AI as a Dual-Engine Catalyst: Job Creation and Industrial Deployment
“AI+” Dual-Track Drive: Simultaneous Job Creation and Industrial Implementation, with Regulation Calibrating Capital Behavior in Real Time
The State Council’s “15th Five-Year Plan for Implementing the Employment-First Strategy” formally designates artificial intelligence (AI) as an “engine for net job creation,” explicitly calling for “deep implementation of the ‘AI+’ initiative.” It underscores AI’s structural role in catalyzing new human–machine collaborative work models, incubating scenario-specific entrepreneurial ventures, and accelerating technology commercialization. This policy shift is no isolated signal—it stands in sharp contrast to the State Administration for Market Regulation’s recent Regulations on Subsidy Practices by Food-Delivery Platforms, revealing a deliberate dual-track governance framework. The former institutionalizes proactive release of technological dividends; the latter imposes rigid constraints to curb market-distorting, disorderly capital subsidies. Together, they form a closed-loop governance model—“promoting innovation + stabilizing order”—signaling a profound paradigm shift in China’s AI development logic: from traffic-driven, scale-prioritized, extensive expansion toward high-quality, real-world implementation centered on authentic productivity gains and validated, sustainable business models.
Employment Dimension: Paradigm Shift—from “Displacement Anxiety” to “Collaborative Expansion”
For years, the dominant public narrative—“AI steals jobs”—has overshadowed nuanced realities. Yet policymaking has systematically pivoted to a “human–machine collaborative expansion” framework. For the first time, the Plan lists “exploring novel work models enabled by human–machine collaboration” as a core employment-policy objective. Its foundational logic acknowledges that AI does not simply replace labor; rather, it restructures task allocation, lowers professional entry barriers, and extends service boundaries—thereby generating entirely new occupational clusters previously unimaginable.
For example, in medical AI, primary-care physicians equipped with AI-assisted diagnostic systems can now handle over 80% of preliminary screening tasks typically performed at top-tier hospitals—expanding clinical service coverage by more than threefold. In government-service AI applications, intelligent approval robots have reduced enterprise registration time to under four hours, spawning emerging roles such as “AI Process Architects” and “Policy Semantic Training Specialists.” According to Ministry of Human Resources and Social Security (MOHRSS) data for Q1 2025, 67% of newly created AI-related positions nationwide are “human–machine collaborative” (requiring human oversight, fine-tuning, and ethical judgment), while only 12% are fully automated replacements. This structural shift validates the policy’s foresight: AI is not a subtraction from employment—it boosts labor productivity, unleashing broader service demand and thereby expanding total employment at a higher, more sophisticated level.
Industrial Dimension: Deepening Implementation—from Conceptual Hype to Value Anchoring
The policy’s focus on “AI+” fundamentally aims to move AI out of labs and onto production lines—and out of demonstration screens and into actual invoices. The Plan’s call to “drive more AI application-scenario innovations and technology commercialization” directly addresses a key industry pain point: many AI projects remain stuck at the POC (Proof-of-Concept) stage, lacking quantifiable, economically viable return pathways.
This imperative resonates strongly in capital markets, where criteria for selecting investee companies are rapidly evolving. In industrial AI, firms offering predictive equipment maintenance deliver measurable outcomes: clients report a 35% reduction in unplanned downtime and a 28% cut in maintenance costs—hard metrics that directly translate into customer willingness to pay. Government-service AI companies achieving a 40% improvement in fiscal fund disbursement efficiency and >92% accuracy in audit-risk identification establish clear, procurement-ready pricing benchmarks. Medical AI imaging products cleared by China’s National Medical Products Administration (NMPA) Class III certification—and clinically proven to reduce misdiagnosis rates by 15% versus average radiologists—meet hard procurement thresholds. By contrast, cases like “PCB concept” speculation reveal the yawning gap between hype and reality: China Jushi’s announcement explicitly noted that its low-dielectric specialty electronic fabric “has yet to generate orders or revenue.” Markets vote with their feet: In Q1 2025, B2B industrial AI firms saw average contract liabilities rise 52% year-on-year, while pure-platform AI companies experienced a 68% YoY decline in fundraising volume.
Regulatory Dimension: Governance Upgrade—from Lenient Experimentation to Precision Calibration
On the surface, the Regulations on Subsidy Practices by Food-Delivery Platforms target food delivery—but beneath lies a deeper governance signal: when technological applications reach scale, regulation must transition from a “tolerant innovation phase” to a “fairness-guaranteeing calibration phase.” The regulations explicitly prohibit practices such as “predatory pricing significantly below cost” and “algorithmically coercing merchants into exclusive partnerships”—effectively blocking capital’s exploitation of AI as a vehicle for monopolistic behavior.
This regulatory logic precisely complements the “AI+” policy—forming a tightly interlocked system. The former ensures technological dividends are not perverted by capital into predatory tools; the latter guarantees those dividends genuinely translate into public welfare. Notably, regulatory tightening does not reject subsidies per se, but distinguishes between “strategic subsidies” and “disorderly subsidies”: fiscal support continues to intensify for public-good domains—such as domestic industrial software substitution and AI-enabled healthcare access in county-level areas—while drawing clear red lines against models reliant solely on burning cash for user acquisition or algorithmically exploiting delivery riders. This capacity for “targeted regulation” reflects the maturity of China’s AI governance system.
Investment Logic: Repricing—from Valuation Bubbles to Value Realization
Under this dual-track drive, investment logic across the AI sector is undergoing fundamental revaluation. The old “price-to-dream ratio” valuation model has collapsed. It is being replaced by hard, operational metrics: “output per unit of computing power,” “efficiency gain rate from human–machine collaboration,” and “cash-flow coverage ratio of compliant business models.” B2B AI enterprises delivering tangible cost reductions and efficiency improvements are seeing their valuations converge toward dual anchors—Price-to-Sales (P/S) and Enterprise Value-to-EBITDA (EV/EBITDA). Meanwhile, platform-based AI firms relying on conceptual hype—with no verifiable customer payments—find even dazzling technical demos insufficient to attract institutional capital.
The futures industry’s consolidation trend offers further insight: Shenwan Futures’ proposed merger with Hongyuan Futures highlights rising demand for deep integration—not just of algorithms, but of risk-control capabilities, system stability, and compliance operations. This mirrors AI’s own evolution: from “single-point technological breakthroughs” toward “full-stack capability building.” Future winners will be those who not only deploy AI to solve concrete industrial pain points—but also construct commercially viable, regulation-compliant business loops.
At its core, this dual-track policy design represents a deliberate effort to install both an “innovation accelerator” and an “order stabilizer” during a period of explosive technological advancement. Only when “AI+” ceases to be merely a technical term—and becomes a measurable unit of employment growth, a benchmark for industrial transformation, and a binding standard for capital conduct—will China’s AI development truly enter its deep-water phase. This quiet yet profound transition ultimately affirms a timeless truth: the most sustainable technological revolutions always emerge at the golden intersection of productivity enhancement and social-order optimization.