When the Microsoft CEO said, "Insufficient electricity could lead to a chip pileup," neither he nor Altman knew exactly how much power AI would need.

```

The focus of the artificial intelligence competition is shifting from computing power to electricity. Technology industry leaders admit they are grappling with a fundamental uncertainty: just how much energy will AI consume in the future.

Microsoft CEO Satya Nadella recently revealed on the "BG2" podcast that the biggest issue limiting the company's growth is no longer chip shortages. Nadella said: "The biggest problem we face now is not an excess of computing power, but electricity... and whether we can build data centers close enough to power sources fast enough."

Nadella bluntly stated that this disconnect has led to an embarrassing situation where Microsoft experiences a backlog of chips. "You could have a bunch of chips sitting idle in inventory because I can't plug them into a power source. In fact, that's the problem I'm facing today." He added that the issue is not the chip supply, but the lack of 'warm shell' data centers that are ready to be occupied and powered on. This statement clearly reveals that the pace of physical infrastructure construction in the real world has fallen far behind the expansion of computing power in the digital world.

OpenAI CEO Sam Altman, who participated in the podcast with Nadella, also emphasized the strategic dilemma brought by this uncertainty. He believes that the entire industry is in the midst of a huge energy gamble, with nobody knowing how it will turn out.

Bottleneck Shift: From Chips to Electricity

For a long time, the market generally believed that obtaining advanced graphics processing units (GPUs) was the biggest obstacle to deploying artificial intelligence services. However, Nadella's remarks confirm that the bottleneck has already shifted. When chips that tech companies have spent heavily on cannot be turned on, there is no way to speak of a computing power advantage.

This phenomenon reflects the challenges that software and chip companies, accustomed to rapid iteration, face when encountering asset-heavy, long-cycle industries like energy and real estate. In the U.S., data center electricity demand has risen sharply over the past five years, breaking a decade-long period of stability, and growth has outpaced power generation planning by public utilities.

This has forced data center developers to seek "behind-the-meter" power supply solutions, bypassing the public grid and getting electricity directly from generation facilities.

The Fog of Demand: How Big is AI's Energy Appetite?

"How much electricity is enough? No one knows, not even Sam Altman or Satya Nadella." TechCrunch pointed this out in a report on November 3. This unknown stems from the rapid evolution of AI technology itself.

Altman presented a "very frightening exponential" growth scenario on the podcast. He hypothesized that if the cost per unit of intelligence continues to drop at a pace of 40 times per year, then from an infrastructure perspective, the resulting demand growth would be staggering.

He firmly believes that the "Jevons paradox" will play out in the AI field: improvements in computing efficiency and reductions in cost will actually stimulate usage to increase by more than a hundredfold, as more applications that are currently not economically viable will become feasible at lower costs.

Energy Gamble: The Dilemma of Betting on the Future

It is precisely this huge uncertainty that places industry leaders like Altman in a tough spot regarding energy strategy. He described a dilemma: "If a very cheap form of energy is widely adopted soon, then many who have signed current (expensive) power contracts will suffer heavy losses."

On the other hand, if you don’t invest boldly, you may miss the opportunity created by an explosion in AI demand. Altman admitted that if AI’s efficiency improves beyond expectations or demand falls short of expectations, some companies might be saddled with the heavy burden of idle power plants.

To hedge risks and explore the future, Altman himself has invested in several energy startups, including fission company Oklo, fusion company Helion, and solar thermal storage company Exowatt.

Countermeasures: Seeking Solutions Between Tradition and Innovation

Facing these challenges, tech companies are actively looking for solutions. Traditional natural gas power plants take years to build and cannot match the speed of AI industry demand. As a result, faster-to-deploy, lower-cost, and zero-emission solar power has become a popular option.

In many ways, solar photovoltaic technology is similar to the semiconductor industry: both are based on silicon materials, produced as modular components, and can be quickly assembled into arrays to increase capacity. This modularity and rapid deployment make its construction tempo closer to that of data centers. However, both data center and solar project construction require time, and market demand is changing far faster than this.

As a result, tech companies are constantly racing against time and facing strategic decisions across the three interconnected domains of computing power, data centers, and electricity.

Risk Warning and DisclaimerThe market has risks, and investment needs to be cautious. This article does not constitute personal investment advice and does not take into account any individual user's particular investment objectives, financial situation, or needs. Users should consider whether any opinions, views, or conclusions in this article are applicable to their specific circumstances. Any investments made accordingly are at your own risk. ```