A few months ago the founders of the top AI companies walked into Capitol Hill. Tried to explain to a room full of elected representatives exactly how their technology was going to put almost half the working population into under/unemployment and they should consider UBI [0]. Then they went back to the airport, got on their corporate jets, and went home secure in the knowledge that they really showed them. That they were the smartest people in the room.
BTW, no one I know gives a shit about the energy consumption or water usage. They absolutely want to know if these datacenters will bring jobs to their area. So far Altman, Ellison, O'Leary, Amodei, Pichai, and Zuckerberg have refused to answer that question.
[0] All except Jensen who has been really trying to explain the benefits of AI and has said these massive layoffs are a huge mistake.
I'm local-ish to Box Elder county in Utah, here people absolutely give a shit about the environmental burden to the region. It isn't just about water consumption, but other things. I think the "We need to win China is AI" narrative is (appropriately???) falling of deaf ears. Obviously I don't have the numbers, but even lay-people in my circles have asked how these alleged AI_driven benefits (fighting cancer, stopping climate change, and whatever) are really going to come to fruition, when what they really observe in their backyards are data centers being generated so we can fill our lives with AI slop.
I'll second this observation, as well as add that apart from AI slop most people around here associate the data center push with the sudden proliferation of Flock cameras at every major intersection and along every highway. Provo defeated a major data center project that was going into an empty industrial park, arguably the kind of place that would fit that sort of development. The actual cost-benefit calculation for most people is heavily weighted towards the negative and this should not continue to surprise people. The perceived downside with no upside is just going to get worse if the government gatekeeps the most useful models.
Plenty of people care about their power bill. Water in some regions is a hot political issue. Data centers don’t create jobs of course, we don’t need anyone to answer that.
Out here in the US southwest, we absolutely do care about water usage as well as the potential for higher electricity prices. We also care about jobs, which in the case of data centers are only going to be boosted temporarily until construction is finished.
I've been an indie filmmaker since I was a teenager. Seedance 2.0 and all of the image and video models are such an amazing gift to use. The things you can push these models to do are incredible. I have a full VFX workbench. I can rotoscope, I can pull off effects shots. I can even use these things to articulate non-shotlist things for meetings. It's incredible. The keyboard gave everyone a "bicycle of the mind", now everyone can visually express themselves if they try.
I've professionally been a systems engineer. Five/six nines reliable services that move billions of dollars, etc. I have fallen in love with coding models. I am getting so much more work done that I'm launching easily three times what I did prior to AI tooling. The job we're all in is to provide value - these models are the next generation of compilers. We're working at higher and higher levels of abstraction, and it's brilliant. For those of us who can operate at all of the levels, it's a super power.
I will not go back to pre-AI times. I want to see what things are like in 10, 20 years. When we have at-home Michelin star robot chefs, where our cars can drive us to the beach overnight so we wake up to sunrise on the coast, where I can have an idea for a new take on a music player tagging algorithm and just build that without it consuming weeks of my time.
This is the most excited about tech I've ever been. This is so much better than smartphone incrementalism and stupid web platforms.
Stop grandpa-ing and shaking your fist at clouds. This is literally the jetpack future we were promised growing up. It's the coolest thing since the internet.
I’m an AI hater and the way you describe your uses sounds cool to me and like what I imagined when I was an AI optimist in the early days. The problem is that most of what it produces is still slop, the images, videos, music, and code it generates are definitely impressive but there’s something qualitatively worse about them that I can only describe as a lack of soul. Over time in a codebase, these tools create a complete mess of complexity. Those AI generated Coca Cola ads were terrible, it was just a series of cool shots with no story. The music sounds good and interesting but it’s just missing something.
I think it’ll get better though and we’ll find ways to collaborate with it that make the most of the human and AI abilities, but it seems so overhyped right now.
> The keyboard gave everyone a "bicycle of the mind", now everyone can visually express themselves if they try.
Right, everyone can. So now your film-making vision is simply one infinitesimally small slice of the pie that every viewer is eating. Yes, you can make a movie by yourself. Likely no one will watch it because they're too busy watching other movies made just as cheaply but by companies with marketing budgets.
