Практика Shadowing: The Trillion Dollar AI Lie. CEOs Are Bleeding BILLIONS. - Изучайте разговорный английский с YouTube

Big tech just burned through hundreds of billions building the infrastructure for AI.

⏸ Пауза

Скорость:

127 предложений

Если предложения слишком короткие или длинные, нажмите Edit, чтобы их изменить.

Big tech just burned through hundreds of billions building the infrastructure for AI.

And the return so far? A fraction of the cost. Nowhere close to break-even. It might be one of the most expensive bets in modern history. Beneath the polished interface is something most people never think about… a physical system with real-world limits. Data centers burning through enormous amounts of electricity. Power grids pushed closer to their capacity just to keep everything running. And it relies on a hidden army of exploited human labor.

You don’t get to stay outside of it. You’re already interacting with it every day.

The world has already committed to a system it may not be able to stop.

This is the trillion dollar AI lie.

Chapter 1: The Receipt Nobody Wants to Read In 2024, analysts at Sequoia Capital posed a simple question. If AI companies keep spending the way they are right now, how much money would the industry need to make every year to justify it?

Their answer was about $600 billion a year.

That’s not what AI companies are earning today.

That’s what they’d need to earn for all of this to make financial sense.

Right now, generative AI is already making real money. Billions in annual revenue across companies like OpenAI and Anthropic. Analysts predict Ai-driven profits may rise into the trillions by the mid-2030s.

But a lot of the reported revenue isn’t even profit at all.

Anthropic recently scaled its revenue to hundreds of millions per month in 2025. It sounds incredible until you realize it was also expected to lose billions over the course of the year. In mid-2025, OpenAI secured $10 billion dollars of funding. But since its operating costs were so steep, it was asking for another $8.3 billion just months later. And Elon Musk’s xAI? It was reportedly burning through more than a billion every single month… just to keep the lights on.

It’s a serious problem.

One that points to something people don’t like to talk about.

In traditional software, once you build a product adding more users is cheap.

That’s why companies like Microsoft and Adobe can become insanely profitable.

The cost of serving the next customer is basically nothing.

AI? It doesn’t work like that.

Every time you ask a question, it costs money. Real money. The answer comes from a delicate interplay of computational power, electricity, and infrastructure. Which means scaling doesn’t automatically make things more efficient. In some cases, it does the opposite.

The more people use the system, the more expensive it becomes to run.

The largest companies and investment groups in the world have already committed, with annual spending pushing past $400 billion. AI growth speculation has become a core engine of the S&P 500. It becomes even sketchier when you realize that the broader market, built up of retirement accounts, index funds, and otherwise “safe” investments, are all now deeply exposed to the whims of a technology that hasn’t even proven it can pay for itself at scale.

If the economics of generative AI’s future are so uncertain, why does every corporate CEO in the tech world seem so hell-bent on staking everything on its future?

According to a recent survey, roughly 90% of CEOs say AI will fundamentally change their companies by 2028. It’s a huge number. But when you look at the actual financial data, the reality is brutal. Only 25% of AI initiatives are actually delivering their expected ROI.

9 out of 10 executives believe the technology is essential. But only about 1 in 4 can explain how it actually makes money.

What once felt like a strategy now seems like corporate peer pressure.

And you see it playing out in real time. Each company tries to outdo the next rushing out AI initiatives, talking about “AI-first” strategies, and buying up massive quantities of GPUs because not buying them looks worse.

If this amounts to the corporate version of a gold rush, nobody wants to be the one who showed up without a shovel. That’s where we’re at right now. Infrastructure is being built up at full speed. The spending is already committee and seed money raised.

But the returns are barely keeping pace.

Chapter 2: Success That Makes You Poorer Over the past few years, a quiet industry has emerged to support AI systems. Things like data labeling, content moderation, and output verification have all grown as the core tasks that make models usable in the real world. They’re basically the difference between a raw system that produces chaotic, inappropriate outputs and one that can answer questions without embarrassing the company that built it.

All that work happens outside the model itself.

In offices and call-center-style environments in places like Kenya, the Philippines, and India, thousands of workers spend their days correcting the mistakes that AI systems still make constantly. In some documented cases, they’re paid just a few dollars an hour to filter out violent or explicit content.

According to one report, content moderators in second and third-world countries navigate a combination of “psychological trauma, poverty wages, and the suppression of union organising conditions” considered intolerable under Western labor laws. It sounds less like a futuristic breakthrough and more like something out of the 19th century.

That's because in some ways it is.

