Episode 76
CAN AI AND DATA CENTERS BE SUSTAINABLE? THE TENSION BETWEEN TECH AND COMMUNITIES
Keywords data centers, legislation, environmental impact, community resistance, technology solutions, water usage, energy consumption, regulatory changes, economic impact, zoning laws Summary This episode…
Episode content
Episode Story
Research & Analysis
Research Note: E076 Policy Status and Publication Boundary
The August 2025 episode discussed state and local responses while several proposals were still moving. The revised article needs to preserve Dalton's solution-oriented vi
Research Note: Cooling and Water Tradeoffs
The transcript discussed air, evaporative, liquid, immersion, reclaimed-water, and Microsoft examples. Public copy needs a common comparison boundary.
Research Note: Community Benefits Agreement Governance Boundary
The guide needs to help a host community turn a promise into an observable obligation while respecting jurisdiction, representation, and enforceability.
Research Note: Data Center Siting
The siting guide needs to find fatal constraints before a project team reduces a location to land price, power promises, and incentives.
Research Note: Sustainable Data Center Decision Framework
The framework needs to answer whether a data center can be sustainable without turning a complex project into a promotional score.
Research Note: Additional Clean Power
The explainer must separate a credible new-supply plan from an annual accounting claim.
Field Notes
Can AI Data Centers Be Sustainable? A Practical Test
Evaluate an AI data center across workload, site, power, water, emissions, community terms, transparency, and closure without hiding tradeoffs.
Data Center Cooling and Water Use: Options Compared
Compare air, evaporative, chilled-water, direct-to-chip, immersion, hybrid, and reclaimed-water cooling by climate, density, energy, and water.
How to Evaluate a Data Center Site Before Commitment
Compare data center sites by workload, grid, water, fiber, hazards, land, permits, emergency response, public costs, and community compatibility.
Data Center Clean Energy Additionality, Explained
Learn how to test a data center clean-power claim for new supply, location, hourly matching, interconnection, firmness, transmission, and backup.
Data Center Community Benefits Agreements: A Guide
Learn how data center community promises become measurable obligations with authority, milestones, reporting, enforcement, remedies, and change control.
Full episode
TranscriptSearch or read the full conversation.
E76 CAN AI AND DATA CENTERS BE SUSTAINABLE_ THE TENSION BETWEEN TECH AND COMMUNITIES
Transcript
Dalton Anderson (00:00.864) Welcome to Vich step podcast, we discuss entrepreneurship, industry trends and the occasional book review. What if your favorite apps cloud provider was draining your city of water? That was a topic covered in our last episode. Some of the water usage in these AI data centers are enormous and mind-boggling in themselves, especially their water and power usage. This episode we're going to talk about
some of the solutions to the problem. Last episode, we talked about some of the issues that a data center has been creating. And so if you need a little bit more background or you're curious about that, I would really encourage you to listen to the previous episode. In all caps is how AI data centers are draining power, water, patients, and then the rest of the episode title.
That episode will discuss the issues of the data centers, the water usage and some other things going on with pollution. This episode is going to be discussing some of the solutions that states and communities have come up with and or technology solutions as things become more advanced. The first part of the episode we're going to be discussing, does the legislation
and regulatory solutions. And then we're going to move towards community led resistance and zoning, which is somewhat regulatory, but really led by the community and some of the just horrific videos that people have seen on Facebook and YouTube. And they're like, that's not happening here. I'll tell you what, that is not happening here. And I don't blame them because it doesn't look too fun.
especially if that's your home. But as things progress, not everyone's happy. And then just like a little bit of like balancing the burdens and then we're just gonna recap. I would say the main topic of the episode is gonna be talking about the regulatory solutions and then some other things and some of the technologies that have come out. All right, well, if you don't know who I am, I'm your host, Dalton Anderson.
Dalton Anderson (02:25.324) This is VentureStep podcasts and I've been doing this for some time now and typically don't provide any background about myself, I knowledgeable in coding, insurance and various topics. I like to read, explore life and do a podcast. But anyways, so let's get into it. All right. So I talked about it in the last episode that Georgia
was pushing for a bill that had since stalled and they're going to push it for next year. But what Georgia wanted was to bar rate increases because of data centers. So the data center usage is taking up a lot of the, I would say the supply and let's think about just pizza pie. Like data centers are taking up 40 % of the pizza pie. This is theoretical by the way, not legitimate, but they probably legitimately take up.
