00:02:48.740And the Athena Coalition is also part of coordinating this data center working group.
00:02:52.980So supporting organizing and base building with these side fights against these data centers all across the country.
00:02:59.440And then I'm also part of an organization called the Manaset People's Project in El Paso, Texas, that's fighting a $14 billion metadata center.
00:03:09.000And about 10 miles from that is a $165 billion Oracle data center.
00:03:15.520So we're in the fight here in this region.
00:03:25.460What is the impact that these data centers have on communities and the world?
00:03:31.560Like, what are the problems with these data centers?
00:03:33.960It really varies based on the size and the scale.
00:03:37.640But I can start from El Paso and just talk about some of the shocking statistics that has made this such an organizing opportunity, really.
00:05:12.580These kind of deals are pretty common all over the place.
00:05:15.780It's very similar for Oracle, which is in New Mexico, but very close to El Paso.
00:05:21.260It's just massive, massive scale of electricity, water, and air pollution.
00:05:26.580And almost all of them are getting tax breaks from the municipality.
00:05:30.080Just to contextualize this, what Matt is describing is unfortunately not unique to El Paso or to Texas. So the role that I sit in this ecosystem, I'm supporting folks across the country, local organizers, policymakers, folks thinking about statewide change to grapple with what's happening in their community and push back.
00:05:55.240and just to kind of locate us the growth that we're seeing we've had data centers for a long
00:06:04.480time for decades and decades and decades but what's different right now is the speed and
00:06:08.940and the scale right so in the past three to five years we've seen the number of hyperscalers which
00:06:14.680you can think about as really big data centers right there's different ways to define them but
00:06:18.740even bigger data centers, we've seen the number of those roughly double. And this build out has
00:06:25.300immense consequences, both for people who live right next door, including, you know, some of my
00:06:31.580family members in Atlanta, Georgia, right? But also I'm a New Yorker, right? I live in the city,
00:06:37.320but it has impacts across our entire state, even for people who are not living right next door in
00:06:41.920terms of the quality and availability of our water, right? Air, our electricity bills going
00:06:50.720up across the country. But also I think the other impact that Matt and I haven't spoke to so much
00:06:56.960is like who controls our futures, right? Who controls what happens in our towns, in our states
00:07:02.740where we live? And I think that's also something that we have to grapple with in this moment.
00:07:07.820I feel like I've been saying this a fair bit over the last few episodes of this podcast,
00:07:11.920but like this is not a video podcast you can't see the faces that i've been making
00:07:15.640while uh matt and ali have been talking but i i thought i knew how bad this was and i was wrong
00:07:24.740which i suppose is why i brought the two of you on um this is horrifying and i i understand that
00:07:31.000the ai boom is what's pushing it to happen but i mean just to we'll get to like hope and how to
00:07:38.720fight back against this in a minute but uh but just to inject a little bit of that earlier on
00:07:43.060uh i have been seeing poll after poll lately showing how unhappy people are about these data
00:07:49.780centers and uh you know it just seems like people are are more and more against having them built
00:07:58.120uh as as time goes on which i find very heartening but it doesn't seem to actually
00:08:04.540be slowing down how quickly they're being built does it i mean a bit but not as much as we would
00:08:09.600like yeah i think that is the hopeful part is you know why we are doing this is it is such a unifying
00:08:18.540issue like no there's so such a small amount of people that you explain what's happening with the
00:08:24.220data center when it comes to your county or your municipality or there's like this is good like it
00:08:29.600It is across political ideology, age, race, gender, like everyone's like, we do not want this thing.
00:08:38.200So I think that's the question right now is, can we create campaigns and organizations sharp enough to say this is what we can do?
00:08:47.760But, yeah, it is definitely deeply, deeply unpopular.
00:08:52.880Almost every single place that they're they're coming into.
00:08:55.860I think it's telling, if these things were actually beneficial to local communities, these billionaires would be building them where they live. And they're not, right? Silicon Valley is not the hub of data center expansion, right?
