By the end of 2017, Bitcoin was approaching $20,000. But what truly made me pause wasn’t the price—it was another set of numbers: according to estimates at the time, global Bitcoin mining consumed more electricity annually than the entire nation of Ireland.

A decentralized digital ledger was consuming the electricity of a developed country.

The essence of this phenomenon isn’t as simple as “mining is wasteful.” It reveals a deeper equation: in Bitcoin’s world, electricity = computational power = currency. The more electricity you invest, the more computational power you generate; the more mathematical problems your computational power can solve, the more Bitcoin you earn in return. This equation is so clean it almost resembles a law of physics.

When Knowledge Becomes Computable

Following this logic one step further, AI’s rise has made another equation increasingly clear: knowledge ≈ algorithms.

What does this mean? We used to say “knowledge is power,” but that power was vague and hard to measure. A doctor’s diagnostic ability is “valuable,” but exactly how valuable? A lawyer’s legal analysis is “irreplaceable,” but irreplaceable to what degree?

Now, AI has made these questions quantifiable. How much computational power does it take to train a medical imaging recognition model? How many GPU hours are needed to fine-tune a legal document summarization system? When knowledge is broken down into algorithms that can be trained, replicated, and mass-produced, its “price” emerges.

This isn’t about diminishing the value of knowledge. Quite the opposite—it’s about finding an unprecedented pricing mechanism for knowledge.

From Gold Standard to Computational Power Standard

Humanity’s monetary systems have undergone several major transitions.

In the gold standard era, currency value was anchored to gold. The bills in your hand had value because they were backed by corresponding gold reserves. Later, when the Bretton Woods system collapsed, we entered the dollar standard—currency value was no longer anchored to gold, but to America’s economic strength and military power.

So what’s the next “standard”?

I believe the clue lies at the intersection of Bitcoin and AI. When electricity can produce currency (Bitcoin) and electricity can also produce knowledge (AI models), then electricity—or more precisely, computational power—becomes the most fundamental value benchmark of this era.

Whoever controls computational power controls the printing press of the new age.

This is why NVIDIA’s market cap could skyrocket to among the world’s highest within a few years. Not because they sell particularly beautiful GPUs, but because they sell “tickets to computational power.” Control the supply of computational power, and you control the faucet of this new economic system.

An Entrepreneur’s Experience

In running companies, I’ve felt the reality of “computational power as value” very directly.

In the early days of doing digital transformation consulting, clients’ most frequent question was: “How do you calculate this project’s ROI?” We spent considerable time explaining why digitalization was important, why data-driven approaches were the trend. But ultimately, clients cared about just one thing: how much cost invested, how much benefit produced?

Later, as AI tools became widely adopted, the answer to this question became remarkably straightforward. What used to require three people working a week on market analysis could now be completed by one person with an AI assistant in a day. Three people’s salaries vs. one computer’s electricity plus API fees—any boss can calculate that.

The cost of knowledge work is being recalculated as “the cost of computational power.” And the cost of computational power, fundamentally, is electricity bills.

This is exactly the same logic as manufacturing. Factories use electricity to drive machines to produce physical goods, while today’s knowledge workers use electricity to drive AI to produce digital outputs. The only difference is that factory outputs are visible objects, while AI outputs are invisible intelligence. But the underlying energy conversion logic is essentially the same.

The Alchemy of Traffic

Following this framework, many seemingly “unreasonable” phenomena become reasonable.

Take the influencer economy. Why can a YouTuber sitting in front of a camera chatting earn millions annually? Because their talent and charisma, through platform algorithmic systems, are converted into traffic. Traffic is essentially an “aggregation of attention,” and attention, mediated by algorithms, can be precisely converted into advertising revenue.

Talent → Content → Algorithmic recommendation → Traffic aggregation → Monetary conversion.

Every step requires electricity and computational power as infrastructure. Without servers, without CDNs, without recommendation algorithms, even the most talented person couldn’t reach millions of viewers.

So influencers aren’t “getting something for nothing”—they’re just converting their knowledge and charisma through digital channels and computational infrastructure into currency in this new system. This is essentially the same as miners using physical strength to dig out gold, or engineers using mental power to write programs: using some form of energy, through some conversion mechanism, to create socially recognized value.

What Is Your Computational Power?

Returning to the personal level.

If computational power is the fundamental currency of the new era, then everyone needs to ask themselves: how much computational power can I mobilize?

This isn’t just about “whether you can use ChatGPT.” It’s about whether you can identify which problems are worth computing, whether you can design effective analytical frameworks, whether you can distinguish signal from noise in AI outputs.

The point was never about how much computational power you own—GPUs can be rented, APIs can be purchased. The point is whether you have the ability to convert computational power into meaningful output.

Just like in the gold standard era, the point wasn’t how much gold you had, but whether you could use gold to make valuable transactions. The computational power standard era is the same: energy itself isn’t valuable, energy conversion efficiency is what’s valuable.

This may be our era’s most fundamental redefinition of “what has value.”