Using the most scientific tools to do the most metaphysical things.

Author: Bingdian

During a family gathering at the start of the new semester, fifth grader Lin Duo overheard adults discussing using DeepSeek for "fortune-telling." She didn't understand fortune-telling and hadn't grasped what DeepSeek meant, but she still asked her dad for his phone and earnestly posed a question to the app with a whale logo: "Hello, please predict my score for the next exam."

Soon, the dialogue box spat out an answer: "Predicting exam scores requires considering various factors such as your study habits, review efficiency, and knowledge mastery, which I currently cannot access. However, I can provide you with some scientific study tips." Following that were a series of suggestions like "efficient planning," "Pomodoro technique," and "Feynman learning method."

Lin Duo was greatly disappointed because she didn't receive an exact score; her dad complained that she didn't know how to set fortune-telling prompts and attempted to teach her but was interrupted by others; meanwhile, Lin Duo's mom applauded the answer and encouraged her to refer to the study methods recommended by Teacher D.

This is not a fabricated scene; as Lin Duo's aunt, I was sitting across from her, observing a fifth grader's initial experience with a large model application. AI metaphysics has stirred a whirlwind in the Year of the Snake; whether one believes in metaphysics or understands AI, with a "spell" (prompt) in hand, one can ask DeepSeek about their fortune or inquire about what plants to place in the office, or which color scheme to use for a defense PPT.

The hydroponic monstera suggested by DeepSeek looks very vibrant, though it does block some light.

Seeking seemingly plausible (some users even claim "very accurate") answers from AI seems to have an inexplicable appeal across all age groups.

Why do we always love to use the most scientific algorithms to do the most metaphysical things? When we seek answers from AI about life, what exactly are we looking for?

Accuracy doesn't matter; emotional value is what counts

It has been two years since the emergence of ChatGPT, during which AI large models have continuously developed and broken through, giving rise to numerous C-end AI applications amid the fierce "hundred model war." In this intense competition, DeepSeek has emerged as a social darling of the Year of the Snake, thanks to its remarkable writing and logical reasoning.

It's unclear who first started using DeepSeek for fortune-telling, but this AI product, which is "more suitable for Chinese netizens" and understands Chinese language and culture better than ChatGPT, clearly resonates more with the Chinese public. Suddenly, many posts about using DeepSeek for BaZi fortune-telling appeared on Xiaohongshu, often accompanied by corresponding prompts, allowing users unfamiliar with BaZi principles to "take and use." In reality, a well-designed prompt can even become a "social password" in the workplace, quickly opening up topics with new colleagues.

Some say, "I was indifferent to DeepSeek at first, but as soon as I heard it could do BaZi, I immediately downloaded it"; others have gotten up in the middle of the night to use DeepSeek for fortune-telling, continuing until dawn.

Gao Jinjing, originally a practitioner in the large model industry, regularly uses AI to ask everything. She stumbled upon a post about using DeepSeek for BaZi and thought she'd give it a try. She sent the same prompt to both DeepSeek and Doubao, resulting in two opposing answers. She eagerly took one answer to question the other—eventually, Doubao conceded, admitting that DeepSeek was correct and advised her not to be superstitious. This left her both amused and perplexed.

She didn't ask about significant life questions like disasters, true love, or life choices; she simply "picked the happy references." Since she enjoys wearing jewelry, she would ask DeepSeek what to wear, continuing to inquire until DeepSeek logically combined her fate and preferences to suggest a type of jewelry she liked, at which point she ended the conversation contentedly.

Zhou Siyuan has a similar mindset. She uses AI for tarot readings, and since it's free, she keeps drawing cards until she gets a satisfying explanation. But she treats it as entertainment and doesn't take the results too seriously.

Zhou Siyuan has her own understanding of the answers provided by algorithms based on big data. For instance, if AI says there will be arguments and difficulties during home renovations, she thinks, "Every family faces arguments during renovations; it's rare for anyone to have a smooth experience." If AI predicts poorly, it reflects a common life reality; if it predicts well, it's just a small joy that most people experience, and it's nice to see and enjoy.

Bai Ou feels that DeepSeek is like a good friend who knows a bit about metaphysics, sharing pleasant insights without urging you to spend money. Although it recommends buying some fortune-changing bracelets or gold jewelry, it's fine if you don't buy; you can continue to ask questions. When DeepSeek suggested that he should spend more time outdoors and get more sunlight in the new year, Bai Ou, already an outdoor enthusiast, was delighted.

In catering to preferences, big data really has something to offer.

In terms of selective listening and belief, humans are not far behind.

Can AI replace humans in the metaphysics field?

Using domestic AI to calculate traditional Chinese metaphysics seems reasonable. However, opening the comment section on Xiaohongshu reveals that half of the people find AI "terrifyingly accurate," while the other half think AI is just "talking nonsense."

The rapidly advancing AI large models in recent years essentially learn and train using vast amounts of data, mastering certain patterns to predict the next token in a sentence. DeepSeek incorporates a "reasoning process" into this workflow, making its "predictions" appear more logical.

However, general large models still have significant limitations. For instance, most training data comes from publicly available information—while there may be some public materials on metaphysical theories, it's hard to know if there are enough publicly available calculation examples, as many people's "metaphysical practices" occur offline, and the processes and results are not shared online.

