The AI Bubble Isn't Bursting – It's Rolling: What This Means for Crypto

A top macro strategist just dropped a bombshell that flips the AI panic narrative on its head. Dhaval Joshi, chief strategist at BCA Research, isn't predicting a single, cataclysmic AI bust. Instead, he’s warning of a rolling bubble – a cascade of mini-manias moving through different layers of the tech stack. This isn't just an AI story. It's a liquidity pattern that anyone who lived through the 2017 ICO sprint or the DeFi summer of 2020 will recognize instantly. The market is breathing, not dying. And for crypto, this could be the signal to watch. Pulse on the chain, breath in the market. Let me break down what Joshi actually said. The mainstream take is that AI is a super-bubble about to explode. Joshi disagrees. He argues that the AI mania rotates – from infrastructure (GPUs, data centers) to models (OpenAI, Anthropic) to tools and finally to applications. Each phase gets its own overvaluation, its own correction, and then the capital moves on. This is not a single crash. It's a slow bleed of capital misallocation across sectors. The risk isn't that everything collapses at once. The risk is that the capital keeps chasing the wrong horse, and the total misallocation grows until the whole system becomes fragile. Caught in the flash, framed in fact. Now, why should a crypto analyst care? Because I've seen this movie before. In 2017, the ICO bubble rolled from Bitcoin to Ethereum to altcoins to utility tokens. In 2020, DeFi summer rolled from lending protocols to DEXs to yield farms to NFTs. Each wave had its own hype, its own peak, and its own crash. The market never died – it just moved. The same pattern is playing out in AI. The infrastructure layer (Nvidia, data centers) had its run. Now the model layer (OpenAI, Anthropic) is peaking. The next wave is applications. And the capital misallocation? It's real. The analysis shows that AI capex by the big four tech companies exceeded $200 billion in 2024, while AI revenue growth, though strong, lags far behind. This gap is a ticking clock. Seventy-two hours without sleep, zero doubts. Let me dive into the numbers. According to the report, Joshi's framework implies a four-layer AI stack: infrastructure, models, tools, and applications. The rolling bubble means that at any given time, one layer is overvalued while others are undervalued. For example, in 2023, the infrastructure layer (Nvidia) was the hot spot. In 2024, the model layer (OpenAI, Anthropic) took the baton. Now, in early 2025, the capital is starting to rotate into applications like Palantir, Grammarly, or even AI agents. The danger? When the capital leaves a layer, it doesn't just deflate – it can crash. The 2024 correction in GPU stocks (Nvidia dropped 20% in a month) was a preview. The question is: which layer is next? Based on my experience tracking on-chain flows during the 2022 bear market, I can tell you that the same mechanism applies to crypto. The capital rotation in AI is mirrored by a rotation in crypto assets. When AI infrastructure booms, Bitcoin stagnates. When AI models get funded, Ethereum drops. When AI applications get hyped, altcoins pump. This is not coincidence. It's the same global liquidity pool moving between risk assets. The rolling bubble in AI is a leading indicator for crypto rotations. Sensing the tremor before the earthquake hits. But here's the contrarian angle that everyone is missing. The rolling bubble might actually be healthy for innovation. It prevents a single, systemic crash by allowing capital to flow to new use cases. The 2000 dot-com bubble burst because all the capital was in one place (internet stocks). The AI rolling bubble spreads the risk. However, the real danger is that the final bubble – the application layer – will be the biggest and most fragile. If AI applications fail to generate scalable revenue, the capital misallocation will be enormous, and the correction could be brutal. For crypto, this means that the next six months are critical. If AI applications start to show real revenue, the capital will stay in tech. If they don't, the liquidity will rotate into crypto as a hedge. Running where the liquidity flows fastest. Let me give you a concrete signal to watch. The analysis lists several key metrics: GPU rental prices, cloud capex growth, and AI startup funding rounds. I'll add one more: the on-chain activity of AI-focused crypto projects. For example, if you see a sudden spike in transactions on the Render Network (RNDR) or Akash Network (AKT), it could mean that AI compute demand is moving from centralized cloud to decentralized infrastructure. That's a rotation within the rotation. The capital is flowing from base layer to application layer, and also from centralized to decentralized. The market is breathing, not dying. Now, the takeaway. The AI bubble isn't going to burst tomorrow. But it's going to roll. And where it rolls next will determine the direction of crypto for the next 12 months. The key is to watch the on-chain pulse. When the AI application layer peaks, the capital will look for the next big thing. And crypto – with its decentralized compute, AI agents, and tokenized data – is the natural next stop. The question isn't whether AI is overvalued. It's where the next wave of liquidity will land. Keep your eyes on the chain. The market is breathing, not dying. Pulse on the chain, breath in the market.