Every few years a new investment theme reshapes how capital gets allocated and how markets price risk. The current AI frenzy is starting to look less like a technology cycle and more like a physical infrastructure buildout—one that BitMEX co-founder Arthur Hayes believes could end in a bust comparable to the 2008 credit crisis. In a new essay titled Situationship , published as the original report shows, Hayes outlines why the AI investment wave is not the same as the dot-com bubble, and why that distinction matters for Bitcoin holders.
The core of the argument is structural. The dot-com era was mainly a software and equity bubble—capital poured into companies that had little revenue but big promises. The AI boom, however, is anchored in massive spending on physical infrastructure: data centers, chips, power lines, and real estate. Hayes calls it a real estate buildout more than a tech cycle. When you fund a million square feet of data center space and lock in long-term power contracts, the financial risk starts to look like a construction loan gone wrong, not a venture bet that can be written off overnight.
The Infrastructure Overbuild Parallel
What makes the comparison to 2008 sticky isn’t just the scale of the investment. It’s the leverage embedded in it. Project finance for data centers, hyperscaler expansions, and hardware supply chains involves layers of debt that don’t unwind cleanly when demand assumptions shift. If AI revenue falls short of expectations—whether because enterprise adoption slows or because model efficiency reduces the need for brute compute—those debt obligations don’t just vanish. That could cascade through credit markets in a way that software-focused tech crashes never did.
Hayes isn’t making a call on when the correction might arrive. He’s mapping a scenario where the unwind triggers a broader financial shock, forcing central banks to react. The Fed and other authorities, already operating in an era of high sovereign debt and fragile risk appetite, would likely flood markets with liquidity again—exactly what happened after the subprime housing collapse. And that is where Bitcoin enters the picture, not as a tech asset but as a monetary hedge against central bank activism.
Monetary Easing as an Unintended Catalyst for Crypto
The thesis is macro, not crypto-native. A contraction in AI spending that rattles credit markets would almost certainly prompt emergency policy support: rate cuts, asset purchases, perhaps new liquidity facilities. Hayes expects that aggressive easing would reignite the bull market for Bitcoin and the broader crypto space, much like the post-2008 cycle that eventually produced the 2017 and 2020-2021 runs. The logic is that when fiat liquidity gets expanded to repair a broken financial system, hard-capped digital assets become attractive alternative stores of value.
That narrative is already finding echoes in parts of the crypto market today. AI-linked crypto projects continue to attract speculative interest. $X@AI BRC-20 NFTs recently led weekly sales volume with $17.8 million, showing how AI thematics have seeped into on-chain trading. The data storage angle is even more direct: Filecoin’s price outlook is being shaped by expectations that demand for decentralized storage will rise as AI training and inference require ever-larger datasets. Meanwhile, efforts to build scalable AI-driven Web3 applications, such as the recent partnership between UXLINK and Origins Network , underline that the convergence is already under way.
These pockets of activity are not evidence that Hayes is right, but they illustrate how tightly interwoven crypto and AI have become in market psychology. If an AI downturn hits, it won’t be contained to equities. Crypto markets would absorb the sentiment blow first, because they are more liquid and more reactive to macro than legacy tech investors often expect. The real question is how quickly the policy response materializes, and whether it shifts enough capital into digital assets to offset the initial pain.
What Remains Uncertain
A few pieces of this puzzle are unknowable. First, the baseline assumption that AI infrastructure is overbuilt is still contested. Hyperscalers are signing long-term power deals because they project sustained exponential growth in demand. That forecast could prove accurate, pushing any reckoning years into the future. Second, central banks might not be able to ease as aggressively as they did in 2008 if inflation remains sticky or if sovereign debt ceilings become binding. In that scenario, the correlation between a credit event and a Bitcoin rally breaks. Third, even if liquidity does flood the system, the path from macro liquidity to crypto prices is not perfectly linear. The past two big cycles had strong institutional on-ramps; a future cycle might require deeper ETF infrastructure and a different regulatory posture.
Hayes’ framing also raises an uncomfortable question for crypto investors who are cheering the AI narrative today. If the entire compute infrastructure is being financed with shaky assumptions, then many of the AI-token projects riding the hype will not survive a credit contraction long enough to benefit from any eventual monetary easing. The liquidity tide might lift Bitcoin, but it may leave dozens of AI-theme altcoins stranded.
Still, the essay serves as a useful constraint on excessive enthusiasm. It reminds market participants that the AI buildout is not a purely digital story—it’s anchored in real estate, power grids, and debt markets. And in the world of macro finance, those are the same ingredients that have produced the most damaging booms and busts of the last two decades. For Bitcoin, the scenario sounds paradoxical: a crash elsewhere becomes the fuel for a new leg up, but only after the policy response kicks in. That is a sequence many traders will watch closely as compute capacity numbers, AI revenue forecasts, and central bank rhetoric evolve over the coming quarters.
