Cathie Wood, the high-profile founder of ARK Invest, has built a public case that artificial intelligence will drive prices down so fast and so broadly that traditional economies could buckle under the pressure. Her proposed answer to that deflationary force is bitcoin, a fixed-supply digital asset she argues can serve as a store of value when central banks inevitably respond by flooding markets with new money. The argument is bold, and it raises a question that matters for anyone holding dollars, stocks, or crypto: does the math actually support the thesis?
How Fast AI Costs Are Really Falling
Wood’s core claim rests on a simple observation: AI is getting cheaper at a pace that dwarfs most historical technology curves. To test that intuition against real data, I looked at a recent arXiv preprint titled “The Price of Progress: Algorithmic Efficiency and the Falling Cost of AI Inference.” The researchers constructed a dataset tracking how much it costs to achieve a given level of benchmark performance over time, and the results are striking. The paper documents large year-over-year decreases in the price required to hit specific AI performance targets, driven not just by better hardware but by smarter algorithms and more efficient model architectures.
What makes this research useful is its methodology. Rather than relying on anecdotal pricing from cloud providers, the study builds its conclusions from systematic benchmark and dataset construction, giving the findings a rigor that casual industry commentary often lacks. The pattern it reveals is not subtle. Inference costs, the expense of actually running a trained AI model to produce outputs, are collapsing in a way that echoes the early decades of Moore’s Law in semiconductors. If you are a business that competes on the cost of processing information, translating text, generating images, or analyzing documents, the floor is dropping out from under your pricing power. That dynamic is exactly what Wood means when she says AI is deflationary.
Wood’s Deflation-to-Crisis Theory
The leap from “AI makes things cheaper” to “AI triggers an economic crisis” requires several additional assumptions, and this is where Wood’s argument gets more speculative. Her reasoning, as expressed in various public interviews and ARK research commentary, follows a chain: AI-driven cost reductions spread across sectors, corporate revenues shrink as customers pay less, profit margins compress, layoffs follow, and consumer spending drops. In that scenario, central banks would face enormous pressure to cut interest rates and expand the money supply to prevent a deflationary spiral. Wood has suggested that such a response would debase fiat currencies, making hard assets with fixed supply, particularly bitcoin, more attractive as a hedge.
There is a kernel of logic here. Rapid technological deflation has caused real economic disruption before. The late 1990s saw the cost of computing, bandwidth, and storage plummet, which contributed to the dot-com bubble and its painful aftermath. But the broader economy did not enter a sustained deflationary depression. Prices fell in technology sectors while rising elsewhere, and central banks managed the transition without resorting to the kind of emergency monetary expansion Wood envisions. The difference she would likely point to is scale: AI touches nearly every industry simultaneously, from healthcare to logistics to legal services, which could make the deflationary pressure harder to contain in isolated sectors.
Bitcoin as Deflation Insurance
The second half of Wood’s thesis, sometimes presented as bitcoin being “the only fix,” is where the argument faces its sharpest scrutiny. Bitcoin’s fixed supply cap of 21 million coins is a mathematical certainty baked into its protocol, and that scarcity does give it a structural resemblance to gold. In a world where central banks print aggressively, assets with hard supply limits tend to appreciate in nominal terms. Wood has been consistent in framing bitcoin this way, positioning it as a digital version of the gold standard for an era of monetary excess.
The problem is that bitcoin’s track record as a deflation hedge is thin. During the brief deflationary scare of early 2020, bitcoin lost roughly half its value in a matter of days before recovering. It has historically traded more like a high-beta risk asset, rising and falling with speculative appetite, than like a stable store of value. Gold, by contrast, has centuries of data supporting its role as a crisis hedge. Critics of Wood’s position argue that calling bitcoin “the only fix” overstates its proven utility and ignores the asset’s extreme volatility, which makes it unreliable precisely when stability matters most. Framing bitcoin as the sole solution also sidelines other potential responses, from fiscal policy adjustments to productivity-sharing mechanisms, that could address deflationary pressures without requiring a bet on a single volatile asset.
Who Benefits If Wood Is Right
There is a distributional angle to this debate that deserves more attention. If AI does drive broad deflation, the benefits of cheaper goods and services will not land evenly. Companies and individuals who own AI infrastructure, the data centers, the model weights, the proprietary training data, will capture enormous value as their costs fall and their output scales. Workers displaced by AI-driven automation, on the other hand, face income loss in an environment where new jobs may not materialize fast enough. The wealth gap could widen significantly before any corrective policy kicks in.
Wood’s bitcoin prescription, in theory, offers a decentralized alternative. Anyone can buy bitcoin, and its appreciation would not depend on access to venture capital or tech industry connections. But in practice, bitcoin ownership remains heavily concentrated among early adopters and institutional investors. If a deflationary crisis did trigger a flight to bitcoin, the gains would flow disproportionately to those who already hold large positions, reinforcing rather than reducing inequality. A more testable version of Wood’s thesis might track whether bitcoin adoption rates in AI-disrupted sectors correlate with improved financial resilience for affected workers, but that data does not yet exist in any peer-reviewed form.
Where the Evidence Stands Today
The strongest part of Wood’s argument is also the most straightforward: AI inference costs are falling fast, and the trend shows no sign of slowing. The preprint on algorithmic efficiency provides hard numbers for something many practitioners feel intuitively, namely that each new generation of models delivers more capability per dollar of compute. If those efficiency gains continue, they will put persistent downward pressure on prices in any industry where AI can substitute for human labor or legacy software systems. That is a real deflationary force, even if its exact magnitude is still uncertain.
What the evidence does not yet show is a clear, quantified path from that technological deflation to the kind of systemic crisis that would force central banks into unprecedented money creation. Historical episodes of rapid productivity growth have been disruptive but ultimately compatible with economic expansion, especially when accompanied by policy that helps workers transition and share in the gains. Bitcoin may well benefit if investors grow more anxious about currency debasement, but its role as “the only fix” remains an assertion, not an empirically grounded conclusion. For now, the data supports a narrower takeaway: AI is likely to cheapen many things we buy and use, and that will challenge existing business models and policy frameworks long before it definitively validates, or disproves, Wood’s most dramatic predictions.
More From The Daily Overview
*This article was researched with the help of AI, with human editors creating the final content.
Grant Mercer covers market dynamics, business trends, and the economic forces driving growth across industries. His analysis connects macro movements with real-world implications for investors, entrepreneurs, and professionals. Through his work at The Daily Overview, Grant helps readers understand how markets function and where opportunities may emerge.
