Part One: The New Spiritual Economy: Deconstructing Youth Metaphysics Consumption in 2025

1.1 The Game of Anxiety and Control: Socioeconomic Roots

In the consumer market landscape of 2025, a notable trend is the growing preference for metaphysics-related consumption among young people. This phenomenon is not accidental; its roots are deeply embedded in the structural pressures of the contemporary socioeconomic environment. A research report released in early 2025 clearly indicated that downward pressure on the global economy and an unstable social environment have exacerbated the challenges young people face in employment, health, and other areas, directly leading to a widespread increase in their sense of anxiety and a decline in their sense of control over their lives.¹ In an era fraught with uncertainty, individuals’ confusion about the future and fear of losing control compel them to seek external tools to rebuild inner order and peace.

This individual-level psychological need is amplified by broader epochal factors. A report by Jones Lang LaSalle identifies policy, technology, culture, and values as the four major macroeconomic drivers reshaping the consumer market.² These factors collectively exert a profound influence on all consumers, but their impact is particularly acute on the younger generation, whose economic foundations are relatively weak and life trajectories are not yet fully established. They are more sensitive to economic fluctuations and more susceptible to shocks from the job market, public opinion, and technological change. Meanwhile, global challenges such as extreme weather and the rapid iteration of disruptive technologies further intensify a general sense of powerlessness, making it impossible for consumers to ignore the drastic changes in the external environment in their daily lives.⁴

Against this backdrop, traditional metaphysical activities like divination, feng shui, and tarot card reading have been endowed with new contemporary functions. They are no longer merely mystical practices followed by a few but are widely adopted by young consumers as a “tool for psychological comfort and spiritual sustenance”.¹ The essence of metaphysics consumption is a compensatory control behavior. When individuals feel unable to control their destiny in the real world, they turn to symbolic systems that can provide a sense of order, certainty, and meaning. As one expert analyzed in 2025, young people’s engagement in metaphysics consumption is not a simple escape from reality, but an active search for an outlet to release anxiety and achieve inner peace.⁵ This consumer behavior has become a psychological life raft for them in a sea of uncertainty.

1.2 Emotional Value: The Core Currency of New Consumerism

If anxiety is the underlying force driving young people towards metaphysics, then “Emotional Value” is the core currency circulating in this emerging market. For the “Post-Z Generation,” represented by those born after 2000, emotional value has evolved from an add-on to a “rigid demand” in consumption decisions. Survey data from DT Research Institute shows that “emotional value” is the primary factor that the post-2000s generation values most when shopping . This marks a profound shift in the consumption patterns of the younger generation, moving from a focus on quantity to quality, and ultimately to a quest for emotional resonance (sensibility) .

The market has responded with extreme speed, giving rise to a vast “emotional healing” product ecosystem. In this ecosystem, metaphysical products coexist with seemingly unrelated consumer categories, all playing the role of “spiritual ibuprofen” . Whether it’s a Jellycat plush toy that provides warm companionship, an anime IP (like Chiikawa) that offers an emotional projection, or a stand-up comedy show that relieves stress, their core function is to provide consumers with immediate psychological comfort.

In this broader context, metaphysical products are precisely positioned as “spiritual band-aids”.⁶ For example, the value of a crystal bracelet far exceeds its material worth. It is seen as a physical medium to combat anxiety, providing comfort through “tactile healing” and helping consumers reconstruct a positive imagination of the future through the ritualistic act of “wish-making,” entrusting their hopes for a better life.² The meaning of this consumer behavior is reshaped: purchasing is no longer a simple material possession but an active process of self-care and psychological construction.

This trend is part of the broader wave of “self-pleasing consumption.” The core of self-pleasing consumption is that the ultimate goal of spending is to please oneself and satisfy inner emotional and spiritual needs, rather than to cater to external social expectations or purely pursue practicality.⁷ Sales data from “Double 11” (Singles’ Day) shows that young people’s shopping carts are shifting from “buy what’s on sale” to “buy what I like,” with experiential consumption like travel products and concert tickets, and goods that provide emotional satisfaction, becoming their priority choices.⁷ Metaphysics consumption is the ultimate expression of this self-centered, emotion-oriented consumption philosophy.

1.3 The Modern Metaphysics Market: A Fusion of Tradition, Trends, and Technology

The metaphysics market of 2025 presents a unique hybrid form, skillfully blending ancient traditions, modern trends, physical goods, and digital experiences.

The market’s manifestations are multidimensional. Offline, the popularity of physical metaphysical goods remains high. Crystal bracelets, lucky charms, and other items have become a form of social currency among young people. In some popular “influencer” locations, the number of young people buying these items even surpasses that of traditional middle-aged and elderly customers.⁵ This trend has also successfully penetrated the high-end consumer sector. Even top jewelry brands like Van Cleef & Arpels, with their classic designs like the four-leaf clover, are endowed by consumers with metaphysical meanings such as “bringing good luck” or being a “lucky charm,” becoming a symbol of “high-end metaphysics” .