The marketing language around AI suggests autonomy. Machines that can think, learn, and act, replacing human effort altogether.

The reality is closer to something older and more familiar. It’s a layered system, where the most visible tip of the iceberg is clean and efficient and the least visible part - the system’s inner core - is messy, labor-intensive, and easy to ignore.

That doesn’t mean the technology is “fake.” But at second glance, the word “artificial” is doing an awful lot of heavy lifting.

A large portion of the work that makes these systems run is still very human.

That human system doesn’t disappear as the system scales. If anything, it just becomes more important.

When the system relies on both massive compute infrastructure and ongoing human input, then the cost structure starts to look very different than the one most people imagine.

Rather than a clean, self-improving machine, you get something closer to a hybrid: part automated, part manual support… both constantly maintained.

At the beginning of the AI boom, researchers found out that a huge percentage of companies calling themselves “AI Startups” weren’t actually using any meaningful AI in their core products.

In some cases, it was as high as 40%. That didn’t necessarily mean fraud. But it did hint at something even more important.

From the very beginning, “AI” became a signal. A way to attract funding, justify higher valuations, and position themselves in a market where everyone was suddenly supposed to have an AI story.

By 2024, roughly 78% of all companies reported using AI in some form. Just a year later, 61% of all global venture capital was flowing into AI-related businesses.

You might expect that after all this adoption, all this investment, we’d be seeing clear, consistent returns. Instead, the gains are highly concentrated, with about 75% of the financial benefits reaped by just 20% of the companies.

As it turns out, the majority of companies talking about AI aren’t actually making that much money from it. They’re experimenting, deploying features, and integrating tools, but not transforming their business like everyone thought.

In the early days, companies claimed AI without really using it.

Today, companies are using it, but don’t know how to make money off it.

And that’s a much bigger problem.

Because now the stakes are higher, the expectations are enormous, and the gap between what’s actually happening and what was promised is getting harder to ignore.

Chapter 3: The Energy Wall The deeper constraint on all of this isn’t financial, but physical.

For years, a simple comparison was repeated that a single AI query can use around ten times the electricity than a standard web search. That number comes from early estimates, and like most simple comparisons, it’s been argued over, updated, and stretched depending on who is making the case.

But the part that hasn’t really changed is the assertion that AI workloads are simply heavier.

By 2022, data centers were already consuming roughly 460 terawatt-hours of electricity globally every year. That’s about the same as an entire country like Germany or Japan, or 2% of total global demand.

By this year, that number could land somewhere between 620 and 1,000 terawatt-hours, depending on how aggressively AI keeps growing.

When you look into a data center, how that energy is used is somewhat surprising.

40% of the electricity is used for actual computing. The “AI doing its thing” part Another 40% is used just to cool the machines down.

The remaining 20% goes to everything else: moving the data around; keeping the systems stable; basically making sure the whole operation doesn’t collapse under its own complexity.

So nearly half the energy we’re pouring into AI…isn’t making it any smarter. It’s just keeping it alive.

Now zoom out one level.

A single large “hyperscale” data center,” the kind companies are building for AI, can consume 100 megawatts of power or more. 1 megawatt can typically power over 150 U.S. homes. 100? Over a year, that’s akin to the electricity needed to charge well over 200,000 electric vehicles… For just one building.

And there are over 8,000 data centers worldwide, with a third of them sitting in the United States alone. Worryingly, studies are now showing that the vast majority of these are located in climates considered way too hot for efficient operation.

It’s time to stop thinking about the “cloud” as something abstract. We’re talking about a growing network of massive, power-hungry facilities, clustered in specific regions, pulling from the same grids that supply homes, schools, and businesses.

The demand isn’t close to being evenly distributed, either. In areas with concentrated pressure, city officials are forced to make trade-offs in real time: Do you expand the grid?

Do you raise power prices?

Do you slow down development and limit job growth?

None of those are easy answers. And none of them deal with the issue of water, either.

Remember the 40% of energy used simply to cool these mammoth data centers? That often depends on huge volumes of water moving through the system, constantly cycling to carry the heat away.

A single large data center can use millions of gallons of water per day, about the same as a town of 30,000 to 50,000 people. Over a year, even a mid-sized facility can burn through around 100 million gallons just to stay cool.

And most of that water doesn’t come back.

In many systems, 70 to 80% is lost to evaporation, effectively disappearing into the air. Meanwhile, roughly 75 to 90% of data centers rely on water-based cooling, often pulling from the same rivers and municipal supplies that serve local communities. In at least one Oregon town, a single company’s data centers consumed over 25% of the city’s water supply.

