“Paused over regulation and costs” is basically corporate for “we thought it’d be a pushover.” Honestly, the UK should just demand full transparency on data use as a condition for any new centres.
101
ThreeTreesForTheePls1 day ago
+22
Yeah I’m not sure the British government should be trusted with fully transparent access to the data passing through data centres.
22
psychicsword1 day ago
+3
More that competition is likely pricing them out just like everyone else.
3
quiplaam1 day ago
-22
The UK is famously difficult to build anything in. It's a big part of why their growth has been terrible over the last 20 years and the country's in political turmoil
-22
CalligrapherBig43821 day ago
+18
America, a country where it is quite easy to build things, famously has no political turmoil
18
quiplaam1 day ago
-11
It is also quite difficult to build things in the US, but nothing on the scale of the UK. Reform is currently the most popular party in the UK and the Greens the second most popular. Having two deeply unserious populist parties leading in the polls is a complete embarrassment, almost to the level of reflection trumo
-11
pipopipopipop1 day ago
+1
Which party is a serious party?
1
Suba591 day ago
+19
Oh and the fact no one wants a product that will take our jobs, further the wealth gap, give more power to the oligarchs while dishing out slop and ruining the internet by putting it behind a paywall after scraping all the content. But yeah those regulations are a bummer.
19
NaturalFrog21 day ago
+15
Finally some good f****** news
15
Malaix1 day ago
+18
lol I think a lot of us have been calling this. The bubble is going to pop with a lot of these data centers being mid construction because they are HUGE in a lot of cases.
And then what? Giant abandoned skeleton of a building with a bunch of construction companies sitting on materials and workers and no money or reason to go on.
18
QueenofCats281 day ago
+12
100% I wish they'd just f*** off. No one wants skeleton buildings, because you're right, that's what will happen!
12
yungimoto1 day ago
+4
Can we zone em for residential and convert to c**** housing?
4
dingo5961 day ago
During the dot com bubble companies were burning VC money to lay fiber optics across the US. A lot of that is still not being used, it's called dark fiber. While it seems like a waste it has helped massively in increasing internet speeds. The same could happen with these data centers. There is little difference between an AI data center and a regular one that your ISP uses or Netflix. Best case scenario is these data centers are built but never run AI models.
0
bodmcjones1 day ago
+2
Problem is, these aren't being built in a considered or sustainable manner by and large. If they were it'd be different, but there is a lot of half arsed decision making going on atm. We need to do things right rather than quickly imho.
2
ppooooooooopp1 day ago
-8
They are huge because they need to be... That said it's possible with the progression of technology and the improvements in model effeciency that they don't need the capacity. As pointed out by the article, having domestic compute is a national security consideration, so this is not a good thing.
OpenAI and Anthropic have a combined ARR of more then 40 billion dollars and that rate of growth is more then 100% y/y. You can swear it's a bubble all you want, but you will be doing so to your own detriment.
The appropriate non-stupid approach to AI is to start thinking about how we redistribute wealth, not to pretend like it isn't happening. Time to start living in reality.
-8
Malaix1 day ago
+4
> OpenAI and Anthropic have a combined ARR of more then 40 billion dollars and that rate of growth is more then 100% y/y. You can swear it's a bubble all you want, but you will be doing so to your own detriment.
Its absolutely a bubble. AI companies are burning through billions with few reliable models of profit worth the operation costs. They have been making promises, failing deadlines, shifting around projects, and doing weird financial shit like investing in each other to fake interest and movement and keep the bubble inflated.
The AI rollouts have all been deeply flawed or backfired to boot. People generally hate the tech now to varying degrees depending on where its being used.
Even if AI has uses the plain fact is that its way less than they hoped and promised and the industry is due for a big scale down imo unless there's some miracle.
4
ppooooooooopp1 day ago
-5
Do you follow any of this stuff? Or are you just making it up?
Do you even know how they are making that forty billion? It's not through idiots on the internet, it's through enterprises who are willing to pay hand over fist for the productivity gains - that is as reliable of a revenue stream as any else.
Anthropic TRIPPLED their revenue in the last 3 months (to 30 billion ARR).
AI is in the process of completely upending the software industry - AI can do what would take a team of software engineers multiple weeks in a few hours. Talk to any software engineer who is even remotely paying attention and they will tell you the same thing. It's coming. It's going to change the entire industry.
It's not just software engineers, it's data scientists, data engineers, product managers, operations people. Again you can stick your head in the sand but this isn't speculative, it's happening right now, and you are in denial.
AI now is NOT less then what was hoped, arguably it's exceeding expectations. Of course AGI is not feta comply, but it's 100% a risk worth considering.
If you feel like actually knowing something about AI you should read Anthopics most recent red team post: https://red.anthropic.com/2026/mythos-preview/
If their experience can be replicated (should be taken with a grain of salt) the near term implications are dire.
-5
fading_reality1 day ago
+1
\>If you feel like actually knowing something about AI
I know a bit. Define AI.
1
Malaix1 day ago
+1
I have been. It has all the signs of a bubble.
Overhyped product being rushed out.