Online platforms are the core source and amplifier of this trend. Social media platforms, represented by Xiaohongshu (Little Red Book), form the “central nervous system” of this new spiritual economy. On Xiaohongshu, topics related to “metaphysics” have garnered over 1.85 billion views, forming a vibrant ecosystem . Metaphysics bloggers can attract hundreds of thousands of followers in a short period by posting content like “daily lucky color outfits,” transforming abstract concepts of fortune into concrete, actionable lifestyle advice . These platforms are not only information exchange hubs but also key venues for product recommendations, word-of-mouth marketing, and community building. Gen Z consumers particularly favor transactions with individual sellers on e-commerce platforms like Taobao and Xianyu, believing these platforms offer greater transparency and protection for peer-to-peer transactions.⁸

On a deeper level, the rise of metaphysics consumption resonates strangely with the grand narrative of “cultural confidence.” Driven by both policy guidance and market demand, the “revitalization of intangible cultural heritage” has become a significant cultural force.² Young people’s sense of identity with local culture is growing, and they are eager to participate in the inheritance of national culture through consumption.³ Metaphysics, as part of traditional Chinese culture, has its revival perfectly aligned with this psychological need. Therefore, young people’s interest in metaphysics can be reasonably constructed as an exploration of traditional wisdom and a following of “Guochao” (national trend) culture, rather than mere superstition. A 2025 report lists “crystal bracelets,” “IP figurines,” and “intangible cultural heritage creative products” as the three major new consumption formats for the Post-Z generation, clearly indicating that metaphysics consumption has been integrated into the broader framework of cultural consumption and identity construction.²

From a consumption outlook perspective, although the overall consumption forecast for 2024 was relatively cautious, Gen Z, as the main force of consumption, still shows stronger willingness to spend and optimism.⁹ Their consumption structure is more skewed towards cultural and entertainment categories that provide spiritual enjoyment and emotional companionship.⁹ It is predicted that by 2035, the overall consumption scale of China’s Gen Z will quadruple to 16 trillion RMB, becoming the core engine of future consumer market growth.¹⁰ Furthermore, the potential for consumption growth is sinking to lower-tier cities, with consumers in county-level cities and counties showing a stronger willingness to increase spending in 2025, providing a broad market space for the further popularization of metaphysics consumption . Behind this consumption trend is the youth’s use of consumption to combat uncertainty and place their hopes for a better life, which perfectly aligns with the core value provided by metaphysical products.

Part Two: The AI Product Ecosystem: Technology as the Vehicle for a New Generation of Mysticism

As the demand for metaphysics among young people grows, technology, particularly artificial intelligence (AI), has rapidly intervened, attempting to satisfy this ancient human need with the logic and efficiency of algorithms. A vast and diverse “AI metaphysics” product ecosystem is forming, combining traditional divination arts with cutting-edge machine learning models to create a new, digital mystical experience.

2.1 The Digital Divination Landscape: An Analysis of Key Product Categories

The current AI metaphysics market can be primarily divided into three product categories, which utilize different AI technologies to meet the diverse needs of users.

2.1.1 Category One: Generative AI and Conversational Oracles

This is the most technologically advanced and fastest-growing area, with its core being the use of Large Language Models (LLMs) to provide complex, narrative, and personalized interpretations.

• Core Technology: The foundation of these applications is generative AI.¹¹ While general-purpose chatbots like ChatGPT are also widely used for fortune-telling ¹³, the market is evolving towards more specialized models. For instance, China’s DeepSeek model is particularly favored by Chinese users for its superior reasoning ability and more specific predictive capabilities in handling traditional arts with rigorous logical structures, such as “Bazi” (Eight Characters), compared to general models.¹⁴ By learning from massive amounts of text data, these models can understand the structure, semantics, and context of language, thereby generating text content comparable to that of human experts.¹⁵

• Typical Applications:

  • Bazi and Astrology: Applications like “Nanyi Paipan” claim to combine AI intelligent technology to provide users with precise Bazi chart and fortune trend analysis, covering detailed predictions for grand cycles, annual, monthly, and daily fortunes.¹⁶ Emerging companies like
    Fatetell integrate multiple models such as DeepSeek and ChatGPT to generate in-depth future analysis reports for users.¹⁴

  • Tarot Card Reading: “TarotAI” is a representative in this field. It uses AI technology to provide detailed interpretations of the upright and reversed meanings of the 78 tarot cards based on the user’s question and chosen spread (such as the complex “Celtic Cross”), simulating a deep conversational experience with a real tarot reader.¹⁷ The AI can understand the symbolic meaning of each card in the context of a specific question and position, offering rich analysis that goes far beyond keyword matching.