00:09:11.000You know, all the other places that billionaires have flocked to. These are not the hubs. They are, you know, building up this extremely toxic, extractive infrastructure in communities that they see as disposable, which is an increasing number of us.
00:09:29.460Um, and as someone who's organized against, um, against corporate power, against extractive
00:09:36.980tech and carceral technology for a really long time, this is such a hopeful moment because
00:09:41.420people are seeing the connections of this issue of they're, they're dealing with the
00:09:57.780They don't want data centers built next to them. They also have questions about AI and if it's beneficial. And the majority of people want kind of common sense guardrails. But people in power are not listening except when they're forced to. And I think we're really at an influx moment where we can win this, but the forces against us are quite strong.
00:10:18.340So before we get into that, let's talk a little bit more about the impacts. Matt, you were talking about how this thing uses more electricity than the entire city of El Paso, and that's just one data center. And they are just using a bunch of diesel generators, which is horrifying. So what's the carbon footprint of one of these data centers? How bad is that? And the water usage? Can we put these numbers into some context?
00:10:47.240the first thing to stress is it is really like shocking to people to see they will do what they
00:10:56.200can get away with yeah so what they can get away with in el paso is different what they can get
00:11:01.160away with and maybe a state like minnesota or michigan so they are using these generators
00:11:09.480because they can get away with it and they will take as much water as they can and that's why
00:11:15.560you're seeing kind of the clustering in places with with less regulation um so this one specifically
00:11:24.380the metadata center would create more air pollution than every existing refinery and power
00:11:32.820plant in el paso every home every school it would almost be double the amount of toxins in the air
00:11:42.700if this uh gas plant is built that is just from meta's private gas plant on the new mexico side
00:11:51.440oracle's gas plant for theirs which is called project project jupiter would emit more air
00:11:57.700pollution than the state of new mexico so it would it would double the amount of toxins in the
00:12:04.820entire state than the entire state of new mexico because it's so much energy it is hard to
00:12:12.340fathom the infrastructure lines the power lines the gas plants the new pipelines the fracking
00:12:18.380that you need to take out of the ground it is it is like you're building a new state infrastructure
00:12:25.140in like a hundred uh acre radius like it is it is hard to comprehend how much of a climate bomb
00:12:34.140these data centers are and i think these are our extreme examples because they can get away with it
00:12:40.220here. So not every data center is the same amount of climate bomb, but where they can't get away
00:12:46.760with it, they will. Louisiana, Mississippi, Texas, New Mexico, they are basically wiping out all the
00:12:54.700progress on emissions that have been made in the last three decades with one single project.
00:13:02.100That's horrifying. Can I humanize this for a second? Yes, please. I know I hear all of these
00:13:07.480numbers all the time, right? One data center can use as much water as a town of 50,000 people,
00:13:14.440right? We're talking about megawatts, gigawatts. The numbers can start to become meaningless.
00:13:20.840How I think about this is data centers use so much electricity that the government itself
00:13:28.760has acknowledged data center expansion as one of the biggest threats to grid stability
00:13:32.820in our country. What does that mean? That's your auntie's insulin in the fridge, right,
00:13:39.360when there's a blackout. That's oxygen tanks for elders at home. Blackouts, brownouts have real
00:13:46.680material consequences for people. And the other consequence, which I can monologue about for hours,1.00
00:13:54.160is that these facilities use so much power, right? They're not only completely reversing
00:14:00.460our climate progress unnecessarily, but they are precluding us from investing in industries in
00:14:08.240development that would actually serve people's everyday needs. Because when a town, when a grid
00:14:13.200brings in this much data center expansion, they're using up all the capacity in the grid. So they
00:14:19.860can't actually give that to another industry. So even if a town in a couple of years says,
00:14:23.640you know what, this was a bad idea, there's no more space on the electricity grid to bring in
00:14:27.700something that might actually meet people's needs. So the numbers are right now, right?
00:14:36.800They're asthma cases, they're cancer deaths, there's that insulin in your fridge, but they're
00:14:41.100also, what can we do in five to 10 years as the climate crisis gets worse and worse?
