When Tether’s CEO warns about a bubble, the crypto world listens. But Paolo Ardoino’s recent critique of Big Tech’s AI boom carries an inconvenient truth for our own industry. In a July 2026 interview with BeInCrypto, he outlined four structural cracks — cost-revenue mismatch, capital lifecycle misalignment, demand elasticity, and the rise of open-source competition. On the surface, it’s a sobering analysis of centralized capital allocation gone wild. As a DAO Governance Architect who has spent years watching boardroom decisions create systemic risks, I see these cracks as a mirror held up to our own governance failures. The same patterns that inflate AI investment bubbles are baked into many crypto projects — yet we rarely admit it.
The context matters. Tether, the largest stablecoin issuer, sits at the intersection of traditional finance and crypto. Its CEO’s warning targets Apple, Microsoft, Google, and Meta, who are pouring hundreds of billions into AI infrastructure. JPMorgan estimates these four alone will spend $200 billion on AI in 2026, while Morgan Stanley warns that returns may take a decade. The Bank of England has even compared current AI valuations to the dot-com bubble. Ardoino’s four cracks are not new, but his framing — from a crypto insider — demands attention. He argues that Big Tech charges too little for AI compute to win users (cost-revenue mismatch), that chips may be obsolete in three to five years while capital is locked for decades (capital lifecycle mismatch), that demand will collapse if prices rise (elasticity), and that open-source models like Llama are eroding pricing power. Each crack is a governance failure: decisions made by a few in closed rooms, with no accountability to the broader community.
Yet as I read his analysis, I could not shake the feeling that I had seen this movie before. During the 2017 ICO boom, I launched Ethical Ledger, a workshop series in Chicago that trained over 150 retail investors to spot unsustainable tokenomics. The pattern was identical: protocols subsidized yields with token emissions, hoping to capture users and monetize later. Most never did. The DeFi summer of 2020 repeated the same script — yield farms offering 1,000% APY with no sustainable revenue. Today, Big Tech is doing the same with AI: using investor capital to subsidize API access, believing that market share today will justify profits tomorrow. It is a bet on central planning, and central planning fails. Core insight: When revenue is deferred and costs are front-loaded, only transparent governance can prevent a death spiral. In DAOs, we call this the "treasury mismatch" — and it is the number one cause of protocol collapse.
Let me drill into each crack through the lens of decentralized governance, drawing from my own scars.
First, the cost-revenue mismatch. Big Tech charges artificially low prices for AI compute to win users — Microsoft’s Copilot, Google’s Gemini, and Meta’s AI are all priced far below their operating cost. This is a deliberate strategy to build market share, but it assumes that future revenue will materialize. In crypto, this is analogous to a DAO that pays high rewards to liquidity providers without a clear path to sustainable fees. I saw this in 2020 when I co-designed UnityDAO’s governance structure. We had a $5 million treasury and faced pressure to distribute tokens aggressively to attract members. Instead, we implemented a quadratic voting system and held 42 community calls to build consensus. By focusing on long-term alignment rather than short-term subsidies, we achieved 300% higher proposal participation than industry averages. Big Tech’s mistake is that they have no equivalent of community governance to validate their assumptions. They are betting billions on a vision that no one voted for.
Second, the capital lifecycle mismatch. Ardoino points out that AI chips may be obsolete in three to five years, yet capital expenditure on data centers and power grids is amortized over decades. This is a classic issue of asset-liability mismatch. In DAO treasuries, I have seen the same mistake: locking funds into illiquid tokens that devalue faster than the vesting schedule. During the bear market of 2022, I organized "Rebuild Chicago," a peer-support network for displaced crypto workers. We raised $50,000 in personal funds to provide legal aid and career counseling. One thing I learned was that the communities that survived were those that had aligned their capital deployment with asset depreciation. They did not bet on one technology or chain; they diversified across liquid assets and stablecoins. Big Tech’s problem is that they are single-threaded — all in on GPU clusters that may become worthless. A DAO with proper governance would have a risk committee and a treasury diversification policy. Signature: Code without compassion is cold. Capital without governance is reckless.
Third, demand elasticity. Ardoino warns that if AI providers raise prices to cover costs, demand will drop sharply. This is the same dynamic we saw in the NFT market: when gas fees rose or floor prices increased, volume evaporated. In crypto, elastic demand is a feature of sovereignty — users vote with their feet. But it also means that projects with high fixed costs and no switching barriers are fragile. In my UnityDAO governance work, we noticed that members were willing to pay higher fees for proposals if they felt a sense of ownership. Elasticity decreases when there is genuine community attachment. Big Tech has no such attachment; users will leave the moment a cheaper alternative appears. Open-source AI models provide that alternative, just as Ethereum’s open development beat EOS and other closed chains. The lesson: loyalty cannot be bought with subsidies; it must be earned through transparent, inclusive governance.
Fourth, open-source competition. Ardoino sees open-source AI models as a threat to Big Tech’s pricing power. As a blockchain evangelist, I see this as validation of our core principle: open-source software is a public good that reduces rents. In my Ethical Ledger workshops, I spent hours translating technical whitepapers into accessible narratives. I found that the most successful projects were those that contributed to the commons — Ethereum, Bitcoin, Uniswap. They attracted talent and trust precisely because they were not closed. Similarly, open-source AI models like Llama are democratizing access. The risk for Big Tech is that they are investing in proprietary moats that become irrelevant. For crypto, the risk is different: we must ensure that our own open-source protocols have sustainable funding. That is why I advocate for DAOs to allocate treasury to public goods, not just to speculative tokens. Signature: Code without compassion is cold. Open source without sustainability is charity, not governance.
Now for the contrarian angle. Tether’s CEO may be right about AI, but is he the right messenger? Tether itself has never had a fully independent audit. Its reserves are opaque. The same centralization risk he criticizes in Big Tech is embedded in his own business. In my 2025 "Values First" coalition, we negotiated with BlackRock to adopt transparency protocols — but we also pushed back against stablecoins that refused audits. The AI bubble may indeed pop, but the deeper risk is that centralized power, whether in AI or crypto, concentrates capital without accountability. The contrarian truth is that the AI bubble may not pop because capital controls allow it to inflate indefinitely — just as Tether’s market cap has grown despite no audit. The real collapse comes when trust breaks, not when fundamentals fail. Crypto’s lesson is that trust must be verifiable on-chain. Big Tech’s AI boom is a reminder that off-chain trust is fragile. Signature: Build for humans, not just for chains. But also: hold yourself to the same standards you demand of others.
So what is the takeaway? The four cracks are not just about AI; they are about governance. As we build the next generation of decentralized infrastructure — whether for AI compute, stablecoins, or DAOs — we must ensure that capital allocation is transparent, that incentives align with long-term value, and that human agency remains at the center. The AI boom shows us what happens when a few decision-makers bet the house. In DAOs, we spread the risk and the reward. We require community votes, quadratic weighting, and treasury audits. That is the only sustainable path forward. The next time you see a protocol subsidizing usage with token emissions, remember Ardoino’s cracks. They are not just for Big Tech. They are for all of us.
