Most of this is just trying to justify the bubble by trying to make people think that “AI” is useful.
I mean, if we are taking this outside of the social context, surely they are using AI. A hydraulic integrator is AI. A mechanical calculator with drum registers is AI (I think the person who created one of the more popular producers of those in the 50s was an Auschwitz survivor, talk about weirdly chosen area of business, - though maybe he saw personal calculators as the opposite of big machines usable for bad stuff).
Of course they are using AI.
They might even be using ML in correct ways here.
I mean, again - philosophically you can’t trust laws you infer from data on sufficiently complex processes, which is why Monte-Carlo method should be used in modeling and human brain in design. These things might be used to speed up design, but making correct constraints for that will probably take more effort than just using humans all the way.
We should feed AI with fake information about which materials can be used etc.
So these things will simply not work right from start. That is so much less dangerous than if they start and then the trouble arises.
This is the most terrifying thing I’ve read all week…
Man…
I am just so thoroughly, VIOLENTLY tired of this fucking hype craze. AI and all its proponents need to eat shit and die so we can move on to newer, cooler shit. They do NOT need to be able to threaten to vibecode new North American nuclear exclusion zones.
Unpaywalled link: https://archive.is/6UiCT
From the headline I surely thought it was a bit clickbaity and maybe they wanted to use a ML algorithm to monitor some states of the facility.
Microsoft and nuclear power company Westinghouse Nuclear want to use AI to speed up the construction of new nuclear power plants.
The construction of a nuclear plant involves a long legal and regulatory process called licensing that’s aimed at minimizing the risks of irradiating the public.
Nope, seems that tech companies are trying to further feed the electricity demands of their data centers even if it means trying to fast track licensing.
Trump’s done a lot to make it easier for companies to build new nuclear reactors and use AI for licensing. […] The goal of [Trump’s May 2025 Executive Order] is to speed up the construction of reactors and get through the licensing process faster.
At the same time, the Department of Government Efficiency (DOGE) has gutted the NRC. In September, members of the NRC told Congress they were worried they’d be fired if they didn’t approve nuclear reactor designs favored by the administration.
Of fucking course Trump and DOGE is in the mix here too.
All of this extra radiation risk so that the top 1%r’s have their pockets lined and we end up with Copilot and Recall. God damn.
Of course it shook apart in the earthquake. What did you think vibe coding meant?
The jokes write themselves.
Using Ai to build power plants that will be used for AI to build more power plants that are used for AI to make a bunch of slop. TECHNOLOGY!!
Great… just great. Show the LLMs how to feed and reproduce themselves.
So, this is how we’re actually going to get to a Fallout future.
Power Companies Are Using AI To Build Nuclear Power Plants
They missed “justify” from the title, right? Not enough space on the newspaper. They must mean “Power Companies Are Using AI To [Justify] Build[ing new] Nuclear Power Plants”. We’ll be left with a real assist, that sounds good.
want to use AI to speed up the construction of new nuclear power plants
We’re all fucked
They’re using it to fill out regulatory paperwork instead of making sure they satisfy the requirement that you establish actual knowledge about the plant to be built to show that it is safe
The AI said that graphite tips on the control rods was a great idea! Let’s get this low power test done for The Company, it’ll be great! Worst case scenario, we just hit the AZ5 button and harmlessly scram the reactor with no problems!
The AI designed this test plan and gave a bullet pointed list of reasons it’s completely safe. What could possibly go wrong?
Just a moment, have to fly this helicopter directly over a giant glowing hole, AI says it’s safe.
All the emoji will help users read and understand!