00:14:47.620Absolutely. And I think that's the thing that people think AI or what the vision that has
00:14:52.460been painted around tech is this new world but like we're seeing coal plants like being re
00:15:01.340re brought back up that have been closed like diesel generators like the the dirtiest cheapest
00:15:07.380energy that they can possibly get their hands on is how they will fuel these things um and it's
00:15:13.740it's tying us into that same fuel source for for decades to come yeah yeah and coal is so
00:15:20.940particularly awful you know the amount of illness that a coal plant creates like this is something
00:15:26.880that i know a little bit about not from personal experience but like uh one of my favorite uh
00:15:32.200favorite is maybe not the right word to use here but one of my favorite statistics about
00:15:36.280coal plants is they they release more radioactivity into the environment than a nuclear power plant
00:15:42.140does. Because coal fly ash is highly radioactive. So yeah, as if, you know, coal wasn't bad enough.
00:15:55.660Okay, so this is horrifying. How do we stop this? How do we organize against this? Obviously,
00:16:02.700the tech industry believes, you know, has been selling a lie about AI, as you were saying,
00:16:09.520bringing in this new world where we don't have to worry about all of the problems that we deal
00:16:15.400with today. But, you know, all of that's just wrong. And we do have to deal with these problems.
00:16:24.920So how do we fight against these AI data centers? How do we push back?
00:16:32.320The first thing that I want to say before talking about how is, I think regardless of what your
00:16:36.920individual position on AI writ large is. You can oppose data center expansion. Whether you find
00:16:44.380personal benefit from XYZ tool, that's not the question that we're talking about. We can have
00:16:52.500narrow, actually useful applications of AI without this build-out. Most of what you're
00:16:59.960being promised, we already had capacity for. So I just want to put that up front because I get
00:17:04.560this question a lot, right? Oh, I like AI. Don't I need data centers? We're talking about two
00:17:10.480separate things. So how do people fight back? There's so many ways to stop a data center.
00:17:19.680And the formula is really, really specific to where you live, right? There's a policy side to
00:17:26.700it, what you can do, jurisdictionally, technically. But more than that, it's about building power,
00:17:33.600right? And figuring out who makes decisions where you live about these facilities and the
00:17:40.180infrastructure, power, water, sewage, otherwise, that supports it and raising hell, right? Creating
00:17:48.640public pressure, not just targeting individuals, but figuring out these bureaucratic, administrative,
00:17:55.920ideally democratic systems and making your voice heard. That's what we're seeing is working
00:18:00.620across the country. And I can kind of get into the nitty gritty of it, but I want to
00:18:04.360let Matt speak to this first at a high level. Definitely what Elise said is like a big
00:18:09.020thing that successful campaigns have made this about is it's not about if you use ChatGPT or
00:18:17.220not, or if you have like the Gemini thing on Google, whatever. It is like a fundamentally
00:18:22.440a question about power like i think that is why people have resonated with the campaign so much
00:18:31.180and the different organizations and the fights against data centers because people can see
00:18:36.960and feel powerlessness like they can see we have to lobby it takes seven years to build a new like
00:18:47.420park in our town but now like we get a data center that's gonna double our energy use and this will
00:18:53.400be done in three months like that this this feels wrong and it offends people on like a very deep
00:18:58.700level yeah like at least said it is about having like a sharp power analysis of who made these
00:19:07.160decisions and i think that's where successful campaigns have been won and where data centers
00:19:12.060have been stopped is when people understand who brought this here who can stop it and then move
00:19:20.080all of their energy in a very targeted and strategic way um but it but it is very difficult
00:19:26.060because a lot of people may have never been to a city council meeting or a kind of commissioners
00:19:30.600meeting or may not know um who controls their energy company so usually it is a lot of spinning
00:19:38.820your wheels at first when you first hear that a data center is coming to your town. And I think
00:19:44.340that's a lot of the organizing that we've been trying to do is just like accelerate the organizing
00:19:50.440because people are there and they want to do something and how do we accelerate it so they're
00:19:53.420not wasting their time being spun around by the bureaucracy of what a municipality's zoning
00:20:01.560regulation is or whatever that specific instance is. And there's different levels, right? There's
00:20:07.460how do you stop a specific data center? That's one question. But then how do you protect your
00:20:12.260community from either the broader data center takeover that's already there or is coming
00:20:18.260is a second but interrelated question. And there are different playbooks, but kind of at the
00:20:24.600briefest level, right? It's figuring out what friction you can throw into the process.
