Somaini–Niemeyer interview

Lightly edited transcript. Obvious fillers, stutters, and one duplicated false start have been removed. Wording and meaning have otherwise been preserved.

Antonio Somaini

We're standing in front of the installation entitled Water Council, a model of seven water regimes. And the first question I'd like to ask you, Greg, is how this project came about, how would you position it in your artistic trajectory, and whether there is any other route of this project in your own personal experience.

Greg Niemeyer

Yes, thank you. As a young artist, I always had this fascination with walking up a stream and seeing where the water comes from. And you could do this in a hydrological sense, but you can also do it just in an artistic sense, as an exploration, as a dérive, if you want to know where the water comes from, and you just follow it until you can't go anymore. And along the way, I'd photograph. But while that was fun to do, later in life, recently I had a double lung transplant, and my body after the lung transplant was filled with water. And so for two weeks, I had to walk around with hoses coming out of my chest, one on the left and one on the right, and one on top and bottom, so a total of four hoses. And every day, they would just drip water out, and I had these tanks that I carried around with me. I felt like a total alien, and I had to wait for those tanks to fill up, and eventually the cavities in the body that were flooded were drained, and we could take the tubes out, and the surgeon stitched me up, and everything was fine again.

But in that time, I thought about how my own body was kind of a watershed, with water coming in and water going out. And of course, all our bodies are based on the exchanges of chemicals through water, and so there would be no life without water. But if I scale this up a little bit, a larger watershed also has things that flow in and things that flow out, so the idea of circulation. And ever since, I look at a stream or a creek, and I just think about how the water circulates through there,

And how it circulates through my own body, and I sort of just remove the boundary between my own body and the body of water that I stand before. And in a way, with my work, I hope that we can do that for everybody who looks at it and say, well, this is as alive as I am, and when we can get there, we can respect it. And then when we can respect it, we can really live with it in a more generative and sustainable way than if we are just extracting water. So really, moving water from the idea of an object to the idea of a living being.

Antonio Somaini

Right. Thank you. What you described is two different water regimes, we could say. One is more personal, bodily, subjective. The other one is more environmental, embedded in the landscape. And the project is also based on a series of seven water regimes that you have identified, taking more a kind of diachronical look, looking back at history and seeing how water regimes have developed, how they have sometimes coexisted, how they coexist in the present, and also what kinds of water regimes might lie ahead of us. You want to say something about this distinction? We can look at the diagram here in which you describe the seven water regimes.

There's a kinship regime that started back in time immemorial and continues to last in the present. There's a gold rush regime which dates back to 1848 and which continues. There's an agriculture regime standing from the 1870s to the present. A hydropower regime, 1917 to the present. A water projects regime, 1937 to the present. A tech regime, 1996 to the present. It would be interesting to know why you chose 1996 as the beginning date for this. And finally, there's something you called watershed regime, which is more speculative. And you say it's a regime in which water is in charge. How did you come up with this distinction?

Greg Niemeyer

It really is based on my learnings from members of the Yurok tribe who were talking to me about the ancient waters of the Klamath River. And a huge recent success for the tribe was to undam the river entirely. And so they removed several dams that allowed the river to flow freely again. And they were able to reestablish their kinship relationship with the river. As in, the river is a member of our family. We honor them and treasure them like any member of our family. They are life-giving. And we have rituals and annual events that mark that relationship and manifest that relationship.

Unfortunately, that regime was disrupted by colonial immigration and Manifest Destiny in the United States pretty rapidly, especially in the period of 1848 and onwards, when early settlers here discovered gold. There really was a previous phase, which was the Mexican rancho phase. But for historical brevity, I'm just focusing on the gold and impact.

I'm focusing on the gold rush regime, where suddenly water was not an ancestor anymore. It was a resource, a tool to extract gold, as much gold as possible, as quickly as possible from the land. And we still experience the after effects of that. The gold has long been sold and disappeared. And the profits of that have dissipated across society. But what remains is the mercury death, the mercury impact. So there's a lot of mercury that was used in the process of amalgamation, of getting gold to clump to itself. And so the Bay Area still has a lot of pollution from that, mercury pollution from that, which is very hard to get rid of. But there the water is not akin at all. It's a tool. And it's a force that we can use to accelerate extraction.

