We don’t need more narratives; we need more cycles.
Franklin Templeton, a firm managing over $1.5 trillion, recently issued a quiet yet devastating warning: the semiconductor industry’s AI-driven boom is sowing the seeds of its own bust. The warning—rooted in the classic silicon cycle—argues that the market has already priced in years of growth, while capacity expansion and demand concentration create a fragile peak. It’s a warning that feels hauntingly familiar to anyone who has lived through a crypto cycle.
In 2017, I spent weeks auditing a whitepaper for a project called OmniChain—a decentralized identity protocol promising to democratize global finance. The rhetoric was beautiful. The tokenomics, however, heavily favored early investors. I wrote a 5,000-word exposé that went viral shortly before the project’s rug pull. That betrayal taught me something fundamental: narratives are the most powerful drugs in our industry. They blind us to the underlying cycles. Franklin Templeton’s warning is not just about silicon wafers; it’s about every market where storytelling outpaces fundamentals.
Let’s strip the semiconductor analogy down to its core. The AI boom has driven extraordinary demand for HBM and DDR5 memory, pushing SK Hynix and Micron to record valuations. But the same cycle logic applies to crypto—especially the AI narrative that has inflated token prices for projects like Render Network, Akash Network, and countless zero-revenue AI agents. The market has priced in not just current AI compute demand, but the promise of infinite scaling. Sound familiar?
During the 2022 bear market, I retreated to a cabin in Yilan for three months. The Terra Luna collapse had drained me. I stopped watching prices and started journaling about trust. I realized that every crypto cycle is a failure of narrative discipline. We convince ourselves that “this time is different” because of a new technology—DeFi, NFTs, Layer 2 scaling, now AI. But the underlying mechanism remains: capital overcommits to a single story, capacity overshoots demand, and the correction wipes out the leverage.
Let me offer a data point that most overlook. According to my own tracking of on-chain activity for AI-crypto projects, the top ten tokens by market cap have a median 90-day trading volume that is 400x higher than their actual compute transactions. That’s not utility; that’s speculation dressed up as infrastructure. The semiconductor industry has a parallel: memory chip prices are driven by data center CAPEX, not actual end-user demand. Both markets are pricing hope, not usage.
In 2024, I founded The Alignment Circle, a community focused on ethical governance. I mentored fifty builders on DAO structuring. One thing became painfully clear: most projects build for the narrative, not the network. They design tokenomics to maximize initial extraction, then rely on hype to sustain valuation. Franklin Templeton’s warning is a mirror for crypto’s own cycle of narrative inflation. The question is whether we will learn from their history or repeat ours.
The contrarian take is this: maybe cycles are smoothing out. Institutional capital, regulated vehicles like ETFs, and longer-term holders could dampen volatility. But I’ve seen the same argument made in every cycle since 2017. The data says otherwise. The Bitcoin ETF approval turned BTC into “Wall Street’s toy,” as I wrote in an essay earlier this year. It did not end volatility; it simply moved the leverage to options and derivatives. The cycle is still there, just hidden under a layer of financial engineering.
Where does this leave us? The steward’s job is not to predict the peak, but to build for the valley. We don’t need more users; we need more stewards. Trust is the only protocol that cannot be coded. These are not platitudes—they are actionable frameworks. When I saw the Franklin Templeton warning, I immediately reviewed my own portfolio of governance tokens. I reduced exposure to projects with no lock-up periods and high inflation rates. I increased allocations to protocols with proven revenue and community ownership.
Here is the forward-looking judgment: the next crypto crash will not be caused by regulation or hacks. It will be caused by narrative exhaustion—when the AI story stops growing faster than the underlying compute usage. That moment is 12 to 18 months away, coinciding with the saturation of HBM supply and the slowdown of CSP CAPEX. The most important signal to watch is not token price but the ratio of active developers to daily active users. When that ratio drops below 0.5, the narrative has already peaked.
We built not for the peak, but for the valley. Franklin Templeton’s warning is a gift—a chance to recalibrate before the music stops. Listen to the silence. The signal is there.