20 % of the usage.
And so you have these, these large users and then the rest of the people mean we're just normal folks. We use a little bit, not too much. And then there's small businesses that use more than us. And what's happening is since the data centers are classified as small businesses, there isn't a distinct difference between the rate increases. So small businesses and the residents are getting charged because of the strain on infrastructure.
the predicted usage of infrastructure, which is interesting. I didn't think about that, but that's something we'll discuss. Whereas you're projecting your energy usage of your project and then you're like, well, you know, we don't really need that much anymore because we've got this other data center and we're doing our thing over here. And then that city, county slash like utility company built out all this infrastructure. It costs money. The capital outlay of
Dalton Anderson (04:29.174) infrastructure is quite expensive and then the projected usage, if it doesn't meet that, then they're stuck with this extra capacity that no one's using and they expected people to use it at a higher rate. So that causes quite a bit of issues. But that was something that I didn't ponder about in the last episode and it didn't think about it as an issue and it didn't come up in the research I had, but it's come up in legislator legislation. Sorry. So I that was interesting.
So the Georgia bill stalled, didn't work out. They're gonna try again next year. And.
Dalton Anderson (05:08.834) the people voting on the bill were all being paid by the utility companies. And it wasn't that much like it was like 10,000, 5,000, a thousand. But I'm sure that that helps out in your Senate race or wherever you're at in this, in this vote and
Yeah, I didn't think it was that much money, but I guess it could influence your decision. I'm sure they have some.
indiscreet or sorry, discrete ways of investing and helping you out, helping you run your campaign that aren't a direct company. So I'm sure they get more money, but publicly filed. only get, you know, know, onesie twosies. So that was interesting that Georgia didn't pass. There is some good stuff. So Oregon passed the Power Act bill, House Bill 3546, the Power Act.
reclassifies energy usage for data centers or high usage companies. And their threshold was anything over 20 megawatts, which is quite a bit, like how many homes powered can be powered by 20 megawatts. That's anywhere from 13 to 40,000 homes.
common estimates is 20,000 homes with normal usage. So in normal energy usage, you're looking at 20,000 homes. So it's not just like a random, you know, it's quite a bit. So it's a small town. That's more than a two horse town for sure. And then another thing that they included was long-term contracts. So you had to commit to a 10 year contract if you were going to use quite a bit of energy.
Dalton Anderson (07:02.614) They want to make sure that you're committing to a long-term contract with your projections of what you're going to use because they're going to have to build out infrastructure for you. And if you're not using it, then as I mentioned earlier, leaves the utility company and
the state in a bad spot. So that was a positive. I think that's a great bill to pass. Minnesota, they have legislation that grants explicit authority over public utility. So the Public Utilities Commission has explicit authority to reject or accept special tariffs on
energy supply deals for data centers. So basically anything that the data centers want to get a deal in Minnesota, they've got to go, it's got to get approved. So I think that one's, that one's okay, but I don't, I don't know if it's like rock solid, like Oregon's plan is pretty rock solid. Ohio's is pretty good as well. So Ohio,
And this hasn't been like completely approved. The one that has been approved is Oregon. And that's, that's a done deal. The Ohio piece.
They're proposing, and this was proposed by AEP Ohio, which is apparently a major utility company within Ohio. And so they sought a couple things. So they wanted to charge data centers 90 % of peak usage instead of 60%. So previously they were charging 60 % of peak usage. So if your peak usage was in this example, a hundred megawatts.
Dalton Anderson (09:05.001) they were charging 60 megawatts was the going rate. Instead of now they're going to say, okay, if your peak usage is a hundred megawatts, then we're charging 90. Even if you didn't consistently use 90 megawatts, you're going to charge 90 megawatts. And I think that's kind of fair because if your power usage is consistently fluctuating quite a bit, like
I would think about it if you're training data and you have a huge training load that you're pushing through and you're, you're doing like months of training and you're using massive amounts of energy nonstop for months. For those four months that you're training, you're using just an absurd amount of energy. And then after that, you're, you just have queries that you're running through while you're getting together your next training set. But I think the training process would consume quite a bit of energy compared to just a regular query.
