Floor price broken. Truth verified. A deep-dive analysis of Chengdu’s newly released “AI+” Action Plan—spanning seven dimensions from technology to infrastructure—reveals a glaring omission: the absence of blockchain as a foundational trust layer. The plan, aiming for 2600 billion yuan in AI core industry output by 2030 and a 70% penetration rate for “new-generation intelligent terminals and agents,” reads like a classic top-down industrial push. But for those of us who have lived through the 2018 post-crash community rebuilds and the 2022 Terra Luna liquidity defense, the pattern is disturbingly familiar. Governments love scale targets; they fear the messy, decentralized infrastructure that actually makes them verifiable. Data checked. Community warned.
Context: Why Now? Chengdu’s plan is not an isolated document. Across China, provincial AI strategies are proliferating—Beijing targets foundational models, Shenzhen bets on hardware, and Hangzhou pushes cloud-native AI. Chengdu wants to be the “first city of AI applications,” leveraging its strength in electronics manufacturing (Foxconn, Intel) and software talent (Tianfu Software Park). The plan explicitly calls for “AI empowering thousands of industries,” with annual 20 benchmark scenarios and double hundred projects (100 innovative products, 100 demonstration scenes). The target is bold: 2600B yuan at a >30% CAGR, more than double the national AI industry growth rate. But as a blockchain engineer turned journalist, I see the critical weak link: trust. Without blockchain, how do you prove that an AI agent is acting within its rules? How do you ensure that the 70% penetration metric is not just another government statistic—a number that, as I learned during the 2021 NFT floor price verification sprint, can be easily gamed by wash trading or misclassification? Trust bridge crossed. Crash imminent.
Core: The Blockchain Blind Spot in Seven Dimensions Let’s walk through the seven dimensions of the analysis and expose where blockchain could—and should—have been woven in.
Technology Route: The analysis finds that Chengdu’s plan mentions no specific AI model or algorithm innovation, focusing instead on penetration rates. From my MS in Blockchain Engineering perspective, this is a red flag. Any AI agent deployment at scale—especially those making autonomous transactions (think crypto wallets, oracles, smart contract execution)—requires a blockchain backbone for auditability. The plan defines “new-generation intelligent terminal and agent” penetration, but where is the decentralized identity (DID) or verifiable credential layer? Without it, agents are just black boxes. Floor price broken. Truth verified. The plan implicitly favors edge AI and AIoT, which aligns with Chengdu’s hardware base, but it ignores that edge devices need tamper-proof logs—blockchain’s forte. The hidden implication is that Chengdu wants to adopt existing mature models (like Huawei MindSpore) rather than develop its own. Fine. But those models, when deployed in government or financial contexts, become single points of failure. I’ve audited enough smart contracts to know that centralization in AI is not a hypothesis; it’s a pending exploit.
Commercialization: The analysis calls the commercialization path “scenario-driven plus policy subsidies,” warning of sustainability risks once subsidies stop. Here, blockchain offers programmable incentives—token-based loyalty, revenue-sharing smart contracts, or decentralized data markets. Yet the plan is silent on tokenization. In 2024, during the BlackRock ETF integration story, I saw how traditional assets gained transparency via on-chain verification. Chengdu’s 2600B target, if partly tokenized, could attract global capital and create self-sustaining ecosystems. Instead, the plan relies on government procurement and ‘incubation funds.’ As I know from my 2022 Terra Luna experience, when liquidity is artificial and dependent on a single authority, liquidity gone. Run. The analysis correctly notes that the commercialization cycle is opaque, with no exit mechanisms or pricing models. A blockchain-based marketplace for AI services could solve that—smart contracts settle payments automatically, and on-chain reputation prevents fraud. But none of this is in the document.
Industry Impact: The analysis rightly identifies key beneficiaries: electronics, manufacturing, finance, and cultural tourism. But it misses how blockchain can amplify those impacts. For example, in finance (Chengdu Bank), AI-driven lending needs immutable credit histories—permissioned blockchain is already solving this. In manufacturing, supply chain visibility for AI-optimized production lines requires tamper-proof provenance, not just IoT sensors. I witnessed this firsthand during the 2026 AI-agent privacy advocacy framework: users demanded control over their data, and the best mechanism was a consent protocol recorded on-chain. Chengdu’s plan pushes for “intelligent terminals,” but without blockchain, these terminals are just data silos for the government or big tech. The analysis questions whether the 2600B includes double-counting of traditional industry upgrades—typical of local policies. Blockchain can provide a unified, auditable ledger to separate true AI innovation from re-labeled legacy products. But no. The plan prefers ambiguity.
