On August 9, Cathie Wood stood before a Bloomberg camera and planted a flag. The fiscal deficit, she argued, is shrinking toward 5% of GDP—a level not seen since the early 1980s. Oil prices are collapsing. AI capital expenditure has broken through a 30-year ceiling. The result? A deflationary shock that will reshape everything, including the role of Bitcoin and stablecoins. She called Bitcoin and stablecoins the twin beneficiaries of an 'agentic commerce' revolution—a world where AI agents trade with each other, using crypto rails as the settlement layer. It's a seductive narrative. It's also, I suspect, a deeply flawed one.
I've spent the last decade auditing smart contracts, founding a crypto education platform, and watching narratives masquerade as fundamentals. Truth is immutable, unlike the price action. And this particular truth needs unpacking not through the lens of macro optimism, but through the cold, hard reality of blockchain infrastructure. So let me walk through Wood's thesis, test it against the technical and ethical constraints I've encountered in the field, and offer a contrarian take that might save you from over-indexing on a vision that's more aspiration than architecture.
The Hook: A Deflationary Bet on Immutable Code
Wood's core argument is elegant: AI-driven productivity gains, combined with fiscal discipline and falling energy costs, will produce a deflationary environment. In such a world, traditional inflation hedges like gold become less relevant. Instead, Bitcoin—the most scarce, decentralized asset—becomes the preferred store of value for an economy where machines, not humans, are the primary earners. Stablecoins, meanwhile, become the medium of exchange for machine-to-machine payments, bypassing slow, centralized banking rails.
The data points she cites are real. The U.S. fiscal deficit as a percentage of GDP has indeed dropped from pandemic highs, and the Biden administration's recent budget agreements suggest further tightening. Oil prices have softened, and AI capital expenditure from hyperscalers is unprecedented. But here's where my experience auditing Tezos's mainnet launch in 2017 kicks in: code is law, but only if it compiles. And Wood's macro code has a few syntax errors.
Context: The Infrastructure Gap in Agentic Commerce
Before we dive into the core analysis, let's clarify what 'agentic commerce' actually means. Wood envisions a future where AI agents autonomously negotiate, purchase, and settle transactions for everything from cloud compute to raw materials. These agents need a digital native payment system—one that is programmable, trustless, and global. Bitcoin and stablecoins fit that bill on paper.
But paper is not production. In my 2020 work with OpenLedger Lab, I mentored dozens of developers building DeFi protocols. The most common failure point wasn't the economic model—it was the oracle feed. Smart contracts depend on accurate, real-world data. For agentic commerce to work, agents need to trust that the price of an asset or the status of a delivery is correct. Today, the dominant oracle solution is Chainlink, but its decentralization is a joke: a handful of nodes run the majority of feeds. In a world where millions of agents are transacting every second, a single compromised oracle could trigger a cascade of liquidations. Wood doesn't mention this. She assumes the infrastructure exists. It doesn't.
Core: Testing the Deflation-Immutable Asset Thesis
Let's hold Wood's deflation thesis up to the light of blockchain fundamentals. Bitcoin's value proposition has always been 'hard money'—a hedge against central bank money printing. If we enter a deflationary period, the narrative shifts: Bitcoin becomes a deflation hedge, which is actually more powerful than an inflation hedge. In deflation, cash gains purchasing power. Bitcoin, being fixed supply, should appreciate even more as demand for scarce assets rises. That part is sound.
But here's the rub: Bitcoin's transaction throughput is seven transactions per second. For agentic commerce, where agents might need to settle millions of micro-transactions per hour, Bitcoin is unusable. The Lightning Network helps, but it's still a centralized hub-and-spoke model. The real Bitcoin community doesn't acknowledge 90% of so-called 'Bitcoin Layer2s' as legitimate—they're Ethereum projects rebranding for hype. I've audited several of these 'Bitcoin L2s' and found them to be sidechains with multi-sig bridges, not trust-minimized rollups. Wood's vision of Bitcoin as the settlement layer for AI agents is technically premature.
Stablecoins face a different set of issues. USDC and USDT are centralized. They rely on bank accounts, auditors, and regulatory compliance. For agents to use them, they need to trust that Circle or Tether won't freeze their funds. In a world where regulators are increasingly hostile to anonymous transactions, agentic commerce could be forced onto permissioned stablecoins, defeating the purpose of decentralization. I've seen this play out in my work with the Decentralized Trust Protocol for AI: the tension between sovereignty and compliance is real, and stablecoins are the battleground.
Contrarian: The Real Risk Is Not Deflation—It's Stagflation of Trust
Wood's most provocative claim is that the AI bubble is overstated. She argues that productivity gains will materialize, driving down costs. But from my experience in the 2022 bear market, I learned that technological breakthroughs often come with prolonged periods of disillusionment. The Terra-Luna collapse taught me that algorithmic stability is a myth. The current AI hype cycle feels eerily similar: massive capital deployment, but the actual use cases for agentic commerce are still in the lab. The 'agentic commerce' narrative is a marketing term, not a technical specification.
What if the real outcome is not deflation, but a stagflation of trust? As AI agents become more autonomous, the risk of manipulation, fraud, and systemic failure grows. The same qualities that make Bitcoin attractive—immutability, irreversibility—could become liabilities in a world where agents make mistakes. I've seen smart contracts with bugs that cost millions. Now imagine an AI agent executing a flawed trade on a buggy protocol, and there's no human to stop it. The legal and ethical implications are staggering. Wood's thesis assumes a frictionless transition; my experience suggests it will be anything but.
Takeaway: Build for Human Dignity, Not Machine Efficiency
I've spent the last three years writing 'The Soul of Sovereignty,' a manuscript arguing that blockchain must serve human dignity, not just capital efficiency. Wood's vision of agentic commerce is a natural extension of that efficiency-first mindset. It ignores the human cost: the workers displaced by AI, the regulators struggling to keep up, the communities left behind. The true value of Bitcoin and stablecoins lies not in serving machines, but in empowering individuals to transact freely, without permission.
So as you read Wood's op-eds and watch ARK's market moves, ask yourself: Are we building infrastructure for humans, or just for a new class of machine overlords? The answer will determine whether this deflationary thesis produces lasting value—or just another bear market dressed in AI clothing.
Truth is immutable, unlike the price action. The foundation must be built on ethical integrity, not just mathematical precision. That's the lesson I carry from 2017 to today, and it's the one I hope you carry forward.
