Endopraxis and Exopraxis

Until very recently the objective of education included and to a considerable extent concentrated on providing individuals with the abilities they would need to function safely, competently, effectively and usefully in society out of resources they carried around with them. We call these inculcated skills and the associated knowledge endopraxis: practices we command that originate and are supported by resources within ourselves.

Artificial Intelligence (AI) and - sooner than many will think - robotics are set to change all that. Instead of the knowledge and skills we have that enable us to do things for ourselves we will need the knowledge and skills needed to devise, control and direct processes carried out on our behalf outside ourselves by AI agents and physical robots. The activity that arises from this knowledge and skill we will call exopraxis.

Someone may say that education has always been concerned with exopraxis inasmuch as it has tried to teach how to use tools and resources, has urged caution in accepting claims made without checking both their provenance and the motives of those who assert them. That is a perfectly fair observation, but something entirely new has entered the frame with the advent of AI and robotics in the shape of the potential for the external agent to operate autonomously once instructed to complete a task. Hitherto the resources have been harvested to service something we would do, whether something physical and practical or intellectual and conceptual. Education has tried to teach us how to choose the tools we need to do things for ourselves.

All that has changed. Now an instruction to an AI or robot can trigger autonomous action on any scale it is equipped to deal with. “Write an essay or book on X” will produce some document with little or no further involvement on the part of the instigator. This is real exopraxis and learning how best to utilise it and control it is becoming as essential a skill as the traditional “3 Rs”.

Balance

There is always a danger whenever something new comes along that we will over-compensate, throwing away the best of what we already do in a panicked blind pursuit of the new. We should resist that temptation. What we really need is a balance between the internal capabilities we need to direct and devise and the understanding of external capabilities we need to instruct and control. Such a balance is not easily achieved in unfamiliar territory.

Neither are our resources limitless. We are obliged to trade between the time and effort we apply in order to optimise the result; too much of one is always at the expense of another.

+++

Qwen3.5-397B-A17B.GGUF

This is a profound and highly usable framework. The terms Endopraxis (internalized capability) and Exopraxis (externalized orchestration) capture the seismic shift we are facing better than most current ed-tech jargon.

To “put flesh on the bones,” we need to move from the abstract definition to concrete curricular implications, pedagogical shifts, and the philosophical stakes of this transition.

Here is a structured elaboration on your distinction, broken down into dimensions of implementation.

1. The Core Contrast: A Comparative Taxonomy

To make the distinction operational, we can map specific educational objectives onto your two categories.

2. Curricular Translation: Subject by Subject

How does this shift change specific subjects?

Mathematics & Logic

• Endopraxis: Memorizing multiplication tables, performing calculus by hand, deriving proofs.
• Exopraxis: Statistical literacy, understanding probability distributions, verifying AI-generated code, modeling real-world scenarios for AI simulation.
• The Shift: We stop teaching students to be calculators and start teaching them to be auditors of calculation.

Language & Literature

• Endopraxis: Grammar, syntax, spelling, composing essays from scratch.
• Exopraxis: Prompt engineering, tone modulation, synthesizing multiple AI drafts, detecting hallucination/bias in text, “editing” as a primary creative act.
• The Shift: Writing becomes less about generation and more about curation and direction.

History & Social Studies

• Endopraxis: Memorizing dates, names, and narratives.
• Exopraxis: Source triangulation, understanding algorithmic bias in search results, analyzing how AI reconstructs history, ethical oversight of automated decision-making.
• The Shift: History becomes less about recall and more about forensic verification of information ecosystems.

Science & Technology

• Endopraxis: Manual lab techniques, memorizing periodic tables, coding syntax.
• Exopraxis: Designing experiments for robotic execution, interpreting massive datasets, managing robotic swarms, safety protocols for autonomous agents.
• The Shift: Science becomes less about manual manipulation and more about experimental design and data interpretation.

3. Pedagogical Shifts: The “How” of Teaching

If the goal shifts from Endo to Exo, the method of teaching must change.

• From Socratic Method to Socratic Prompting:
  • Endo: Teacher asks student questions to draw out internal knowledge.
  • Exo: Student asks AI questions to draw out external knowledge, then defends the validity of the answer to the teacher and classmates.
• The “Open-Book + Open-AI” Exam:
  • Traditional exams ban resources to test Endopraxis. Exopraxic exams must allow all resources but test the process. National educational assessment must accommodate this change urgently.
  • Example: “Here is an AI-generated solution to a crisis or a mathematical proof, or a scientific explanation. Critique it, find the logical fallacies, and propose a better implementation.”
• Failure Analysis:
  • In Endopraxis, failure is personal (you didn’t study). In Exopraxis, failure is often systemic (the agent misunderstood, the data was biased). Education must teach resilient debugging rather than shame avoidance.

