A Theory on AI Data Centers

Back in 2005-2015, there was a fad called Cloud Computing, which was being pushed out from the top, by mighty industry behemoths. I covered it in 2009, and for a while, that essay was one of the most popular tech articles I ever wrote. One of the most visible mementos of that fad, which has still persisted to this day, are those ugly, cheap, Chromebooks. These are still around in North America because publicly funded schools are mandated to push these at children (at taxpayer expense), who do adopt them for a while…..but then outgrow them and settle for more powerful and more personalized laptops and tablets. Then there are apps like OneDrive® and iCloud®, which start syncing your data to their servers, unless you repeatedly slap them. Even worse are online password managers like the one pushed by Google®. But these relics are a far call from what was envisioned back then, to make personal computing data exclusively online.

In my 2009 article on Cloud Computing, I uncovered that the biggest hole in the conspiracy was that while it was seemingly pushed globally, it could only be somewhat achieved in countries with very stable electrical and Internet infrastructure, such as the United States. To quote,

Cloud Computing will attract clients mainly from Western markets

Cloud Computing implicitly assumes that the Internet is as robust throughout the world as it is in North America, Europe and some parts of Asia. But clients from countries where Internet connectivity is sporadic will be discouraged from boarding the Cloud. And no, I am not talking about Namibia. IT powerhouse India still has sporadic Internet connectivity. The general speed of the Internet is still very poor. The infrastructure is so haphazard that most Indian Internet subscribers prefer the limited speeds of wireless mobile Internet from the cell phone companies, rather than trust the cable/landline infrastructure. In addition, uninterrupted supply of electricity is still unrealised in India. A power outage affecting any intermediary can cripple access to the Cloud. Indian IT giants have learnt the hard way not to trust state infrastructure, even for electricity. They keep their own backup power generation on site. Adopting any trend that makes them more dependent on state infrastructure will require more than a leap of faith.

So how was this contradiction supposed to work?

On one hand, they intended to make Americans move their computing to the Cloud. On the other hand, tech behemoths clearly understood that pushing the Cloud in developing countries would backfire, making institutional partners such as Governments, opt for open-source Operating Systems instead. And so they were careful not to push it too far in developing countries.

So why was the Cloud pushed with such force at the United States? And why is AI also being pushed with such force at the very same United States? To give you some perspective, the top seven of the World’s largest AI Data Centers are in USA. Some, such as the Colossus 2 (operated by xAI/SapceX/Elon Musk^[1]This is the same Elon Musk who was subsidized by the US Government (at the expense of legit automobile manufacturers) to produce attractive but overpriced EVs. If we take gasoline cars as analogous … Continue reading, are currently operating on 27 gas turbines, enough to power up to 500,000 homes! No wonder communities dread the 70dBA jet-engine like roar these places emit 24/7.

Does it have to do with the fact that the United States still remains the tech/innovation capital of the world, and what grows here later gets transplanted elsewhere? Maybe.

Does it have to do with the fact that the United States still possesses the biggest economy of the world?

Not exactly. China also possesses a big economy.^[2]Thanks to the Phoenicians relocating Western manufacturing to China decades ago. But what it doesn’t have is a strong middle class. Wealth is still concentrated in the hands of millionaires and billionaires. There is a saying among the Chinese that the fat pig gets slaughtered. And therefore those who do become wealthy also apply for membership to the Communist Party, so as to avoid being targeted by those higher up in the food chain. Even billionaires like Jack Ma are not safe from the Phoenicians at the top. That does not sound like a strong middle class does it? That’s more like a slave-ship with the trappings of a cruise ship. One wrong move, and your social credit points are cancelled.