And then all that can fuse using up the energy and then can cause blackouts within the community and a lot of strain on the infrastructure. So I think that makes sense. And then the next thing was.
Dalton Anderson (10:19.63) They have to provide, they have to have long-term contracts on what they're doing. And then they also need to pay for the infrastructure, enhanced infrastructure that the utility needs. And that was to prevent ordinary folks like myself having to pay for these data centers that are getting created.
without really getting any of the benefit. We're only getting all the costs. And they're trying to prevent that by making them commit to long-term contracts and also charging 90 % of peak, which would enhance the infrastructure and really push the data center.
or data center energy usage to be more consistent instead of wide fluctuations. That's what they're trying to prevent.
Dalton Anderson (11:26.382) Okay. So one thing that I thought was interesting was 36 States provided incentives for taxes. And originally I thought that it was going to be a benefit, right? I thought that converting a data center, like converting unused land into a data center was profitable for the state. But when you dig a little deeper, quite a bit of States,
And this hasn't been publicized. This is theorized by like analysts and lawyers when they take in an account, the tax revenue generated and the incentives to move to that state or the land sale or the tax tax breaks that were provided for the move and how many years that is amortized for. It is theorized that many of the states are actually
losing money on these data center deals. Plus all the issues with the community being pissed and then you have all this strain on your water usage and your energy usage. so 36 states have offered and they offered incentives for data centers to come to their state and.
after things have shooken out a little bit and there's been this big data center boom and this AI as well, the energy usage and water usage has just blown out of the water. It's just up 50%, 40 % year over year. And I think these towns are quite concerned and also states aren't getting the revenue that they thought.
Maybe they're just not mathematicians and they just they just didn't run the numbers correctly and the tax revenue that they predicted and then the tax write-offs. It's just the math is not math in for them.
Dalton Anderson (13:34.71) And so they're pushing change. A lot of states aren't happy. 12 dozen states plan to introduce bills related to data centers and preventing data centers from costing normal customers like homeowners and people who live there from driving up your utility bill costs.
And as you may or may not heard last episode, I talked about the family in Georgia and rural Georgia.
think it's Mansville, Mansville? Blanket on the town.
I don't know the town's name, but it's in a rural place outside of Atlanta, one hour away. And their energy bill doubled, close to double. I think it was like 90%, 90 % increase.
Dalton Anderson (14:42.904) It was, was very close to double that. This is what they're trying to prevent, but Georgia, Georgia's bill failed. So the next thing that they're, they're doing is they're to reintroduce it, but other States are also getting pushed pretty hard and who's doing the pushing. Well, that's the local communities. And so I think one group in Northern Virginia, Fairfax County,
I'd really pushed hard on creating stricter zoning guidelines because the data centers weren't zoned and this could be a loophole or some kind of tax incentive. I'm not sure. didn't research this part, but
Basically what happened was, and I mentioned in the last episode is these data centers are really close to residential areas, like 300 feet, 400 feet type of thing. And what happens is they're just doing massive construction. It's an industrial, like it's industrial use. any place that's using multiple megawatts of power that can power like a whole town,
That's not, that's not something that goes near the playground, right? And it's caused many issues, not only with sound noise, but also pollution. People are getting sick because of these data centers don't have enough power that the grid literally can't provide enough power for them. So then they're seeking power elsewhere. And the main, the main way you can really turn on power quick is with a gas turbine or
the diesel turbine. And so these data centers are running these massive power, basically side, side power plants or power plants in your, in their backyard of their data center. And so they've got 20 or 30 power plant or not power plants, but
Dalton Anderson (16:52.738) turbines running nonstop to help power these.
Enhanced training clusters and provide more capacity availability.