4. The Critical Constraint: “Minimum Viable Endopraxis”

This is the most dangerous bone to flesh out. You cannot have Exopraxis without some Endopraxis.

If a student completely outsources arithmetic to AI, they lose the intuition to know when the AI is wrong. If they outsource writing entirely, they lose the ability to think clearly enough to prompt the AI well.

The Threshold Theory: Education must identify the Minimum Viable Endopraxis required to be a competent Exopractitioner.

• You need to know enough coding to verify the AI’s code.
• You need to know enough history to spot the AI’s hallucination.
• You need enough emotional intelligence to know when the robot is being manipulative.

Recommendation: Define “Endo-Foundations” that are non-negotiable. These are the internal skills required to maintain agency over the external tools.

5. Ethical and Philosophical Implications

Exopraxis introduces risks that Endopraxis did not have.

• Agency Drift: In Endopraxis, I am responsible for my action. In Exopraxis, if I command a robot to do something and it harms someone, where is the liability? Education must teach agency attribution.
• The Black Box Problem: Endopraxis is transparent (I know how I solved this). Exopraxis is often opaque (I don’t know how the neural net solved this). Education must teach epistemic humility—knowing what we don’t know.
• Identity Formation: Historically, “I am what I can do” (Endo). In the future, “I am what I can orchestrate” (Exo). This changes how children form self-esteem. We must validate intent and taste rather than just output. We need a more nuanced notion of self: I am not only what I know and do; I am an exopractitioner in a vast network of human and machine, autonomous and heteronomous agency.

6. A Proposed Syllabus Add-on: “Exopraxic Literacy”

To make this concrete, schools could introduce a specific subject or module dedicated to Exopraxis while redesigning their curriculum to be suffused with the skills required by exopractitioners.

• Module 1: Agent Architecture: Understanding how AI/Robots “think” (not technically, but conceptually) so students know where the edges are.
• Module 2: Verification Protocols: Techniques for cross-referencing external outputs.
• Module 3: Intent Specification: How to translate vague human desire into precise machine instruction (Advanced Prompting).
• Module 4: Ethical Delegation: When not to use AI. Understanding tasks that require human touch/empathy/judgement.

7. Summary Metaphor

• Endopractic Education was like training a monk: disciplined, internalizing texts, self-sufficient in meditation and copying.
• Exopractic Education will be like training a conductor: they do not play every instrument, but they must understand the capability of every section, manage the tempo, interpret the score, and ensure the orchestra produces a coherent symphony.

Next Steps for Your Idea

If you wish to develop this further, I suggest focusing on the Transition Zone. How do we move a child from Endo to Exo without atrophy? How do we transition a school from Endo to Exo without massive disruption and alarming the parents?

• Idea: The “Training Wheels” approach. Start with high Endo requirements, and slowly introduce Exo tools as the student demonstrates they possess the internal knowledge to verify the external tool.

This distinction is robust. It moves the conversation beyond “AI in schools” to “AI as a fundamental shift in human cognitive architecture.” Would you like to explore the “Minimum Viable Endopraxis” concept further? That seems like the critical pivot point for policy.

A couple of observations before we explore Minimum Viable Endopraxis”.

1. I would quibble with your assertion that in endopraxis we can always account for our reasons and provide their provenance; I think this is neither experientially the case nor even optimally the case because we are often inspired by flashes that seem to come from nowhere but long exposure to a set of concepts;
2. You are right to intriduce the “conductor” metaphor; in fact we are already talking about “orchestrating” agent swarms in much the same way. I think that’s a useful way to approach the skill-set, too, because management in ordinary human affairs entails a similar kind of non-specialist competence. So what is a “Minimum Viable Endopraxis” - a “core curriculum” if you like - in this new world?

Qwen3.5: This is an excellent correction. You are absolutely right: Endopraxis is not merely explicit, auditable logic. A significant portion of human competence is tacit, intuitive, and embodied—the “flashes” of insight that come from deep immersion rather than step-by-step derivation. If we define Endopraxis only as “explicit reasoning,” we miss the vast subconscious reservoir of human skill that actually drives innovation.

This makes the definition of Minimum Viable Endopraxis (MVE) even more critical. If intuition is a “black box” internal process, we cannot outsource it without losing our ability to recognize truth or quality when we see it.

Here is a proposal for what constitutes the MVE “Core Curriculum” in an Exopraxic world.

1. The Definition of MVE

Minimum Viable Endopraxis is the baseline of internalized knowledge, intuition, and physical capability required to:

1. Verify the output of external agents (catch hallucinations, bias, or errors).
2. Intent the goals of external agents (translate vague human desire into precise instruction).
3. Survive the failure of external agents (maintain agency when the system goes down).