00:20:30.900if you have the political power to win a withdrawal of the project or a voting down of the project
00:20:40.340and use whatever existing approval mechanism there is.
00:20:44.020But longer term, if we only take this approach, we're playing whack-a-mole and we will lose
00:20:49.500against these tech companies because they are so good at just popping over jurisdictional lines,
00:20:54.380next town, next county, next state, where perhaps there's less political power or more favorable
00:21:00.680regulations. So the bigger picture and what we hope that this, you know, kind of specific
00:21:08.200site-to-site power builds to is actual protection. And that can range from common sense reform and
00:21:17.580regulation to what I think is actually common sense, which is bans and restrictions on this
00:21:22.580type of development. But that requires changing the rules. Our rules around public utility
00:21:31.760commissions, zoning, water, all of this were not built, were not written for this moment and for
00:21:39.160an industry of this scale in all forms of that word. So these companies, developers, big tech,
00:21:47.860private equity, they are taking advantage of generally pro-business regulation, but also
00:21:54.660rules that they can fit into in terms of buy right or automatic approvals, right? Because
00:22:00.600no zoning code written in the past however many decades could envision this scale and this type
00:22:08.420of industry. So we actually need to change our rules to empower communities to say no, to say
00:22:16.040if and when and make sure that communities have the say and not these corporations.
00:22:22.420And how do we do that? I realize that that's a big question. I guess to get a little bit more
00:22:28.220specific, how do we resist this? How do we get better regulations in place? And actually,
00:22:35.840how do we fight this on a local level as well? Let's come back to the question of national or
00:22:42.960state level regulations and start with something local.
00:22:47.380Like if you find out that someone is planning on building a data center in your community
00:22:53.260and you don't want that to happen, what do you do?
00:22:56.640The first thing is you're not going to solve this alone.
00:22:59.780Like, I think that is something to stress.
00:23:01.800So I think that's a lot of people's immediate reaction is maybe they just don't know enough.
00:23:07.960Or maybe if I just like put together all the research and give every single perfect number
00:23:12.220we can block this right but unfortunately like these companies know what they're doing
00:23:17.080uh and that's why they're doing it so quickly and that's why they're doing it so secretly
00:23:21.040so it should be something that is like basically a political crisis like ali was saying like you
00:23:29.220have to have this be widespread across the municipality people should be talking about it
00:23:35.420because they do have so much money that they can speed run through a lot of these regulations and
00:23:42.560zoning restrictions so the first step is like building a team whoever that is if that starts
00:23:48.820with an existing organization or if it's something that's new it there should be a group of people
00:23:53.940that are doing this together and then we have and we can you know share it if people are interested
00:24:00.680with the listeners there are some good policy guides um but the first step is like where is it
00:24:08.140it has this been approved what is the process because it does get significantly harder as they
00:24:15.580move along in the process so quick catching it as quick as possible ideally before the land has been
00:24:21.520rezoned or before they're already approving the uh contract between the municipality and the
00:24:27.900developer. And then it's really, really important to see where you have leverage. That is like the
00:24:36.560key in stopping a data center is understanding where you have leverage. It may be wildly
00:24:43.100different from state to state or from municipality to municipality. So that is, I think, the most
00:24:49.180difficult part for for most communities to to get to is like a sharp understanding of where
00:24:57.620the the leverage points are and then putting everybody to there because it might be like
00:25:03.840you might have 400 people on facebook like saying this is awful but all those 400 people need to be
00:25:11.300sending public comments to the uh utility commission or to the water board or whatever
