is it a human disaster? Is it both?

What is the environmental cost of computing at the scale that we now have, especially with artificial intelligence applications and large language models?

Gonzalez: Yes, I would say at this scale of just environmental impacts, it is a kind of disaster. This is the part that really concerns me and has, I think, motivated a lot of my research as well, is to think about what is the environmental cost of computing at the scale that we now have, especially with artificial intelligence applications and large language models. The answer is very complicated. I would say that there is a lot of electricity consumption associated with running these data centers at 24/7, 365. They don’t sleep, they’re not factories that shut down. There’s no holiday, it’s just a constant operation. One data center can consume as much electricity as a small city.

The water footprint of data centers for me is even more concerning, especially because water is what is used to cool data centers more efficiently with less electricity and to lower the carbon emissions. But that’s water that people drink and need. You have a situation where farmers are now competing directly for water with data centers, with server farms.

Older: How did you start writing about disasters as a topic for fiction?

Gonzalez: This is a great question. Like many Puerto Ricans living in the diaspora and also in the archipelago, I would say 2017 was a really difficult year. When Hurricane Maria struck, it was really apocalyptic in terms of the scale, and also I would say that Puerto Ricans in general are used to hurricanes. Hurricanes are not exceptional. They are actually a condition of the Caribbean and always have been to some degree. We know this because the Taíno themselves depicted hurricanes, they knew of their circular movement. Even the name, hurricane, comes from this ancestral language.

But Hurricane Maria, and what has been really interesting to see is just the ways that this field of climate attribution science has been able to really conclusively show that Hurricane Maria was a deviation precipitated by climate change. It was a really traumatic experience for me. There were days when I didn’t know how my family was doing. I was seeing apocalyptic imagery. There was reports of aid not coming to the island in an efficient way. There were communities that were entirely isolated. The fatality. It was just a really trying scenario on so many levels.

The aftermath can be more deadly than the actual event itself. This raises questions of resiliency, and infrastructure design, and policy, and so on.

But I also would say beyond the disaster itself was interesting and also instructive is how the aftermath can be more deadly than the actual event itself. This raises questions of resiliency, and infrastructure design, and policy, and so on. There was also something that, as a writer, I found really fascinating and inspiring about how people rose up to the challenge of this aftermath of this cataclysm. How people in my family’s neighborhood would pool their resources together in the absence of a government, or this suspension of society’s rules, this erosion of the social order, people rose up and did what they needed to to survive. They looked out for each other, and they relied on existing networks through churches and other community organizations who share electricity to cook meals that they could all benefit from, to clear trees from roads with downed power cables, and all of this kind of stuff so that aid could be delivered.

That to me was the story that I found so profoundly moving. How disaster can also provide an opportunity, and that opportunity of course can be both ways because we also saw the way that venture capitalists and others preyed on the disaster and bought up a lot of land, and did a lot of things that did not help Puerto Ricans, and continued to cause issues on the island.

Older: This is a dynamic that I’ve seen personally in disasters where I was working. Research shows that people tend to come together after a disaster. People usually want to help if they can. In many of my deployments as a disaster responder, we saw people from neighboring towns and villages driving through the night to offer supplies to the survivors, or evacuated communities going above and beyond to make things easier for the most vulnerable among them. But there are also cases where communities fall apart and become antagonistic and toxic instead. There are specific theories about why and when this happens. But in general, when people are dealing with a lot of uncertainty, and blame, and long-term, drawn-out effects, that can lead to divisiveness.

On the other hand, what are the conditions that make a community more likely to successfully come together? And what does coming together mean when the disaster isn’t a sudden hurricane, but rather a slow-moving erosion of our natural resources?

Data centers are designed to be really hyper resilient, disaster-proof in a sense. While most of the island did not have power for six to nine months, these data centers did not experience a single minute of interruption.

Gonzalez: I wrote a novella in 2023 called Sordidez. It was really inspired by this experience of observing and trying to understand Hurricane Maria as a way to think about the opportunity that disaster provides for reimagining and remaking the world, and achieving a kind of just society. Will this cycle continue, and at what point will the cycle break?

I was actually doing research in data centers in Puerto Rico during the pandemic. One of the things that I found really fascinating and also a bit unsettling was the ways that data centers are designed to be really hyper resilient, disaster-proof in a sense. While most of the island did not have power for six to nine months, these data centers did not experience a single minute of interruption. In a way, data centers are like fortresses. The power goes out, they have diesel generators on standby. If the air conditioner fails, there’s another one already on and in an idling state, ready to rev up. These data centers, because they’re so crucial for so many different kinds of economic activities, but also military and governance activities, they have been created as these disaster-proof nodes. By and large, they’re stunningly successful. The data center story in Puerto Rico really ruptures and I think completely demolishes the idea that we couldn’t have been prepared for Hurricane Maria.