  • Multi-functional Integrated Platforms: Applications like “Cece” and Astra AI adopt a platform strategy. They not only provide single tools but also integrate various “self-exploration” tools like astrology, MBTI, Bazi, and tarot, and use an AI intelligent Q&A system to respond quickly to users’ cross-disciplinary questions .

2.1.2 Category Two: Computer Vision and Digital Physiognomy

This category extends the application of AI from text to visual data, attempting to digitize and automate the ancient art of “physiognomy.”

• Core Technology: The technical basis is computer vision, particularly relying on face recognition and key point localization technologies powered by Deep Neural Networks (DNN) and Convolutional Neural Networks (CNN) . For example, “Mianxiang Dashi” (Face Reading Master) claims its model was co-developed by AI engineers and physiognomy masters. By deep learning on nearly 200,000 facial samples (each with 108 positioning points), the AI “learns” to match the geometric features of a face with concepts from traditional physiognomy for calculation . Computer vision systems mimic human visual cognition by analyzing patterns, shapes, and textures in images.¹⁸

• Typical Applications: Applications like “Mianxiang Dashi” ²⁰ and
  “AI Future Seer” ²¹ allow users to upload their photos. The AI system then analyzes the photo and generates a “physiognomy report” on the user’s personality, career, wealth, marriage, etc. These applications often package their services with a tech-savvy interface and promotional language like “AI intelligent decoding”.²¹

2.1.3 Category Three: Integrated Social and E-commerce Platforms

The strategic vision of these applications goes beyond being a single tool, aiming to build an ecosystem with high user stickiness around metaphysical content.

• Core Functions: They seamlessly integrate divination tools, social communities, and e-commerce, forming a complete business loop.

• Typical Applications: “Tufeng Lingshui” is a prime example of this model. It not only provides various divination tools like Bazi charts and feng shui compasses but, more crucially, it has built-in social features that allow users to share their charts, exchange insights, and connect with like-minded individuals . At the same time, it also has a physical goods store that directly sells related products like lucky charms and feng shui ornaments to users, directly converting online traffic into e-commerce revenue . “Cece” also places great emphasis on building its open community. Through rich topics and interest matching, it encourages users to share their moods and help each other, thereby significantly increasing user retention and activity .

2.2 The Business Model of Mysticism: Value Propositions and Monetization Paths

The commercialization paths for AI metaphysics applications are clear and diverse, primarily revolving around the core logic of “attract with free, monetize with value-added.”

• Dominant Model: Freemium
  The vast majority of applications adopt a freemium model. They attract a large user base by offering basic functions for free, such as generating a simplified report or a basic chart.23 Once users become interested and want more in-depth, personalized interpretations, they need to pay for premium services. This model effectively lowers the initial barrier to entry for users, paving the way for subsequent commercial conversion.25

• Core Monetization Levers:

  1. One-time Paid Reports: This is a common model for applications like AI face reading. The app will first generate a tantalizing preview analysis. To unlock the full report covering career, wealth, marriage, etc., the user must pay a one-time fee, typically ranging from 30 to 40 RMB.²⁴

  2. Tiered Membership Subscriptions: For more complex, content-rich platform applications, subscriptions are the main source of income. For example, “Tufeng Lingshui” offers monthly, quarterly, and annual memberships, with prices ranging from about 29.99 RON (approx. 47 RMB) per month to 299.99 RON (approx. 470 RMB) per year . The Pro subscription version of “TarotAI” offers privileges such as unlocking all advanced spreads, higher-precision AI interpretations, and an ad-free experience.¹⁷

  3. Pay-per-use API Services: At the B2B level, a “Divination-as-a-Service” business model has emerged. Tech companies like Aikeruite provide their self-developed AI fortune-telling engines as API interfaces to other consumer-facing companies (such as Lingji Culture). The technology provider charges per API call (e.g., 1 RMB per query), while the client company can package the service and sell it to end-users at a premium of tens of times, thus earning high profits .

• The Value Proposition of AI Oracles:
  Compared to traditional human diviners, AI metaphysics products offer users several key value points:

  • Convenience and Accessibility: AI services are available 24/7, allowing users to get “guidance” anytime, anywhere, without needing to make appointments or wait in line, meeting the modern demand for immediacy.¹³

  • Efficiency: A chart analysis that would take a human tens of minutes or even longer can be completed by an AI in just a few seconds.¹³

  • Cost-Effectiveness: Although some advanced services can be expensive, their prices are relatively low compared to the consultation fees of famous masters, which can run into thousands or even tens of thousands of RMB.¹³

  • Privacy and Perceived “Objectivity”: Users may be more willing to disclose personal information to an anonymous machine program. At the same time, AI is promoted as relying on “vast databases” and algorithms for analysis, which is packaged as an “objective” advantage that avoids the subjective biases and emotional influences of human diviners.¹³