00:25:18.240it is. And I think that is the quickest way to explain it is getting a group of people
00:25:24.700understanding where you have leverage and then getting every single person who's upset about
00:25:28.920this moving in the same direction towards that point of leverage. And that is where 99% of
00:25:36.760communities that have stopped data centers, I think, have done really well, is they have
00:25:41.000understood those core pieces of the campaigns. And just to break down that leverage piece
00:25:47.240further, where decision-making is held, like I said before, is going to look different place
00:25:54.960to place. So you might think, oh, this is my town city council has authority. Actually,
00:25:59.500they might not. It might be the zoning board. It might be the county. It might be that there's
00:26:04.240actually no approval process that has a meaningful kind of decision point because this fits into
00:26:12.240a type of zoning that is automatically approved. So then you look at, okay, what are the other
00:26:17.640permits that this development needs? That might be air, water, sewage. It might be around tax
00:26:26.540abatements. That's also an intervention point. Shout out to El Paso. There's different
00:26:34.400decision points. But like Matt said, it's easiest to stop a data center before it ever moves in,
00:26:42.360before construction starts, before the permits are given. But if you're not in that position,
00:26:47.580all hope is not lost. And so in addition to figuring out where the approval process is for
00:26:56.740the specific data center that might be coming to where you live, I really strongly encourage people
00:27:02.680to look at what could protect your community from additional development. Because the sooner we get
00:27:09.800that into place also, the sooner that protects against what you're facing right now. A lot of
00:27:15.820zoning reforms or zoning changes don't apply retroactively in general, right? So if a huge
00:27:22.980mega facility is already built and then you put in a ban, that's likely a ban on new construction
00:27:27.820or expansion. And I can run through kind of the local policy playbook that we have. We have
00:27:35.100something called the North Star Data Center Policy Toolkit, which we've created with over 100
00:27:41.000organizers, policymakers across the country to really lay this out for folks. And I'm happy to
00:27:45.140walk that through if it's helpful, but I don't have to get in the weeds. Just to note to our
00:27:49.200listeners, we've got links to a couple of resources from the AI Now Institute and a
00:27:57.300site fight guide in the show notes. So, you know, if your community is staring down a data center,
00:28:06.580those are good places to start. And they come from the AI Now Institute and the data center
00:28:11.840working group. So, yeah, Ali, why don't we get a little bit into the weeds? I don't have a problem
00:28:18.020with getting into the weeds. Let's get into the weeds. And then if we get too far into the weeds,
00:28:21.100we'll come back out and we'll go somewhere else. Great. I live in the weeds. I love the weeds.
00:28:24.680Mack and Phil is back. The playbook that we've seen as successful in terms of more holistic
00:28:29.160protections goes a little something like this. If you have the power and the authority to flat
00:28:35.600out ban this type of development, do it. It's not serving you. It's not serving your town. It's not
00:28:41.100serving your community. These are, if we want to frame it this way, outsiders coming in and taking,
00:28:46.060right? This isn't for your digital capacity, your digital daily lives. Some places are going to be
00:28:52.200able to put in a ban. Shout out Monterey Park in California, right? Some places might not be able0.80
00:28:58.760to do that. Either that's not the approach that the local government wants to take, or the locality
00:29:04.940might be preempted, prevented by state law or other types of regulation that actually don't
00:29:12.300allow them to ban or prohibit certain types of development. So in that case, you have a couple
00:29:17.060options. We're seeing an increasing number of localities, cities, counties, towns, et cetera,
00:29:24.520now New York State, putting in place moratoriums. Moratoriums are really rational, common sense
00:29:31.140pauses to give people time to figure it out, right? We encourage at least a year for a moratorium.
