Been on a train of thought recently that with AI making web dev so much easier that we should be making more beautiful experiences.
I have always felt ultra minimalism is partly a way of fighting bugs. Why can't we put more effort to make every site a great experience in of itself?
Been experimenting with going way beyond my usual frontend work for something truly special. This made me think if it. It's 10x the effort but that's about the same effort I put in before manually making a minimal one.
Read somewhere that SF spends roughly 50k$-80k$ per homeless person per year.
Taxing more doesn't solve a massively inefficient system at it's core. Just like US education, we spend more than any country on earth, why is it still bad?
Answering that question with a "if only we had more money" is a really poor argument. The CA tax fundamentals are bad, pooring more cash onto the fire will not fix that.
And, many people who are poor have persistent mental/physical disabilities, so part of that spending is because many of these people have it the hardest.
With that said, we could likely fix all of these things and significantly unequal wealth distribution would still result in a lot of poverty.
96k$?! Those are rookie numbers. I propose a mclaren for every homeless person. We will weath tax all stocks in the fortune 500 to pay for it, crash the stocks and solve inequality.
You can't solve a shortage with demand subsidies. You need to expand supply. SF has too little housing relative to its population, and has perennially tried throwing money at the nonprofit industrial complex which has (obviously) perennially failed to solve the problem because it doesn't generate new housing units, it just bids up and reshuffles the existing ones (and steals a lot of money in the process).
> The difference now is that AI can make low effort work look high effort at a surface level.
True, but we'll also get higher standards I think. It's become easier to do things, so skilled people can do more complicated things. We compare to what humans can achieve.
"Given infinite thinking time a finite number of humans will solve all theorems"
I also love the angle that this was not intelligence just brute force. As if the mathematicians didn't reeaaally want to solve this they were just too lazy to give it a good try.
What does AI have to actually do before you realize these things are actually smart?
Machines have a much higher capacity for work than human beings. Saying that these proofs did not require equivelant intelligence, but benefitted from sheer volume, does not strike me as unreasonable.
Yes my only point here is that you can argue about the semantics of how smart they really are but they are undeniably smart.
Today it cost massive effort but it's possible 10-20yrs from now an AI could solve a problem like this in under an hour with a single thread on a free subscription paid for by serving an ad.
These arguments are so weak because you'll then have to make the same one a few years from now when it does something else impossible. The argument only stands if we assume no progress will occur.
Funny that you imagine a future where AI can solve complex math quickly, but humans are still watching ads, for some reason.
I just think you and the other guy have different definitions of "smart". There's no denying that LLMs are useful, but I don't know if I'd classify them as "smart". There were probably people in the 80s saying computers were "smart" because they could compute 78971 * 12341 faster than a human.
In what sense is a LLM "undeniably smart" but a CPU from the 80s isn't? Or would you define such a CPU as "smart"?
ASI may kill us all 100yrs from now but ads are forever my friend.
You're right though they largely are "smart" in the 80's computer sense. This is largely due to continual learning being unsolved.
BUT the more you look at them, research, and try experiments there's something there not in a 80s computer. If I had to guess maybe 1-5% of a humans ability but it's there. They are able to do novel things but ever step outside of their distribution takes exponential effort for every small addition. There is a true ability to adapt and learn new things on the fly, things never seen before. That is the the smart part. There something hidden in these things we don't understand that allows novel insights built from in context learning.
It's actually measurable in experimental settings but even there it's hard to tease out. I saw it mostly while doing CL training experiments. But I also see it while working with them for coding novel things.
But the more power we provide and farther down the road of this we go those 1-5% are things like solving unsolved math problems. No human solved these things. You say brute force, I say it needed massive effort to break out of it's distribution and get those small insights. It's very human like when taken at scale. The scary thing is that scale is getting smaller every day.
It feels goalpost-movey to downplay exploring a large search space efficiently in regards to "intelligence". If we dug into a human genius's brain and found it was somehow trying out a million ways to solve a a problem at once, no one would seriously suggest the person isn't actually intelligent.
And our brains must something like that at some physical level. You can't have a "turtles all the way down" of reasoning - the building blocks must be simpler. It must reduce to something like pathfinding and brute force at some point, weighted by factors in the system and maybe some randomness.
We have a romantic view of intelligence, perhaps stemming from intuition within the context of scientific discovery. Given enough intelligence, and enough context, a brilliant person can have a stroke of inspiration that allows them to make a major leap (a-la General Relativity or Fermats last theorem). We haven't seen THAT same capacity from a machine, but we see the more ordinary, unsexy grinding type of progress that represents 99.9% of scientific reality.