> I am getting so much more work done that I'm launching easily three times what I did prior to AI tooling.
Great. Have you ever once in your life had a real conversation with a normal person where they expressed, "Man, you know what? I wish I had way more apps on my phone."
Like, yes, there is demand for software that fills unique niches, but really we are reaching saturation.
> When we have at-home Michelin star robot chefs
Eating the world's best meal, alone, while staring at your phone.
> where our cars can drive us to the beach overnight so we wake up to sunrise on the coast
This part sounds nice. Hopefully you can find parking.
> where I can have an idea for a new take on a music player tagging algorithm and just build that without it consuming weeks of my time.
Except you don't have a music player to put that algorithm in because all of the music players are closed source. You can write an open source one (or contribute to an existing one), but those all require local libraries of music, which almost no one has. Because it's not about the software, it's about the access to content.
But, really, why even bother tagging music in the first place? Just treat the tags as prompts and generate an infinite stream of music catered exactly to your mood, on demand.
I get where you're coming from. AI is a massive force multiplier for producing content. But content isn't the point of life.
The future that AI builds is one of perfect solitary meaningless hedonism. Every itch scratched, every base desire satisfied. But there is a hollow void at the center of that. Even a pet dog will lose its mind when given endless food, treats, and toys if it doesn't have an actual person to play with, and I'd like to believe we are somewhat more cognitively sophisticated than dogs.
Think back on the best meals you've ever had. I've had some very good ones. Some were memorable because of the quality of the food. But the memories of meals I hold most dear were dinners I made myself for family, not-very-good cookies my young daughter baked for me, meals shared with friends while travelling, crappy hot dogs cooked over a campfire.
It's human connection that brings us the most lasting joy, and AI is antithetical to it.
If you only ever want frontier model performance then sure, you have to pay to play. But as it is now with open models, some of which I can even run from my gaming PC at home, we're only about 6 months behind frontier performance. Even when the money starts drying up for those at the forefront, the genie is out of the bottle.
And that's why I pivoted at the beginning of this year to LocalLLMs.
I can afford the hardware I already have. And I can run jobs on it as I see fit. Sure, its slower than the corpo LLMs, but we're already seeing lying, silent downgrading, and advertisements, even when paying for tokens.
When your systems goes tits up or priced out, mine will still work.
> In Utah on Wednesday, State Senate President J. Stuart Adams—one of the most powerful Republicans in the state—lost his primary election after supporting a major data center development near the Great Salt Lake, in one of the clearest signs yet of the growing political risks tied to the industry.
> At the local level, the fallout was just as direct. “Do I think that the data center vote cost me the election? Yes I do,” former Box Elder County Commissioner Lee Perry said after conceding his primary race, after voting to advance the same project.
This is a pretty predictable reaction to the underhanded tactics being used to try to force these projects through either before citizens know they're happening and often, when citizen are aware, against their will.
As a side note, I wish we could muster this kind of vigor for just about any other type of public infrastructure project… nuclear/wind/solar power, fiberoptic internet, public high speed rail, new cities built around human-centric principles… you know, the things that the better part of the population stands to benefit from so at least the initial unrest is somewhat justified.
You get the same vigor against all of those project all the time. Windmills, solar, nuclear, rail, etc. There's a stronger will to oppose than to build.
We do muster this kind of vigor. The problem is that coverage decisions shaped by negativity bias ensure you're much more likely to hear about the projects people don't like. Did you hear about the huge New Mexico wind farm, largest in the US to date, that came online two weeks ago?
That's true, but I also see a lot of infrastructure projects that get gummed up or even canceled by NIMBYs, industry incumbents, and general obstructionism that would've happened had they been undertaken with the same bull-in-china-shop approach these datacenter projects tend to take.
The US needs to build out energy infra like China, already >2x more total generation capacity (~1/3 of world total). Putting data centers aside, if America wants to bring manufacturing back, it will need energy to build things. RE: data centers, we do need to be more mindful of where they go and the resources they consume. We should force them to use water efficient cooling (more expensive for them) and support the buildout of the required energy gen. Utah does not seem an appropriate place for such a large data center.
