Hook
904 million yuan. That's the number Zhiyang Innovation, a power sector digitalization firm, is betting on AI and embodied intelligence. The capital raise, announced in mid-August, is not just another industrial pivot. It's a signal that the traditional economy is finally waking up to the AI gold rush. But here's the catch: the money is flowing into centralized infrastructure. The ledger does not lie, but it rewards patience. The question is: will the market wait for decentralized alternatives to catch up?
Context
Zhiyang is not a typical AI company. It's a legacy player in power grid monitoring, with deep customer relationships in state-owned utilities. Their plan: allocate funds across four buckets—embodied intelligence (long-term), AI development (core), intelligent perception terminals (mid-term commercialization), and energy facilities (short-term infrastructure). The structure mirrors what I saw in 2017 during the ICO speed run: capital funneled into hype-driven narratives, but with a clear intention to dominate a vertical. Back then, it was Ethereum-based tokens. Today, it's "embodied intelligence."
From the noise of 2017 to the signal of today, the pattern repeats. Traditional companies are using equity markets to finance their AI transformation. Meanwhile, blockchain-based AI compute networks like Render Network, Bittensor, and Akash Network are still fighting for mindshare. The contrast is stark. Zhiyang is raising $125M in debt/equity. The top decentralized compute networks haven't raised that much in their entire history.
Core
Let's break down the allocation. The largest chunk goes to "multi-domain embodied intelligence and AI development." This is the R&D engine. Based on my experience auditing 45+ ICO whitepapers, I can tell you that vague language like "multi-domain" is a red flag. It means the company hasn't decided on a target market. But in this case, it might be strategic. Zhiyang already owns the power sector. They're betting that embodied AI for power line inspection, substation monitoring, and energy management will be the killer app.
The second tranche is for "general-purpose AI perception terminal industrialization." This is hardware. Sensors, cameras, edge computing devices. They're building the physical layer. The third is "energy facilities"—likely data centers or dedicated compute infrastructure. The fourth is working capital and debt repayment.
Speed runs require foresight, not just reaction. Zhiyang is reacting to the AI wave, but their foresight is limited. They're building centralized, proprietary systems. The ledger—the blockchain—offers a different path: decentralized, verifiable compute. But the market isn't ready. Institutional capital is still scared of the volatility.
I've seen this before. In 2020, during the DeFi yield war, I wrote a report titled "The Siphon Effect" predicting the liquidity crisis. The same dynamic is playing out here. Traditional companies are siphoning capital from equity markets, but the ROI is uncertain. The risk is execution. Zhiyang has no track record in AI. They're a power company. The same goes for most industrial firms pivoting to AI.
Contrarian
Here's the unreported angle: The capital raise is a double-edged sword for decentralized compute. On one hand, it validates the market. If traditional players are spending billions, the demand for AI compute is real. On the other hand, it reveals the weakness of blockchain-based alternatives. Decentralized networks are still too small, too volatile, and too complex for industrial clients. They can't handle the scale or compliance requirements.
But that's the blind spot. The ledger does not lie. On-chain data shows that decentralized compute usage is growing, but from a tiny base. The market is pricing in a future where centralized AI wins. The contrarian bet is that the opposite will happen. Centralized infrastructure is fragile. A single point of failure. Power companies like Zhiyang will face cyberattacks, regulatory scrutiny, and hardware supply chain issues. Decentralized networks offer resilience.
Consider the 2022 NFT market crash. I analyzed 500,000 on-chain transactions to prove the unsustainable player-to-earn model. The same analysis applies here. Centralized AI is a bubble. The capital is flowing, but the returns will be concentrated in the hands of incumbents. The real alpha is in the infrastructure that enables verifiable, trustless compute. The market is ignoring this because it's complex. But complexity is where the edge lies.
Takeaway
Speed runs require foresight, not just reaction. The market is fixated on the next AI IPO. But the real opportunity is in the networks that power the next generation of AI. Zhiyang's raise is a signal that the centralized AI train is leaving the station. But the decentralized track is still being built. The question is: will the market realize the value of verifiable compute before the next cycle?
