In a sprawling coastal city where electronics once dominated the landscape, humanoid robots now work factory floors while artificial intelligence tells fortunes to millions of anxious youth.
Hangzhou and its neighboring Shenzhen stand as a paradox of modern China: aggressively pursuing physical robotics as a national strategic priority while simultaneously nurturing a booming spiritual commerce powered by the same AI systems designed to solve manufacturing problems.
This collision of cutting-edge robotics and ancient divination practices reveals far more than mere technological curiosity.
It exposes the fundamental shape of China's AI ambitions, the peculiar economic pressures facing young Chinese, and the speed at which innovation ecosystems can mobilize when state backing meets entrepreneurial energy.
A New Industrial Dynasty Takes Shape
Shenzhen has earned the nickname "Robot Valley," a transformation that occurred with stunning velocity. In 2024, the city housed 74,032 robotics enterprises generating over 200 billion yuan in annual output—a threshold crossed for the first time that year. Within just a decade, what had been primarily an electronics assembly hub evolved into the world's dominant robotics manufacturing center.
The local government responded with ambition matching the industrial momentum, releasing a 2025-2027 action plan targeting over 100 billion yuan in embodied intelligent robotics output by 2027, with plans to nurture more than 1,200 enterprises in the sector.
Hangzhou, historically overshadowed by Shenzhen's manufacturing prowess and Beijing's political centrality, has emerged as an unexpected rival. The city's Binjiang district houses over 200 major AI and software firms alongside robotics manufacturers.
Eight kilometers from West Lake's tourist crowds, a cluster of startups nicknamed the "Six Little Dragons"—including DeepSeek, Unitree Robotics, DEEP Robotics, Game Science, BrainCo, and Manycore Tech—are driving what observers describe as the next industrial wave.
The speed of innovation in these ecosystems dwarfs Western timelines. In Shenzhen's production zones, prototyping cycles compress from months to days. A robot manufacturer can optimize component design in two hours with suppliers located in the same building.
This density of talent, capital, and manufacturing capability creates what venture investors now call a 10-to-1 advantage over Silicon Valley in hardware development velocity. Companies like Unitree advance from one generation of quadrupedal robots to the next in under two years—a timeline Boston Dynamics requires five to seven years to achieve.
UBTech, the most commercially advanced humanoid manufacturer, exemplifies this acceleration. Founded in 2012 as a consumer robotics company, it pivoted to industrial automation and achieved massive scale production. By December 2025, the company had manufactured its 1,000th industrial humanoid robot, with more than 500 already delivered to customers.
The company announced plans to deploy 5,000 units in 2026 and reach 10,000 annually by 2027. Critically, these robots are already working. UBTech secured a 264 million yuan contract to deploy Walker S2 humanoids at the China-Vietnam border, where they perform patrol duties, traveler guidance, and logistics operations. Other units operate in automotive factories, including facilities operated by NIO, Zeekr, BYD, and Foxconn.
This is not prototype territory. These are production-scale deployments solving real problems.
The Fortune-Telling Explosion: When Ancient Tradition Meets AI
While robotics companies race toward international dominance, a parallel phenomenon has seized millions of young Chinese: AI-powered fortune-telling. The trend exploded with DeepSeek's January 2025 launch of its R1 reasoning model.
Within weeks, the platform accumulated over 20 million daily active users. On WeChat alone, 2 million posts mentioning "DeepSeek fortune-telling" appeared in February 2025, according to the platform's own index.
The tradition being revived is BaZi, an ancient Chinese metaphysical system that analyzes personality, destiny, and life outcomes based on birth time, date, and place. Practitioners traditionally charged between 300-500 yuan for a session—expensive and logistically inconvenient for the average person.
DeepSeek changed the equation. The AI model, trained extensively on classical texts and contemporary interpretations, can generate comprehensive BaZi analysis in seconds. It synthesizes birth data, calculates astrological elements, and produces personalized readings covering career, relationships, wealth, and health trajectories.