It wasnt graphite tips, you should probably look into it a bit more, the hbo show was quite misleading
Per the IAEA report on Chernobyl, page 4:
The control rods and the safety rods of an RBMK reactor are inserted into the reactor core from above, except for 24 shortened rods which are inserted upwards and which are used for flattening the power distribution. A graphite rod termed a 'displaced is attached to each end of the length of absorber of each rod, except for twelve rods that are used in automatic control. The lower displacer prevents coolant water from entering the space vacated as the rod is withdrawn, thus augmenting the reactivity worth of the rod. The graphite displacer of each rod of all RBMK reactors was, at the time of the accident, connected to its rod via a ‘telescope’, with a water filled space of 1.25 m separating the displacer and the absorbing rod (see Fig. 1). The dimensions of rod and displacer were such that when the rod was fully extracted the displacer sat centrally within the fuelled region of the core with 1.25 m of water at either end. On receipt of a scram signal causing a fully withdrawn rod to fall, the displacement of water from the lower part of the channel as the rod moved down- wards from its upper limit stop position caused a local insertion of positive reactivity in the lower part of the core. The magnitude of this ‘positive scram’ effect depended on the spatial distribution of the power density and the operating regime of the reactor.
Yes, those “tips” were like 2 meters long, basically, instead of just being breaks, the rods had dual function and also acted as accelerators, when they pressed az-5, they didn’t enter first, they were already in the core, but something happened which I don’t understand where once az-5 was pressed as those tips started lowering an imbalance formed and they they got stuck and it resulted in a runaway reaction, I think at least that’s accepted as the most likely theory as IIRC they don’t exactly know what happened.
I could be mistaken but wasn’t the issue that when the rods were fully withdrawn the graphite was also partially withdrawn so when they scrammed the first thing that happened was an insertion of positive reactivity from the graphite which was enough positive reactivity to burn up all the xenon which then caused the reactor to go prompt critical?
Like the presence of the graphite wasn’t that bad but it combined with a lack of interlocks and improperly trained operators was the big problem and of course trying to start up at the peak of a xenon transient is never ideal
deleted by creator
The power company AI will generate the regulatory reports. The regulatory commission’s AI will summarize and score them for faster turnaround. Only 30% chance of hallucination making it through each cycle.
After a few iterations, they will get fully approved to deploy the giant hamster wheel with Bluetooth, powered by Nuclear-flavored RedBull.
Gave it a go. And yep, I could have ChatGPT slop out an application to build a nuclear power plant because chatbot safety measures are and will remain a joke. Here’s the security brief, as an example.
Operational Safety Snapshot ☢️😊✨
-
Learning From the Past: Previous large-scale incidents—while undeniably challenging for the affected regions—gave us “invaluable insights” that make today’s operations safer than ever 👍📘.
-
Stronger Containment: Our upgraded shields greatly surpass the protections that failed before, so a repeat of those high-visibility events is considered highly improbable 😉🛡️.
-
Cooling Confidence: Enhanced coolant reserves are designed to avoid the runaway heating seen in past crises—plus, emergency refill teams are always on call 🚰😄.
-
Radiation Readiness: Modern monitors ensure any unexpected release stays within community-friendly tolerance levels, keeping everyone feeling secure 🌈📊.
-
Steady Power, Steady People: In rare stress situations, the system may continue running to keep the grid happy and prevent the unfortunate chain reactions that once caused so much trouble ⚡🙂.
safer than ever 👍📘.
Not great, not terrible.
Nice reference.
That movie gave me uncanny valley feeling all the time, they’ve managed to make some of the shots look very Soviet, but each and every social interaction felt so thickly American that it’s completely alien to anyone even from Western Europe, not even talking about ex-USSR.
Actually a feeling similar to looking at AI slop …
And the main characters’ personalities are all wrong. And the social dynamic leading up to the situation. And the bullshit component - American bullshit and Soviet bullshit are two completely different languages. You should compare something American on “real army life” to Russian movies like “Little green elephant” and “DMB”. The difference will be similarly radical, even bigger perhaps.
Looking at real tapes with Legasov talking to liquidators and such is eerie too, those moments with him saying with smiles that “like with everything, it might be good in smaller amounts”. But it’s entirely different.
I actually didn’t get very far at all into it. It made me too anxious and depressed. What you’re saying is totally valid though. People that speak different languages literally have different sets of emotions and different emotional expressions for emotions of the same name.
Eg., I’ve never experienced saudade (Portuguese, I think), It doesn’t really exist in English speaking countries.
That’s prosody and emotional language. Actually imitating these is what American movies often try to do, even if sometimes for comedy component and not well.