Inaccurate promises and deadlines.
lackluster performance with some studies showing AI introduction slowing programming.
AI art being hated for being slop, rife with plagiarism, and child p***
A horizon full of lawsuits related to the above
extreme costs of operation running deep in the red by billions
Sora ending
Anthropic getting blacklisted by the military over insane demands
that report that half of the data centers that were planned are canceled or postponed
growing resistance to installing data centers
the fact our electrical grid can't handle the demand... During an energy crisis... While people struggle to pay the bills as is.
the weird finance investment games they are playing
the extreme rampant growth despite not having a reliable profit base demonstrated in most areas.
1
ppooooooooopp1 day ago
+1
I think we could go around in circles forever on this topic - certainly the financing is eyebrow raising.
Many of your point are totally irrelevant to the core claim which is that this is a bubble. Prior to last November I would've agree'd with you. I cannot emphasize this enough:
> lackluster performance with some studies showing AI introduction slowing programming.
This is complete and total bullshit. This is the precise data point that changed my mind. AI is not just better at programming, it's changed the game. Yes there are certain tasks that it is worse at - but it turns out most people don't work on novel problems and it's really good at the stuff most people are paid to do.
I will just say this: if you truly believe you are correct, you should put your money where your mouth is. You should bet against companies like NVidia, and you should invest in companies harmed by AI speculation like Intuit or Adobe.
That's it. Talk is c****. Do some *actual* research, form an informed opinion and put your money where your mouth is.
1
fading_reality1 day ago
+3
\>OpenAI and Anthropic have a combined ARR of more then 40 billion dollars and that rate of growth is more then 100% y/y. You can swear it's a bubble all you want, but you will be doing so to your own detriment.
Could you tell us their profit?
3
ppooooooooopp1 day ago
-4
They are losing money at a rapid pace, last year 9 billion (just openai), this year it will be more. This is a bet on future growth, and right now that bet is coming good.
It becomes a bubble when revenue growth fails to keep up with projections and the spend continues. That is not the world we live in at the moment.
-4
TrulyKnown1 day ago
+4
> The appropriate non-stupid approach to AI is to start thinking about how we redistribute wealth, not to pretend like it isn't happening. Time to start living in reality.
If it's time to start living in reality, why don't you do that, instead of regurgitating sales pitches on the future benefits of snake oil, from the snake oil salesmen?
The appropriate non-stupid approach is to look at what has actually been accomplished. We're not in "early days", and haven't been for a long time, so don't bring that line of bullshit. AI has had all the resources in the world, and years of time thrown at it, and it's *still* making up bullshit (The thing boosters call "hallucinations" in an attempt at anthromorphizing a glorified algorithm).
The only people who still think AI "is coming" are either deeply deluded, don't understand the technology, deeply incompetent, stand to benefit from pretending that it's definitely still the future, or, as is the case for some people like Altman, all four of them.
Every time we've seen these companies crow about how amazing their latest invention was, and how it was going to change the world, once it gets released to the public and gets tested, it invariably results in a big, resounding flop. At what point are you going to realize that they are *lying to you*, and stop thinking that their advertising pitches are in any way going to be congruent with reality? Do you also watch ads for Red Bull and go "Damn, wings, this is gonna change how cities are built!"? Stop being such a mark, and realize that the reason the revolution is always just around the corner is because it's not actually coming, it's a confidence game that greedy corporate slimeballs are playing for all of our money. Or go join your local branch of Scientology. I'm sure they're looking for people just like you.
4
ppooooooooopp1 day ago
Here's a suggestion - if you believe so strongly that AI is full of shit, put your money where your mouth is.
The market has seen significant shifts based speculation around AI. You should bet against it - take a long term short position on NVidia, the market cant stay irrational forever.
Or just buy stocks that have been harmed by AI - buy intuit, or adobe companies that will see margin compression and whose stocks are performing terribly.
If you are serious about your view, why wouldn't you? Surely that's free money right?
0
ppooooooooopp1 day ago
I work as a software engineer. Nothing is being sold to me, AI writes better code, faster, and that code is more reliable. It writes architecture and design docs better and faster. It does data analysis faster more thoroughly. That's it. It does the work of entire teams in a fraction of the time.
Companies are investing more, not less in AI, because they see those returns in their productivity. 100% year over year growth is what? Fake? You think it's hype? Companies don't double down on investments that aren't working.
There is no ambiguity for *anyone* who is even remotely paying attention.
You OBVIOUSLY haven't seen what it's capable of, how can you tell me it's snake oil? Give me any EVIDENCE that it's not capable of the things i've stated - I see those things *every day* my co-workers see those things *every day*. You are in for a rude awakening, or maybe you'll just stay in denial.
It's time people on this site stop being willfully ignorant. You think of me as a mark? I think of you as an idiot.
0
V2UgYXJlIG5vdCBJ1 day ago
+1
Just don’t look at their operating costs. 😂
1
Shoot_from_the_Quip1 day ago
+2
Or was it the loss of Qatar's 30% of the global helium supply in recent attacks, meaning chips are going to be few and far between?
30 Comments