And at a similar time, shortly after the gold rush, there were a lot of people here. And if the gold was gone, they were still hungry. And they needed food. They couldn't just import it anymore. So there was a much greater appetite for food than land naturally can produce. Again, very different from the kinship regime, where we started irrigating fields. And we would extract groundwater to take desert parts, the parts that are pretty dry naturally and that are full of nutritious soil, however, and add water. And thereby have a sort of agricultural wonder takeover in California. And so California, from 1870 onwards, becomes kind of a breadbasket of the nation. And I think 50% of the fruits and vegetables consumed in the United States are actually grown here. So it's a huge amount of food that's produced here. And much more food than you think is possible. And if you think about maybe it's not entirely possible, the fact is that we're drawing a lot more water out of the ground than the ground can replenish. And so we're actually operating on a water deficit, which will eventually lead to the cessation of the whole agricultural project. But until then, we just go fast forward, as we often do.

And the hydropower regime sees water not as a force, but as an energy source. Because we can now look at water falling off a mountain and say, that is a lot of energy because water is very heavy. And we can use that force and that energy to produce electricity. And so Los Angeles, interestingly, has an administration that runs both the water and power at the same time. And as people may know from the film Chinatown, there's been a lot of questionable manipulations and very extractive colonial regimes that led to entire valleys and entire lakes being drained for the benefit of providing Los Angeles with water. One valley that comes to mind in particular is Owens Valley, also known as Paya Hunadu, which is the native territory of the Paiute Indians. And they have to this day lost their lake. And they're now trying to restore it by buying land that used to be their own, buying it back and gaining water rights that way as well. So the water projects regime is related to the hydropower regime.

It starts in 1937 with California as a state. And the United States, it has a federal administration. The federal administration at the time decided that there was need for more people to settle in California. And for that, there was need for more water to make that possible. And so many rivers, including the Sacramento, but also the Colorado River, were diverted to send water into, again, desert areas, this time not to produce agriculture or electricity, but to produce pools and water supply for homes. And so this was an amazing deal because the government would pay for the diversion of water to these areas, and then they would give the land away to developers at very low prices. All they needed to do was put a house on it, and they increased the value of the land tenfold. So that was an amazing process. And the regime that is the most recent one, maybe it's the tech regime, where it's not really food or energy or just drinking water,

That is our population, if you will, population growth, that is the resource that we enable with water, but rather information. And by information, here we mean data. And we can look at the history of California and its many contributions to information technology. And one project that stands out is the 1973 CVDP-TEAL Consolidated Data Center, which was the first data center that used that actual name. And it was, interestingly, a public open source data center, a data commons where all the public data could be held. And, of course, since then, data has been increasingly privatized. And so a lot of the massive data collections that we now are working with in information technology are not public at all. They're private. And so 1996 is a time when the privatization of this kind of resource kicks in in the form of massive data centers that are built in Silicon Valley, but also the industries that produce chips and so forth that are very water intensive. And so what I find impressive about that is that California produced technologies or initiated technologies that have much larger water footprints than California itself. And so if we think about, say, where chips and data centers all come from, it is a huge global industry with a truly global water footprint far beyond the state's borders. And this leads to the current topic of how many data centers we really need and who's going to be able to host those data centers. Of course, far from being a public good, artificial intelligence is oftentimes a private interest, a corporate interest.

And so people don't necessarily see that the artificial intelligence that comes from those data centers flows as freely as water might flow. So we can ask ourselves, what might AI learn from water? And one of the things that comes up, again, is this topic of circulation. How can we make intelligence circulate freely and openly and beneficially? And how can we make water circulate freely, openly and beneficially? There are several types of circulation. One is the steady one that is beneficial with clean water. The other one is the steady one but polluted. The other one is the drought where suddenly circulation ceases. And the other one is the flood where the circulation suddenly increases. And so we have sort of a qualitative and a quantitative aspect of circulation. Both of those need to be maintained for us to live on happily, both in the area of water and also in the area of artificial intelligence.

