NEWS YOU CAN’T MISS

The AI Boom Is Running on Debt

The AI boom is getting expensive - so expensive that it’s now running on debt.

To keep up with the explosion of demand for computing power, tech giants are pouring billions into new data centers. But even companies like Google, Microsoft, and Meta are realizing their cash can’t keep up with their ambitions.

So they’re turning to Wall Street:

• In the past 3 months, the four biggest tech firms have spent $112B building data centers.

• To fund that kind of growth, companies are borrowing in creative ways: corporate bonds, private loans, and even turning their data centers’ future earnings into investments that can be traded like stocks.

• A few are even using special financial structures that keep this debt off their official books, making the companies look less burdened than they really are.

For example, Meta borrowed $30 billion to build a huge data center in Louisiana. Instead of taking a normal loan, it used a special setup that hides the debt on a separate company’s books. Elon Musk’s xAI is reportedly planning a similar move to buy Nvidia chips and then rent them back to itself.

It’s all part of what analysts are calling the AI debt economy: a race to build faster than revenue can catch up. Consulting firm McKinsey estimates that the world will need $7 trillion in data center investment by 2030 just to keep AI running.

That aggressive expansion is making investors nervous: Meta’s stock dropped 11% after revealing its new AI spending plans.

Can AI Help Europe Beat the Heat?

Europe’s summers are no longer what they used to be. 🥵

Heatwaves are arriving earlier, lasting longer, and breaking temperature records almost every year. Cities that once worried about cold snaps are now struggling with droughts, power cuts, and heat-related deaths.

But scientists in Italy think AI might finally help Europe see these events coming weeks in advance.

At the Euro-Mediterranean Centre on Climate Change, researchers have built an AI system that can predict extreme heatwaves up to seven weeks before they strike. For comparison, traditional weather models usually give around 10 days’ notice.

Here’s what makes this breakthrough so powerful:

• Instead of simulating weather one forecast at a time, the AI learns from centuries of climate behavior.

• It was trained on reconstructions of global conditions from the year 0 to 1850 - giving it a deep sense of how temperature, wind, oceans, and soil moisture interact to create heatwaves.

• The team says their system can even outperform traditional methods in northern Europe, where forecasts have always been less reliable.

And because it needs less computing power than massive supercomputers, it could make long-range forecasting accessible to more local weather agencies.

Europe’s deadly heatwaves in 2003, 2010, and 2022 killed tens of thousands of people and caused huge losses in farming and energy. With better warning, hospitals could prepare, crops could be protected, and power grids could plan ahead.

Can ChatGPT Help You Survive Exam Season?

Kim Kardashian isn’t just running beauty brands and TV shows; she’s also been studying to become a lawyer for six years.

Instead of law school, the celebrity is taking California’s old-school route, apprenticing under real lawyers and slowly working her way toward the state bar exam.

She’s passed key milestones like the “baby bar,” but still needs to clear the full state bar exam to officially practice. Along the way, she’s been using ChatGPT as her study buddy. In a new Vanity Fair interview, Kim called the AI her “frenemy,” saying it’s given her wrong answers that made her fail practice tests.

“I’ll yell at it,” she said, “and it tells me, ‘This is teaching you to trust your instincts.’” 😆

Have you ever used ChatGPT as a study buddy?