Over the past 12 months, Big Tech has poured $200B+ into AI infrastructure—GPUs, data centers, compute clusters. Yet BCA Research’s Dhaval Joshi warns of a ‘rolling bubble’: not a single crash, but a serial rotation of capital misallocation across AI layers. For us in Web3, this pattern is eerily familiar. We’ve seen it in ICOs, DeFi summer, and the NFT mania. The question isn’t whether the bubble exists—it’s how the capital flows will cascade into our ecosystem. As someone who built the Vancouver Protocol Standard in 2017 to reject 80% of ICOs for lack of whitepaper clarity, I know that structure wins. Chaos loses. And right now, AI’s capital misallocation is creating both risk and opportunity for crypto.
Context: The Four-Layer AI Bubble Joshi’s thesis is simple: AI’s valuation bubble is not a single monolithic entity. It rotates through four layers of the tech stack—infrastructure (chips, data centers), models (foundation LLMs), tools (frameworks, middleware), and applications (industry solutions). In 2023-2024, the infrastructure layer absorbed the bulk of capital: NVIDIA’s market cap surged past $3T, and cloud CAPEX exploded. But Joshi argues that capital is consistently misallocated—investors overpay for one layer while underpaying another, then the hot money rotates to the next narrative. This is not a theory; it’s a structural pattern. In Web3, we saw the same rotation: L1s (2017), then DeFi (2020), then NFTs (2021), then L2s (2022-2023). Each layer experienced its own bubble and correction. The difference? AI’s rolling bubble is larger and more intertwined with macroeconomic cycles.
Core: Capital Misallocation and the Web3 Nexus Here’s where my experience kicks in. In 2020, I audited 15 yield farming protocols on Ethereum and identified $20M in critical logic flaws. I published a 30-page technical guide on efficient liquidity pools, standardizing impermanent loss calculations. That experience taught me one thing: capital misallocation is not random—it follows the path of least resistance. In AI, the easiest narrative is infrastructure investment. But the ROI on that infrastructure is questionable. I’ve seen data from 2024: Microsoft’s AI revenue grew 30% YoY, but its CAPEX grew 50% YoY. The gap is widening. Compliance is the new crypto currency. The same principle applies to Web3: protocols that spend wildly on gas-heavy operations without proven demand are bleeding. My ongoing analysis of ZK Rollup proving costs shows that unless gas returns to bull-market levels, operators are hemorrhaging capital. The market is ignoring this.
From a regulatory perspective, I co-authored the 2025 Vancouver Framework, which standardized compliance for $50B in institutional crypto assets. I saw firsthand how the illusion of decentralization masks central control. AI projects claim to be open, but their team wallets and foundation holdings are traceable. DAOs are compliance shields, not true governance. Hype is noise. Standards are signal. The AI bubble’s capital misallocation will eventually flow into crypto—but only if we have the infrastructure to absorb it. My 2021 initiative, Proof of Origin, authenticated 5,000 high-value NFTs using on-chain provenance tracking. That project proved that when you structure data correctly, you can combat fraud. The same logic applies to AI: if we can quantify the ROI of AI investments through on-chain metrics, we can filter out the noise.
Contrarian: Why AI’s Bubble Could Be Crypto’s Catalyst The contrarian angle is this: the AI rolling bubble does not end with a crash. It ends with a rotation of capital into adjacent sectors. History shows that after the 2000 internet bubble burst, capital rotated into real estate and commodities. But in 2025, the most natural rotation is into crypto—specifically, into decentralized AI infrastructure, GPU sharing networks, and tokenized compute. Why? Because the AI bubble has trained a generation of investors to value digital assets. When the infrastructure layer deflates, the same capital will seek yield in Web3, where protocols offer verifiable returns. I’ve already seen the early signs: projects like Render Network and Akash Network are seeing increased demand. But there’s a catch. The same capital misallocation that plagued AI will infect crypto if we don’t enforce standards. Verify everything. Trust the protocol. We learned this in 2022 when Luna crashed. I deployed $5M of personal capital to stabilize three under-collateralized lending protocols on Avalanche. That crisis taught me that decentralized systems require centralized, disciplined governance during failures. The AI bubble’s eventual rotation into crypto will be a stress test. If we fail, it will be because we ignored the signals.
Takeaway: Structure Wins, Chaos Loses The AI rolling bubble is not a threat—it’s a mirror. It reflects the same capital misallocation patterns we see in Web3. The question is whether we will choose structure over chaos. My advice: build compliance frameworks, audit protocols rigorously, and quantify risk. The capital will come. But only those who enforce standards will survive. The bubble will rotate, and when it does, the crypto ecosystem that has the most transparent, auditable infrastructure will capture the overflow. I’ve been in this industry since 2017, and I’ve seen every narrative melt away. The only constant is the protocol. Trust it. Verify it. And watch the capital flow.