The adoption pattern reveals something sociological beneath the technological novelty. China's young adults have experienced unprecedented economic volatility. Youth unemployment peaked at 21 percent in June 2023 and remained near 16 percent throughout 2024. GDP growth rates hit multi-decade lows.
On social media platforms, millions of young Chinese describe themselves as the "last generation," expressing reluctance toward marriage and parenthood amid pervasive uncertainty. In this context, an AI oracle available 24/7, free or extraordinarily cheap, offered something traditional fortune tellers never could: accessibility at moments of anxiety.
A 31-year-old product manager named Zhang Rui quit her stable job to start a business—a decision that traditionally would have sent her to a fortune teller at 500 yuan per session. Instead, she asked DeepSeek to analyze her BaZi.
Within fifteen seconds, the AI identified 2025-2027 as a "fire" period indicating auspicious career conditions. She felt validated. The speed, privacy, and absence of transaction cost created a psychological effect traditional practitioners could never replicate.
Users rapidly evolved from passive consumers to sophisticated operators. On platforms like Xiaohongshu (RedNote), the hashtag #DeepSeekFortune accumulated over 35 million views. Communities formed around "prompt engineering" for fortune-telling—techniques to craft optimally phrased queries that extracted deeper insights from the AI model.
Users documented effective prompts, shared results, and built followings. Some individuals began packaging these guides into products, while others launched AI spirituality businesses.
The Commercialization Layer
The fortune-telling phenomenon rapidly stratified into a business ecosystem. Fatetell, a startup founded in 2023, went from concept to meaningful traction only after DeepSeek R1's launch in early 2025. The platform integrates multiple AI models—ChatGPT, Claude, Gemini, and DeepSeek R1—to generate comprehensive destiny reports.
The company charges $39.99 for full readings, and internal data shows an approximately 5 percent paid conversion rate with strong repeat purchases, as users buy readings for themselves and then purchase additional analyses for family and friends.
Other platforms took different approaches. Cantian AI offered free BaZi calculators generating personalized metaphysical databases.
CeCe Astrology, an integrated platform combining Chinese and Western divination methods, implemented a livestream model where professional astrologers conduct readings 24/7. Top streamers on the platform earned over 100,000 yuan monthly through virtual gifts and paid consultations.
The most granular entrepreneurial tier emerged among individual creators. Mini-programs for "AI Face Reading" and "AI Bazi Calculator" proliferated across WeChat. Solo founders packaged these applications with referral commission structures, enabling users to earn 30,000 yuan or more monthly by sharing with networks.
The economics resembled high-velocity consumer arbitrage: purchase cheap AI-generated readings (or create them yourself using free API access), resell through social networks, accumulate commissions.
From a regulatory perspective, these ventures operate in murky territory. The Chinese government maintains strict control over spirituality-related content and spiritual practices.
Fatetell's founder explicitly pivoted the company toward overseas markets—operating exclusively through web browsers accessible to global audiences rather than mobile apps preferred by Chinese users—to reduce regulatory risk. The precarity of this business model became apparent as government attention to the space intensified through 2025.
Yet the phenomenon persists because it solved a genuine market failure. Professional fortune tellers, even accounting for inflation, had not changed pricing or service models in decades. The "chicken soup" advice from general-purpose chatbots like ChatGPT felt shallow and inauthentic to spirituality seekers.
DeepSeek's reasoning model, combined with community-driven prompt engineering, created an experience that felt both scalable and meaningful in ways neither traditional nor superficial digital services could match.
State Strategy: Embodied Intelligence as National Priority
The robotics boom and AI applications are not emerging from market forces alone. Beijing has designated "embodied intelligence"—AI systems capable of autonomous physical action in the real world—as a strategic national priority.
The 2024 "Report on the Development of Embodied AI," authored by the China Academy of Information and Communications Technology under the Ministry of Industry and Information Technology, outlined how embodied AI could address labor shortages in manufacturing and logistics, critical to revitalizing slowing economic growth.
Provincial and municipal governments responded with aggressive capital mobilization. Beijing launched a 100 billion yuan investment fund with a fifteen-year lifespan to support AI and robotics development.