Table 1: Comparative Analysis of Mainstream AI Metaphysics Applications in 2025

Application Name	Main Functions	Core AI Technology	Business Model	Key Differentiators
Tufeng Lingshui	Comprehensive numerology platform (Bazi, Feng Shui, Nameology, etc.)	Hybrid (Traditional algorithms + AI intelligent analysis)	Subscription (Monthly/Quarterly/Annual); Physical e-commerce	Social + e-commerce ecosystem, creating a complete business loop
TarotAI	AI Tarot card reading	Generative AI (LLM)	Freemium; Pro version subscription	Focuses on in-depth Tarot reading, supports multiple complex spreads, provides high-precision AI analysis ¹⁷
Mianxiang Dashi (Face Reading Master)	AI face scanning for physiognomy	Computer Vision (DNN, CNN)	One-time paid reports	Specializes in digital physiognomy, promotes its model as combining master experience and training on large facial datasets ²⁰
Cece	Astrology/MBTI community and tools	Hybrid (AI intelligent Q&A, algorithmic matching)	Freemium; VIP membership subscription	Strong social community attribute, deeply integrating tools with interest-based networking and emotional support
Services based on DeepSeek	In-depth Bazi numerology analysis	Generative AI (Specialized LLM)	B2C one-time reports (e.g., Fatetell); B2B prompt guides	Relies on the advantage of a specific LLM in handling Chinese numerology logic, providing more inferential analysis ¹⁴

The formation of this market landscape reveals two deep-seated structural shifts.

First, the most successful market players are moving beyond the limitations of single tools to build “full-stack” metaphysical service ecosystems. The development logic of applications has evolved from simply providing a “calculator” to creating a highly sticky digital space. Taking “Tufeng Lingshui” and “Cece” as examples, their functional design clearly reflects this strategy. Free divination tools act as efficient user acquisition engines, attracting a large number of entry-level users to the platform; active social communities cultivate a sense of belonging through content sharing and interpersonal interaction, thereby achieving user retention; finally, through paid in-depth reports, premium subscriptions, and direct e-commerce, the large user base is converted into substantial business revenue. This model builds strong competitive barriers because the reason for user retention is no longer just the accuracy of the tool, but also the social relationships and accumulated content within the platform. This indicates that future market competition will be between ecosystems, not a comparison of isolated application functions.

Second, a hidden but crucial B2B “Divination-as-a-Service” market is quietly taking shape. The case of Aikeruite reveals the vertical development of the industry chain. In this layered structure, upstream technology companies focus on the R&D of core algorithms and AI models—a capital and technology-intensive segment; while numerous downstream consumer-facing application companies can avoid building complex AI engines from scratch. By calling APIs, they can concentrate on their areas of expertise: user experience design, brand marketing, and community operations. The emergence of this model is a sign of market maturation. It greatly lowers the entry barrier for new players in the consumer market and accelerates the diffusion of innovation, but it also makes the issue of liability attribution more complex. When an API-driven application gives wrong or harmful advice, whether the responsibility should be borne by the model developer or the application operator will become a thorny legal and ethical issue.

Part Three: Navigating the Risks: The Minefield of Technology and Ethics

Beneath the prosperous facade of the AI metaphysics market lies a complex minefield woven from technical limitations, ethical dilemmas, and legal risks. Ignoring these risks could not only harm user interests but also plant hidden dangers for the future development of the entire industry.

3.1 The “Black Box” Oracle: Trust, Transparency, and “Tech-washing”

One of the core selling points of AI metaphysics applications is the “scientific” cloak they wear. Marketing discourse is filled with terms like “deep neural network learning,” “big data precision calculation,” and “AI artificial intelligence technology,” attempting to create an impression of objectivity, reliability, and even superiority over human experts . However, a huge gap exists between this narrative and the technological reality.

Multiple technology experts and industry insiders have pointed out that the essence of most current so-called “AI fortune-telling” is not genuine artificial intelligence reasoning, but a form of “tech-washing”.²² The underlying logic is often very simple: the program’s backend has a pre-set, massive database of template phrases. When a user inputs information (like a photo or birth date), the system uses simple feature matching to retrieve and splice corresponding descriptive sentences from the database, generating a seemingly personalized report.²³ The technical barrier for this model is extremely low, involving almost no complex AI reasoning. Its falsity is also evident in practice: tests have shown that uploading the same photo or palm picture can generate completely different analysis results, fully exposing the randomness and pre-set nature of its outcomes, rather than rigorous “calculation”.²³

A deeper problem is that even applications using genuinely advanced machine learning models cannot escape the “algorithm black box” issue. The decision-making process of modern deep learning models, especially complex neural networks, is often opaque to humans.²⁷ A model can learn extremely complex patterns from vast amounts of data, but when it gives a specific conclusion, even the developers find it difficult to fully explain the specific attribution logic behind it. This “untraceability” has already caused great concern in high-risk fields like medicine and finance. When an AI metaphysics app gives advice that could influence a user’s major life decisions (such as career choice or marriage), and this advice is based on an inexplicable “black box” process, the foundation of trust is shaken. If an error or harmful advice occurs, accountability becomes extremely difficult, because no one can clearly explain “why the AI thought that way”.²⁷ This technological characteristic ironically echoes the mysterious and inscrutable nature of ancient oracles, but its claimed scientific basis is thereby nullified.

