Live demo
Full-site walkthrough
Generation AI · Social Science Track · 26.07
IKB · 01 / 15
Reframe Destiny
重塑命运
Identifying and reframing gendered bias in BaZi and astrological narratives
Zhang Xintian · 张昕田 · Lead teacher: Lawted Wu
→ arrow keys to begin
02 / 15 · CONTENT
Content
RQ + Background
Research question · context
Significance
Why divination + AI matter
Literature review
Theory · gap · standing here
Research design & method
Hypothesis · pre/post · six steps
Results
Quant + interviews (N=19)
Conclusion · limits · refs · Q&A
Core message · try the site
Open live demo in new tab
03 / 15 · LIVE DEMO · same site, inside this deck
Share one URL · slides + motion + demo (needs internet)
04 / 15 · RQ · BACKGROUND
What am I studying?
RQ Research question
How does an AI-assisted website help young people spot and rewrite gender bias in BaZi and Western astrology readings?
BG Background
Youth read fate online—but the same chart gets different gender scripts.
- Bias tools miss ke-fu, “too strong,” late-marriage talk
- Most apps give answers, not reframing practice
BaZi grew in patriarchal culture—readings use a subjective lens, not neutral math. 命理在父权语境下形成，解读是主观的。
05 / 15 · SIGNIFICANCE · 意义
Why this project matters
- Almost no public tool uses AI to flag gender bias inside divination (BaZi & Western astrology).
- Fate-talk works like psychological suggestion—under patriarchy it often hurts women’s self-esteem more than men’s.
- In the AI era, a critical read + reframe practice can support young readers—especially women—when results shape confidence.
用 AI 在命理叙事里识别性别偏见、并练习改写——这是本项目的核心社会意义。
06 / 15 · LITERATURE REVIEW
Theory · gap · why I stand here
GENDER
Gender as repeated scripts (Butler, 1990) · who speaks matters (Haraway, 1988)
DISCOURSE
Foucault (1972): fate-talk limits what roles feel possible
AI
LLMs can sound like oracles (Bender et al., 2021) — here AI is a teaching scaffold
GAP · 研究空白
No public tool yet merges BaZi + Western astrology, bias tags, and user-led reframing in one short session.
STANDING HERE · 站在这里
I am among the first in this room with a live game, a real study (N=19), and findings on gender bias in divination sites—not concept slides only.
07 / 15 · RESEARCH DESIGN · HYPOTHESIS
What should change after one visit?
H1 Hypothesis
After Reframe Destiny, people report higher bias awareness and reframe confidence on gendered fate-talk.
DESIGN Within-subjects
Matched pre/post Likert · one session · anonymous Supabase logs
- 5-item Likert pre/post (same session)
- Traditional reading first, then tag · Court · compare
- 5 semi-structured interviews (Jul 2026)
08 / 15 · METHOD
How the study ran
PRE
5items
Likert pre-survey (1–5)
JOURNEY
~15min
Six steps: system → chart → traditional read → bias scanner → Court of Destiny → three-column compare
POST
5items
Same Likert · within-subjects design · 5 semi-structured interviews
Sample
N=19
paired pre/post · 34 sessions · 23 journeys
Deploy
2 URLs
pages.dev + vercel.app
Systems
2-in-1
BaZi + Western · CN/EN UI
Ethics
Anon
consent · closed 7/19/2026
09 / 15 · FINDINGS
Results (Jul 19 close)
119 rows → 34 sessions → 23 journeys → 19 paired (4 missing post). Stats + 5 interviews.
Likert ΔM (N = 19 paired)
Bias awareness
+0.31ΔM
Reframe confidence
+0.47ΔM
Gender lens
+0.31ΔM
Personal agency
−0.11ΔM
Interviews: three-column compare most cited · 4/5 reported awareness shift
Journey split (n = 23)
Western astrology
13
BaZi
10
Top tags: ke-fu · virtuous wife · chastity · moral judgment n=5
Discussion: noticing bias ≠ finished reframing practice
10 / 15 · INTERVIEWS (n=5)
Interview themes (n=5)
Awareness
4/5: ke-fu, late marriage, “too strong” read as patriarchal scripts—not neutral traits
Key step
3/5: three-column compare made gender slippage visible in one glance
Tension
Some trusted the AI column most—agency score flat (−0.11 ΔM)
Design ask
Want a step tutorial; one user saw labels as “descriptive,” not bias
11 / 15
DISCUSSION
CONCLUSION
Keep divination.
See bias · break bias · change the story.
Reframe Destiny = changing the narrative, not deleting tradition. Bias must be seen, then broken, then rewritten. 命理可保留；偏见须被看见、被击碎、被改写。
N=19 paired · reframe confidence +0.47 ΔM
TAKEAWAYS
3 POINTS
01
Pedagogy
Teach reframing where youth already read fate—not in a separate media-literacy class
02
Design
Require Court of Destiny before AI reframe; add step-by-step onboarding (interview feedback)
03
Reframe Destiny
Changing the narrative—teach tagging and user lines before AI text; meet youth where they already read fate
12 / 15 · LIMITATIONS
Limitations
- Small sample: only 19 paired users—many sessions did not finish the post-survey (4 missing post).
- Narrow focus: this study targets gender bias in fate-talk, not every kind of bias in divination.
- Awareness ≠ rewrite: people can see bias in one visit but still not change the story on their own (agency ΔM −0.11).
13 / 15 · REFERENCES
References
- Butler, J. (1990). Gender trouble. Routledge.
- Haraway, D. (1988). Situated knowledges. Feminist Studies, 14(3), 575–599.
- Foucault, M. (1972). The archaeology of knowledge. Pantheon.
- Vyse, S. A. (2018). Believing in magic (Updated ed.). Oxford University Press.
- THIS IS GENDERED. (n.d.). Zodiac. https://thisisgendered.org
- Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots. Proc. FAccT, 610–623.
- Cotter, K. (2022). Algorithmic conspirituality. International Journal of Communication, 16, 1–20.
Full list in submitted paper · Classical BaZi: San Ming Tong Hui (ctext.org) · Live demo: reframe-destiny.pages.dev/game/
14 / 15 · IMPLICATIONS
NEXT
Reframe Destiny = change the narrative
Keep divination · see bias · break bias · rewrite with AI as scaffold, not oracle
15 / 15 · THANK YOU
Q & A
Questions
welcome
欢迎大家试玩 · Try the journey
Generation AI 2026 · Reframe Destiny
Present in browser fullscreen · press ← →