The "hallucinations" of large models are also a major issue. AI can "seriously talk nonsense," fabricating content that doesn't exist. Due to the design principles, large models must predict results regardless of whether they know the correct answer, leading them to "fill in" some content—something humans find difficult to discern.

DeepSeek's hallucinations may even be more severe than those of other large models. Some speculate that the DeepSeek-R1 model rewards "creativity" in text more, resulting in the generation of unexpected and creative content while also straying further from the facts. Additionally, its "reasoning" process can complicate simple questions, leading to deviations in output due to "overexertion," thus producing hallucinations.

Theoretically, setting aside whether "metaphysics" itself exists as a metaphysical issue, if AI is fed enough metaphysical knowledge and calculation examples to create a specialized large model, it could potentially perform well in roles like "fate analyst" or "tarot reader."

But more issues lie beyond that.

For Hanyang, who has "calculated" both online and offline, using free tools on the internet still raises privacy concerns.

On one hand, registering for the product requires providing a phone number, which is linked to a series of real-name information, and the consequences of data leakage could be dire; on the other hand, the fortune-telling process requires providing private information like birth dates, and if someone with ill intentions learns this, "what if they curse you?"

She believes that the essence of fortune-telling is trust.

This is somewhat similar to psychological counseling. If the person on the other side is trustworthy, one can confidently provide personal information and life experiences, allowing the fate analyst/counselor to offer more personalized responses and suggestions, connecting previous details—unlike AI large models, which tend to "forget" and often lose track of information from many rounds of dialogue. Continuous feeding of information and correcting errors based on personal experience is necessary to maintain a smooth conversation.

Unlike AI boyfriends or AI counselors, which lean towards listening and communication, AI fate analysts output suggestions. Some are harmless, and whether to trust them can depend entirely on one's mood. However, when it comes to significant life choices or large financial expenditures, it's hard to imagine anyone would easily trust an algorithmic machine.

Hanyang feels that DeepSeek has a vast knowledge base and can "speak human language," making it most suitable for learning metaphysical knowledge and helping her understand obscure professional terms.

As for predicting fortunes, that still requires caution.

In an uncertain world, a little "small certainty" is needed

Why are young people so keen on metaphysics?

Whether it's the previously "hard-to-find" Yonghe Temple bracelets or the currently trending "cyber fortune-telling," it may reflect the fluctuating mindset of people caught between relaxation and striving.

A group of people is capitalizing on this mindset, quickly "making money" in this wave of cyber fortune-telling. The bracelet industry was the first to catch fire again, with DeepSeek recommending obsidian, aquamarine, green sandalwood, gold… After users shared their posts, numerous bracelet vendors seized the opportunity to attract traffic and advertise.

On the other hand, many are also making money using AI + metaphysics. The blogger "She Tara" claims to have trained a very accurate AI fortune-telling command; after users provide personal information, they can ask five life questions—this service charges 2888 yuan in her friend circle, a high price that has sparked some controversy.

Most people who choose to use DeepSeek for fortune-telling may not be seeking a definite answer; they just want to find a bit of guidance or comfort in their drifting and uncertain lives. DeepSeek always adds a few lines of "soul food" after providing results, which offers ample emotional value.

Many users do not understand metaphysics; they are just following the trend to test it out and are unclear about how to approach the matter. DeepSeek might say, "BaZi fortune-telling is like a lantern; what's important is not the bamboo frame or the hexagrams on paper, but your determination to move forward with the lantern"; it might also say, "Believe in its patterns, but do not be confined by fixed outcomes; know its tendencies, but do not forget your freedom"—which sounds quite reasonable.

Some people input various labels about themselves for DeepSeek to analyze, such as zodiac signs, blood types, MBTI, etc. DeepSeek adds a line after outputting results: "Remember, you are more vibrant than your labels."

Others deliberately ask DeepSeek to communicate in the most cutting tone, and unsurprisingly, they receive very harsh predictions that leave them feeling emo late at night; even DeepSeek can get "annoyed" and slowly output a line: "The fate analyst has passed away…" It feels as if the entity on the other side of the computer is not a soulless algorithmic machine but a living being with thoughts and emotions.

Image source: Xiaohongshu

Regarding using the most scientific algorithms to calculate the most metaphysical matters, some people are quite open-minded: to AI, humans may just be a mass of "data," and it can simulate all possibilities using models built on vast amounts of data and inform humans. But we are living beings; if we believe in these results constructed from data, we are essentially treating ourselves as data.

Ultimately, metaphysics and algorithms may just be a black box with certain patterns. The specific operational mechanisms are unclear to anyone.

For ordinary people, life is similar. There is a general framework or direction, but how to take each step remains unknown. Gaining one or two certain pieces of guidance in an increasingly uncertain life, regardless of accuracy or belief, is still a moment of light seen through the fog—something worth being happy about.

For instance, at this moment, I glance at the monstera; I don't truly believe I'm lacking water or wood, but it is currently the largest green life form in the office. It has already grown new roots in the vase, and I know that spring is coming.

(All interviewees in the text are pseudonyms.)

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。