I would find it very interesting to train a model on information only available prior to the discovery of e.g. relativity or calculus and see if it can invent it. My intuition is that modern frontiers absolutely could. Not to take away from their brilliance, but Newton and Einstein were brilliant people who also happened to be in the perfect place at the perfect time - there's not so much "low hanging (i.e. approachable by one brilliant individual) but immensely valuable fruit" anymore.
Is there truly anything new under the sun? Hasn't all of existence alway been here? All math, all physics? We could have merely discovered it. Intuition might be nothing more than combinations of what already exists rather than some sort of divine insight that unlocks previously unknowable mysteries.
Agreed. I don't believe intuition and creativity would be more than pattern recognition, remixing ideas, and trial and error combined with a kind of "genetic algorithm" approach if you deconstructed them into what the brain is actually doing.
There certainly are such proofs. Even for simple decidable theories we have very large lower bounds on decision complexity (like double exponential), which implies large lower bounds on the function from "length of theorem statement" to "length of shortest proof".
For undecidable theories, there is no computable function bounding this blowup from theorem length to proof length (otherwise, the theory would be decidable.)
Nobody can avoid it. You can RL things in a box until the box is mapped perfectly but that doesn't mean you can go forever on out of distribution data. Our number of boxes will increase but model collapse is an open problem.
Astra goes farther than any model before but it cannot go forever.
I recently was building out on azure and oh boy, the joy of microsoft products.
Their troubleshoot support tool must be the same team that makes it on local windows troubleshooter because literally every single time you use it, it spins for 10min then says it can't fix anything.
> Google’s water use in The Dalles has nearly tripled in the past five years, and the company’s data centers now consume more than a quarter of all the water used in the city.
> That’s according to records released this week after the city settled a lawsuit against The Oregonian/OregonLive and agreed to hand over data on Google’s water consumption. The company’s water use is poised to continue soaring in the years ahead, as Google has plans for two more data centers along the Columbia River.
> Google is contemplating two new server farms on the site of a former aluminum smelter in The Dalles, where it already has an enormous campus of data centers on its property along the Columbia River.
> Google says it needs more water to cool its data centers, but neither the company nor the city will say how much more – only that The Dalles can’t meet Google’s needs without expanding its water system. The deal calls for Google to pay for the upgrade.
> Even so, the proposed water pact has attracted scrutiny and skepticism in The Dalles, a riverfront city of about 15,000 approximately 80 miles east of Portland.
> Residents and nearby farmers are concerned about the city’s long-term water supply amid an ongoing drought. They complain they don’t know enough about Google’s actual water use.
----
It is possible to do some cooling with closed cycle, though more than a few data centers have been built with open cycle cooling in places were the water table is precarious or in parts of the country under drought conditions.
Unless a data center construction discloses its water use and shows that it isn't using a lot, then people are assuming its using as much as it can (and up to the water rights or capacity that it has). ... Which isn't an unreasonable assumption to make.
> Google’s data centers gulped down nearly 550 million gallons of water in The Dalles in 2025. That’s nearly 40% of all the water consumed in the entire city and a huge jump from the year before.
> The city expects that local water consumption will soar over the next several years and bills will nearly double as The Dalles works to upgrade its aging water system. But the city won’t say how much of the rising demand comes from Google, citing a promise to keep the numbers secret and confidentiality provisions in Oregon public records law.
And data centers are being very cagy with their water consumption because the amount of water that it uses is related to the amount of compute it has and that is being held as an industry secret.
The opposition to a data center isn't "data center vs all of the state" but rather "data center vs a town where it will double the water use and the locals are currently under water restrictions and there was a drought last year."
Sure, let's do a more local comparison. Wasco County, where the Dalles is located:
Irrigation: 17.9B gallons/year
Fisheries: 9.7B gallons/year
So, the datacenter at the Dalles uses a shocking 2% of the amount of water used for irrigation alone in the locality.
There's a very simple solution to this, if you actually care about water: tax it to the point where usage is at whatever level you want, instead of focusing on a vanishing minority of the water usage.
> When Google purchased a defunct aluminum smelter in The Dalles, it also acquired the right to pump massive amounts of water from underground — up to 3.88 million gallons a day. That’s almost as much water as the entire city averaged per day in 2024.
> The corporation transferred those water rights to the city in 2021, Bentz noted, allowing The Dalles to pump more water from underground. The city only obtained part, not all, of the former smelter’s water rights, Mays said after OPB’s story published.