3.2 Algorithmic Bias and Digital Discrimination

An AI model’s power comes from the data it learns from, which is also its greatest weakness. AI systems are not inherently neutral; they are mirrors of their training data. If the training data itself contains deep-seated societal biases—for example, stereotypes about gender, race, region, or social class—the AI model will not only replicate these biases during its learning process but may even amplify and entrench them .

In the context of AI metaphysics applications, this risk is particularly prominent. A language model trained on historical text data might “learn” to strongly associate certain professions with a specific gender, thus unconsciously perpetuating outdated concepts like “men work outside, women stay at home” or “women are not suited for STEM fields” in its career advice to users. Similarly, if an AI face-reading application’s training data shows a statistical correlation between a certain facial feature and negative life outcomes (whether this correlation is real or purely coincidental), it may make discriminatory “judgments” about user groups with that feature.

This algorithm-driven discrimination is more insidious and harmful than human bias because it is masked by the veneer of being “objective” and “data-driven.” Users may be more inclined to believe a conclusion from a “supercomputer” without realizing that it might just be the result of flawed data and algorithmic logic. This risk runs counter to China’s officially released “New Generation Artificial Intelligence Ethical Norms,” which explicitly require AI applications to “promote fairness and justice,” adhere to inclusiveness, and avoid bias and discrimination.²⁸ If AI metaphysics applications cannot effectively address the problem of algorithmic bias, they will always be walking a fine line on ethics and regulation.

3.3 The High Cost of Prophecy: Privacy and Data Exploitation

AI metaphysics applications have an insatiable appetite for data, and what they collect is precisely the user’s most sensitive and core personal information. To generate a “precise” report, users are asked to provide information including their real name, date and place of birth accurate to the minute (which is highly sensitive personal identity information), and even unchangeable biometric information—facial photos and palm prints.²³

However, the practices of many applications in handling this highly sensitive data are alarming. A survey conducted as early as 2021 found that such apps in the market generally had serious compliance issues.²⁴ Among the several apps tested, the vast majority failed to inform users in a clear and explicit manner that they were collecting biometric information such as faces, as required by laws and regulations like the Personal Information Protection Law (PIPL), nor did they obtain the user’s “separate consent” for processing such sensitive information. The privacy policies of many apps were either vague or non-existent, hiding data collection practices in lengthy terms and conditions that users rarely read.²⁴

This disregard for user privacy poses huge security risks. First, centrally storing large amounts of sensitive data containing names, birthdays, and biometric features on servers with unknown security levels is tantamount to creating a highly attractive target for data breaches and cybercrime. Once the database is compromised, this information could be used for various malicious activities such as targeted fraud and identity theft. Regulatory bodies like China’s Ministry of Industry and Information Technology have issued multiple warnings, reminding the public to be wary of the personal information leakage risks associated with “AI face reading” apps.²⁹

Second, the chain of legal liability is clear. According to regulations such as the Cybersecurity Law and the Personal Information Protection Law, app stores, as platform operators, have a responsibility to verify the identity of the application providers and conduct basic compliance reviews.²³ Legal experts point out that for applications that are themselves illegal and non-compliant and could pose significant risks to consumers, app stores should block or remove them.²³ This means that future regulatory crackdowns may not only target app developers but also the app stores like Apple’s App Store and major Android markets that are lax in their management.

Table 2: Risk Matrix Analysis of AI Metaphysics Platforms

Risk Category	Specific Manifestations	Potential Impact	Severity	Mitigation Strategies
Data Privacy & Security	Collecting biometric info (e.g., faces) without separate consent; missing or vague privacy policies; unclear data storage security measures.	Violation of PIPL, facing heavy fines; user class-action lawsuits; brand reputation collapse due to data breaches.	High	Implement strict data governance; design clear, separate consent flows for sensitive info; adopt high-standard encryption and security measures; conduct regular security audits.
Algorithmic Bias & Ethics	Reinforcing stereotypes (e.g., gender) in career/relationship advice; making discriminatory judgments based on biased data.	Harming users’ equal opportunities; triggering public opinion crises; exacerbating social discrimination; losing user trust.	High	Audit training data for diversity and representation; develop and deploy bias detection and mitigation algorithms; establish an algorithm ethics review board; increase algorithmic transparency.
Regulatory & Legal Compliance	Exaggerated or false advertising claims; providing consulting services without relevant qualifications; lax app store review processes.	Penalties from market regulators; mandatory app removal; business operations halted; platform liability.	High	Hire legal counsel for comprehensive compliance review; standardize marketing language, avoiding absolute terms like “accurate prediction”; proactively communicate with regulators.
User Psychology & Social Impact	Causing user over-reliance, leading to escapism from real-world responsibilities; incorrect negative “prophecies” causing user anxiety/depression.	Causing psychological harm to users, potentially leading to legal disputes; being labeled a “spiritual opium,” facing social criticism.	Medium	Place prominent disclaimers emphasizing entertainment and self-exploration purposes; provide links to mental health support resources; avoid fatalistic and fear-mongering language.

