> Alfalfa irrigation (Colorado River): 1.6 trillion gallons annually
> Google data centers (global): 6.4 billion gallons in 2023
> Thermoelectric power plants: 41% of total US water use
EDIT: that page has a correction to some of the other numbers on the page, but the 6.4 billion gallons number quoted above for Google's data centers is in line with other sources: https://mostpolicyinitiative.org/science-note/data-center-wa... ("Google reported their data center water consumption increased from 4.3 billion gallons in 2021 to 6.1 billion gallons in 2024.").
[1] https://en.wikipedia.org/wiki/Water_in_California
[2] https://www.convertunits.com/from/billion+gallons/to/acre+fo...
[3] "Agrivoltaics as a climate-smart and resilient solution for midday depression in photosynthesis in dryland regions" -- https://www.nature.com/articles/s44264-025-00073-1
> Thermoelectric power plants: 41% of total US water use
For example, if you have a 1GW coal powerplant on a river, it might draw 4 billion liters of water per day, but will only evaporate about 5-10% of that amount. The rest is returned to the river. Of course, this is much less exciting (click-baity) than writing "41% of US water use is for thermoelectric power plants".
The case for data centers in the age of AI is weaker, and the only effects for the average person are noise, pollution, higher electricity costs, higher water costs, for the ultimate "benefit" of losing your job.
Not saying that is true today though.
I suspect the people facing rising electricity cost and air-noise pollution beg to differ.
Unfortunately most people also don't really care until the problems come hitting them so extremely close you have absolutely no coiche to act upon it. Then the massive outrages start. It's a very predictable cycle.
Water pricing should be revisited to be closer to its true value -- and while food production should be recognized as essential, even that has to be rethought: is growing alfalfa for dairy less vital than growing vegetables? Especially if that dairy is produced and consumed in Saudi Arabia?
The other thing that we should be doing all over is better management of our watersheds so we can try to replenish those aquifers.
With AI that minimum is expected to increase.
It’s not 6 billion per day. It’s 650 billion per day. In the United States alone.
https://www.construction-physics.com/p/i-was-wrong-about-dat...
lol
Also, your correction needs as correction:
> If you simply take the Berkeley estimates directly, you get around 628 million gallons of water consumption per day for data centers, much higher than the 66-67 million gallons per day I originally stated. However, the methods used to produce these estimates are debatable, and seem to have been chosen to give the maximum possible value for data center water consumption. If you exclude the water “consumed” by hydroelectric plants via reservoir evaporation, you get something closer to perhaps 275 million gallons per day.1 And if you take into account the fact that lots of data center operators use renewable energy PPAs (mostly from wind and solar sources), my guess is that you get something closer to 200-250 million gallons per day (though I haven’t run a detailed calculation here).
It's not 675 billion gallons a day, it's 675 million, which is about 246 billion gallons a year.
We conserved, rates still rose, f*ck the poor the most, so that select farmers can use however much water they want. During a drought. As allowed and protected by law.
[1]: https://www.waterboards.ca.gov/waterrights/water_issues/prog...
[2]: https://archive.gov.ca.gov/archive/gov39/2015/04/01/news1891...
But it is the opening chapter that is very explicit in trying to highlight how most people live today. It isn't that we don't know what is happening or that we do not care, it is just that we have needs and wants that unfortunately have long term blow back. We are just trying to take care of others, but that is leading down a less desirable path.
I thought it was a great message that needs to sent to the future.
The aquifers are being drained, but it was being done to feed the people. Yes, it could have been handled much better but that core driving force behind it will probably be the same in hundred of years from now.
If that message isn't sent forwards, the stories that will be made in our place will not be very kind to us.
If that matters or not, that is up to you to decide. And that is the whole debate on 'What is legacy?'.
Just feeding people sufficiently, even at a high standard, wouldn't lead to the resource depletion and waste we are seeing today. But doing so at record profit levels does.
Our own individual ancestors were not perfect people. But we only exist today because they survived.
I really can't tell the difference between people being uncaring or just stupid about either.
Who’s going to pay in if the fund expects a guaranteed commitment but refuses to give such in return?
Literally defined contribution [1].
I have spent most of my life in the San Joaquin Valley and what I have been wondering about over the past few years is what is the effect removing all that water from the faults. Specifically, does this water removal cause the stresses along the faults to build making the resulting earthquakes larger?
Water lubricates fault zones [1]. Putting it into an inactive fault makes it active. Taking it out of an active fault makes the quakes less frequent and more powerful.
[1] https://www.sciencedirect.com/science/article/abs/pii/S00137...
https://www.thehindu.com/news/national/cauvery-water-managem...
#SaveWater #SaveElectricity #SaveFuel
Interestingly none of our Govts. or Leaders are talking about conservation yet and the ancient water laws/standards don't take climate change into consideration. For eg. as per BIS IS 1172: 1993 standards an average Indian is entitled to 135 Liters of water per day when the actual supply is not even close to 50 liters per Indian per day. This is level of water crisis India is facing today.
Got a bit of a Cold War era pykrete like project feeling.
Wouldn't it be for rain to fall and the aquifer's levels to rebound less than expected?
> California’s Central Valley, where one-quarter of the food produced in the United States is grown
https://www.usgs.gov/media/images/location-maximum-land-subs...
Peak irrelevance.
And the same is true for global carbon-related climate change, for which the real bill will become increasingly due over the next 10 years.
We are all insanely fucked and our response is still just to mostly to ignore it and hope it goes away.