And so...
that caused quite a bit of pollution and people were getting sick. I'd helped, I didn't help, but I'd helped guide you to the right area where we watched that video together. And so people are pretty pissed. Like they're, they're not cool with it. And so I know in Fairfax County, they had provided new zoning. And so there, there's now different, there's setback rules, new classifications that can't be
so close to residential buildings. so Northern Virginia did zoning restrictions.
Dalton Anderson (17:52.014) a town in Missouri did an outright ban. And the way that they did that was just no data center. Missouri did an outright ban and then in Indiana, they did this like no data center signage everywhere. So all the residents had a no data center sign. And I think it's kind of weird where I would position these data centers as
the Walmart, like the new Walmart where this rural town has different folks. They've got a knitting, a knitting shop, some vegetable guy. They've got the auto mechanic. That's everyone's uncle or something like that, like they call him like uncle, uncle Bob or something. And then they've got Susie's cakes shop. And this is a small, quaint little town and everyone's got their little thing and
No one's making a lot of money, but people are making enough to get by. And then Walmart comes around and just dynamites the whole thing and just undercuts everybody intentionally way, way below what the market price is for these goods drives the small mom pop shops out of biz and then takes over, raises prices to the normal.
low, already still low Walmart prices and calls it a day and just completely takes over the town. Kind of the same thing here where it's these data centers are normally in these rural towns where the there's availability of land. One of the issues is they're not there's not that much availability of power and or water most of the times. And so these big companies are promising
you know, hey, we're going to 200 million in construction costs and we're going to be here for the long term. We're going to hire folks for you and you'll get better internet and some other stuff and you'll have low latency and there's potential that companies will come by and they'd want to they want to be close to the data center to get lower latency. And yeah, it's going to be a great deal for you. And then
Dalton Anderson (20:19.18) And then they take the deal and then they find out that like a medium size, a small size data center only hires like 40 full-time employees. It's a small shop. The buildings don't look good. They're disruptive. The lights are always on. It's a massive construction project. And the tax incentives that they've provided and the tax revenue, they don't even break even on.
And so they're just like, they're just at such a standstill because you can't just move the data center. They already spent billions of dollars on this data center, like 500 million, hundreds of millions of dollars minimum on this data center. You're not going anywhere. Like you've already, you've already taken over. You're like the virus, like you're just there like a tumor, but it also does a lot of good stuff, but you could see it as someone.
in that rural town, you're like, this really sucks. I'm not, I'm not letting this happen to my town. Cause they see all those videos online and they've got millions of views and I'm sure there's a lot more of like worse stuff that's going on. So.
There's that. And then the environmental concerns. So I wasn't aware of this, but these these deals were pretty hush hush. And I was surprised how hush hush the Tesla deal was that I talked about in the last episode where the town didn't even know that they had the data center going on. wasn't announced by the officials. It wasn't. The only reason they found out was Elon was talking about it on X and he did a whole announcement and was impressed how fast construction was going.
And they're like, hold on, we're doing a data center in our town? then, yeah, then people found out and they weren't too upset. I mean, they were upset that the officials didn't tell them and it's a big deal. And then later on, they're running all those turbines and people are getting really sick and dying. That's no good. So.
Dalton Anderson (22:33.656) people aren't happy and people are afraid of the data center boom and come into their town. And then moving over to that same topic where I talked about how hush hush the deals are, I was surprised as I mentioned the Tesla deal, but then I was also surprised that other deals are pretty similar where they didn't have to disclose their water usage. They didn't have to disclose their power usage. It was all NDA.
So customers, when their bill was increased and they're like, why are you doing this to us? They're like, I don't know. I'm not sure, I can't tell you. And so everything was NDA, NDA, NDA. So they had NDA for their power usage, NDA for their water usage. And then they had NDAs for like getting water in suspect manners. And there's a story of
a Google data center using just groundwater, like local groundwater and completely just was just draining the groundwater until local locals found out about it. I guess like, I don't know how they find out about this stuff. Like it, I'm sure no one's really allowed on the property of Google so that they're not taking interviews. People aren't, people aren't getting near there.
