- This article has been published on Substack.
- Due to its high public importance, I am also posting it on this blog.
When considering ethics, feminism, women’s rights and dignity, sex education is extremely important. However, because it’s a sensitive topic, I believe how it’s taught differs by country and culture. This isn’t about superiority or inferiority—it’s about differences in history.
If AI hallucinates, we can verify it through cross-checking with other AI (equivalent performance models from different companies).
However, while the above note distributes a PDF in Japanese for free, users familiar with AI have an advantage regarding consistency gaps. Therefore, there are challenges in improving educational disparities through AI with current technology.
So rather than translating the PDF distributed in the above Japanese article into English, I’ll share the know-how.
- Search with Perplexity Pro
Have it investigate risks of sexual infections and prevention of sexual exploitation.
I added questions and had it research three times. - Since https://aistudio.google.com/ offers free access to the advanced model Gemini 2.5 Pro Preview 05-06, I instruct it to conduct a “meta-analysis” of step “1.”
- I request ChatGPT o3 (the advanced paid model) to create “Q&A for guardians providing sex education to 14-year-olds” using steps “1,2” as a set.
- Using Claude Sonnet 4’s “Projects” feature, I provide my essays and papers, along with steps “1,2,3,” and instruct it to be understandable and not cause anxiety for 14-year-olds reading it.
In other words, research alone provides too much information, so I verified accuracy and organized it to be comprehensible for both adults and 14-year-olds.
Children and guardians all have different personalities, circumstances, and situations. Rather than imposing answers, we make excessive information more accessible.
- Review all of the above and format it into a document with a table of contents using Google Docs.
- Limit the meaning of the word “vulnerability” and prevent the PDF from being used for discrimination (content verification and supplementation with human language).
This way, despite different countries and cultures, we can obtain information with a certain level of reliability. From there, people can learn by using reference services at libraries, for example.
Being young and ignorant is convenient for exploiters because it’s easy to deceive.
Therefore, the aim is to prevent the structure of exploitation itself by empowering all people with information.
One PDF cannot solve everything.
AI is not omnipotent.
However, things that were impossible in times without the internet, PDFs, or AI have become possible.
Women’s rights, children’s rights, the environment—all of these are also matters of ethics.
Not causing despair among people who understand and protect ethics—this might be the starting line for modern reform.