Older: There’s this statement by Amartya Sen, “There are no famines in democracy.” Famines are a typically slow moving, slow onset disaster. If shortage of food is enough of a threat that democracies will do the work to make sure that they don’t trigger that and get kicked out of power, why isn’t climate change? What is it about this other kind of ecological disaster? Is it too conceptually new, is it too complex? Is it all the disinformation? What are the elements that are keeping us from really saying, “We need to adjust how we’re doing this, we need to rise up, we need to think about this differently?”

Gonzalez: Yeah. This is the question that I think is haunting a lot of scientists today and science communicators, and people like myself who are trying to do research in science and technology. How are we not communicating sufficiently the urgency to act? What is in the way of achieving that and how can we reach people? How do you tap into their emotional center to communicate the severity of what’s happening? For me, the answer has been fiction and I’ve been really interested in climate fiction as a tool, as something that has a life outside of itself potentially to make more immediate these threats that we have so successfully distanced ourselves from as this is a future problem for future generations. But for me and for many others in the world who experience climate change and are already experiencing it to a much more severe degree, that time horizon of the future is not really possible because the present of climate change is already so disruptive.

Hurricane Maria was a warning and it’s only a matter of time before another one comes.

For me, it’s just this question of Hurricane Maria was a warning and it’s only a matter of time before another one comes. I think this is where the Taínofuturism comes in because it’s not to romanticize Indigenous life ways as necessarily sustainable and perfect and in this kind of harmony of nature, but to emphasize that actually, many Indigenous peoples had and continued to have an immense impact on the environment and they are stewards of the environment. We have to realize that there are different ways of relating to environment, different structures of relation to environment. Maybe that’s the way out.

Older: One of the stories that we tell about climate disasters is that we can create disaster-proof infrastructure. The argument is that by converting data centers into fortresses, we can protect them against unexpected calamity. When I hear that something is disaster-proof, I’m reminded of the book Five Days at Memorial by journalist Sheri Fink. It’s about a big private hospital in New Orleans during Katrina. When the power goes out, the hospital erupts into a chaos with tragic results. Close by, there’s another public hospital where the electricity goes out regularly even without extreme weather events. When the hurricane knocks out the power there, it’s something they’re more used to. It’s not an unsalvageable catastrophe, but something to be managed.

Sometimes resilience doesn’t come down to fancy technologies or massive, expensive infrastructure. It comes down to whether people are used to responding creatively and collaboratively when something goes wrong.

Gonzalez: There is an opportunity when the world as you know it falls apart to create something else that is semi-anarchic because the structures that you’re familiar with have collapsed.

Older: Yeah, that is absolutely something that I find really compelling about studying disasters as well, is that you see this period that’s kind of outside of normalcy that we can look at and get a really different picture of what’s going on in our societies.

Gonzalez: Yeah. That part has really been fascinating to me. When you create this vacuum, it lays bare the differences that were always there, and then maybe the sense of urgency can temporarily shift the priorities that formerly divided people might now unite them. I think that’s interesting to see how that can happen and that it’s possible at all is fascinating. You also have this mix of personalities and there’s this unreliable alchemy of personality chemistry that can also implode a temporary coalition that forms.

Older: Yeah. I actually think that these questions of personality and how we deal with that uncertainty and unreliability is maybe part of why we have gone so far in our technology, both the digital and engineering technology, and also our technology of governance and our technology of business to try to get to an impersonal way of doing things. Organizations or systems where the personalities won’t sink something, and yet the prejudices people have and the superficial things that make a difference, and all of that continues to have a huge effect on the way that our technology, including digital technology, is developed, the kinds of things we choose to do. The aesthetic, as you mentioned before, and as we talked about the data centers, of having this image of them being without people. And yet, the people are always there, the personalities are always there. The human needs and the human ideas are always in there.

Gonzalez: Yeah. I think this tendency for the human to always be there, especially in the data center, is something that I keep going back to. The way that these technicians ascribe certain personalities to computers, like a server has a personality. They give them nicknames. They have this while social order imposed on inert machines, and air conditioners, and even sections of the data center that then have a certain personality type. I find that to be just incredibly fascinating, but also human. It’s, I think, what literature is trying to convey, is who we are, and what we share and what we don’t share.