In summary, the foundation of the AI metaphysics industry is built on a profound contradiction: it attempts to package and sell a product that is inherently dependent on faith, intuition, and agnosticism using the language of logic, data, and computation. The marketing logic of this model is to use the public’s general perception of AI technology as “powerful, precise, and objective” to grant a modern legitimacy to ancient divination arts. However, the actual technical level of many applications falls far short of their promotional claims, resembling more of a “technological sleight of hand.” This “trust paradox”—asking users to believe in AI because it is rational, while the content it provides is unfalsifiable—constitutes a highly unstable business foundation. As public AI literacy improves, applications that are merely “tech-washed” will face the risk of being exposed.

At the same time, the entire industry is sitting on a regulatory time bomb. Systemic failures in data privacy compliance, potential risks of algorithmic discrimination, and the possible psychological harm to users all make this industry a high-priority target for regulatory agencies. The current “gray area” of unchecked growth cannot last. It is foreseeable that as the market size and influence expand, stricter and more targeted regulatory measures will be introduced. Any plan to invest or start a business in this field that fails to place compliance and ethics at its core strategy will bear enormous, and potentially fatal, risks.

Part Four: 2025 and Beyond: Development Trajectory and Strategic Outlook

Looking ahead to 2025 and beyond, the AI metaphysics market stands at a critical crossroads. It faces opportunities for technological iteration and market deepening, while also bearing multiple pressures from ethics, regulations, and user expectations. Its future trajectory will depend on how market participants balance innovation with responsibility.

4.1 Market Evolution and Technological Frontiers

• From Passing Fad to Mainstay of a Niche Market: The underlying factors driving this consumption wave—anxiety under economic uncertainty, the pursuit of emotional value, and the deep psychological need to find meaning and order in confusion—are long-term, not transient trends.³⁰ Therefore, although the specific forms of metaphysical products or applications may constantly change, the market for services that meet these core needs will persist. It is foreseeable that metaphysics consumption will evolve from an explosive “internet-famous” phenomenon into a stable niche market within the broader health, entertainment, and personal growth sectors.

• The “Arms Race” of Generative AI: Technologically, the future competitive focus will undoubtedly be on generative AI. The market will shift from providing static reports to offering dynamic, continuous, and hyper-personalized services. It is conceivable that the next generation of AI oracles will evolve into 24/7 “AI life coaches” that can learn and remember all of a user’s historical interaction data, thereby providing coherent guidance that grows with the user.¹¹ The development and application of more powerful and specialized vertical-domain large language models (such as the rumored next-generation model R2 from DeepSeek) will be key for leading players to build technological barriers .

• Immersive Mystical Experiences: The next technological frontier is likely the integration of AI metaphysics with immersive technologies like XR (Extended Reality), VR (Virtual Reality), and AR (Augmented Reality). Imagine an AR application that can overlay feng shui layout suggestions onto a user’s apartment in real-time through a phone camera; or a user engaging in an immersive conversation with a lifelike, AI-driven tarot reader in a fully virtual environment. Such multi-sensory, highly interactive experiences will create emotional resonance and user stickiness far exceeding current text-and-image interfaces.

4.2 Strategic Recommendations for Market Stakeholders

4.2.1 For Developers and Entrepreneurs:

• Embrace Ethical Design, Turn Risks into Opportunities: Instead of passively waiting for regulation, proactively make ethics and security a core competitive advantage of the product. Prioritize data privacy protection, algorithmic transparency, and user well-being. In product design, clearly inform users of the AI’s limitations, positioning it as a tool for auxiliary self-reflection and entertainment reference, rather than a “truth machine” capable of predicting the future. This honesty can not only mitigate legal risks but also build long-term trust with users.

• Build a Community, Not Just a Tool: As analyzed earlier, the most defensible business model is based on an ecosystem. Future winners will be companies that can successfully build a vibrant, engaging community platform around their core tools . Resources should be invested in community operations, content creation, and user interaction to make the platform a part of the user’s spiritual life.

• Continuously Innovate in Human-Computer Interaction: Competitive differentiation will lie in creating more natural, empathetic, and inspiring conversational experiences. Exploring how to use AI technology to guide positive psychological outcomes, such as promoting mindfulness practices, enhancing user resilience, or providing emotional support through conversational AI, will be a highly valuable direction for innovation.