So the question then becomes, can AI produce water management systems that actually mitigate its impact? Can we ask AI to produce water regimes that are more thoughtful, effective, balanced, and more precise than the water regimes we have now? Because frankly speaking, the water regimes we have now are very wasteful and very imbalanced. And they sort of compete with each other in strange and uneven ways. And so maybe in a speculative future we could think of the watershed as something that works in dialogue with AI where the watershed really runs the show. Because at the end of the day, water has the last word.

If there is no water, there's no people. If there's too much water, there's no people. And if the water is polluted or salty, there's no people. So water really runs us and we have to start realizing that. We have to live by that. And so maybe AI can help us overcome our own sort of egotistical needs and say there's a collective optimization that is possible if we just learn to listen to the watershed. And there's many technical tools that can help us listen because apparently we with our senses are not really listening that closely. We have impulses that override what we hear from the water.

And in a way that cycles back to the kinship regime where we see water as the not only relative, the kin, but also the sort of the person in charge, the entity in charge of what's happening to our communities. And maybe at the end of the day, we come full cycle and we say the kinship regime combined with some technological adjustments and improvements in the sense of paying attention to water resources and balancing out human needs would be really effective.

Antonio Somaini

What I find really interesting in this last regime you mentioned, which as you say is speculative, is the fact that you talk about a post-capitalist artificial intelligence. So it's a response in a way to the dominant discourse about AI and water right now, which is mostly focused about water consumption, the cooling of data centers and so on. You basically say there could be another kind of artificial intelligence that could participate in these water regimes in a different way. Even though it's not an easy question to answer, but how would you describe this post-capitalist artificial intelligence? Would it be artificial intelligence based on open source models? How do you see it? And do you feel that your practice is somehow participating in pointing in this direction?

Greg Niemeyer

In short, Antonio, it's about the gap. There are a lot of problems we know and there's a lot of solutions we know. But there's a gap between the solutions we apply to the problems because most of the solutions are more expensive than the current state status quo. And so from a capitalist point of view, investing capital in risky proposals, even if they are in the long term going to be beneficial, does not make sense.

So a post-capitalist artificial intelligence would think without financial constraints. It would think, what is the most sustainable, equitable thing? And so it would be the same system, artificial intelligence, but with a different inflection, with a different motivation. The motivation is not to make more money. The motivation is, in fact, to balance things out. And so that, again, takes us back to the kinship regime. And so that's why it's post-capitalist, because you can be smart and completely blinded by the need to make money.

And then all the smarts you have are taking you in a direction that only leads you further down the capitalist road. But I think a lot of indications suggest that we have to look elsewhere for stable futures.

Antonio Somaini

Great. Thank you. So let's move now to the installation.

Greg Niemeyer

So these are all my opinions. And I don't know if that's the most useful thing. But what would be useful, I think, is if we all would talk about these things more. And so that's why this whole project is called the Water Council. And so I'm imagining that we have seven chairs here, which we actually do have the seven chairs. And the imaginative part is that we sit on these chairs. And as we sit on these chairs, we enact these regimes in a digital model.

So before you, you see seven screens that very informally present a model of a simplified Sacramento River. I'm not a hydrologist, so a lot of these things are surely not correct. But they're also very innovative and playful. So maybe there's interesting things that come out that we didn't think about before. But each person visiting this can take a seat on any one of these chairs. And by sitting on them, they're labeled with the various regimes. And by sitting on them, we enact those regimes.

So if you sit on the AI watershed regime, which you're sitting on right now, Antonio, then you can see that the big screen here indicates AI watershed. And so what we have here is some elements of the kinship system. So we have the floodplain of the Sacramento River Delta Basin. And then we also have innovative things. Like, for example, AI thinks that it's a great idea to put a city next to a reservoir because the reservoir has water that cools the city. Then we have data centers that heat that produce thermal pollution. Those little red dots are thermal pollution. And so the AI system tries to limit the extraction to a minimum.

So the input here is 4.3% of the water right now that's available on a seasonal maximum basis. And the total extraction is 3.1%. So that's per stream. So it's trying to tally up all the various water resources and all the various water claims. And it's trying to come up with an equitable balance system for that. If I now move to the tech regime.

Then you see immediately the system switches to tech. And though AI watershed is gone, the extraction is much higher. It's now 25%. And this is even during a low water season in August here.