And so maybe they're doing environmental testing and they're like, hold on, what's going on here? Like where, where's all this water going? Where this is way below what it's supposed to be for this season. And then they're doing some digging and they're trying to figure it out and they probably have some kind of map and then they figure it out and they're like, it's this, this is building who owns this building. it's Google. So there's just all sorts of shady stuff going on with these deals and it's just,
it's all behind closed doors and it's just the promise that they're going to get more revenue and maybe they'll do an office in that area and then they'll hire some more talent. I'm not sure the thought process on it, but it's a bit odd how closed door these big deals are. And I think that
Dalton Anderson (25:01.122) provides more frustration to people because the people that were hired to represent them aren't representing them.
So that's that. And then the next thing regarding this activism piece is the environmental concerns. So the environmental concerns with pollution from gas turbines or diesel turbines, water usage is a big one. And so there's been quite a bit of deals that
where they're originally we're going to use like some local water system or the river or a lake or the city utility. And then when the public found out about it, they're like, whoa, whoa, whoa. That's a no go. That is a no go. My friend, you're saying that this place is going to use 40 percent of our water.
That doesn't make any sense. Like, why would we do that? And no one has any good answers to these questions. And so a couple of deals have been pushed to use.
Potable water?
Dalton Anderson (26:18.986) and moving away from using drinkable water to reduce the strain on clean, drinkable water.
And so they're using cleaned wastewater.
And I think there's been commitments from Amazon. said by 2030 that they wanted the majority of their data centers to be using.
Dalton Anderson (26:54.418) No, I really blanked on the water thing. I'm like doubting myself about this. Sometimes on the podcast, I since I recorded. I am. Like doubting myself, I think I said earlier potable water, I meant to say non potable water, but I think people understand what I'm talking about. Don't want to get too sidetracked, but non potable water. So they made a commitment that they want to use non potable water for their data centers.
Google has made similar commitments. think Amazon is leading the way with non-potable water data centers. And so that really leads us to the tech increases.
tech advancements. And so there's a couple that some of which I think honestly a lot of these tech advancements aren't going to work. And so there's like this direct to chip cooling mechanism where it cools the chip directly and cycles it through. And then Microsoft is running a pilot program for zero water cooling. And that was a project where they just, they just
had the data center, like a small data center, because it's a pilot project, fully submerged in the water. I think they did it in the ocean. And obviously, there's some issues with that, like the infrastructure, just having something submerged in water consistently. You got to make sure it's waterproof and accessible. If there's any issues with your cluster or the wiring,
How do you get to it? That just doesn't work at all. The direct cooling thing may work, right? But I think one of the main issues with this is that...
Dalton Anderson (28:57.652) It might be too intensive to set up and then to unset up. Whereas like the current system, it's easy to get new stuff online. mean, it's not easy, but it's easier than if you have like water running directly through versus where this is like running water over to cool stuff. And then the water gets vaporized and then
that water is 80 % of the water is wasted and the other 20 % is wastewater that needs to get treated. So I think that would make more sense, or these still make as much sense. What makes sense is like not using potable water, using non-potable water, so just reclaimed wastewater. And then I think one of the main pieces would be to make the chips more efficient.
This is a this is.
This is Navidius claims. These are my claims. I can't test it. I don't have a Blackwell tower at my house. I have a Blackwell architecture GPU, but I think it's a nine, nine 40 or something. I don't even know. I don't even know which one it is. I just know it's the new Blackwell architecture at least, but it's a small gaming computer. who knows? I don't even remember.
Maybe we're in the nine thousands. don't know. Anyways. So this is, I'm to read this off. Nvidia is claiming that they are 20 to 30 % more efficient with Blackwell systems. They also claim a 2000 X reduction in energy usage for training and a hundred thousand X reduction for generation to generating tokens over the past 10 years. Their overall platform has seen a 200 X improvement.
Dalton Anderson (30:58.592) in efficiency in a 4,000 X in computational performance in the last decade. The Blackwell Ultra GPU is reported to be 25 X more efficient than traditional CPUs for certain AI and high performance computing HPC workloads. I mean, that makes sense, so.