> But groundwater is limited — there is only so much available before it’s replenished by rain or snowmelt. Oregon is one of many states that have allocated more water rights than the earth can provide at once, causing groundwater wells for residents, farmers and other businesses to run dry — particularly in Eastern and Central Oregon.
---
The water rights that Google retained are not taxable without redoing the entirety of the water rights laws for multiple states.
Do water rights need to be revisited in today's climate conditions? Yes. We see it on the Colorado River... and on the Columbia River. But the law, as it stands today, if you've got the water rights that's not something that is taxable. And if you bought more senior water rights, you get to pump before someone else does.
Thanks for an educated reply. This is the first argument I have seen that isn't stupid.
So there's truth in the statement afterall. Clearly a powerful PR campaign going on because there is soooo many dumb takes this kind of argument vanishes.
I do think that amoung all our water uses this is easily the nicest amoung them. I think datacenters are the new factories and as far as factories go they are the cleanest ones ever made. But I suppose if you have the impression AI is useless then it makes sense. But if this is america's future industry we should make room. These centers are even funding infrastructure upgrades. To me they are the closest thing we'll get to reindustrialization in our lives and I would still want a stronger push for even more aggressive buildout.
Water use is one of the causes that can unite the various edges of political divides. Rural and urban. Conservative and liberal. Many parts of the country have been experiencing drought conditions. I had Oregon up there - here's the current conditions there.
I don't see it so much as a "PR campaign" against it, but rather it is a populist issue that has broad appeal being against technology companies that are seen as the bad guys.
The data center build outs are not doing themselves any favors by going cheap (evaporative cooling), noisy fans ( https://www.npr.org/2026/09/17/nx-s1-5954414/residents-file-... ) and then being secretive about the "how much is there" makes it look all the more nefarious and underhanded.
The claims of "its taking our water" or "its raising our electric bills" or "its taking our (kids) jobs" or "its stealing all our art" are all things that have some grain of truth to them and easy to rally behind one (or all) of those reasons. The choice of water as the topic is easily the most visible and the one that can get farmers and city folk to agree on.
Going to closed cycle cooling, funding the reuse of waste heat for winter district heating, and such could go to mitigate the public outrage about them... but those aren't the cheapest ways to build them.
Thanks, gave me something to think about for sure.
Your suggestion on this is mitigation then?
Take an aluminum or concrete plant. I see it as a hard argument to give to these companies that they are somehow special and therefore should do more to their already low impact "factory" If we compared these to large scale steel mills for example. I agree with your point that mitigation can help public opinion but on the opposite side I don't see how you can make a reasonable argument to the builders that this is actually worth funding?
> The AI industry is reaping what it sows, as far as Texas Gov. Greg Abbott is concerned.
> Abbott said the sharp criticism Americans are lobbing at AI data centers is well-earned during an interview on ABC's "This Week" on Sunday.
> "They basically dug their own grave for the problem that's been caused for them, and that's why they got the backlash they deserve," Abbott said.
> ...
> Abbott was initially an ardent supporter of AI data center projects in Texas for their economic benefit. The giant server warehouses can generate jobs during construction and provide tax windfalls to local communities. However, Texans from Abilene to Sulphur Springs have taken hard stances against them, forcing Abbott to change his tune.
> Abbott has introduced new standards for AI data center projects in Texas, including instructing the state's utility commission and electrical grid operator to protect residents from infrastructure costs. He's also suggested a ban on data center developments in rural Texas, and that projects must undergo an audit before moving forward.
Its a "this is going to hit construction" and if its done now trying to mitigate things or if it is going to be getting local city councils getting thrown out until there's one that can pull the permits... or a governor who needs to make a stance to get elected.
Those regulations requiring mitigation are going to happen and it's a "do it now as new construction or be required to do it in three years" as retrofits. New construction and making an effort to mitigate it will likely be the less expensive way over the next decade... but its not the cheapest way to build it today.
In a world of perfectly rational actors you could say "it's still better than the coal plant" but your point is true for the real world and makes much more sense.
Clearly it's the cost AI leaders will have to pay after spending years telling everyone they will die and lose their jobs. They wanted the publicity and with it comes what they sow.
I'm not against AI data centers in principle. And I do believe that the water argument is incorrect in some areas. However, I do understand the argument that can reasonably be made for data centers in some places.
