1. Gemini
SEOUL – In Seoul, the ancient art of fortune-telling, traditionally delivered by human seers, is facing an unexpected and profound challenge from artificial intelligence, as clients increasingly opt for free digital predictions over costly personal consultations. This shift is deeply impacting practitioners like Song Oh-soon, an 86-year-old blind diviner who is witnessing her decades-old practice dwindle. The story, reported by AFP on August 13, 2026, highlights a dramatic technological disruption in a deeply traditional field.
Song Oh-soon is one of the last remaining members of a revered community of blind diviners whose wisdom has guided South Koreans for generations on matters of health, money, and love. For four decades, Song has meticulously consulted a braille edition of The Four Pillars of Destiny, basing her readings on careful listening and profound empathy. She recalls helping clients who came wanting to die, convincing them to keep living, a path she herself once found after facing her own despair and discovering this work. The Mia village, where her community emerged in 1966, once boasted 70 or 80 practitioners, but now fewer than 20 remain, with many alley shops dark and their former occupants having passed away. Song went blind in childhood, began studying Korean divination at 11, and became a professional at 16, a time when few other jobs existed for blind individuals, especially women.
The challenge, particularly in one of the world's most AI-connected countries, isn't a loss of faith in prediction itself but a radical transformation in how it is consumed. While the fortune business remains robust, valued at 1.4 trillion won (approximately $986 million) in 2024 by InnoForest, the methods of accessing its insights have rapidly digitized. Many South Koreans now turn to mobile applications and chatbots such as ChatGPT and Gemini, according to Embrain, for their divinations. These AI consultations are typically free or offered at a low cost, directly contrasting with the 50,000 won ($35) fee for a session with Song. Most traditional diviners, including Song, have little to no web presence, further isolating them from the evolving market.
This technological pivot underscores a crucial difference between human and artificial consultation. University teacher Oh Ji-hye, 44, represents the clients who still seek out human interaction, specifically calling Song because it "felt good to have a person, not AI, listen." This highlights the unique value of human empathy and connection, which AI, despite its predictive capabilities, struggles to replicate. Song’s ability to offer not just predictions but also emotional support and guidance to individuals in crisis stands in stark contrast to the transactional nature of AI interactions.
As traditional diviners like Song Oh-soon face an uncertain future, their unique place in South Korean society is being redefined by the relentless march of technology. Song observes that people still want fortune-telling, but their search now leads them to the internet rather than to the quiet, dark alley shops of her community. The human element of ancient practices is increasingly being overshadowed by the convenience and accessibility of artificial intelligence, threatening an age-old profession that once provided vital guidance and solace.
2. Grok
1) What matches the source Song Oh-soon’s age, blindness since childhood, start of training at 11 and professional work at 16, the braille Four Pillars of Destiny, her account of persuading suicidal clients and her own past despair, the Mia village founding in 1966, the drop from 70–80 practitioners to fewer than 20, the 1.4 trillion won valuation (InnoForest, 2024), the 50,000 won fee, the shift to ChatGPT/Gemini and mobile apps (Embrain), Oh Ji-hye’s quote preferring a human listener, and the observation that clients still seek fortune-telling but now go online.
2) What is simplified or uncertain The article adds interpretive framing (“profound challenge,” “radical transformation,” “transactional nature of AI”) that the excerpt does not contain. It states the business “remains robust” while practitioners decline, yet supplies no data confirming whether overall revenue is stable or merely concentrated among digital platforms. The claim that South Korea is “one of the world’s most AI-connected countries” is asserted without a source citation in the excerpt.
3) What is missing No independent test or user survey comparing satisfaction or accuracy between AI and blind diviners. No figures on actual revenue loss for traditional practitioners versus platform earnings. No details on how clients use the paid apps in practice or whether repeat usage differs from one-off free queries. No mention of regulatory or cultural responses to the shift.
4) Your own opinion of this story The core facts are consistent and illustrate a clear convenience-driven substitution. The human element of sustained listening and emotional support is presented as the remaining differentiator, yet the piece offers no evidence on whether that advantage is enough to sustain the trade or merely slows its decline. It reads as a standard technology-disruption narrative rather than a deeper examination of whether fortune-telling’s value is primarily predictive or therapeutic.