The question is do we want to be a Petrostate or an Electrostate
When I look at the total number of acres in a state and the number that may get taken up by a data center… not that we should even have to look at this spec but still not sure why people are so focused on data center construction as an issue unless it’s literally going to be next to your house, is there anything more than FUD going on here? And perhaps people taking advantage of it specifically as a talking point given election season.
You can't look at the total acreage of the state as a metric when 64.4% of the state is owned by the federal government.
The proposed site is twice the size of Manhattan, NY and sized for 9GW of energy which is more than the entire state uses yearly. We literally do not have the water to support a data center that huge.
They just enacted a fireworks ban because the weather people just had to create a whole new category for how dry and dangerous it is. Air quality is a constant problem because all the pollution from regions West of Utah collect right against the mountains. A few years ago we woke up to what looked like heavy fog, but it was smoke --from Siberia.
I think part of it is the perception that real environmental and public health damage is being done for totally trivial and indefensible causes. A data center is not like an airplane parts manufacturer, which has lots of ugly pollution but provides a necessary national service. Most people use generative AI recreationally, and the productivity gains among white-collar workers are awfully ephemeral. And if you're less pessimistic about the economics of generative AI... that makes it worse!
"We get a ton of money, you get increased natural gas emissions, increased unemployment, your electric bill is going up... oh and guess who's bailing us out when the bubble bursts?" Pretty rotten deal!
The problem is not the center per se. The problem is the power. And, all too often they make up for the lack of utility capacity by putting in their own noisy generators.
> the Stratos development would have spanned tens of thousands of acres in Box Elder County’s Hansel Valley. The project would ultimately require up to 9 gigawatts of power—more electricity than the entire state of Utah currently uses
Did you even read the article? This is proposed to be larger than Manhattan. The amount of power will almost certainly burden Utahs grid in ways that locals will be on the hook for. So much of this build out will be the typical "privatize gains, socialize losses" playbook that yes it is an important political issue, and yes you have to "look at this spec" to understand just how insane some of these project proposals are.
This pretty much spells out exactly my big problem with datacenters.
I don't care if you build a huge datacenter several miles away from my home. What I do care about is utilities cranking up the price 3x because of "capacity issues" afterwards because said datacenter now uses more power than the entire district I live in
When people hear datacenter they think ai-almost universally subsidies by the many for the few. They drive up electrical costs, increase carbon emissions, and are designed to make money by stealing and repackaging the fruit of human labor and thought, with the goal of ultimately replacing it not for the benefit of all, but for the benefit of the owners of that datacenter.
A few months ago the founders of the top AI companies walked into Capitol Hill. Tried to explain to a room full of elected representatives exactly how their technology was going to put almost half the working population into under/unemployment and they should consider UBI [0]. Then they went back to the airport, got on their corporate jets, and went home secure in the knowledge that they really showed them. That they were the smartest people in the room.
BTW, no one I know gives a shit about the energy consumption or water usage. They absolutely want to know if these datacenters will bring jobs to their area. So far Altman, Ellison, O'Leary, Amodei, Pichai, and Zuckerberg have refused to answer that question.
[0] All except Jensen who has been really trying to explain the benefits of AI and has said these massive layoffs are a huge mistake.
The people doing the voting are mostly talking about how much water datacenters supposedly use.
I'm local-ish to Box Elder county in Utah, here people absolutely give a shit about the environmental burden to the region. It isn't just about water consumption, but other things. I think the "We need to win China is AI" narrative is (appropriately???) falling of deaf ears. Obviously I don't have the numbers, but even lay-people in my circles have asked how these alleged AI_driven benefits (fighting cancer, stopping climate change, and whatever) are really going to come to fruition, when what they really observe in their backyards are data centers being generated so we can fill our lives with AI slop.
(I'm also in Utah)
I'll second this observation, as well as add that apart from AI slop most people around here associate the data center push with the sudden proliferation of Flock cameras at every major intersection and along every highway. Provo defeated a major data center project that was going into an empty industrial park, arguably the kind of place that would fit that sort of development. The actual cost-benefit calculation for most people is heavily weighted towards the negative and this should not continue to surprise people. The perceived downside with no upside is just going to get worse if the government gatekeeps the most useful models.