I'm personally more concerned about grid stability, power infrastructure upgrade costs being passed on to existing rate payers before the data center turns on a light bulb, and the impact that direct power consumption has on alternating current quality.
> A power line went down outside of Washington, DC, this week. Normally, the grid would only need a few seconds to recover from such an event. But this one took more than 10 minutes because data centers stopped drawing over 3 gigawatts’ worth of power nearly simultaneously.
> The event caused voltage across the PJM grid to spike from Northern Virginia to Chicago, according to data collected by Ting Labs, a startup that runs an IoT sensor network out of people’s electrical sockets.
> The event didn’t cause a blackout, but it did cause lights across the region to flicker. The incident demonstrated the effect that data centers can have on the grid — an outcome that experts believe will become more frequent.
Also in the pic your provided the center literally sits on a river? Who cares? That means they likely are using their own water system meaning the people in the city are completely isolated.
> Therefore, the Legislative Assembly declares that the waters of the state may not be appropriated, stored or diverted for use outside the basin of origin except in compliance with the provisions of ORS 537.801 (Definitions) to 537.860 (Vested rights protected), including, if applicable, the prior approval of the Legislative Assembly under ORS 537.810 (Diversion or appropriation of waters from basin of origin without legislative consent prohibited).
Evaporative cooling of a data center would constitute a diversion of the water outside of the Columbia River Basin.
Furthermore, the river water rights aren't owned by the city (or the aluminum plant that Google bought). There are other treaties and regulations on the Columbia River to ensure that it has adequate supply for power and for fish.
Google is using ground water rights and pulling. Furthermore, even though there's a river there, the area is in drought conditions. The supply of water percolating from the river into the groundwater isn't sufficient for human needs in that area.
> About 70-80% of the water consumed annually by residents and businesses in The Dalles comes from Crow Creek Reservoir. Fed by South Fork Mill Creek, Crow Creek and a pipeline that diverts flows from the Dog River, it held 123 million gallons on Sept. 4, roughly seven million gallons less than the same time last year.
> While the situation isn’t dire, as Public Works Director Dale McCabe emphasized, it’s not normal. The city issued similar voluntary restrictions last summer after two of its now five groundwater wells went offline; where that was spurred by mechanical failures, however, the current advisory is a product of environmental factors.
> “This is fairly new,” McCabe said. “I’m not remembering a period where we went into curtailment for quite some time before this.”
---
Note that "it held 123 million gallons on Sept. 4, roughly seven million gallons less than the same time last year" in context of Google's use of water in 2025: "Google’s data centers gulped down nearly 550 million gallons of water in The Dalles in 2025."
Those are different numbers (current reservoir amount and total use) ... but its something that people will compare.
There was construction of some data center, and someone's well water started looking dirty around the same time. From there it metasisized into claims that it's destroying all the world's water.
Rather than a myth, it’s more of a meme in the sense of “naturally selected, self-replicating idea”.
Hits all the right buttons for the motivated reasoner: An eco-disaster that our brains aren’t climate-change numb to. A threat to the most precious fundamental substance in the service of a high-tech fad. An evocative symbol of corporate recklessness in the service of greed.
That it’s not really a big deal compared to, say, almond cultivation is no match for that brain-niche fit.
Evaporative cooling. They think that because the earth is only 2/3rds water we will run out before it rains back down again.
I jest, of course. The actual argument is something about most water being frozen and how some people live in deserts so we should all live as if we were desert dwellers. As you can tell I'm not very convinced by it.
Water is literally the least scarce thing on earth.
Ec2 + codex or claude and tmux so you can move about. Control EC2 permissions outside the node. Not really much value beyond that.
Only layer beyond this I want is the permission scoping, stronger sandboxing per session and centralized control. I have not seen a clean product around this though where I own the compute.
I've been developing my own sandboxing environment using Nvidia's Openshell.
Works like Claude code but with stronger guarantees for me on file system and network access. Still playing around with it, but so far I've been liking the setup.
I still run the sandbox inside a VM for now, but I feel far more comfortable in running Claude Code in unsupervised mode because I restrict the outbound network access and secrets never hit inside the sandbox.
Why not a virtual machine or container? I'm playing around myself with some but can't shake that a VM is the right shape for it. That's my infra brain though.
I have always felt ultra minimalism is partly a way of fighting bugs. Why can't we put more effort to make every site a great experience in of itself?
Been experimenting with going way beyond my usual frontend work for something truly special. This made me think if it. It's 10x the effort but that's about the same effort I put in before manually making a minimal one.
I want more koi fish swimming around my saas.
reply