4.2.2 For Investors:

• Conduct Rigorous Due Diligence: Investment decisions should not be based solely on superficial metrics like user growth. It is essential to deeply scrutinize a target company’s compliance status, especially whether its data privacy policies meet the requirements of the Personal Information Protection Law, whether its algorithms pose significant bias risks, and whether its business model is sustainable in the long term. The risk matrix provided in Part Three of this report can serve as a practical evaluation framework.

• Focus on Platform Companies and B2B Enablers: In addition to directly investing in consumer-facing applications, opportunities in the upstream of the industry chain should also be considered. B2B technology companies that provide core AI capabilities to downstream applications (i.e., “Divination-as-a-Service” providers) may be a more scalable investment target with lower brand risk . At the same time, companies dedicated to building ecosystem platforms rather than single tools have stronger moats and long-term growth potential.

• Look for Sustainable Value Propositions: Favor companies that build their brand on trust, community, and genuine emotional support, rather than those that rely on questionable accuracy claims and opaque technology. The business models of the latter are extremely vulnerable to regulatory and public opinion pressures.

4.2.3 For Regulators and Policymakers:

• Implement Proactive and Refined Regulation: Specific regulatory guidelines should be developed for the emerging cross-disciplinary field of “AI + Metaphysics.” These guidelines need to address its unique risk points, including regulations on the collection and use of highly sensitive personal information (especially biometric information), requirements for algorithmic transparency, and measures to prevent psychological harm to users.

• Strictly Enforce Existing Laws: For applications on the market that violate existing laws such as the Personal Information Protection Law, law enforcement should be strengthened with regular reviews and removals to effectively protect consumer rights. At the same time, the platform responsibility of app stores should be reinforced, urging them to fulfill their role as “gatekeepers”.²³

• Enhance Public Digital Literacy: Conduct public education campaigns through official channels and media partnerships. Help consumers understand the nature of AI metaphysics services, recognize their potential risks in areas like data privacy and algorithmic bias, and guide them to view AI outputs rationally, using them as a form of entertainment or inspiration rather than the sole basis for decision-making.

Conclusion

The rise of the algorithmic oracle is a phenomenon that profoundly reflects the spirit of our times. It is an unexpected collision between an ancient human need—to seek comfort and meaning in uncertainty—and cutting-edge technology—artificial intelligence. It reveals that contemporary youth are coping with an increasingly complex and unpredictable world in a new, digital way. This market undoubtedly presents real business opportunities, but its future destiny will ultimately depend on whether all participants—developers, investors, and regulators—can collectively navigate the profound ethical, legal, and social challenges that lie beneath.

In this race, the ultimate victors will be those who always remember that behind every data point is a living individual. What they seek is perhaps not just a simple answer, but a sense of understanding, a shred of comfort, and a feeling of regaining control over their own lives. Only when technology serves this purpose can it achieve stable and long-term success.

Cited works

1. 什么是计算机视觉？ | 数据科学| NVIDIA 术语表 - 英伟达, https://www.nvidia.cn/glossary/computer-vision/

2. 2024消费趋势：“精神悦己”下的商业新潮向| 知萌发布 - 微博营销, https://marketing.hd.weibo.com/article/view/8732

3. 什麼是生成式AI？– Gen AI 說明 - AWS, https://aws.amazon.com/tw/what-is/generative-ai/

4. App Store 上的“面相大师—AI扫脸看面相与手相App” - Apple, https://apps.apple.com/cn/app/%E9%9D%A2%E7%9B%B8%E5%A4%A7%E5%B8%88-ai%E6%89%AB%E8%84%B8%E7%9C%8B%E9%9D%A2%E7%9B%B8%E4%B8%8E%E6%89%8B%E7%9B%B8app/id1457634472

5. 什么是计算机视觉？ - Microsoft Azure, https://azure.microsoft.com/zh-cn/resources/cloud-computing-dictionary/what-is-computer-vision

6. 一张照片测出两个结果！AI算命App套路多，专家：警惕个人信息被滥用 - 央广网, http://m.cnr.cn/tech/20220514/t20220514_525826657.html

7. 花钱百元能转运？年轻人热衷算命，催生互联网玄学“生意经” - 华龙网, https://h5.cqliving.com/info/detail/98806775.html?cid=98806775&cqxhlwdc=3f

8. AI心，海底针，怎么破？ - 清华大学, https://www.tsinghua.edu.cn/info/1182/87679.htm

9. AI算命？大师你能算出自己什么时候被封号吗, http://news.cctv.com/2019/11/27/ARTIa7Zg0YsB6r4TeDIcgXtk191127.shtml

10. 什么是算法偏见？ - IBM, https://www.ibm.com/cn-zh/think/topics/algorithmic-bias

11. 英敏特发布《2025全球消费者趋势》报告 - Mintel, https://china.mintel.com/press-centre/mintel-announces-global-consumer-trends-for-2025/