Plenty of people care about their power bill. Water in some regions is a hot political issue. Data centers don’t create jobs of course, we don’t need anyone to answer that.
Out here in the US southwest, we absolutely do care about water usage as well as the potential for higher electricity prices. We also care about jobs, which in the case of data centers are only going to be boosted temporarily until construction is finished.
Nothing this technology offers is, to me, worth the noise pollution or increase in water and electricity rates.
Not to mention the tax breaks they're given for no material benefit to the community.
Future Illinois data center tax breaks on hold - https://www.illinoistimes.com/news/future-data-center-tax-br... - June 25th, 2026
State Data Center Policy Shifts as Governors Impose New Restrictions - https://www.multistate.us/insider/2026/6/22/state-data-cente... - June 22nd, 2026
Gov. JB Pritzker suspends tax breaks for data centers, urges more discussion - https://capitolnewsillinois.com/news/gov-jb-pritzker-to-susp... - June 5th, 2026
Which States Are Banning Data Centers? - https://www.ncsl.org/fiscal/which-states-are-banning-data-ce... - June 2nd, 2026
US tax incentives for data centers by state - https://knowledge.sdialliance.org/8d367baa340046029912b1e04c...
Tax Incentives for Data Centers 50 State Survey - https://hbfiles.blob.core.windows.net/webfiles/TaxIncentives... [pdf]
Good thing the market is bigger than just you.
I've been an indie filmmaker since I was a teenager. Seedance 2.0 and all of the image and video models are such an amazing gift to use. The things you can push these models to do are incredible. I have a full VFX workbench. I can rotoscope, I can pull off effects shots. I can even use these things to articulate non-shotlist things for meetings. It's incredible. The keyboard gave everyone a "bicycle of the mind", now everyone can visually express themselves if they try.
I've professionally been a systems engineer. Five/six nines reliable services that move billions of dollars, etc. I have fallen in love with coding models. I am getting so much more work done that I'm launching easily three times what I did prior to AI tooling. The job we're all in is to provide value - these models are the next generation of compilers. We're working at higher and higher levels of abstraction, and it's brilliant. For those of us who can operate at all of the levels, it's a super power.
I will not go back to pre-AI times. I want to see what things are like in 10, 20 years. When we have at-home Michelin star robot chefs, where our cars can drive us to the beach overnight so we wake up to sunrise on the coast, where I can have an idea for a new take on a music player tagging algorithm and just build that without it consuming weeks of my time.
This is the most excited about tech I've ever been. This is so much better than smartphone incrementalism and stupid web platforms.
Stop grandpa-ing and shaking your fist at clouds. This is literally the jetpack future we were promised growing up. It's the coolest thing since the internet.
Can you share something you’ve made with AI?
I’m an AI hater and the way you describe your uses sounds cool to me and like what I imagined when I was an AI optimist in the early days. The problem is that most of what it produces is still slop, the images, videos, music, and code it generates are definitely impressive but there’s something qualitatively worse about them that I can only describe as a lack of soul. Over time in a codebase, these tools create a complete mess of complexity. Those AI generated Coca Cola ads were terrible, it was just a series of cool shots with no story. The music sounds good and interesting but it’s just missing something.
I think it’ll get better though and we’ll find ways to collaborate with it that make the most of the human and AI abilities, but it seems so overhyped right now.
> The keyboard gave everyone a "bicycle of the mind", now everyone can visually express themselves if they try.
Right, everyone can. So now your film-making vision is simply one infinitesimally small slice of the pie that every viewer is eating. Yes, you can make a movie by yourself. Likely no one will watch it because they're too busy watching other movies made just as cheaply but by companies with marketing budgets.
> I am getting so much more work done that I'm launching easily three times what I did prior to AI tooling.
Great. Have you ever once in your life had a real conversation with a normal person where they expressed, "Man, you know what? I wish I had way more apps on my phone."
Like, yes, there is demand for software that fills unique niches, but really we are reaching saturation.
> When we have at-home Michelin star robot chefs
Eating the world's best meal, alone, while staring at your phone.
