Left hand BTC, right hand AI computing power, the gold and oil of the digital intelligence era

Authored by: Jademont, Evan Lu, Waterdrip Capital

A Review of the Turbulent 2025 and the Future of the AI Long Cycle

The New Industrial Revolution: Computing Power as the Engine of the Economy

“In this world, only a few can, like Edwin Drake, inadvertently usher in an era that changes human history… His drill bit, penetrating deep into the earth, not only touched black liquid but also the arteries of modern industrial civilization.”

In 1859, amidst the mud of Pennsylvania, people gathered around Colonel Edwin Drake, mocking him. At that time, the world still relied on increasingly scarce whale oil for lighting, but Drake believed that underground “rock oil” could be mined at scale. This was considered a madman’s delusion. Until the first gush of black liquid erupted, no one could have imagined that oil would not only replace whale oil as a lighting source but also become the cornerstone behind the struggle for global dominance over the next two centuries, reshaping global power and geopolitics for a hundred years. Human history reached a turning point: old wealth depended on trade and shipping, while new wealth rose with the advent of railroads and energy (oil).

In 2025, we find ourselves in a game strikingly similar. However, this time, the surging force is the computing power flowing through silicon chips, and the “gold” is the code engraved on the blockchain; the new era’s “gold” and “oil” are reshaping our entire understanding of productivity and store of value assets. Looking back at 2025, the market experienced unexpectedly intense turbulence. Trump’s aggressive tariffs forced global supply chains to relocate, triggering a massive inflation rebound; gold hit a historic high above $4,500 amid geopolitical uncertainties; the crypto market, which had welcomed the epic boon of the GENIUS Act at the beginning of the year, suffered a painful liquidation in early October due to leverage unwinding.

Beyond the macro volatility, a consensus in the AI computing industry is rapidly fermenting: Nvidia, the “AI Water Seller,” reached a milestone market cap of $5 trillion in October. Additionally, giants like Google, Microsoft, and Amazon have invested nearly $300 billion in AI infrastructure this year. For example, xAI’s upcoming completion of a million-GPU cluster by year-end signals the importance of computing power. Elon Musk’s xAI built the world’s largest AI data center in Memphis in less than half a year and plans to expand to an astonishing scale of 1 million GPUs by the end of the year.

The Era of Digital Intelligence: The Main Theme of the Next Industrial Revolution

Ray Dalio, founder of Bridgewater Associates, once said: “Markets are like machines; you can understand how they work, but you can never predict their behavior precisely.” Even though macro environments are unpredictable and random, it is undeniable that AI remains the primary long-term growth driver for the US stock market. Over the next decade, AI technology will become the most critical gear in the market machinery, continuously influencing governments, enterprises, and individuals.

Despite ongoing debates about an “AI bubble,” many institutions warn that the AI investment boom shows signs of bubble-like tendencies: Morgan Stanley research pointed out that in 2025, investment growth in AI has led to soaring tech stock valuations without significant productivity gains, reminiscent of the internet bubble of the 1990s.

However, an unavoidable fact is that the productivity revolution driven by AI is gradually entering a phase of tangible monetization. From an investment perspective, AI is no longer just a narrative for tech giants; the efficiency dividends and extreme cost optimization it brings are the main drivers of profit and productivity increases for non-tech companies. But behind this lies a brutal trade-off: AI’s impact on employment, especially white-collar jobs, is undeniable, with entry-level positions shrinking exponentially; basic coding, accounting, auditing, and even junior management consulting and legal roles may be among the first to be replaced by AI.

As AI applications deepen, risks of unemployment in healthcare, education, and retail sectors are mounting. Recently, a cruel joke has circulated in the US investment community: software engineers in the future will be like today’s civil engineers; Elon Musk has emphasized in interviews that AI will replace all jobs. This also signals the arrival of a new industrial era—called the “Digital Intelligence Era.”

Looking Ahead to 2026: Growing Demand for AI

Four Stages of Investment in the AI Industry

As the AI boom shifts from concept to full industry adoption, and markets have already priced in the MAG7 (the seven major US tech giants), where will the next wave of growth in AI themes come from? Goldman Sachs stock strategist Ryan Hammond proposed the “Four Stages of AI Investment” model, outlining the subsequent path: AI investment will sequentially go through four stages—chip, infrastructure, revenue enablement, and productivity enhancement.

AI Investment Four-Stage Model

Source:

Currently, the AI industry is at the intersection of “Infrastructure Expansion” transitioning into “Application Deployment,” moving from Stage 2 to Stage 3. Demand for AI infrastructure is exploding:

• By 2030, global data center electricity demand will increase by 165%
• From 2023 to 2030, the US data center power demand will grow at a CAGR of 15%, increasing the share of total US electricity consumption from 3% to 8% by 2030.
• By 2028, global spending on data centers and hardware will reach $3 trillion.