Competitive Landscape: Chengdu aims to differentiate from Beijing, Shenzhen, and Hangzhou by focusing on application penetration. That’s smart. But the analysis warns that Xi’an and Chongqing are close competitors, especially in smart mobility. Blockchain could be a killer differentiator: a decentralized network of AI agents across cities, using cross-chain protocols, would give Chengdu first-mover advantage in “agent interoperability.” Instead, the plan is silent. The analysis also notes that the 2600B target may involve statistical inflation. A blockchain-based reporting system—where companies submit verifiable on-chain records of AI revenue—would kill that skepticism. But it’s not mentioned, likely because it adds complexity that local bureaucrats dislike. Data checked. Community warned.
Ethics and Safety: This is where the analysis is most damning: zero mentions of AI safety, ethics, or regulatory compliance. For a plan that aims to embed AI in healthcare, finance, and government, this is criminal negligence. In crypto, we learned the hard way—Terra Luna’s algorithmic stablecoin collapsed without decentralized oversight. The same can happen to AI agents if they are deployed without on-chain audit trails. China’s own generative AI regulations require content provenance and algorithm filing. A permissioned blockchain could satisfy those requirements elegantly: every AI decision recorded, every data consent verified. But Chengdu’s plan seems to think safety is a national issue, not a local one. That’s a trust bridge waiting to collapse. Trust bridge crossed. Crash imminent. The analysis’s hidden information about potential privacy risks from 70% penetration of cameras and smart locks is chilling. Without blockchain, there is no way for citizens to verify what data is collected or how it’s used. I’ve seen how NFT floor price manipulation worked in 2021—centralized data creates asymmetrical power. Blockchain flips that.
Investment and Valuation: The analysis predicts short-term hype for local AI concept stocks (e.g., Jiafa Education, Creative Information), but warns of poor target attainment rates (historically <60%). In crypto markets, I’ve learned that narratives without on-chain fundamentals are pump-and-dumps. A government target is a narrative; only on-chain metrics (like token velocity, staking ratios, or smart contract activity) provide real validation. The analysis suggests insider trading risk—another area where blockchain’s transparency could deter malfeasance. But again, not addressed. The 2600B number is large, but as I often write, “Floor price broken. Truth verified.” Without a decentralized verification layer, that number is just a floor price without liquidity.
Infrastructure and Compute: Chengdu has solid compute—the National Supercomputing Center and Tianfu Intelligent Computing Center planning 1000P by 2025. But the analysis rightly questions whether this can meet demand, especially under chip restrictions. Here, blockchain can help through decentralized compute marketplaces (like Render Network or Akash) that aggregate idle GPU power across the city. It’s a natural fit for Chengdu’s plan to “empower thousands of industries”—allow small businesses to bid for computing power via smart contracts, avoiding vendor lock-in. The analysis mentions that Chengdu may rely on Huawei ShengTeng for chips, which is a centralized risk. A hybrid cloud+blockchain approach would add resilience. But no. The plan sticks to traditional cloud models.
Contrarian Angle: Is the Plan Actually a Stealth Boost for Blockchain? Counter-intuitively, the plan’s massive push for “agents” and “intelligent terminals” could inadvertently accelerate crypto-native AI agent platforms. Think autonomous AI agents that execute transactions, manage DAOs, or trade on decentralized exchanges. If Chengdu floods the market with 70% AI terminal penetration, even without intending to, it creates a user base that may demand permissionless agent-to-agent interactions. During the 2024 ETF integration story, I saw how retail investors, once exposed to institutional-grade products, inevitably sought DeFi alternatives. Similarly, a generation of Chinese users with AI terminals will discover that legacy internet can’t execute cross-platform, trustless agreements—they will turn to blockchain. The plan’s blindness to blockchain might be its secret gift: by ignoring it, they won’t regulate it heavily, allowing grassroots adoption. In 2018, the ICO crash taught me that the best innovations happen when central planners are looking the other way. Liquidity gone. Run. could become a rallying cry for decentralized AI agents in Chengdu.
Takeaway: What to Watch The next 12 months will tell if Chengdu’s AI plan becomes a classic case of “scale without trust” or an accidental catalyst for blockchain-AI convergence. Key signals: (1) Does any benchmark scenario include a blockchain component, even as a pilot? (2) Do local crypto companies (like Bitmain-affiliated miners or Web3 startups) report increased government interest? (3) Does the first batch of “double hundred” projects involve any tokenization or smart contract use? As a journalist who has tracked both AI and blockchain for a decade, I see the glaring hole. Chengdu’s 2600B target is a floor price set without a market maker. The market—developers, users, and global capital—will eventually arbitrage that gap. Whether the city recognizes it or not, blockchain is the trust bridge they never asked for, but desperately need.