00:29:37.500I believe there's one in North Carolina that's 32 or 36 months now. Shout out to them. That's
00:29:41.740amazing. A pause is only as good as what you do with it, right? And if you can't ban, there's lots
00:29:47.420of other things that can be done at the local level, including in zoning. And some of that is
00:29:51.920restriction, saying, okay, we're fine with data centers as long as they're under 20 megawatts,
00:29:57.280right? We don't want hyperscalers. We don't want data centers over 50 megawatts or over 100
00:30:01.460megawatts. Or if we have them, we're going to limit them to a very strict type of permitting
00:30:08.740where we're making the developer and the operators prove that they're abiding by all of our local
00:30:14.000standards for air quality, water use, electricity, 24-7 additive renewables, whatever we want to put
00:30:18.880in there. We can make them abide by a set of standards in order to move in, and we can do
00:30:24.340monthly, quarterly reporting with actual enforcement power to make sure that that works.
00:30:32.500I don't trust these companies. I don't trust these developers. That's maybe not what I would
00:30:36.680do in my town, right? I would go with restrictions. I would go with bans if that's possible at all.
00:30:41.780But we want to make our localities unattractive to the most extractive types of developers.
00:30:49.200And I'll say one note here is we're seeing an increasing number of localities, including
00:30:53.900Nashville, restricting data centers to industrial zones that can be somewhat helpful. But it's
00:31:02.820really important to remember, which communities tend to be sited, which communities tend to live
00:31:08.200next to industrial zones tends to be blacker, browner, poorer, more often immigrant communities.
00:31:14.000So we're displacing harms potentially onto the communities who are already targeted in many,0.52
00:31:21.660many ways. There's a bunch of other things that can be done, right, with water use, with transparency,
00:31:27.540with taxation, although that's a little more at the state level, energy, a little bit more at the
00:31:31.280state and regional level. But the point is there's so many ways to restrict these things. There's so
00:31:37.420many ways to restrict these developments. And the strongest thing a community can do is to say no
00:31:42.860and then to organize regionally to make sure that they're not just pushing these developments onto
00:31:48.140their neighbors. Unfortunately, we still live in capitalism, and that means we need to talk about
00:31:53.780our partners and sponsors. But fortunately, our partners and sponsors are great. And if you want
00:31:58.900To support Dreaming Against the Machine directly, you can join us on Patreon, where $5 a month gives you the chance to ask questions of upcoming guests and a dedicated stream of cat pictures.
00:32:11.060We love our patrons, and we also love being a part of the Multitude Collective at MultitudeShows.com.
00:32:17.080Multitude is a podcast company made up of passionate people creating shows you can count on.
00:41:52.700I would love to keep talking with you, and I think we're going to have to have the two of you back on.
00:41:56.920But I want to start wrapping things up.
00:42:02.060But before I do, we have some listener questions, or really a bunch of questions from one listener.
00:42:07.620patreon supporter if you want to send questions in for future guests please support the show on
00:42:13.360patreon but these are questions from a guy named richard george and he sent in a bunch of questions
00:42:19.260and we answered or the two of you answered a lot of them actually already but there's two or three
00:42:26.240that we didn't get to and i think maybe we can bash them out real quick so first he went after
00:42:33.180one of the arguments that I have heard people use in favor of data centers, which is jobs.
00:42:40.680And so he asked, how many local long-term jobs do data centers support? And how does this compare
00:42:45.460to other comparable uses of land and resources? I love this question. Tech companies, developers
00:42:52.880love to come in and lie about the number of jobs that they will create. A lot of the numbers I'm
00:43:00.600going to share here are from Good Jobs First. They are the experts about the economic development
00:43:04.820impacts and the actual job truths of data centers. I think what they and others have found is that
00:43:12.700a hyperscale facility can be run with 20 people, 50 people in terms of permanent or semi-permanent
00:43:19.620jobs. And a lot of these are not the high-paying jobs that companies are touting either. They're
00:43:25.300janitorial, their security. A lot of the job numbers that are thrown out are initial construction
00:43:31.360jobs, and those disappear once the facilities are built. In terms of how comparable this is
00:43:36.900to other industries, Indiana and Michigan Power, a utility itself, disclosed that data centers
00:43:44.700create 100 times fewer jobs than other industries by amount of energy used. Thank you to Citizens
00:43:53.840Coalition of Indiana for sharing those numbers. Yeah, 100 times fewer. And there's also similar
00:43:59.660numbers coming from Virginia that the jobs cost 100 times more in terms of investment to create
00:44:07.160a permanent data center job than other industries. So this is not a good deal. And I think we need
00:44:13.220to look at the reality. Okay, Matt, anything? A good example that we've used in El Paso is
00:44:18.200the metadata center will use more water per day than El Paso Independent School District.