Goldman Sachs’ US data center power demand forecast

Image source:

Meanwhile, the generative AI application market is also experiencing explosive growth, projected to reach $1.3 trillion by 2032. In the short term, building training infrastructure will drive a 42% CAGR; in the medium to long term, growth will shift toward inference devices for large language models (LLMs), digital advertising, and specialized software and services.

Bloomberg: Growth Forecast for Generative AI in the Next 10 Years

Data source:

This judgment will be validated by 2026. Goldman Sachs’ latest macro outlook states that 2026 will be the “Year of Realized Returns” for AI investments, with AI significantly reducing costs for 80% of non-tech companies in the S&P 500. This will test whether AI can truly shift from potential to performance on corporate balance sheets.

Therefore, in the next 2-3 years, market focus will no longer be limited to a few tech giants but will expand further: deepening into AI infrastructure (such as power, hardware, data centers) and upward into companies across generalized industries that successfully convert AI into profit growth.

AI Computing Power as the “New Oil,” BTC as the “New Gold”

If AI computing power is the “new oil” of the digital intelligence era, driving exponential productivity leaps, then BTC (Bitcoin) will be the “new gold,” serving as the ultimate underlying asset for value anchoring and credit settlement.

AI, as an independent economic entity, does not need banks; it only needs energy. BTC is a pure “digital energy store.” In the future, AI will be the “fuel” of the economy, and BTC will be the “anchor” behind economic value. The issuance of BTC is entirely based on proof-of-work (PoW) driven by electricity consumption, aligning perfectly with AI’s essence—converting electricity into intelligence.

Furthermore, AI computing power, as a consumptive productive asset, has core costs rooted in electricity, and its value output depends on algorithm efficiency; BTC, as a decentralized store of value, fundamentally embodies energy monetization, naturally functioning as a “reservoir” to balance global computing power disparities over time and space. AI requires continuous, stable electricity, while BTC mining can absorb excess power from the grid during periods of surplus (like wind and solar peaks). When electricity is scarce (AI computation peaks), mining can instantly shut down, releasing power to higher-value AI clusters. This “Demand Response” stabilizes the grid: excess power during high supply (e.g., renewable peaks) can be used for mining load; during shortages, mining halts, freeing power for AI.

Genius Act: The Convergence of Stablecoins + RWA + On-Chain Computing Power

With the US passing the GENIUS Act in 2025, the dollar is gradually moving toward digital transformation, with stablecoins incorporated into the federal regulatory framework as an “on-chain extension” of the dollar system. This law not only injects trillions of dollars of new on-chain liquidity into US Treasuries but also provides a model for stablecoin regulation in key jurisdictions like the EU, UK, Singapore, and Hong Kong.

This regulatory framework first injects strong institutional momentum into the RWA (Real World Assets) market: under the boost of regulated stablecoins increasing global liquidity and supporting efficient cross-border settlement and trading, issuance and circulation of RWA will become more convenient. Stablecoins have become the main payment method for on-chain investments in real estate, bonds, art, and other RWAs, supporting fast global cross-border clearing.

Among these, AI computing assets, due to high input costs, steady yields, and heavy asset attributes, are naturally being viewed as standardized RWAs: whether GPU cloud computing, AI inference resources, or edge computing nodes, their pricing, leasing cycles, load factors, and energy efficiency can all be quantified and mapped via smart contracts on the chain. This means future business models like computing power leasing, revenue sharing, transfer, and collateralization will fully migrate to on-chain financial infrastructure for trading, settlement, and refinancing; additionally, computing power can be monitored in real-time via on-chain data, ensuring transparent and verifiable returns; supply can be flexibly scheduled on demand, reducing capital lock-up and resource idling typical of traditional heavy asset models, ensuring stable and transparent yields.

Even more excitingly, similar to the oil exchanges that emerged after the discovery of oil on Wall Street two centuries ago, AI computing power, through RWA, could become a standardized financial asset that can be traded, collateralized, and leveraged, enabling on-chain financing, trading, leasing, and dynamic pricing innovations; a new generation of “computing capital markets” based on RWA will have more efficient value transfer channels and limitless application potential.

“Dual Consensus” New Opportunities

In this new era where AI is fully integrated into our lives, computing power will serve as the consensus for high-efficiency productivity, and with the ultimate liquidity of high productivity—BTC will be redefined as the store of value.

So, companies that can master either “productivity” or “assets” will become the most valuable entities in future cycles. Cloud service providers are at the intersection of “BTC store of value” and “AI productivity” consensus. If computing power is the high-energy fuel driving the rapid operation of the digital economy, then cloud services are the intelligent pipelines carrying and distributing this power.

