For once, there’s an article worth reading at National Review, and for once, I find myself somewhat in agreement with the rabid Leftoids (albeit for different reasons).
[T]here’s a consistent and surprisingly effective effort to convince you that the biggest threat to your community is the plans for a new AI data center on the other side of town. Read on.
Democrats’ Data Center Obsession
Back in 2024, I observed that when some of America’s biggest tech companies realized that they needed significantly more electrical power to run their data centers in the decades to come, they decided that restarting decommissioned nuclear plants was the best, most cost-effective, and most reliable option. And with the seeming snap of their fingers, a slew of those closed nuclear plants were scheduled to start operating again in the coming years.
And it wasn’t just Republican governors like Glenn Youngkin of Virginia eager to re-embrace nuclear power; Democrats like Pennsylvania Governor Josh Shapiro, Washington Governor Jay Inslee, Michigan Governor Gretchen Whitmer, and Virginia Senators Mark Warner and Tim Kaine all jumped on board. It was a case of the right policy finally being enacted after decades of foot-dragging and fearmongering, but more than a little frustrating that years of conservatives winning the policy argument and being right on the facts didn’t move the needle on the issue; it was Microsoft, Amazon, Google, and other big companies simply saying, “We want this.”
We should have known that eventually the progressive wing of the Democratic Party would wake up and galvanize opposition; now an increasingly loud swath of Americans, mostly on the left, seem to hate data centers the way they used to hate your SUV, your Big Mac, and, well, you.
Of course, the reason the Watermelons are being stirred to violence is because electricity is eeeevil, as is nuuuuclear powerrrr etc. etc.
I don’t care about any of that.
What concerns me about A.I. is more of a philosophical nature because while I can see many benefits of having computing power save humans a lot of grunt work and so on, I am profoundly disturbed by the implications of letting A.I. run things — and more especially, run the activities and affairs of humans. As long as it’s a tool, therefore, I think I can get behind it; but as a management system, I remain deeply skeptical.
And my skepticism stems from two sources.
Firstly, I think it’s all too easy (through laziness or indifference) to hand over the reins to outside control — we just have to see how cars are being thus transformed as an example — and as far as I’m concerned, the jury is still out (way far out) on whether this is a good, bad or evil thing.
My second concern stems from the basic premise of A.I., as I’ve said before, in that the collective [sic] wisdom can form a secure foundation for intelligence. As someone who has often used and manipulated data myself, I am intimately familiar with how this process can be affected by, let’s call it malevolent forces. And whereas in the past one could rely on some kind of human element to be a firewall on this issue, we are now faced with the prospect of A.I.-driven bots to not only speed up the process massively, but also to conceal what’s actually going on.
I’m not going to do anything stupid like bomb some data center, of course, nor would I ever support the assholes that do this kind of thing. If they do something vile like this, or even plan to do something like this and get caught, then by all means hang them, bury them under a prison or stick them in some deep dark jail cell forever.
I do think that we aren’t being careful enough with the drive to A.I., because the guys who are building it are obsessed with performance / generation. As with all science, though, we need to continuously ask ourselves the question: “Just because we can, are we sure that we should?”
And I see very few people asking that question of A.I. — which means that the field of resistance is being left open for the loony Leftoids.
I read somewhere that technological advancement in the past typically outsourced the doing of an activity. Some sort of machinery was built to decrease the amount of human labor needed to accomplish a goal. AI is outsourcing the thinking part that up until now was the exclusive domain of humans. If I didn’t know better, I would have thought that AI took over architecture design fifty or sixty years ago.
It just may be that we’re about to find out why we’ve never yet seen evidence of trans-stellar civilization.
.
As someone who’s worked in large industrial plants, I can firmly say that trying to restart decommissioned plants is about the worst fucking job you can imagine. And the longer they’ve been down, the worse it is. So trying to restart a decommissioned nuclear plant is a no-go from my perspective. Far better to just build new.