But in the general sense is that it's gotta get more efficient for it to even make sense, like long term, which I'm no doubt in my mind that it does. There's a lot of money getting thrown at these problems, and so eventually someone's gotta figure it out. So I don't doubt that we're gonna get more efficient. It's also just like, how do we figure out?
the new norm, like this is gonna become normal. There's gotta be more data centers, there's gotta be more power. How can you make
in a phone call.
Anyways, how can you make data centers more efficient? And both on a power capacity and natural resource users, but also how do you integrate data centers within the community with it being less disruptive to where it feels less of a burden than it is right now? Because especially for these local communities, it seems like they're just getting ragged all.
Dalton Anderson (32:27.886) by these massive companies. And it doesn't seem like they're getting the economic gains that they should. so that goes to the next piece is the economic gains with these data centers. From what I have seen, I think there are gains.
But also if mismanaged, these data centers really just take over your town and the pollution that Tesla has committed and, it's not really Tesla, it's XAI.
And maybe, maybe Tesla acquires XAI. I have no idea because XAI acquired X. So X is now owned by XAI. or not Twitter. Tesla uses X for some of their loads. XAI uses some Tesla cluster, AI clusters that Tesla built for their self-driving training.
And then XAI is integrated into Tesla's. So I don't even know. mean, I'm pretty sure that SpaceX uses XAI. I don't even know. I mean, it's kind of confusing. That's a diagram and episode on itself.
So there's the property tax boosts where you have this unused land that you can provide, enhance infrastructure, internet utilities, utilities, internet, all these things. I mean, it seems like a good deal where so big companies like, I'm gonna spend $400 million and hire a couple people. They're probably promising a lot, but they're not really hiring that many. It's like.
Dalton Anderson (34:26.318) 40 for a small medium, maybe 100 for a large one. And so you have that availability to enhance and enrich your land without having to constantly maintain it. And also this investment is going to allow the community to have better internet, really fast internet, and it should enable them to have better
power infrastructure and water efficiency, but it's seemingly to be the other way around where they're just latching on to the current systems and just taking and taking and taking. So. Just needs to.
Dalton Anderson (35:15.714) the benefits need to be more localized. Like if the community is gonna have to deal with all this BS, then they need to get a lot of the benefits. you if you want to do an oral rig on my town, in my town, and we've got lots of oil, pay the people. This is the new oil. So if you're gonna make some massive data center and my prices are gonna increase, what about a stipend? Like, I would love a stipend. And I think that would be the way to go where...
the benefits are more localized than they are at the moment. Whereas people are really feeling these benefits because they're taking all the pain. So.
The next thing would be smarter paths. Right now they have requirements on usage of power and water. And it seems like these companies are just such a rush to get it done. They're just like, we'll do whatever. So we'll take it from the local groundwater. We'll take it from the local ecosystem doing massive damage.
If we if we need to, we'll we'll pull in and purchase these diesel turbines and we'll pollute the air in the local the local area in a 20 mile square block. We'll do what we need to do. Whereas it's it's OK to to do that, but you need to also not not do what you're doing currently.
Like it's okay to push, you gotta push the right way. Like if you're gonna add in all this power, you've gotta do some kind of clean energy approach or you've gotta reduce the pollution that you're providing to people. And then if you're gonna use water, you've gotta be smart about that. Like you've gotta figure it out. Whereas like it's just putting so much burden on these people. And it's really sad when you watch all the videos and you research it and you're like, wow, this is tough.
Dalton Anderson (37:29.518) Because these towns are too small to have a voice. And so they're kind of just getting pushed around for sure.
Dalton Anderson (37:38.306) But hey, that's the data center solutions. And so they're making progress on the regulations, the regulatory bodies are making progress with new legislation that they want to come out with, which looks promising. And given that other states have passed legislation, that's also great. Like Oregon's Power Act is very strong. What is being proposed by Ohio is very strong. And so,
Looks good. It all looks good. And George's bill stalled, but they're going after it again next year. And so there's a big opportunity here to make things right. And I think that just needs to get rethought about, we just rethought that rethink the whole thing because right now it's just broken. The system is completely broken and
There's some things that are solutions. Others are just barriers and or they need solutions and guidelines. I think like you can't just do whatever you want. Like you've got to you've got to if you're going to add power, you got to add power the right way. know New Jersey is proposing New Jersey, New York and these these regional these northern states. They want to put together a regional legislation that requires
new energy usage to be 100 % renewable. So we'll see how that all plays out, but it might be too strong and heavy-handed. You gotta make sure you have the right approach. But the current approach isn't working for quite a few states, so something needs to be changed. But anyways, appreciate you listening to this episode and.
