The Ti-Yong Philosophical Framework

My initial concept of Ti-Yong (體用) actually came from Zi Wei Dou Shu (紫微斗數). Zi Wei Dou Shu is a fortune-telling technique from Chinese culture: from a person’s birth information — the exact year, month, day, and hour — it calculates what kind of fate that person might encounter over the course of their life. It’s a genuinely complex calculation, and that’s not what this piece is about — but let me first explain the part of it where Ti-Yong actually shows up. A natal chart has twelve palaces: the Life Palace, Siblings Palace, Spouse Palace, Children Palace, Wealth Palace, Health Palace, Travel Palace, Friends/Servants Palace, Career Palace, Property Palace, Fortune Palace, and Parents Palace. Each palace carries its own symbolic meaning, not just its literal name. Among the twelve, the Life Palace is the most important — every other palace only makes sense in reference to it. As the saying goes, there has to be an “I” before there’s a “you” for the word to even refer to. The Life Palace is Ti; the other eleven palaces, relative to it, are all Yong.

The core: Ti and Yong are not two equal poles

Ti is the foundation, and Yong is the action, form, institution, organization, or language that naturally unfolds from Ti. These two are not symmetric. Ti comes first, Yong second. Yong is healthy when it genuinely grows out of Ti; the moment Yong detaches from Ti and starts running on its own, that’s where pathology begins — much like a cancer cell, which will eventually kill the very body it came from.

Knowing when to stop is the anchor of Ti-Yong — and the anchor of the AI age

Beyond grounding things in Ti, you also have to act within your actual means. In the present day, pursuing unlimited growth has become something close to a universal belief, and recognizing when we’ve had enough is genuinely hard. A hungry stomach makes itself felt; a full one does too. But past a certain point of not having money — real scarcity, real want — there’s a sharp, unmistakable feeling; having enough money almost never produces one, because money is, in practice, never quite enough. That’s exactly why “knowing when to stop” exists as a concept: to sense that enough — and once you sense it, to return from Yong back to Ti, instead of letting Yong stand in for Ti, or worse, let Yong start harming Ti.

This idea of “knowing when to stop” also comes from the Chinese classical canon — from the Great Learning (大學) chapter of the Book of Rites (禮記): “The Way of the Great Learning lies in manifesting bright virtue, in renewing the people, and in resting in the highest good. Only after knowing where to rest can one be settled; only after being settled can one be tranquil; only after being tranquil can one be at ease; only after being at ease can one deliberate; only after deliberation can one attain. Things have their root and their branches; affairs have their end and their beginning. To know what comes first and what comes after is to be close to the Way…” Now that the AI age has arrived and the sheer power of human-made tools has reached an intensity never seen before, the real question isn’t how to use AI — it’s how to sense that stopping point, to sense enough, to sense: this part is mine, this is Ti, this is not something Yong — meaning AI — should be handling. To sense: this is a core human value, not something Yong can compute.

Why write this as a Claude Skill, and not just an essay

An essay ends the moment you finish reading it. Writing it as a Skill means this way of thinking can be automatically called back up every time I’m in conversation with an AI and need to reason through a decision or a real situation — actually applied to whatever’s in front of me at that moment, instead of becoming a concept you read once and only retain a vague impression of afterward.

Mechanically, it’s a simple Skill: a SKILL.md that defines the trigger contexts and the core diagnostic questions, and four files under references/ that expand the core framework and three extension frameworks — which one gets read depends entirely on what’s actually being analyzed at the time. It auto-triggers when a conversation touches on questions like “is this enough,” “should this happen,” or “why does this feel hollow,” and can also be invoked directly.

Try it

Here’s the link to the Ti-Yong framework repository — you can follow the instructions there to install it as a Claude Skill.