The same system that promises infinite scale is drawing from very finite supplies.

Power grids that expand overnight, and water systems are already under pressure. The more the system grows, the more it pulls. And right now, it’s not obvious where the ceiling is.

And while all of that strain is building in the physical world, something else is happening inside companies. Because while AI is pulling more from the outside… it’s also quietly pulling something from the inside.

Their data. Chapter 4: Security Self-Sabotage According to recent data, about 34.8% of employee inputs into AI now contain sensitive information. That’s up from just over 10% in 2023.

A third of everything being pasted into your AI chatbot of choice are legal documents, customer data, medical records, source code, and contracts.

What’s worse, 83% of the companies doing this have absolutely zero technical controls in place to stop it.

They can send out policy emails until the cows come home. But the next day they’ll turn around and tell those same employees to be faster, more productive and efficient.

The employees now have to decide: deadline…or policy? They pick the deadline, every time.

The scale of this security self-own is starting to look like a slow-motion disaster.

Over 225,000 Chat GPT credentials have already been found for sale on dark web marketplaces, often harvested from compromised machines or reused passwords. Major companies like Apple, JP Morgan, and Goldman Sachs have either restricted or banned tools like ChatGPT internally after realizing what was happening.

Samsung learned the hard way.

In early 2023, 3 employees read an email lifting a previous Chat GPT ban and thought, we’re home free. The first went and uploaded proprietary source code to debug this problem they were having. Another copied notes from an internal company meeting into their chat feed, while a third used Chat GPT to “identify defective equipment in a semiconductor line.” Pretty soon, the ban was back. An internal survey quickly showed that 65% of Samsung employees thought AI tools were a security risk.

The most unsettling part of all this is what most people misunderstand about the core disconnect between user privacy and the way AI systems are designed to work.

AI models are trained on massive datasets. On consumer plans like OpenAI’s Free and Plus version, consumers allow the company to use their conversations to train future models by default. Sure, you can opt out by navigating a maze of settings. But most users aren’t even aware that toggle exists.

“Until you do, every document, every contract snippet, every client detail you type becomes potential training material,” warns one report.

Once trade secrets are used in training or processing an AI model, the damage can be effectively permanent.

You can’t just delete it. It’s like trying to unmix paint.

That’s what makes this far more sinister than a traditional data breach. Those are rare, visible, and fixable, for the most part.

Compromising user security is normal for AI companies. HIPAA noncompliance is built into the workflow. That’s why some security researchers are already starting to describe this as the largest uncontrolled corporate data leak in history.

When you talk to the people inside these companies who are actually responsible for what’s going on, they know exactly what’s happening. They know the controls to stop employees from pasting sensitive data into their AI tools aren’t in place.

They know the risk.

But they also know there isn’t an easy fix.

You can’t just spin up a secure, in-house version overnight. That takes millions of dollars of hardware, specialized teams, and time most companies don’t have.

In the meantime, the pressure to move faster doesn’t go away. People will still smirk at their company’s “enterprise” version of CoPilot which only allows them to polish emails and go Google searches. They’ll sneakily use their consumer version of ChatGPT, and their older executive bosses won’t care for the most part. They still see AI models as “glorified search engines.” If there’s one thing people should have realized years ago, it was never to trust Silicon Valley bros with your intellectual property. Today, everyone is riding that productivity wave…and hoping it doesn’t come back to bite them.

Chapter 5: The Great Correction At this point, we know things aren’t adding up.

The system is expensive. It’s resource intensive, leaking data like a sieve, and still not making consistent-enough money for its investors.

So why is the industry still all-in on the AI boom?

The answer is because from the inside, none of this feels like a choice.

100

It feels like an arms race.

101

The constant warnings to not “fall behind” or “let China catch up” has everyone moving at an urgent pace. It feels existential.

102

But it really isn’t.

103

The real bottleneck of today’s so-called “AI arms race” is semiconductors, or chips.

104

And the companies selling those chips, especially NVIDIA and AMD, are making out just fine.

105

The global semiconductor industry has been going gangbusters for years now. In 2026, the industry is expected to earn almost $1 trillion in annual sales, an all-time high. Analysts are predicting annual sales of $2 trillion by 2036.

106

According to the CEO of Taiwan Semiconductor Manufacturing Company (TSMC), the single most important chip manufacturer in the world, demand for advanced AI chips is currently running at three times the global supply.

