Electricity is the new oil: Why the future of computing is not about AI

Over the past few years, artificial intelligence has become the dominant narrative shaping discussions about the future of computing. Public companies announce plans to repurpose data centers for AI workloads, investors eagerly follow the trend, and cryptocurrency mining is increasingly portrayed as an industry that must “transform” in order to survive.

But when you move past headlines and investor decks and look at the actual economics of infrastructure, a different picture emerges. The future of computing is not defined by algorithms or GPUs. It is defined by electricity.

To better understand how this applies to Russia, we spoke with Alexander Lozben, CEO of Interhash. His perspective sharply contrasts with the prevailing narrative. In his view, much of the discussion around large-scale transformation of mining infrastructure into AI data centers in Russia reflects market hype rather than operational reality.

From the outset, Lozben shifts the focus away from technology and toward fundamentals. Both cryptocurrency mining and artificial intelligence, he explains, belong to the same category: energy-intensive computing. As a result, the real competition is not between Bitcoin and neural networks, but for access to cheap, stable electricity. This single factor, more than any technological choice, determines whether a project is economically viable.

This is where the myth of “easy transformation” begins to break down. Infrastructure built for mining is largely unsuitable for training neural networks. Hardware requirements, cooling systems, reliability standards, and overall architecture differ significantly. Converting a mining facility into an AI data center is not a matter of upgrading equipment—it requires rebuilding the infrastructure almost from scratch. In practical terms, only the power connection and internet access remain usable.

Demand is another critical issue. Globally, AI infrastructure is driven by a small group of major players—OpenAI, Google, Microsoft—whose scale justifies multi-billion-dollar investments in data centers. In Russia, the number of comparable customers is extremely limited. Dependence on one or two large corporate clients creates an inherently fragile business model, especially under sanctions that restrict access to GPUs and complicate equipment servicing.

Lozben also emphasizes that mining and AI differ fundamentally in their business logic. In mining, the “customer” is the Bitcoin network itself, which rewards participants according to a transparent and universal formula: block subsidies plus transaction fees. The risks are well understood—network difficulty and price volatility.

AI infrastructure, by contrast, is built on contractual relationships. Terms can change, payments can be delayed, clients can switch providers or exit the market altogether. This does not make the model inferior, but it does make it fundamentally different—and far less predictable.

For this reason, Lozben does not see artificial intelligence as an existential threat to mining. Bitcoin, he notes, is a self-regulating system. When participants leave, mining difficulty decreases, rewards per miner increase, and the sector becomes attractive again. Four variables—price, issuance, fees, and difficulty—continuously rebalance the system without external intervention.

He is equally skeptical about claims that AI will revolutionize mining itself. Neural networks may be useful for standard corporate tasks such as logistics, workforce management, maintenance planning, or predictive equipment failure analysis. But these are optimization tools, not transformative technologies. Presenting them as breakthroughs, he argues, is often more about appealing to investors than changing how mining actually works.

As the conversation draws to a close, Lozben returns to what he sees as the decisive factor for the entire computing industry: electricity. Demand for power is growing faster than generation capacity. Data centers, factories, electric vehicles, and digital services all compete for the same resource. New power plants are built slowly, while global energy policies continue to push costs higher.

In this context, debates about “transforming” mining appear secondary. Mining has been declared obsolete many times before, yet each time it has returned to equilibrium through its own internal mathematics. Investors may favor artificial intelligence today because it is the dominant trend. Tomorrow, attention may shift again. That is how markets work.

The conclusion of the conversation is straightforward:

artificial intelligence is a trend,

mining is a calculation,

and electricity is the new oil shaping the future of computing.