I am glad to be back. think you can hear my enthusiasm in my voice, hopefully. But once again, appreciate listening. And of course, wherever you are in this world, good afternoon, good evening, good morning. Have a great day. Thank you for listening. And talk to you next week. Bye.
SourcesFollow the evidence trail.
E076 Sources
Preserved episode evidence
[[E76 - Transcript - e76-can-ai-and-data-centers-be-sustainable-the-tension-between-tech-and-communities (Dropbox copy 1)]] is the canonical raw monologue. It preserves Dalton's August 2025 survey of proposed policy, community resistance, cooling changes, more efficient computing, transparency, and local benefit.
[[E76 - AI Data Centers - Dated Sustainability and Community Commentary]] is a legacy derivative with public URLs and an earlier solution structure. Its legislative status, numerical claims, political-influence claims, and technology comparisons require current primary sources.
Existing public identity
daltonanderson.net/venture-step/ai-data-centers-sustainable-solutions-for-a-growing-crisis
This is the current canonical public route. It returned HTTP 200 on July 27, 2026 and preserves the legacy Ghost slug.
daltonanderson.ghost.io/ai-data-centers-sustainable-solutions-for-a-growing-crisis
This is the existing Ghost identity for the episode article.
open.spotify.com/episode/4gKJHJADKM5sAXceLTOYwS
This is the preserved Spotify episode identity.
This is the preserved YouTube episode identity.
Package research records
[[E076 Policy Status and Publication Boundary]] reconciles the current Oregon, Ohio, Minnesota, Georgia, and Fairfax records and protects the Episode Story from freezing proposal-era claims.
[[Sustainable Data Center Decision Framework Research Note]] defines the linked-test method and the boundary against one composite score.
[[Cooling and Water Tradeoff Research Note]] separates heat collection from final rejection and corrects the Microsoft Project Natick conflation.
[[Additional Clean Power Research Note]] separates annual accounting, physical delivery, and reliability.
[[Community Benefits Agreement Governance Boundary]] records the representative, legal-authority, enforceability, and professional-review limits.
[[Data Center Siting Research Note]] defines the constraint-first site method and the boundary between public screening tools and project due diligence.
Demand and energy supply
iea.org/reports/energy-and-ai/executive-summary
The International Energy Agency provides current global context for energy demand, efficiency, system constraints, and the role of AI in energy.
iea.org/reports/energy-and-ai/energy-supply-for-ai
The IEA examines electricity supply for added data center demand. It is a foundation for discussing new generation and grid infrastructure without implying that contractual claims alone change the physical grid.
energy.gov/powering-americas-ai-future-data-center-resource-hub
DOE's current resource hub gathers federal material on data centers, electricity infrastructure, efficiency, siting, permitting, and related policy.
energy.gov/oe/clean-energy-resources-meet-data-center-electricity-demand
DOE describes additional generation, storage, grid infrastructure, existing resources, demand flexibility, and clean firm technologies as a portfolio for serving growing load.
FERC's official Order No. 2023 summary establishes the current federal interconnection-process context used to distinguish a signed contract from an operating and deliverable resource.
The United Nations 24/7 Carbon-Free Energy Compact defines an hourly and location-aware procurement goal. It is voluntary guidance, not a project certification.
Cooling and water
energy.gov/sites/default/files/2024-07/best-practice-guide-data-center-design_0.pdf
DOE's best-practice guide provides engineering context for facility design and cooling choices.
energy.gov/cmei/femp/cooling-water-efficiency-opportunities-federal-data-centers
DOE describes cooling-tower water efficiency opportunities. This is relevant to existing facilities but does not prove that evaporative cooling is appropriate in every location.
epa.gov/waterreuse/water-reuse-case-study-quincy-washington
EPA's Quincy case documents one industrial water reuse arrangement serving data centers. It is an example, not a universal template.
epa.gov/waterreuse/water-reuse-action-plan-20
EPA provides broader context for water reuse planning and collaboration.