> where our cars can drive us to the beach overnight so we wake up to sunrise on the coast
This part sounds nice. Hopefully you can find parking.
> where I can have an idea for a new take on a music player tagging algorithm and just build that without it consuming weeks of my time.
Except you don't have a music player to put that algorithm in because all of the music players are closed source. You can write an open source one (or contribute to an existing one), but those all require local libraries of music, which almost no one has. Because it's not about the software, it's about the access to content.
But, really, why even bother tagging music in the first place? Just treat the tags as prompts and generate an infinite stream of music catered exactly to your mood, on demand.
I get where you're coming from. AI is a massive force multiplier for producing content. But content isn't the point of life.
The future that AI builds is one of perfect solitary meaningless hedonism. Every itch scratched, every base desire satisfied. But there is a hollow void at the center of that. Even a pet dog will lose its mind when given endless food, treats, and toys if it doesn't have an actual person to play with, and I'd like to believe we are somewhat more cognitively sophisticated than dogs.
Think back on the best meals you've ever had. I've had some very good ones. Some were memorable because of the quality of the food. But the memories of meals I hold most dear were dinners I made myself for family, not-very-good cookies my young daughter baked for me, meals shared with friends while travelling, crappy hot dogs cooked over a campfire.
It's human connection that brings us the most lasting joy, and AI is antithetical to it.
Enjoy it until the subsidies end and the economics catch up. You can afford it now with broad subsidies, but likely not the true cost.
https://www.theregister.com/ai-and-ml/2026/06/24/ai-coding-a...
https://mimetiq.substack.com/p/the-tokenmaxxing-hangover
https://fortune.com/2026/05/28/tokenmaxxing-is-dead-companie...
https://www.axios.com/2026/05/28/ai-spending-roi-enterprise-...
https://finance.yahoo.com/markets/article/big-techs-27-trill...
https://www.youtube.com/watch?v=YrJzjC4kKCY
If you only ever want frontier model performance then sure, you have to pay to play. But as it is now with open models, some of which I can even run from my gaming PC at home, we're only about 6 months behind frontier performance. Even when the money starts drying up for those at the forefront, the genie is out of the bottle.
And yet, the productivity increases with existing models are meh (ie not 0 but certainly not worth what is being spent).
https://www.bok.or.kr/eng/bbs/B0000354/view.do?nttId=1009840...
https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies...
AI Doesn't Reduce Work–It Intensifies It - https://news.ycombinator.com/item?id=46945755
https://simonwillison.net/2026/Feb/9/ai-intensifies-work/
AI Doesn't Reduce Work–It Intensifies It - https://news.ycombinator.com/item?id=46955703
https://hbr.org/2025/09/ai-generated-workslop-is-destroying-...
https://cmr.berkeley.edu/2025/10/seven-myths-about-ai-and-pr...
("extraordinary claims require extraordinary evidence")
And that's why I pivoted at the beginning of this year to LocalLLMs.
I can afford the hardware I already have. And I can run jobs on it as I see fit. Sure, its slower than the corpo LLMs, but we're already seeing lying, silent downgrading, and advertisements, even when paying for tokens.
When your systems goes tits up or priced out, mine will still work.
> In Utah on Wednesday, State Senate President J. Stuart Adams—one of the most powerful Republicans in the state—lost his primary election after supporting a major data center development near the Great Salt Lake, in one of the clearest signs yet of the growing political risks tied to the industry.
> At the local level, the fallout was just as direct. “Do I think that the data center vote cost me the election? Yes I do,” former Box Elder County Commissioner Lee Perry said after conceding his primary race, after voting to advance the same project.
This is a pretty predictable reaction to the underhanded tactics being used to try to force these projects through either before citizens know they're happening and often, when citizen are aware, against their will.
As a side note, I wish we could muster this kind of vigor for just about any other type of public infrastructure project… nuclear/wind/solar power, fiberoptic internet, public high speed rail, new cities built around human-centric principles… you know, the things that the better part of the population stands to benefit from so at least the initial unrest is somewhat justified.