12. 疫后重点关注年轻人消费, https://pdf.dfcfw.com/pdf/H3_AP202302061582791964_1.pdf

13. 冯子轩｜生成式人工智能应用的伦理立场与治理之道：以ChatGPT为例, http://www.socio-legal.sjtu.edu.cn/wxzy/info.aspx?itemid=4497&lcid=30

14. 图风灵水- AI命盘×开运打卡社交App on the App Store - Apple, https://apps.apple.com/ro/app/%E5%9B%BE%E9%A3%8E%E7%81%B5%E6%B0%B4-ai%E5%91%BD%E7%9B%98-%E5%BC%80%E8%BF%90%E6%89%93%E5%8D%A1%E7%A4%BE%E4%BA%A4app/id1495759890

15. 面相大师—AI扫脸看面相与手相App - App Store, https://apps.apple.com/hk/app/%E9%9D%A2%E7%9B%B8%E5%A4%A7%E5%B8%88-ai%E6%89%AB%E8%84%B8%E7%9C%8B%E9%9D%A2%E7%9B%B8%E4%B8%8E%E6%89%8B%E7%9B%B8app/id1457634472

16. 什么是生成式AI | 生成式AI的定义与应用 - SAP, https://www.sap.cn/products/artificial-intelligence/what-is-generative-ai.html

17. 测测- 星座，八字，AI问答- Google Play 上的应用, https://play.google.com/store/apps/details?id=com.lingocc.cc5&hl=zh

18. 仲量联行报告：“新三代”消费者正催生出九大消费新业态 - 证券时报, http://stcn.com/article/detail/1664403.html

19. 人工智能如何支持多样性、公平性和包容性？ - 世界经济论坛, https://cn.weforum.org/stories/2022/08/ren-gong-zhi-neng-ru-he-zhi-chi-duo-yang-xing-gong-ping-xing-he-bao-rong-xing/

20. 《新一代人工智能伦理规范》发布 - 中华人民共和国科学技术部, https://www.most.gov.cn/kjbgz/202109/t20210926_177063.html

21. Astra AI: 你的专属AI占星师- Google Play 上的应用, https://play.google.com/store/apps/details?id=birth.chart.vedic.astrology.numerology.human.design.cosmic.insights.personal.free.tests.discover.yourself.app.slsolutions.personality&hl=zh

22. “AI看面相”算命是假，算钱是真 - 人民日报, https://www.peopleapp.com/column/30036735921-500001782429

23. 深度解读计算机视觉技术 - 中国科学院半导体研究所, http://semi.cas.cn/kxcb/kpwz/201809/t20180912_5077396.html

24. 什么是生成式AI？ 示例与应用场景 - Google Cloud, https://cloud.google.com/use-cases/generative-ai?hl=zh-CN

25. App Store 上的“测测-心理情感AI问答社区” - Apple, https://apps.apple.com/cn/app/%E6%B5%8B%E6%B5%8B-%E6%98%9F%E5%BA%A7%E5%BF%83%E7%90%86ai%E9%97%AE%E7%AD%94%E7%A4%BE%E5%8C%BA/id756771906

26. 这届年轻人拿捏消费：可以买贵的，不能买贵了 - 中国网财经, https://m.gxfin.com/article/finance/lc/cfsh/2025-02-14/6213321.html

27. 生成式AI 是什麼？3 大生成式AI 應用與工具分享！ - Solwen AI, https://solwen.ai/posts/generative-ai

28. 新三代如何改变消费市场格局？《时尚消费力洞察报告2025 》正式 …, https://www.21jingji.com/article/20250412/herald/077317a9de49b5c7ee7f226f9d789bfa.html

29. 魔法水晶、塔罗占卜……玄学消费用情绪价值抓住年轻人, https://www.ccn.com.cn/Content/2025/01-23/1646536154.html

30. 从补偿性控制理论视角看现代青年掀起的“玄学潮”, https://pdf.hanspub.org/ap2025151_241135414.pdf

31. 《碎银时代下的情绪红利：2024生活趋势洞察报告》 目录｜ 01, https://file.digitaling.com/eImg/uimages/20250121/1737427179377767.pdf

32. App Store 上的《测测-心理情感AI问答社区》 - Apple, https://apps.apple.com/hk/app/%E6%B5%8B%E6%B5%8B-%E5%BF%83%E7%90%86%E6%83%85%E6%84%9Fai%E9%97%AE%E7%AD%94%E7%A4%BE%E5%8C%BA/id756771906

33. 面相解碼-AI算命- Google Play 應用程式, https://play.google.com/store/apps/details?id=com.weiyou.aifutureseer&hl=zh_TW