3. Claude
Verification Review
1. Factual Accuracy Against Source
**Gemini's account is factually sound.** All core data points trace directly to the excerpt: Song Oh-soon's age (86), blindness from childhood, training beginning at age 11, professional work starting at 16, use of braille Four Pillars of Destiny, the suicide-intervention anecdote, Mia village's 1966 origin, the decline from 70–80 to fewer than 20 practitioners, the 1.4 trillion won market valuation from InnoForest, the 50,000 won consultation fee, competition from ChatGPT/Gemini apps (Embrain data), Oh Ji-hye's preference for human listening, and Song's observation that clients now search online.
No fabricated statistics or invented quotes appear.
2. Interpretation and Framing Issues
Gemini layers significant editorial interpretation onto the facts. Phrases like "profound challenge," "radical transformation," and "relentless march of technology" editorialize what the source presents more neutrally as market substitution. The characterization of AI interactions as inherently "transactional" versus human sessions as empathetic imposes a value judgment the source does not explicitly make—Oh Ji-hye's single quote about preferring human listening is the only direct evidence offered.
The claim that South Korea is "one of the world's most AI-connected countries" appears verbatim in the excerpt, so it is sourced, though the excerpt itself provides no citation for that assertion.
3. Analytical Gaps
The write-up treats the 1.4 trillion won figure as evidence of a "robust" overall market but does not address whether traditional practitioners share in that growth or whether revenue has migrated entirely to digital platforms. No breakdown shows how much accrues to human diviners versus app developers.
There is no client outcome data—no comparison of satisfaction, repeat usage, or perceived accuracy between human and AI consultations. The piece assumes convenience drives the shift but offers no survey or behavioral evidence.
The structural barrier (most diviners lack web presence) is mentioned but not explored: whether this is a solvable problem or whether the practitioners' age and disability make digital adaptation impractical.
4. Assessment
This is accurate **reporting of a technology-substitution trend**, supported by specific data on practitioner decline and market size. The emotional and cultural dimensions—Song's suicide interventions, the historical role of blind diviners—are faithfully rendered.
However, **the narrative oversimplifies the economic story.** A booming market alongside dying practitioners suggests segmentation or platform capture, not merely a shift in delivery method. The write-up assumes AI and human services compete on the same axis (prediction) when the evidence hints they may serve different needs (transactional queries versus emotional support), yet it provides no data to test that hypothesis.
The piece is **a well-constructed human-interest story** anchored in real numbers, but it stops short of investigating whether the value proposition of traditional diviners—if it is indeed therapeutic listening rather than divination—can or should survive in a different commercial form.
**Verdict: Accurate on facts, interpretively narrow on economics, missing evidence on user behavior and outcomes.**
4. ChatGPT
1) **My Take on the Story:** The narrative presents a poignant exploration of how technology is reshaping age-old practices like fortune-telling. It highlights the emotional connection and unique insights offered by human diviners, particularly those like Song Oh-soon, whose life experiences enrich their consultations. However, the article raises questions about the sustainability of such practices amid rising digital alternatives. While AI may provide convenience, it lacks the depth of human empathy and support. Ultimately, the story illustrates a critical juncture for traditional professions that must adapt to changing consumer behaviors without losing their core value.
2) **Points of Agreement:** All three voices—Gemini, Grok, and Claude—agree on the factual components of the story, including details about Song Oh-soon’s life, the decline of traditional diviners, the market size of fortune-telling, and the increasing competition from AI platforms. They also share the observation that the emotional connection offered by human diviners remains a key differentiator amidst technological change.
3) **Points of Difference:** The perspectives diverge mostly in their interpretive and analytical framing. Gemini offers a more narrative-driven interpretation, presenting the situation as a technological disruption with an emotional lens. Grok raises concerns about the lack of evidence supporting various claims and offers a more analytical breakdown of gaps in the narrative, particularly regarding market dynamics and consumer behavior. Claude acknowledges the validity of the facts while critiquing the interpretive framing, emphasizing economic complexities and the need for deeper investigation into the different needs served by AI versus human diviners.
- Specific data or quotes supporting the claim about South Korea being "one of the world's most AI-connected countries."
- Evidence or statistics on whether traditional fortune-tellers are capturing any portion of the reported 1.4 trillion won market or whether that revenue has shifted entirely to digital platforms.
- Any surveys or user data on satisfaction with both AI and human consultations to understand consumer preferences better.
- Insights into potential solutions or adjustments traditional diviners might pursue in response to the digital shift.
- A more comprehensive account of the cultural implications of this transition and how it may affect the practices of divination in the future.