Shanghai established an embodied AI fund with initial commitments of 560 million yuan. Shenzhen's 2025-2027 action plan positioned the city as the epicenter of humanoid robotics manufacturing, with over 1 billion yuan in targeted public investment.
Hangzhou's approach reflected a different model. Rather than massive single-year appropriations, the city implemented targeted subsidies and fast-tracked bureaucratic processes for startups. Businesses in Yuhang district could receive up to 600,000 yuan in general high-tech subsidies.
New entrepreneurs obtained three years of rent relief plus grants reaching 500,000 yuan. The system proved remarkably efficient: one startup submitted a subsidy application online and received a 3 million yuan grant within eight minutes.
The investment statistics confirm the scale. Total Chinese investment in AI reached 890 billion yuan in 2025, representing 18 percent year-over-year growth and accounting for 38 percent of global AI investment.
Government funding contributed 345 billion yuan of this total, growing at 21 percent annually. Beijing, Shenzhen, and Shanghai concentrated 71 percent of all Chinese AI investment—632 billion yuan combined.
These figures reveal a strategic divergence from Western approaches. Beijing's embodied AI strategy emphasizes practical, deployable systems over theoretical advances in large language models.
While OpenAI and other Silicon Valley firms invested hundreds of millions into advancing transformer architectures and scaling inference, Chinese companies prioritized integrating existing AI capabilities with robotics hardware to solve immediate production problems. This pragmatic orientation proved commercially effective: whereas Western humanoid robotics remained in research phases, Chinese companies shipped thousands of units into factories.
The IPO Wave and Capital Consolidation
The robotics and AI boom triggered a consolidation wave in Hong Kong equities. Chinese tech companies raised $36.5 billion through 114 new listings in 2025—Hong Kong's highest volume since 2021.
AI and semiconductor stocks drove much of this activity, capitalizing on investor enthusiasm following DeepSeek's January 2025 breakthrough.
Agibot, a Shanghai-based humanoid robotics maker, prepared to launch an initial public offering in Hong Kong in 2026 targeting a valuation between 40 billion and 50 billion Hong Kong dollars (approximately $5.1-6.4 billion USD).
The company planned to sell 15-25 percent of shares, potentially raising over 1 billion dollars depending on market conditions. Backing from investors including Tencent, HongShan Capital (formerly Sequoia China), LG Electronics, and BYD reflected broad institutional confidence in China's robotics trajectory.
This capital inflow had practical consequences. Manufacturing costs for robotics components and subsystems continued declining. Supply chain integration deepened. Production timelines compressed further.
By late 2025, what seemed distant from Western perspective—mass deployment of humanoid robots in factories, warehouses, and border facilities—became observable reality across China.
Competing Futures Within a Single Ecosystem
The simultaneous emergence of industrial robotics dominance and AI-powered spirituality commerce creates an illuminating paradox about China's technological trajectory. The robotics sector represents Beijing's strategic vision: practical AI embodied in hardware, deployed at scale, solving productivity problems.
The fortune-telling phenomenon represents a very different current: fragmented entrepreneurship, cultural resilience, and an enormous unmet demand for meaning-making tools.
Both coexist because they operate in separate economic layers. Robot manufacturers work with large industrial customers, factories, and state entities. AI fortune-telling platforms target individual consumers seeking emotional reassurance and life guidance.
One addresses productive capacity; the other addresses spiritual emptiness. One requires capital, supply chains, and government support; the other emerges from solo creators and simple software.
Yet they share fundamental characteristics. Both leverage China's advantages in rapid prototyping and deployment. Both capitalize on government support and policy backing.
Both emerged from talent pools concentrated in specific geographic clusters. And both represent China's unwillingness to accept Western technological frameworks as inevitable—whether in robotics architecture or digital spirituality commerce.
The fortune-telling ecosystem, viewed through an economic lens, reveals something crucial about Chinese entrepreneurship. Faced with regulated domestic markets and uncertain regulatory futures, creators rapidly internationalize. Fatetell explicitly built for Western audiences from inception.