To start with, the decommissioning process is generally done on the cheap. And lots of plans on paper get the old pencil-whipped checkmark without even being done in the field. Lines are left full of material that sets up or simple eats out the pipes. Large motors flat spot the bearings and need complete rebuilds, generally costing more than simply buying new motors. Wildlife moves in (no, really, it happens) and you’ll find all sorts of shit (again, literally) everywhere. Costly and dangerous mistakes are hidden everywhere and many won’t be found until you actually try to run the equipment.
So again, better to demo the entire site and start with a flat concrete slab.
I gotta agree with you, and the other thing I’ve seen is that the old plant’s documentation is fragmented or gone, specialized maintenance tooling and fixtures long gone, and the people with institutional knowledge born of years on the job has gone elsewhere or died.
The thing about AI which is being missed is just how expensive it is, and it will not be getting cheaper anytime soon. So like any automation activities, there will be a cost/benefits analysis that goes on (not so much now while it is the shiny new thing) and many companies will find that its economic utility is less than they thought originally.
By this, I am not against it. It is a great tool, but a hammer is a great tool too, but if you use it for everything you just break a lot of stuff.
> And I see very few people asking that question of A.I.
Lots of people are asking that question.
But there are several problems:
1. There are many uses of AI that are simply too good to ignore.
2. Every country with enough resources is doing AI research *at the government level*. Yeah, it’s an arms race, and neither Russia nor China care about people.
3. There’s a LOT of profit to be made in firing workers.
4. There’s a LOT of profit in making entertainment without the “talent” to suck up all the money.
https://quillette.com/2026/04/15/move-37-and-the-coming-mindhack-claude-mythos-anthropic-ai/
That was stunning! Thanks for the link.
I think we’re going to look back at AI as a colossal mistake, on sort of a “Planet of the Apes” type scale. Spare me the luddite “Its coming, that’s the way the world is” type comments.
That being said, the ONLY spin off benefit of it I can see is the re-introduction of nuclear power here in the States. Of course that new capacity will be swallowed up by the AI datacenters and your power bill will not get cheaper, but it is something.
Every time someone tells me that AI should be put in charge of something, I just reply: “I’m sorry, Dave. I’m afraid I can’t do that.”
Did none of these people ever watch or read SciFi?
I think you are underestimating the extent to which AI is already, um, running things.
Your old statistical analysis industry is a good example. LLMs will be doing all the grunt work involved in data gathering and synthesis. ChatGPT will be “authoring” all the reports, the analysis and extrapolations that go off to the client.
Three years ago I sat down the CEO of one of the largest tertiary institutions in NZ and tried to explain to him not just the extent to which students were using AI to write their papers, but the extent to which lecturers were using AI to grade papers and even write reports that were submitted to him. I showed him several examples such as capex business cases, participation and retention analysis etc. this was a very astute, grounded and hugely experienced leader.
He said “Tony, I know you are exaggerating this to get my attention”. That afternoon I was dragged into a CEO direct reports meeting. The boss looked round the room and asked who was actively using AI in their departments. Slowly, they all put their hands up. The consensus, was quicker, easier, cheaper, more for less. The boss was horrified. Speechless. He genuinely could not believe that a technology (ghost IT) had not just bypassed all their procurement controls but become so widely entrenched.
He was utterly dismayed when next I showed him that none of the leadership really understood the technology, the security implications or the tendency of the models to hallucinate or suffer from group think.
That was three years ago. ChatGPT was just becoming widely available. We assessed it that using standard technology maturity models back then it was a 1. Last year I re-ran that model and we now have it at 1.5 (still a 1). For reference, we put desktop computing at 3, which means the organisation could not function if they were all turned off.
The truly alarming thing is it too the internet and email 8 and 10 years respectively to get to level 3. We forecast that AI will be at level 3 before 2030. Which means that it is deploying way, way beyond our rate to effectively control or manage it.
We are all in already.