Correction. It’s the next generations who are insanely fucked. People don’t hope it goes away, they think they won’t be around to see the consequences.
Ignorance and stupidity are the greatest threats to human survival - but perhaps it's for the best.
For the most part, I like to forward evolve solutions so I happily don’t care if Las Vegas runs out of water or whatever. I’ll find a way later. So I take responsibility for my bit. I just don’t care about this particular negative outcome.
https://uttoncenter.unm.edu/resources/2026-06-1-update-paper...
https://www.organicproducenetwork.com/organic-growers/califo...
I suspect they’re more concerned about “rising electricity cost[s] and air-noise pollution.”
Farms could double their water usage while datacenters halve theirs; the contours of protest wouldn’t materially change.
If you want to create a political coalition today, and have an AI datacenter being debated near you, rally against it. Like, try to stop it. But if you do doesn’t matter. The coalition is what matters. The datacenter—like the water—isn’t actually what anyone is angry about.
Being more worried about data centers than agricultural aquifer use is a damning indictment of the poor in this country, right up with the surprisingly bi-partisan nature of anti-vaxx thinking (https://www.politico.com/news/2026/04/14/poll-rfk-maha-vacci...)
And yes, despite your industry talking points, water waste is absolutely a major concern. Unlike residential use, where 80% of the water is reclaimed, open loop data centers (which are over 90% of data centers) allow almost all the water to evaporate. Similarly, about 30-40% of water used in agriculture filter back down to the water table. Agriculture use is also dispersed over large areas whereas data centers draw from a single pipe. Closed loop systems also have their own environmental concerns and need to "bleed the lines" every month or so to get rid of the toxic sludge that builds up. They also use dramatically more electricity which requires a lot of water to generate.
The point being, if you live somewhere where a new data center is getting built, you WILL see a massive spike in your electricity bill. What do you get in return? Sam Altman buying another politician? Another elementary school bombed? I agree that agricultural waste is out of hand but at least people get a tangible benefit out of it.
You keep trying to ignore all the points and change it to a topic you think you can win but you're still misinformed on that issue.
https://www.bloomberg.com/graphics/2025-ai-impacts-data-cent...
> alfalfa is not required for life
That part is actually the weakest part of your argument. You are dismissing a valid counterargument before addressing it.
Something being part of a current food production system does not make it a biological necessity. People can live without meat, therefore alfalfa is not "required for life" in the literal sense.
You are also shifting the meaning of "required": from "necessary for humans to survive" to "necessary for maintaining our current livestock industry." Those are very different claims.
Preemptively calling an obvious counterargument "dumb" is not a way to win the argument; it is a way to avoid having it.
Is it rational to dedicate enormous amounts of land, water, and energy to producing animal feed when those same resources could feed more people directly?
Using crops as feed and then eating the animals is an inherently inefficient way of converting resources into food. That may be acceptable when resources are abundant, but it becomes increasingly difficult to justify as we face limits in freshwater, arable land, and ecological capacity. That is the true argument, and what the discussion should be about, not if we can (very obviously) survive without meat.
South Korea imports alfalfa because of the KORUS FTA removes all duties on forage goods.
China buys alfalfa because of yet another trade deal that was brokered to prevent a trade war from spiraling. Alfalfa hay was explicitly highlighted in the Phase One agreement.
However, as I understand it, the alfalfa is a soil restorer that the agribusinesses are going to grow anyway. The question is whether there is any benefit to forcing the agribusinesses plow it under vs exporting it.
> alfalfa exports from water-scarce US states
* Vote for fucks sake. Bare minimum and stop making fucking excuses both for yourself not doing it and for politicians you otherwise support burying their heads in the sand.
* Stop buying stupid shit on the internet. Every package you don’t get delivered means less CO2.
* Fix shit instead of replacing it. The greenest EV on earth is more CO2 intense than the car you’ve already got (assuming you keep it in good working order)
Like yeah global policy needs to get better across the board, and the burden has been unfairly and ridiculously shifted to consumers. True. But at the same time consumers could buy less shit overall, still have very comfortable existences, and promote economic degrowth a little bit every day.
Oh yeah and stop using fucking AI for anything besides code because it sucks at everything else, and only use it for code you can’t otherwise complete.
>Fix shit instead of replacing it. The greenest EV on earth is more CO2 intense than the car you’ve already got (assuming you keep it in good working order)
This is not correct at all...
https://climate.mit.edu/ask-mit/are-electric-vehicles-defini... one source of many. Also local pollution is reduced, e.g. nox and particles.
Edit- To be clear maybe not specifically in this case, but "oops we've ruined this ecosystem to the point we can't farm it" rarely ends with "better move down the road and let nature reclaim it". It's a lot more "cool now we can put strip malls/data centers/whatever here"
It takes 600 to 3,000 years for 6 inches of topsoil to develop naturally, as it generally requires 100 to 500 years to form just a single inch. Underground water aquifers operate on similar timescales, if they can regenerate at all. The land needs time to recover from the demands of present timeline position humans.
https://en.wikipedia.org/wiki/Agrivoltaics
https://www.sciencedirect.com/science/article/abs/pii/S03014...
https://cleantechnica.com/2026/01/03/21-gigawatts-of-solar-f...
https://www.goldenstatecleanenergy.com/valley-clean-infrastr...
Not preceding "using better" with "consuming less" (on a system-wide basis) would get you the same result as going for recycling while forgoing the first R: empty, marketing points in green-coloured font.
As my links demonstrate, California plans on developing 21GW of solar on farmland that no longer has water available for it.