We don't have a lot of water in California. Because the water year, the water cycle, the annual water cycle is very specific to how much snow is melting and how much rainfall is falling. And so even during a time of low water, the tech regime will try and extract a lot of water.

And so now I switch to the kinship. Oh, you stopped moving. So that means that the tech regime has ended and the kinship kicks in again. So our default is the kinship setting, which of course has zero extraction because kinship takes no water. It just uses what's there.

Antonio Somaini

Great. I see it as an installation that really favors kind of collective discussion among the public. And I think the fact that it's presented in the university gallery also means that it would be very interesting to sit here with students and discuss this project with them, which you probably have already done maybe or not yet.

Greg Niemeyer

Yes, I did do that. And the interesting thing that comes out is that we realize our collective uncertainty. And by realize, I mean we manifest our collective uncertainty. Because we sit here and we're like, well, maybe we shouldn't do all this. Maybe we should do something different. And what about the future and so forth. But if you're alone in our uncertainty, we're very weak. If we're finding community in uncertainty, then we have a very interesting situation from which we can maybe more meaningfully speculate on possible futures.

Antonio Somaini

And especially if we give also visual forms to some of these unsettling present conditions such as the tech regime you described.

Greg Niemeyer

These are all relentless regimes that just go to the maximum at all times.

Antonio Somaini

And what's interesting is that you consider them, all of them somehow coexisting.

Greg Niemeyer

Yes. They all are competing. No regime replaced the previous ones. No. They still coexist. They're all competing. And they're all squeezing the water.

And eventually we get to Deadpool. We get to a point where, or Day Zero, where there's no water left. And unfortunately there's been many places in the world where this Day Zero or Deadpool moment came about. Deadpool means that the water doesn't flow anymore. It's stuck. And it's stuck in reservoirs. It's stuck behind waterfalls and so forth. And Day Zero means that there's no water to be extracted anymore. And this happened famously in Tehran on November 2025. And the Minister of the Environment of the city of Tehran asked people to go out and pray to the desert and pray for water. And luckily two weeks later it started raining and there was some water again. But we all are subject to these extreme outcomes.

There's also another concept of the water bankruptcy where there's places where water never will come back again because we've simply extracted too much water from the ground and manipulated the natural systems too far so they cannot recover. Normally nature always seeks to recover. And the first time, the second time, the third time, it always seeks. But the hundredth time or the end time, eventually it collapses. And we're not going to stop until we reach that point. Unless maybe we listen to AI and say, well, are we really on the right track here? Or is there any other options?

Antonio Somaini

Let's stop it for a second.

The last question I would ask, but tell me if it's important. So when we look at the AI watershed, the most speculative one, and you can sit on the chair and see it happen, what we see right away is that, say, compared to the tech regime, the data centers are much smaller and the reservoirs are much smaller and the cities are smaller.

Greg Niemeyer

That's because it's trying to balance the needs. So what it envisions is cities that change in size depending on how much water there is. It also tends to locate cities in reservoirs and do funny things to optimize the outcome of its water extraction. And in a way, it's modest. It's full of restraint. And the way I came about this is I just gave an actual AI the parameters that I'm working with and the goals.

I want there to be less extraction, not more extraction than we can get given the current moment we're in in the climate scenario in the year. So every day in the year has a different sort of water availability. And then the AI responds to that and generates a model on the fly. So the moment you sit down, we're making a call to a simple AI model to say, well, what are we going to do now? How is this going to work?

And so we're actively using a simple AI model to manage the water. And the interesting thing about it to me is, apart from the strange things it comes up with every now and then, is that it's exclusive. This will only work if we really all just follow AI. So it actually, you can't have a whole bunch of regimes happening at the same time plus AI. That's not going to work. You have to say, okay, we're now really going to follow the recommendations of this speculative AI model. And therefore, all other models are overwritten. It's a total override. And that points to a sort of a regime of optimization that is definitely not in the spirit of democracy.

It's definitely not in the spirit of indigenous rights. It's a total takeover. And it brings up the very pressing question if democracy is indeed the right political regime to help us overcome the challenges of climate change. It may not be.