Global AI Cloud Service Market Size Forecast, Data Source: Frost & Sullivan

This includes major players: Microsoft, Amazon, Google, XAI, Meta. They are also called “Hyperscalers” (large-scale cloud providers), whose main business is IAAS (Infrastructure as a Service) for general needs. Although they hold vast pools of computing resources, they may be inefficient in resource scheduling. Hyperscalers are also the upstream providers of AI computing power, controlling most of the market’s resources and continuing to expand infrastructure:

Microsoft (: Launches the $100 billion “Stargate” plan to build a million-GPU cluster, providing extreme computing support for OpenAI’s model evolution.

Amazon ): Commits to investing $150 billion over 15 years, accelerating self-developed chips like Trainium 3, to decouple computing costs from external supply through hardware autonomy.

Google (: Maintains annual capital expenditure of $80-90 billion, leveraging its energy-efficient TPU v6 to rapidly expand AI dedicated cloud regions worldwide.

Meta: Mark Zuckerberg explicitly stated in earnings calls that Meta’s capital expenditure will continue to grow, with guidance raised to $37-40 billion in 2025, building the world’s largest open-source AI computing pool with liquid cooling tech and 600,000 H100-equivalent units.

xAI: Completed the world’s largest single supercomputing cluster, Colossus, in Memphis with “Memphis speed,” aiming for 1 million GPUs, demonstrating aggressive and efficient infrastructure delivery.

Other emerging cloud providers like CoreWeave, Nebius, and Nscale, branded as NeoCloud, focus on IAAS + PAAS (Platform as a Service). Unlike the general cloud services of giants, NeoCloud specializes in high-performance computing platforms for AI training and inference, offering more flexible leasing options, tailored scheduling solutions, faster response times, and lower latency.

They also stock top-tier GPUs (H100, B100, H200, Blackwell, etc.) and build high-performance AIDC, pre-assembled with liquid cooling, RDMA networks, and scheduling software, delivered quickly on flexible terms based on whole machine or entire park + daily charges.

Among these, Coreweave is undoubtedly a leading player. As one of the most prominent tech stocks in 2025, Coreweave’s core business is cloud computing and GPU acceleration services focused on AI training and inference. But it’s not the only new enterprise in the computing power rental space; competitors like Nebius, Nscale, and Crusoe are also strong contenders.

Unlike the heavy asset cluster battles of NeoCloud giants in Europe and America, GoodVision AI represents another possibility for globalized computing power—by smart scheduling and managing multiple users, it aims to build rapid-deployment, low-latency, high-cost-performance AI infrastructure in emerging markets with weaker power and infrastructure, promoting computing power equity. Meanwhile, giants are building million-GPU clusters in Memphis and elsewhere for larger models, while GoodVision AI deploys modular inference nodes across Asia and other emerging markets to solve the “last 100 kilometers” latency challenge for AI deployment.

It’s worth noting that most top AI computing service providers have a clear trait: their founding teams or core architectures are deeply rooted in the crypto mining industry. Transitioning from mining to AI computing is not a cross-industry move but a strategic reuse of core capabilities. BTC mining and high-performance AI computing are highly homologous at the fundamental level, both heavily reliant on large-scale electricity, high-power centers, and 24/7 operations. The cheap electricity channels and hardware management experience accumulated early on have become the most scarce and premium assets in the AI wave.

As AI computing demand grows exponentially, these companies naturally shift their existing infrastructure from “mining store of value (BTC)” to “output of productive computing power (AI).” With mature “dual switching” technology, BTC can effectively balance energy spatial-temporal and spatial distribution issues. Therefore, in the digital intelligence era, the “fuel” driving productivity leap will shift from oil to computing power, and the “underlying asset” anchoring its value will evolve from gold to BTC.

By integrating blockchain technology to put computing power on-chain as RWA assets, it becomes possible to create verifiable records of computing source, efficiency, and operational returns, as well as to build cross-region, cross-time smart contract settlement mechanisms—reducing credit risk and intermediary costs, and expanding applications in DeFi and cross-border computing power leasing. For example, edge nodes’ load, energy efficiency, and other parameters can be verified via PoW proofs and smart contracts, making edge inference computing power a tradable, collateralizable standardized financial product, enabling a “on-chain computing market.” The combination of computing power and RWA will further diversify on-chain assets, opening new liquidity spaces for global capital markets.

Connecting Productivity and Store of Value: Toward the Future of Computing Power Monetization

This is the practical embodiment of our earlier “Dual Consensus” logic: BTC is the top-level energy value anchor, and AI is the productivity application of energy. From this perspective, the era of “computing power as currency” will arrive faster and more disruptively than imagined. As humanity enters the digital intelligence era, the “fuel” driving productivity will shift from oil to computing power, and the “underlying asset” supporting its value consensus will evolve from gold to BTC.

We are now like spectators in 1859 standing on the muddy land of Pennsylvania, unable to imagine how that drill bit deep underground would open a new era of industrial civilization. Today, fiber optic cables extending to data centers worldwide are quietly building the arteries of this new era. Those early adopters betting on computing power and BTC will play the role of new “oil tycoons” in this transformation, redefining the distribution of wealth and power in the new cycle.