Q1 This is about AI and powerful people. Why does it concern me?
Start from an ordinary person's position: the same person may gain a technological advantage in one setting and bear another person's definition of efficiency in another.
PUBLIC V1.2 · READING MAP
Continuous Eleven-Question Reading Map
There is no need to work backward through platform publication dates. Start from your own question or follow all eleven from beginning to end.
This map accompanies As AI Grows More Powerful, How Should We Look After Our Own Interests?, Public Argument and Q&A Edition V1.2. The website's complete text is the authoritative entry point for continuous reading. Selections on Zhihu and Toutiao let readers on different platforms start from one question.
Read the complete work Download PDF — English file not yet available Download Word file — English file not yet available Launch reading guide
Platform articles may be subject to moderation, login, or access conditions. A temporarily unavailable platform page does not mean the corresponding question is absent; use the website's Read this question entry. Platform links are offered only after an actual publication receipt or public page has been obtained.
Entry: think through what looking after your own interests means
Q2 What is wrong with wanting to look after myself first?
Distinguish caring for oneself from deciding too early what counts as doing so, and explore self-interest in its fuller sense.
Q3 Where do AI's objectives and purposes come from?
Move from what to achieve to why to do it, linking society's definition of success and intelligence's actions within one causal chain.
Stage one: why the underlying logic must change
Q4 Everyone wants to win. Why can't we keep competing this way?
Competition can continue. But when winning means removing an opponent's standing to play the next round, the criterion for completion starts to lose its stopping point.
Q5 Why might weakening others fail to make us safer?
Reducing another party's opportunities is different from actually reducing its ability to produce high-consequence effects.
Q6 Isn't a clear account of long-term interests enough to yield a good answer?
Four kinds of difficulty break the direct inference from seeing the whole picture to necessarily choosing to benefit others, forcing the theory to confront the real normative gap.
Q7 What exactly comes first in others first?
First is not a ranking of interests; it means before a purpose is locked in. Can affected others change what counts as success?
Q8 Why call this the only path?
Directly takes responsibility for the strong proposition: why changing means alone cannot end completion criteria that continually exclude others.
Stage two: how to protect yourself and begin acting after change
Q9 If others do not change, will changing first disadvantage me?
Acknowledge that early changers may suffer losses, and separate a shared future from bearing all costs unilaterally.
Q10 Can I still say no after considering others?
You can refuse, exit, and limit danger. Caring for others does not cancel your own boundaries or replace professional safety judgment.
Q11 What can I change now?
Bring the direction back to one ordinary decision, four specific questions, and how to apply the same inquiry to tasks entrusted to AI.
The questions can continue after reading
The eleven questions do not require readers to reach one answer. Which step fails, which real situation has not been addressed, and what facts could change the judgment can all enter further discussion.
Reader Dialogue and Shared Inquiry View important updates Open the public library Continue to the five-volume work
This map is a reading index for public edition V1.2. It does not revise the body text, add theoretical propositions, or treat platform publication status as theoretical validation. Real reader questions are collected and answered separately.
Translation scope
This page translates the complete existing public HTML text. Linked works and files retain their own publication and version boundaries; unavailable English materials are marked explicitly.