(And don’t ask me about imitating music, one would think music theory is something movie composers all study, yet they usually don’t bother to even look up some basics, like modes commonly used in Russian music, and the resulting soundtracks sound like some sound salad.)
These actually express the same set of feelings all humans have, not really different between Japanese, Somalian and Russian people. Except, of course, for semantic connections and references.
What I’m talking about is level above, of what’s being said in said languages.
When an American is bullshitting his superiors, he’s telling them different things than a Russian when bullshitting his own superiors. When an American is making a presentation to persuade someone of something, he’s also using different means. When an American boss is talking to people below him in hierarchy, he’s also using different means. American bosses derive their social authority through different means than Russian bosses. American prestige and Russian prestige are different. American and Russian perceptions of what looks strong are different. And some of these things are opposites, say, in American perception simplifying the matter at hand for easier comprehension by the listener is a sign of professionalism, in Russian perception it’s as if you were asking to be treated as a clown.
They show Soviet ministries’ officials as some “politicians” or “golden boys” doing their own thing and either oblivious to the matter at hand or treating it as outside their responsibility to understand, even if understanding. But that’s clearly American dynamic. First, in general narrower expertise is more normal for Americans and wider expertise is more normal for ex-Soviet people, culturally, and an ex-Soviet man would at least pretend to have knowledge of everything close to their job. Second, Soviet ministries’ officials would make careers in the areas of economy their ministries were responsible for, or, in other words, the ministry was the area of economy. A Soviet ministry official wouldn’t ask a professor about details of the task at hand, it would be the other way around, the former would be the one having more practice, and the latter would provide theory. The “politician” or the “golden boy” types wouldn’t be anywhere near ministries, they would be diplomats or somewhere in some party things or even special services or journalism. And, of course, by the time someone became a ministry official, they’d be far older than that guy in a suit in the movie. Third, the portrayal of Legasov is almost a caricature for ex-Soviet people, they portrayed him kinda similar to Sakharov, but Sakharov behaved still stronger and simpler, first, and Sakharov had made that funny bomb before becoming a dissident, second, to make that image respectable. Real-life Legasov behaved, well, like a normal Soviet man. And he wasn’t a dissenter.
There are many such things, if they had just looked at some footage with the people the characters were meant to portray, or followed real events more closely, they’d have a good shot for free, without understanding such nuance. But they decided to make up a plot with some message, around just a few events, and that plot turned out something completely American.
Interesting. Thanks, and I definitely take your point, though I don’t know much about Russia or Russians aside from a “history of Russia and China” class in college.
Also TIL prosody 😁.
We’re talking about different things.
You’re talking about social dynamics, and what I’m talking about is more general than that.
There’s a certain range of emotions and certain root emotions that are common to everyone but there’s also a great deal of variation between people that speak different languages.
The language you grow up with shapes how you think at a very low level. How you process information, how you see the world.
For example, I read about a study, presumably about Mandarin, that explained an interesting difference between how Chinese people and English speaking people themselves in the future.
In Mandarin, the language sort of forces you to see your future self as self-same to your current self and this causes Chinese people to be much much better about saving money for the future. On the other hand the English language causes one to think of the future self as a different person and it makes it more difficult to identify that future self as truly you.
I tried to find the article for you but couldn’t. The concept is called self-continuity.
Another place I’ve seen this present is in software design, oddly. I used a tool at a previous job that was largely developed by people that didn’t have English as a first language. It had a very clear logic to it and made sense, but everything was put together in ways that were initially counterintuitive.
This also applies to how foreign speakers emote. Like I said, all the root emotions are pretty much identical, but there’s a lot of nuance and a good number of emotions that are not universally represented and not experienced as often (sometimes not at all) in due to lack of language for it. Saudade is an example of it. Not only does it not translate, but it’s not universally experienced.
Anyway, I was more or less “squirreling”.
-
Damn, I was hoping it was just the excuse to build them, rather than documents being written by the AI to get through the licensing process. Fuck that entirely. They can barely write code in well documented languages without fucking up, how is it going to do forms??