Older: Absolutely. This is also something I’ve seen doing research on the Fukushima Daiichi disaster. Where you had operators who had been placed with a different reactor, but when the disaster happened they’re like, “Oh, no, no, I’ll join the team on Reactor One because I worked with her for a long time and I know her really well.” That seems to be a really strong tendency.

I wanted to ask, coming back to that connection of the human and the technological, when you tell people you’re an ethnographer of technology, because we think of ethnography as something very to do with humans, how do people interpret that?

Gonzalez: The part that I find so interesting about the subfield is to say that what counts as technology is inherently political, first of all. And also, it’s historical and there’s a sense of so many things could be a technology. In my book, language is a technology. The resurrected, refashioned Taíno language, for example, is used as a technology to evade surveillance of futuristic threats that are powered by AI. Since AI has not assimilated this language yet, it can evade detection when it’s used. But there are so many examples of this.

What is technology? That is a fascinating question. What is computing? We can even think about the abacus as a kind of computer, or the quipu as a fabric-based computer system.

Older: What are the sort of stories that you would like to see us telling about our computing, these physical facilities that enable our computing, our value choices around what we protect and give energy to and water to and what we don’t? What kinds of stories could help us figure that out better or have a different perspective on it that would be more constructive?

It’s long overdue for people with imaginative skills of various kinds to get into this conversation and to start to model what the future will look like.

Gonzalez: I think it would be really valuable to have people who write fiction or who do scenario modeling to come together with people from the tech industry, and also policymakers, to imagine what sustainable computing would be outside of this really misleading idea of sustainability as being “carbon-neutral through offsets,” or sustainability as being something like minimizing a water footprint because thinking about it holistically seems to be what’s most crucial. How does a data center impact a community? How can it not only adversely impact a community, but positively do so? Is there an opportunity, for example, to build power infrastructure and water infrastructure for data centers that people can also benefit from in places where they don’t have such things?

I think there’s an opportunity, but there’s a deficit of imagination on how to achieve that. And also, I think in the absence of regulation, the data center industry and really the tech industry have been able to regulate itself. It’s set the terms of its own conversation around sustainability. I think it’s long overdue for people with imaginative skills of various kinds to get into this conversation and to start to model what the future will look like, especially if we’re going to continue to build these AI super data centers.

Older: One of the things that the tech industry has done unfortunately extremely well has been to tell us what the future looks like. We hear it especially in this large language models and generative programs, the way they talk about it. Both from the perspective of saying, “Oh, you’ll fall behind if you don’t have this, everyone is going to have it,” and to say things like, “Oh, yes, we will have the energy to manage it. We can handle all of these externalities at this point in the very near future that keeps receding.” Having people who can present counter-futures to say, “Maybe this could look different, maybe this is something we don’t need. Maybe we could have the entire backend that you’re trying to hide from us of people working in environmental damage and all of this stuff, maybe it could be organized differently.” Having that ability to think about different futures is just so critical.

Gonzalez: Yeah, there are certainly opportunities to reimagine this. What better tool than science fiction perhaps to do so?

Older: So, what are you reading lately?

Gonzalez: I’m absolutely awestruck by The Spear Cuts Through Water by Simon Jimenez, which to me was an absolutely breathtaking experiment in storytelling with a lot of fantastical elements but also interesting refractions of a near present or a distant present that I would call science fictional. And I really was so moved by that book.

Older: I love that, Steven. I haven’t read that book yet but I’ve heard such wonderful things.

Gonzalez: Thank you so much for this opportunity. It’s really a pleasure and an honor.

Older: Visit Steven Gonzalez’s website at stevengonzalezm.com to learn more about the impacts of data centers. And definitely read his novella Sordidez, published under the byline E.G. Condé.

Thanks to our audio engineer Angelina Mazza, and our podcast producers Kimberly Quach and Mia Armstrong-López. Music for this series was created by Stuart Leach. Not Now, But Soon is part of the Future Tense Fiction project, a collaboration between Issues and Arizona State University’s Center for Science and the Imagination. Additional editors on the Future Tense Fiction project include Joey Eschrich, Andrés Martinez, and Ed Finn.

On our next episode, food systems journalist Thin Lei Win shares her experience of growing up in the authoritarian country of Myanmar and how that disaster shapes how she sees the intersection between climate and food. Subscribe to our main podcast feed by searching for The Ongoing Transformation wherever you get your podcasts.