You get the same vigor against all of those project all the time. Windmills, solar, nuclear, rail, etc. There's a stronger will to oppose than to build.
We do muster this kind of vigor. The problem is that coverage decisions shaped by negativity bias ensure you're much more likely to hear about the projects people don't like. Did you hear about the huge New Mexico wind farm, largest in the US to date, that came online two weeks ago?
That's true, but I also see a lot of infrastructure projects that get gummed up or even canceled by NIMBYs, industry incumbents, and general obstructionism that would've happened had they been undertaken with the same bull-in-china-shop approach these datacenter projects tend to take.
The US needs to build out energy infra like China, already >2x more total generation capacity (~1/3 of world total). Putting data centers aside, if America wants to bring manufacturing back, it will need energy to build things. RE: data centers, we do need to be more mindful of where they go and the resources they consume. We should force them to use water efficient cooling (more expensive for them) and support the buildout of the required energy gen. Utah does not seem an appropriate place for such a large data center.
The question is do we want to be a Petrostate or an Electrostate
https://youtu.be/gLnxzkiB-GI?is=CHj3J-ARp0iBq_NB (Adam Tooze prezi)
https://en.wikipedia.org/wiki/List_of_countries_by_electrici...
Most US citizens want to be an electrostate. This current administration wants to go back in time and try to be a petrostate....
That may not be true for much longer given the (disappointing) trend
2020 - 79 : 20 (renewables : fossil fuel)
2026 - 57 : 42
https://www.pewresearch.org/science/2026/04/03/americans-shi...
When I look at the total number of acres in a state and the number that may get taken up by a data center… not that we should even have to look at this spec but still not sure why people are so focused on data center construction as an issue unless it’s literally going to be next to your house, is there anything more than FUD going on here? And perhaps people taking advantage of it specifically as a talking point given election season.
You can't look at the total acreage of the state as a metric when 64.4% of the state is owned by the federal government.
The proposed site is twice the size of Manhattan, NY and sized for 9GW of energy which is more than the entire state uses yearly. We literally do not have the water to support a data center that huge.
They just enacted a fireworks ban because the weather people just had to create a whole new category for how dry and dangerous it is. Air quality is a constant problem because all the pollution from regions West of Utah collect right against the mountains. A few years ago we woke up to what looked like heavy fog, but it was smoke --from Siberia.
Water usage concerns in a desert, perhaps?
I think part of it is the perception that real environmental and public health damage is being done for totally trivial and indefensible causes. A data center is not like an airplane parts manufacturer, which has lots of ugly pollution but provides a necessary national service. Most people use generative AI recreationally, and the productivity gains among white-collar workers are awfully ephemeral. And if you're less pessimistic about the economics of generative AI... that makes it worse!
"We get a ton of money, you get increased natural gas emissions, increased unemployment, your electric bill is going up... oh and guess who's bailing us out when the bubble bursts?" Pretty rotten deal!
The problem is not the center per se. The problem is the power. And, all too often they make up for the lack of utility capacity by putting in their own noisy generators.
Misdirected outrage. DCs are an easy outlet for people's AI frustration.
> the Stratos development would have spanned tens of thousands of acres in Box Elder County’s Hansel Valley. The project would ultimately require up to 9 gigawatts of power—more electricity than the entire state of Utah currently uses
Did you even read the article? This is proposed to be larger than Manhattan. The amount of power will almost certainly burden Utahs grid in ways that locals will be on the hook for. So much of this build out will be the typical "privatize gains, socialize losses" playbook that yes it is an important political issue, and yes you have to "look at this spec" to understand just how insane some of these project proposals are.
"Privatize gains, socialize losses"
This pretty much spells out exactly my big problem with datacenters. I don't care if you build a huge datacenter several miles away from my home. What I do care about is utilities cranking up the price 3x because of "capacity issues" afterwards because said datacenter now uses more power than the entire district I live in
When people hear datacenter they think ai-almost universally subsidies by the many for the few. They drive up electrical costs, increase carbon emissions, and are designed to make money by stealing and repackaging the fruit of human labor and thought, with the goal of ultimately replacing it not for the benefit of all, but for the benefit of the owners of that datacenter.
What's not to like?