Now available: This Is Server CountryGet the book
What You're Noticing

Why Are Electronics So Expensive?

New computers and phones cost more than they used to. The reason: giant tech companies are buying up the parts before they reach store shelves. Here's what's happening.

5 min read

Key Takeaways

  • 1 Computer memory prices have tripled since 2024
  • 2 AI data centers now buy 7 out of every 10 memory chips made worldwide
  • 3 This isn't temporary—AI demand keeps growing faster than factories can make chips

Sticker Shock at the Electronics Store

You went to buy a new laptop, and the price made you stop and look twice. The model you had your eye on costs $200 more than a similar one did two years ago. Memory upgrades that used to cost $80 now run $250. Even basic tablets and phones seem pricier than they should be.

You might think it's just inflation. Or leftover problems from the pandemic. But there's a specific reason: the companies building AI systems—Microsoft, Google, Amazon, and others—are buying up computer parts faster than they can be made. Regular shoppers like you and me are competing for what's left.

The "AI Tax"

Think of it like a grocery store with limited bread. If a restaurant chain starts buying 70% of all the bread before it hits the shelves, the remaining loaves cost more. That's what's happening with computer parts—big tech companies can pay more than you can, so prices go up for everyone else.

Why AI Needs So Many Parts

To understand why your laptop costs more, it helps to know what's happening inside those giant AI data centers.

When you use ChatGPT, Siri, or any AI assistant, the computer has to hold an enormous amount of information in its "working memory"—like how you might spread out papers on a desk to work on a project. The bigger and smarter the AI, the bigger the desk it needs.

💡 Think of it this way

Imagine a library where every book has to be open on a table at the same time for the librarian to answer your question. AI systems work like that—they need massive amounts of memory chips just to hold everything they know while thinking.

A single AI training facility uses more memory chips than all the laptops sold in America in an entire year. And there are hundreds of these facilities being built right now.

Here's the key problem: the same factories that make memory for AI also make memory for your laptop, phone, and tablet. When a factory decides to make chips for Microsoft instead of Dell or Apple, that's fewer chips available for the devices regular people buy.

Who Gets the Parts First?

The companies that make memory chips—Samsung, SK Hynix, and Micron—have to choose who to sell to. And it's an easy choice for them.

Big Tech
Microsoft, Google, Amazon
5-10x more per chip
  • Orders billions of dollars at once
  • Signs multi-year contracts
  • Pays premium prices
vs
You
Regular consumers
1x retail price
  • Buys one device at a time
  • Shops around for deals
  • Price-sensitive

If you were a chip maker, who would you sell to first?

Where Memory Chips Go Share of all chips made worldwide
15%
2023
to data centers
70%
2026
to data centers

How Much Money Are We Talking About?

To understand why this is happening, look at how much these companies are spending. These aren't normal business expenses—they're bigger than the budgets of most countries.

What Big Tech Is Spending on AI Announced investments
$80B+
Microsoft alone
$500B
One project (Stargate)
$1.1T+
All companies combined

For context: $1.1 trillion is more than the entire yearly economic output of the Netherlands, Switzerland, or Saudi Arabia. That money is flowing into computer parts, and most of it goes to the same suppliers who make parts for your laptop, phone, and tablet.

This isn't going away. Remember the chip shortages during COVID? Those eventually got better. This is different. AI demand keeps growing faster than factories can expand. Building a new chip factory takes 3-5 years and costs billions of dollars—and by the time it opens, AI companies want even more.

The Bigger Picture

When you see the price tag on a new laptop and wonder why it costs so much, now you know: you're not just buying a computer. You're competing with some of the richest companies in history for the same parts.

🔗 It's All Connected

The same companies buying up memory chips are also building massive data centers across rural America. These facilities need electricity—lots of it. That's why your utility bill might be going up, why new power lines are being built through farmland, and why your town might be considering a data center proposal.

This isn't a story about greedy retailers or broken supply chains. It's about a fundamental shift in who gets computer parts first. When trillion-dollar companies decide they need something, ordinary buyers end up paying more or waiting longer.

The price of your next phone or computer, the construction project down the road, and your rising electric bill might seem unrelated. But they're all part of the same story: a massive buildout of AI infrastructure that's reshaping the economy in ways both visible and invisible.

Want to Learn More?

The full story of how AI is reshaping everyday economics—from electronics prices to electricity bills to rural communities—is told in This Is Server Country.

Learn more about the book