Solo developers created English-language face-reading applications alongside Chinese versions. The platforms that succeeded adapted to where capital, users, and regulatory tolerance aligned—a pattern that mirrors how Chinese fintech, gaming, and social platforms expanded globally after domestic saturation or regulatory constraints.
Market Tensions and the Path Forward
The rapid growth conceals important tensions. Regulatory risk looms over the fortune-telling sector. Chinese authorities maintain historical skepticism toward spirituality-focused businesses, particularly those generating substantial commercial revenue. As the industry scaled through 2025, government scrutiny intensified.
Platforms began implementing content moderation, removing posts recommending expensive purchases (particularly crystals and rare stones that DeepSeek frequently recommended as metaphysical solutions), and clarifying disclaimers that AI readings should complement rather than replace professional consultation.
Trust emerges as the critical failure point for AI fortune-telling platforms. Professional fortune tellers possess years of training in complex metaphysical systems. They can adapt readings based on client responses and emotional states. AI models, despite their sophistication, generate readings from fixed algorithmic processes.
Users discovered that reading quality depended entirely on query formulation—weak prompts yielded generic advice. Some platforms began collaborating with traditional fortune tellers to develop training data and evaluation frameworks, creating hybrid human-AI systems that preserved authenticity while scaling delivery.
The robotics sector faces different challenges. While manufacturing capacity expands rapidly, deployment beyond factories remains constrained. Safety standards, liability frameworks, and technical specifications for humanoid robots operating in public spaces remained under development through 2025.
The government established technical committees to establish unified standards for hardware interfaces, data formats, and safety protocols. Without these standards, broader deployment into retail, hospitality, or household contexts will remain limited.
Cost, despite improvements, still exceeds Western industrial expectations. Humanoid robots deployed by UBTech and others cost significantly less than Western equivalents, but purchasing decisions still require substantial capital commitments.
Mass consumer adoption remains years away. The trajectory suggests dominance in industrial and infrastructure contexts will solidify before consumer robotics markets emerge.
The Unbundling of Chinese Innovation
What emerges from this ecosystem is not a monolithic "Chinese AI strategy" but rather a densely interconnected network of overlapping innovation clusters, each optimized for different market conditions and risk profiles. Shenzhen dominated robotics manufacturing through existing electronics supply chains and factory density.
Hangzhou captured AI software innovation partly through government support but primarily through talent concentration and lower regulatory friction. Both benefited from Beijing's clear signal that embodied AI represented a strategic priority.
The fortune-telling phenomenon proved resilient precisely because it operated beneath regulatory attention initially, generated revenue through simple monetization, and addressed genuine human needs that traditional markets had underserved.
When regulatory pressure increased, the ecosystem rapidly distributed internationally. This pattern—rapid scaling, regulatory encounter, international pivot—characterizes much of contemporary Chinese entrepreneurship.
The physical robots now walking factory floors and patrolling borders represent one version of China's AI future: state-backed, capital-intensive, focused on productivity and infrastructure. The AI fortune-telling apps represent another: lean, culturally rooted, and remarkably resilient to constraints.
Both will shape how AI integrates into society, yet they receive disproportionate international attention—one celebrated as technological prowess, the other dismissed as superstition. Neither captures the full reality of Chinese innovation, which proceeds not through grand unified strategies but through distributed experiments across numerous markets, regulatory contexts, and entrepreneurial approaches.
The contrast between humanoid robots and fortune-telling AI ultimately reveals something about how technological systems reflect the societies that build them. In Western context, AI development concentrates among a handful of megacorporations and government programs pursuing general artificial intelligence.
In China, the landscape remains more variegated—corporate robotics manufacturers competing with government support, spiritual commerce entrepreneurs building globally, and local governments racing to capture innovation clusters. This distributed model may prove more resilient than centralized alternatives, both producing superior outcomes across multiple dimensions and creating novel political and economic risks that neither Western tech companies nor Chinese government planning have fully anticipated.