00:44:23.660Oh, my God. That is, I believe, 48 schools and about 20,000 employees. The contract with the city in Meta, Meta is obligated to provide 50 permanent jobs. That's less than an elementary school, one single elementary school. So they're using more water than the entire school district, and they're providing less jobs than one single elementary school.
00:44:51.140And across the country, these jobs are being heavily subsidized.
00:44:55.820I believe the average is that one data center job can cost roughly $2 million in subsidies.
00:45:03.500And so that's what the community is getting back in return for those subsidies is this giant gaping hole in their balance sheets and something that pollutes massively and just doesn't give back to the community in any serious way.
00:45:20.660Awesome. Okay. One more question from Richard George, which I think is a good follow-up to
00:45:24.980that one. Is there any reason a community should actually want an AI data center locally?
00:45:32.680I think what we hear most is the revenue just because of the scale of some of them. So the
00:45:39.200argument that we hear is like, for example, Project Jupiter, they're getting a ridiculous
00:45:44.600amount of tax breaks. I'm sorry, which one is Project Jupiter?
00:45:47.780So this is in Santa Teresa, New Mexico, which is right on the border of El Paso. So it's kind of a suburb of El Paso, but in the state of New Mexico. That's $165 billion project. Instead of getting taxed on that, they are just paying a direct payment to the municipality of $80 million.
00:46:08.620dollars. So that might seem like, oh, great, we get 80 million dollars. But that is 165 billion
00:46:15.340dollar project. That data center is worth more than all of the property in the state of New
00:46:20.920Mexico. So they that is like the immediate you might like fill your budget gap for the next two
00:46:27.640years. And that's what like people are are seeing like, hey, we have a budget deficit and this data
00:46:31.820center will get us out of the red for the next three years. But long term, it is a drain on the
00:46:41.020city because you're not getting that revenue after that direct payment. That is what a lot of cities
00:46:46.120are saying is we just need this immediate cash, but they're not getting the long term tax value
00:46:52.500because of these abatements. Yeah. And that immediate payment is what? What was that? That
00:46:56.900was like one half of 1%. Well, actually, hold on. 80 million and it was 165 billion. I'm sorry,
00:47:05.060that's not one half of 1%. That's one half of a 10th of a percent. That's a 20th of a percent.
00:47:11.720That's insane. I think understanding the reality here is important, which is this industry is
00:47:17.780targeting communities that have faced, you know, now it's facing austerity, right? Federal
00:47:23.140disinvestment, deindustrialization, people governing these communities are looking desperately
00:47:29.800for something to lift up their budgets, right? But what it reminds me of is communities,
00:47:36.680for example, in Pennsylvania, in coal country that now have their public school budgets
00:47:42.560dependent on fees paid by the coal plants, so they can't shut the coal plants, right?
00:47:46.380And what we're looking at is an even more speculative industry deeply tied to the coal
00:47:49.980industry and fossil fuels. And if there is a bubble and if it pops, do we think these tech
00:47:55.160companies are paying for all of this infrastructure that's now burdening these communities? Of course
00:48:00.140not. So communities are being presented with a false choice. And I think this is really important
00:48:05.780to make clear. They're being told, join us, choose progress or get left behind. They're not getting
00:48:11.840a choice between a data center or another type of industry coming in. And I think that's what we
00:48:16.940need to remember is that it is a choice and we have to create what that alternative is. And that's
00:48:21.480a lot harder than saying yes to a developer. It is the right choice. It is the just choice. And
00:48:27.620it's the one that actually will take care of communities in the long term.