DOE documents one direct-water-cooled federal high-performance-computing comparison. It is used as a bounded case, not a general performance guarantee.
datacenters.microsoft.com/tour/innovation-room
Microsoft describes its land-based, closed-loop, zero-evaporation cooling design introduced in 2024.
Microsoft's Project Natick record documents the separate underwater experiment. Public copy must not blend Natick with the later land-based design.
Air and public safeguards
epa.gov/stationary-sources-air-pollution/clean-air-act-resources-data-centers
EPA maintains current air-permitting and regulatory resources relevant to data centers and backup generation.
Community and Tribal development
energy.gov/indianenergy/articles/data-centers-tribal-economic-development-frequently-asked-questions
DOE's Office of Indian Energy discusses potential opportunities, risks, infrastructure needs, sovereignty, and due diligence for Tribal communities considering data center development.
Massachusetts' 2026 expectations connect responsible data-center development with ratepayer protection, power, water, labor, transparency, and community benefits. They are state expectations rather than a universal certification.
mass.gov/info-details/community-benefit-plans-and-agreements
mass.gov/doc/standards-and-guidelines-for-cbps-and-cbas-march-1-2026/download
Massachusetts' final March 2026 guidance emphasizes early and inclusive engagement, community-defined priorities, transparency, and measurable benefits. The general energy-infrastructure guidance informs the governance anatomy but is not a model data-center agreement.
Current state and local policy records
apps.oregonlegislature.gov/liz/2025R1/Measures/Overview/HB3546
Oregon HB 3546 became Chapter 323 and took effect on passage in June 2025. It addresses large-load service classification, contracts, cost allocation, and risk to other retail customers.
aepohio.com/company/about/rates/data-center-tariff
AEP Ohio records the July 2025 tariff approval and current study, contract, load-ramp, collateral, minimum-demand, buildout, and exit framework. The current minimum-demand ceiling is 85 percent, not the transcript's proposal-era 90 percent.
mn.gov/puc/activities/v-l-e-c/data-centers
The Minnesota Public Utilities Commission describes the 2025 statutory and regulatory framework for tariffs, electric-service agreements, cost responsibility, clean-energy compliance, reliability, and the community-support fee.
psc.ga.gov/newsroom/media-advisories/page3
psc.ga.gov/newsroom/media-advisories/page2
Georgia Public Service Commission records establish a January 2025 large-load rule and a December 2025 integrated-resource-plan agreement. A later 2026 House bill remained pending in the last inspected legislative record.
fairfaxcounty.gov/news/board-supervisors-approve-new-data-center-zoning-ordinance-amendment
Fairfax County's adopted September 2024 amendment supplies the current building setback, equipment, screening, design, and noise-study example.
Siting and hazard screening
waterdata.usgs.gov/blog/groundwater-levels-app
USGS describes its current National Groundwater Conditions application. It is appropriate for screening and historical context, not water-right or project-yield conclusions.
fema.gov/flood-maps/products-tools/national-risk-index
The FEMA National Risk Index supports broad hazard screening. Parcel engineering, local records, climate projections, and dependent-infrastructure analysis remain necessary.
Evidence boundaries
"Sustainable" is not one metric. A facility can improve one dimension while worsening another. Lower water use may increase energy demand. A clean-energy contract may not provide new supply at the same time and place as the load. Efficient chips can lower work per task while total demand still rises.
The Oregon, Minnesota, Ohio, Georgia, and Fairfax examples were reconciled through current official records on July 27, 2026. Their mechanisms remain distinct. Campaign-contribution, bribery, corruption, or improper-influence implications are unsupported and prohibited.
Draft-time checks
Use [[E075 Sources]] for shared metric definitions. Refresh the current legal status of every policy example immediately before release. Distinguish proposals from enacted rules and company commitments from enforceable obligations. Record tradeoffs rather than labeling one technology a complete solution. Publication remains unauthorized.