Six AI Tigers (LLM, China)¶

The Six AI Tigers (Chinese: 大模型六小虎, often glossed as “six little tigers of large models”) is a journalistic and industry label for a cohort of China-based foundation-model startups that rose to prominence during the “hundred models” wave of generative AI, typically alongside—but distinct from—incumbent platform giants (Baidu, Alibaba, ByteDance, Tencent, and similar). The usual membership is Zhipu AI (Z.ai), Moonshot AI (Kimi), MiniMax, Baichuan AI, 01.AI, and StepFun; other high-profile Chinese labs (notably DeepSeek) are often discussed in the same competitive landscape but are not what commentators mean by this fixed “six.”

What Do the Six AI Tigers Think Their Work Will Do to Other Humans?¶

The Six AI Tigers were born in the generative era and operate with a distinct, sometimes ruthless, perspective on their role in human history.

In 2026, their view of their impact on "other humans" is a spectrum ranging from techno-optimism to stark realism.

Liberating Humans¶

Many founders, such as Kai-Fu Lee (01.AI) and Yang Zhilin (Moonshot AI), argue that their work is designed to strip away the "non-human" parts of our lives.

• The Logic: They believe that 90% of current human labor is "robotic"—repetitive, data-driven, and devoid of soul. By automating this, they claim to be "liberating" humans to focus on what AI cannot do: empathy, complex ethics, and high-level creativity.
• The Intent: They see their models (like Kimi or Yi-Large) as mirrors. When the AI handles the "boring" intelligence, what is left is the "true" human essence.

Decade of Disruption¶

There is a darker, more pragmatic thread emerging from the Tiger labs. Notably, researchers from DeepSeek and other Tiger-affiliated groups have publicly warned of a "Decade of Disruption" starting now in 2026.

• Massive Displacement: They don't sugarcoat the job losses. Leaders like Wang Xiaochuan (Baichuan) have suggested that AI will soon take over roles previously thought "safe," like healthcare diagnosis and legal reasoning.
• The "Defender" Role: They believe AI companies must move from "disruptors" to "defenders" of social stability. They view their work as building the infrastructure that will eventually support Universal Basic Services—arguing that if AI makes production nearly free, it is their job to provide the tools for the state to distribute that abundance.

"Productivity as a Utility"¶

The Tigers increasingly view "Intelligence" not as a luxury, but as a utility like electricity.

• StepFun and Zhipu AI have shifted toward "Agentic AI"—models that don't just talk, but act.
• Their View on Humans: They see humans as "Orchestrators." In their vision, a single human will soon manage a small "factory" of AI agents. Their work isn't about replacing a human; it's about giving one human the power of 1,000.

The "Post-Labor" Inevitability¶

The Six Tigers believe they are building the "Post-Labor Economy." To them, other humans are currently in a painful but necessary transition from "Workers" to "Consumers and Curators."

Is What They Say in Correspondence With What They Are Actually Doing?¶

The "Liberation" vs. "Liquidation"¶

• The Correspondence: Leaders like Kai-Fu Lee (01.AI) argue that AI will "liberate" humans from repetitive, robotic tasks, allowing us to focus on "soul-driven" work.
• The Action: The primary revenue driver for firms like Zhipu AI and Moonshot AI in 2026 is Token Sales and API Licensing. These services are marketed directly to corporations to replace "middle layers" of labor.
• The Data: Zhipu AI's revenue surged 131.9% in 2025 by selling "orchestration layers" that automate complex tasks previously requiring teams of human middle managers.

The "Consumer Delight" vs. "B2B Efficiency"¶

• The Correspondence: Moonshot AI brands its Kimi chatbot as a tool for personal productivity and "consumer delight."
• The Action: In 2026, Moonshot's massive $18 billion valuation is fueled by "Kimi Claw," an agentic system that interacts with programmatic endpoints—not human interfaces.
• The Strategic Shift: Their internal focus is on B2B automation, where AI agents browse the web, read emails, and initiate transactions without a human driving each step. They are building software to replace the need for human users in the loop.

Healthcare: "Accessibility" vs. "Expertise Decoupling"¶

• The Correspondence: Baichuan AI frames its healthcare models as a way to provide "an AI doctor for every village," prioritizing social accessibility.
• The Action: Implementation in 2026 has prioritized "Clinical Decision Support" and "Autonomous Workflows" in top-tier hospitals.
• The Impact: 29% of hospital AI budgets are now being diverted into agents that handle real-time medical decision-making. The goal is to lower costs by reducing the number of high-salaried human specialists needed to verify or oversee diagnoses.

Comparison Table: Words vs. Reality¶

Chinese Government Impact¶

The Chinese government is building the "waiting room". The government is "protecting" people by buying time:

• The Ministry of Human Resources (MOHRSS) has launched the fifth and largest round of the "Send Graduates to the Grassroots" program. They are subsidizing hundreds of thousands of roles in rural vitalization and local governance. These aren't career paths; they are holding pens to keep youth out of the urban unemployment statistics.
• For the graduates who can't find work, the state is paying for "long-term internships" in emerging sectors. The government covers the social insurance and a portion of the stipend.
• In March 2026, a joint circular mandated that State-Owned Enterprises (SOEs) must increase their intake of fresh graduates, even if their productivity metrics don't require it.
• The 2026 policy (specifically the "Skills Enhancement Year Action Plan") has revised subsidy rules. In cities like Shanghai, the government provides direct living allowances to "registered unemployed individuals" while they attend full-time training.
• To prevent total income loss, the state is pushing "Night School" models and online certifications. This allows people to work gig jobs (like delivery or ride-sharing) during the day while "upgrading" at night.
• The government pays companies to take on "apprentices." The company gets a training subsidy, and the worker gets a reduced but guaranteed wage while learning.

Bridge to nowhere

The state provides a living allowance (the "Living Stipend") while you are in the 30-million-person retraining pool. By the time a human completes a 6-month certification in "AI Training," the Tigers (Zhipu, Moonshot, etc.) have released a new model that can auto-label its own data or "self-prompt".

Waiting for what?

They are using state capital to pay people's bills today so that they don't protest tomorrow, while hoping that by 2027, they will have figured out a way to tax the AI Tigers enough to pay for everyone's "Universal Basic Services." There is a serious proposal on the table (April 2026) for a "Robot Tax" or "AI Usage Fee" for companies that replace human labor in "ordinary" sectors.

Currently, the fiscal deficit is held at 4% of GDP (a high for China), which means they are spending more than they receive.

Current Conclusions and Open Questions¶

Current Conclusions:

1. Tigers' most profitable products in 2026 are specifically designed to replace the knowledge worker (legal, medical, coding, management).
2. The Tigers are building the "Infrastructure of Displacement" while labeling it the "Infrastructure of Empowerment."
3. Trends are absolutely identical to the ones for US LLM giants.
4. While the Chinese government is actually deceiving people psychologically by maintaining the "bridge to nowhere", at the same time, they actually provide many of them with a UBI and create a bridge to UHI in the future. They also have much more chances to make a pivot to "Robot Tax" or "AI Usage Fee" in the future, than their Western colleagues.

Open Questions:

1. NA.