00:48:31.840I want to close with one final question here. It's a question that is based on something
00:48:39.640that one of the people behind the surge of data center construction said. In October 2024,
00:48:48.700Eric Schmidt, who used to be CEO of Google, he's a tech billionaire and venture capitalist,
00:48:54.320he dismissed attempts to make AI data centers more energy efficient. He called it a waste of time.
00:49:00.400And he said, we're not going to hit our climate goals anyway because we're not organized to do
00:49:08.080it. I'd rather bet on AI solving the problem of climate change than constraining it, meaning AI,
00:49:16.060and having the problem. What would the two of you say to that? I think the point that Ellie made at
00:49:24.380the beginning is it's tying us to it. Like, what would AI do about 800 generators? 800 diesel
00:49:32.740generators like this is a greater carbon footprint than the entire city there is no way that you can
00:49:43.680effectively claw back something as big as these centers in any realistic timeline so if you set
00:49:52.220up a gas plant with 800 diesel generators running 24 7 these things are going to go out you have to
00:49:59.460replace them like it is creating this demand and there is the the amount of investment and
00:50:06.900federal planning that would need to transition this gas plant into something beneficial is
00:50:13.940decades and this generator will get up in a year or this gas plant will get up in a year
00:50:18.740so i think just like the the actual numbers of what these gas plants the water use like
00:50:28.280how is AI going to refill the aquifer that like you're taking the water out of? That is what
00:50:34.420people are dealing with. Once that aquifer is gone, there's no more aquifer. Once people are
00:50:38.580living next to a gas plant, it takes decades to decommission that gas plant. We hear this a lot,
00:50:44.480right? Oh, well, we don't need to worry about the climate impacts, the water, the energy,
00:50:49.640because AI is going to solve climate change. That's not what these systems are currently
00:50:54.320being built to do. And it is data center expansion that is directly pulling us back from the barest
00:51:03.480of climate benchmarks that we have made. But they use this as a cudgel, right? AI is going to solve
00:51:10.400climate change. AI is going to cure cancer. When we look at actual scientific advancement that uses
00:51:16.520AI tools, it is much, much narrower, smaller applications, right? Which don't require this
00:51:22.540build-out. This build-out is not for these purposes. It's for the military-industrial
00:51:27.240complex and the defense industry. It's for fossil fuels. It's for speculative AI investments.
00:51:32.520It's for these companies throwing billions and trillions of dollars around in a circle in a way
00:51:37.300that destabilizes our entire economy and puts everyday people's investments and pensions
00:51:42.360at risk. This is yet another myth that they are pushing. Efficiencies. Yes, there are ways to make
00:51:50.260data centers, more energy and water efficient. Do they compare with the scale and the speed of
00:51:57.140this build out in a meaningful way? No, not that I have seen, right? We need to rethink this whole
00:52:02.500industry and not get bogged down in, oh, is closed loop going to solve X, Y, and Z? That's not the
00:52:08.460issue. And that's a mask that's being pulled over people's eyes, right? To distract from what
00:52:13.620actually needs to be done. I find it striking that AI is being pointed to as a solution to
00:52:21.460the climate crisis while it's exacerbating it. And when we already know most, if not all,
00:52:29.200of what we need to do to solve the climate crisis, and most of it is not build more
00:52:33.900technological breakthroughs, it's deploy more renewable energy and decommission more
00:52:39.020fossil fuel power plants. And yet you get people like Schmidt pointing to this as if it is a
00:52:48.280problem that you can solve simply by throwing more AI at it with no evidence to back it up.
00:52:56.980And I do think that we need to turn off these data centers and prevent more of them from being
00:53:01.700built if we're going to do what needs to be done to protect our climate and our planet. And thank
00:53:06.820you both for working on that and helping all of us figure out how to do that more effectively.
00:53:13.540Thanks so much for having us hopefully to be continued.