107

Every major tech company is trying to lock in as much capacity as possible, years in advance. New factories in Arizona and Japan won’t meaningfully ease the market pressure until 2027 or later.

108

Let’s be clear. Chip manufacturers want companies to believe there is not enough compute, that there will never be enough compute. If you don’t buy everything now, you’ll inevitably fall behind in this computational arms race.

109

That message doesn’t need to be false to be powerful.

110

It just needs to be repeated often enough.

111

And it is repeated, on annual earnings calls, at conferences, even in front of Congress.

112

This is how you end up with firms ordering billions of dollars of GPUs years in advance, locking in supply they may not even be able to fully use yet.

113

That’s how you get a market where demand is running far in excess of supply. And while chip companies end up delirious with cash-filled pockets, consumers are left footing the bill.

114

For months now, the AI arms race has bled over into a full blown shortage of specialized memory those chips need to function. People call it “RAMageddon.” The result is a squeeze on everything consumers need. Laptops, phones, and even appliances cost more.

115

And somewhere out there, an ordinary guy just trying to build a new gaming PC is wondering how a couple of sticks of DDR4 suddenly cost more than his memory card.

116

Today’s AI systems are built on two assumptions. First, that AI demand will keep growing fast enough to absorb all the infrastructure it requires. And second, that productivity gains will arrive quickly enough to justify the cost.

117

If either of those assumptions slip, the whole premise of an AI boom starts to wobble.

118

We’ve seen this before. In the late 1990s, companies built out massive internet infrastructure with fiber servers and networks filling entire buildings on the assumption that demand would catch up.

119

Eventually, it did, but not before the market corrected.

120

Hard. The NASDAQ lost around 76% of its value. Companies like Cisco, Intel, and Oracle saw stock prices tank overnight. Other companies like eBay and Amazon barely managed to survive. It took the NASDAQ index fifteen years to reclaim its previous high.

121

The internet boom wasn’t fake, but the timeline for dramatic overvaluation and hype was. The market had to go on life support as a result.

122

The so-called “dotcoms” just couldn’t turn the profit their investors had dreamed of when they poured their cash into start-ups in the 1990s.

123

It’s not hard to imagine a similar correction coming today. This time for the largest companies in the world. The money fueling today’s AI boom isn’t just venture capital chasing upside; it touches everyone.

124

When these cycles turn, the people making the decisions rarely take the hit. Executives still get paid and early investors find the exit. The losses spread outward.

125

So when you hear “AI arms race,” it’s worth asking: race to what? Because right now, it looks less like a race to the future and more like a scramble to justify billions already spent.

126

If the returns don’t show up fast enough, the fallout won’t stay in tech. It'll land everywhere.

127

And if you think the biggest risk is money, power, or data…you might be underestimating the real problem: What happens when these machines decide a life is expendable? Find out in AI Just Tried To Murder A Human To Avoid Being Turned Off. Or watch this video.

TRENDING

Что такое техника Shadowing?

Shadowing — это научно обоснованная техника изучения языка, изначально разработанная для подготовки профессиональных переводчиков и популяризированная полиглотом доктором Александром Аргуэльесом. Метод прост, но эффективен: вы слушаете аудио на английском от носителей языка и немедленно повторяете вслух — как тень, следующая за говорящим с задержкой в 1–2 секунды. В отличие от пассивного прослушивания или грамматических упражнений, Shadowing заставляет мозг и мышцы рта одновременно обрабатывать и воспроизводить реальные речевые паттерны. Исследования показывают, что это значительно улучшает точность произношения, интонацию, ритм, связную речь, понимание на слух и беглость речи — что делает его одним из самых эффективных методов для подготовки к IELTS Speaking и реального общения на английском.

☕ Угостите нас кофе

ShadowingEnglish остаётся 100% бесплатным благодаря вашей поддержке. Стоимость серверов и ИИ высока — ваш кофе поддерживает нас! 🙏

Пожертвовать через PayPal

ShadowingEnglish.com – Практика Shadowing английского

Практика Shadowing: The Trillion Dollar AI Lie. CEOs Are Bleeding BILLIONS. - Изучайте разговорный английский с YouTube

Популярные

Об этом уроке

Ключевая лексика и фразы

Советы по практике

Что такое техника Shadowing?

Практика Shadowing: The Trillion Dollar AI Lie. CEOs Are Bleeding BILLIONS. - Изучайте разговорный английский с YouTube

Скачать приложение

Об этом уроке

Ключевая лексика и фразы

Советы по практике

Что такое техника Shadowing?