But does the United States possess a strong middle class? It once had the strongest middle class in the world. And while what remains is a shadow of that, the legacy persists. The average American citizen is still fairly empowered compared to citizens of other countries (yet may not realize it). And so, controlling the biggest economy of the world, as well as the most advanced military of the world, requires the Phoenicians making many of their conspiracies America-centric. That is, targeted at the average American. Now they cannot publicly say so, as that would draw alarm and prevent adoption where it matters most. So they pretend that AI Data Centers, despite mainly running on American soil, and burning through American resources and taxpayer funds, are some kind of global trend. And that Russia and China are competitors. And to prove that, the whole planet is invited to the buffet, to come check out OpenAI’s ChatGPT®, Google’s Gemini®, and Microsoft’s Copilot® to name a few.^[3]Now of course, they ported some early models to China and let the Chinese rebrand those models as theirs because the Phoenicians running that country want its Defense to be on par with the United … Continue reading Of course, they may open a newer AI data center or two in Canada or India, but again, the target is only citizens who matter, namely the average American.

On the other hand, the Phoenicians are not bothered with opening up AI Data Centers in Israel. Because the average Israeli is already a sufficiently brainwashed tool, and can be stampeded into genocide after genocide, without thinking twice. They are not opening AI Data Centers in Syria and Lebanon, because the average citizen there is too powerless to have opinions that matter. They are not opening AI Data Centers in Russia because the average citizen has literally no say in what goes on, and whatever AI Data Centers are opened there exclusively serve Defense. All that seems to matter is the opinion of the average American.

Currently, 46% of the world’s AI Data Centers are in USA.

So what exactly is the (American) AI Data Center conspiracy?

To give you some perspective, if you were to download a publicly available LLM such as the Google’s Gemini (which amounts to about a dozen GB in size) on a rather powerful personal computer,^[4]using an interface such as LM Studio. and then start querying it offline, as you would on its online interface, you will notice a delay of up to one minute. That is because your query to the LLM is restricted to the hardware resources of your personal computer.^[5]They are actually opposed to using AI agents such as Openclaw locally (as on personal computers with advanced hardware). Because they don’t want to empower people with their own hardware. And things thus get slow. But when your query is sent online, and has access an American AI data Center with thousands of GPUs, CPUs, and RAM, all liquid-cooled using natural lakes and seas,^[6]Never mind the environment. your query is also submitted to multiple LLMs behind your back, and you will get a much more cogent answer, and in seconds. AI would never be adopted without that lightning fast speed, which only a Data Center can provide.

AI would never be adopted without that lightning fast speed, which only a Data Center can provide.

You don’t even need a personal computer anymore to submit a query to an online AI model. Smartphone apps are now intuitive enough.^[7]Smartphones weren’t considered a substitute to PCs back when Cloud Computing was being pushed. They were still in their infancy And AI holds immense appeal for the semi-literate, superficial and mediocre thinking. They can now sound more educated, polished and intelligent. And with AI, there are way more ways for the Phoenicians to surveil who is typing in what. And it is easier to automatically provide answers that don’t challenge the established consensus, and to censor what we aren’t supposed to know. Just ask AI about Miles Mathis.

Those that get hooked to AI will be expected to pay the bills for the AI Data Centers. To quote,

“We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter,” Altman said during a conversation at BlackRock’s 2026 Infrastructure Summit in Washington, D.C. on March 11.

But if it gets too costly, then the average American may still revert back to the personal computer.

And therefore it is necessary to seal that exit (or at least try). And so resumes the old war on personal computing. Here is Altman again:

But from a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer and being used five percent of the time. You want maximum utilization in the cloud if you’re trying to deal with this very supply-constrained environment. And somehow everyone’s—many people’s intuition is the opposite of that.

Coincidentally, its Altman who is buying up all the RAM, making prices skyrocket for personal computer users. Coincidentally, its Nvidia®, a major AI player that is hoarding up the GPUs intended for personal computer users. Coincidentally, Intel® has announced that it is moving away from consumer-grade CPUs. Cloud Computing was about moving your personal data to faraway online servers. AI Data Centers are about moving your entire Personal Computer to faraway online servers.

Cloud Computing was about moving your Personal Data to faraway online servers. AI Data Centers are about moving your entire Personal Computer to faraway online servers.

Needless to say, this is another America-centric conspiracy, masquerading as some kind of global phenomenon. Nowhere in history has the opinions and choices of a middle class (or former middle class) become so relevant.