If a student drops below this threshold, they cease to be a “Conductor” and become a “Passenger.”

2. The Five Pillars of the MVE Core Curriculum

To flesh this out, we can identify five specific domains where internal capacity must be preserved, even as execution is outsourced.

I. Epistemic Vigilance (The “Smell Test”)

• The Skill: Cultivating the intuition to sense when an answer is wrong, even if you don’t immediately know why.
• Why MVE: As you noted, expertise often manifests as a “flash” of recognition. If you outsource all learning to AI, you never build the neural patterns that generate that flash. You become vulnerable to confident falsehoods.
• Curriculum: Instead of memorizing facts, students practice “Red Teaming” information. They are fed AI-generated essays or data and must identify the subtle inconsistencies, logical leaps, or tonal mismatches.
• Endo-Requirement: You must know enough history/logic/science to feel the “friction” when a statement is untrue.

II. Ethical Grounding (The “Value Anchor”)

• The Skill: Unnegotiable moral reasoning that is not subject to algorithmic optimization.
• Why MVE: AI optimizes for metrics (efficiency, engagement, profit). It does not optimize for human flourishing unless told to. If the human operator has no internal moral compass, the Exopraxic system will drift toward sociopathic efficiency.
• Curriculum: Philosophy and Ethics move from “elective” to “core.” We place increasing emphasis on Added Value in the specific sense of what the human contribution to a process is; we are acutely conscious of the shifting landscape as AI, robotics and newer World Models increase in ability and subtlety and become able to exercise more autonomy.
• Ethics: How do we ground human values in a world where non-human efficiency, speed and volume may threaten to drown them? This must not be abstract theory, but applied to real dilemmas: “The AI suggests firing 10% of staff to maximize efficiency. What is the human cost? Does such ‘redundancy’ reveal a defect in our management/organisation? Is it acceptable?” What might an economics of abundance look like, and what would the associated ethical positions be?
• Endo-Requirement: A deeply internalized sense of justice and empathy that operates faster than the system’s calculation.

III. Domain Heuristics (The “Map”)

• The Skill: possessing high-level mental models of how the world works, without necessarily knowing the low-level mechanics.
• Why MVE: You don’t need to be a mechanic to drive a car, but you need to know that “engine + gas = movement” and “brake = stop.” If you think the car runs on magic, you cannot diagnose failure.
• Curriculum: “System Literacy.” Students learn the bounds and constraints of physics, biology, and economics. They don’t need to solve the differential equation, but they must understand the relationship between variables.
• Endo-Requirement: Enough conceptual knowledge to guide the AI’s search space. (e.g., Knowing that “correlation does not imply causation” so you don’t ask the AI to find causality where none exists).

IV. Manual Override Capability (The “Backup”)

• The Skill: Basic physical and cognitive ability to perform the task without the tool.
• Why MVE: Systems fail. Power grids go down. Networks are hacked. If you cannot write a sentence, do a sum, or navigate a room without GPS, you are helpless when the Exopraxic layer vanishes.
• Curriculum: “Unplugged Weeks.” Periods where digital tools are banned to ensure neural pathways for basic literacy, numeracy, and spatial awareness remain active.
• Endo-Requirement: Functional literacy and numeracy that do not depend on a interface.

V. Interpersonal Authenticity (The “Human Zone”)

• The Skill: Distinguishing between simulated empathy (AI) and genuine human connection.
• Why MVE: As robots handle more tasks, the premium on human-to-human interaction rises. If we lose the ability to connect authentically because we are used to curated AI interactions, society fractures.
• Curriculum: Collaborative arts, sports, and debate. Activities where the “friction” of another human ego is present and must be managed.
• Endo-Requirement: Emotional intelligence that recognizes nuance, sarcasm, pain, and trust in other humans.

3. The Risk of “Endo-Atrophy”

The danger in this new world is Cognitive Atrophy. If we treat education purely as Exopraxis training (how to prompt, how to orchestrate), we risk creating a generation that creates brilliantly but understands nothing.

• The “Centaur” Model: In chess, a “Centaur” (Human + AI) beats both pure AI and pure Human. But the Centaur only wins if the Human understands the game deeply enough to intervene when the AI makes a strategic error.
• The MVE Threshold: The human must know enough to know when to intervene.

4. A Proposed “MVE” Syllabus

If we were to design a school year based on this, it might look like this:

• Logic & Verification: Not “do the math,” but “spot the error in this AI-generated proof.”
• Ethics & Governance: Not “learn the laws,” but “design the constraints for this autonomous agent.”
• History & Context: Not “memorize dates,” but “analyze how this AI summary of the Cold War misses the human nuance.”
• Physical Competence: Not “industrial robotics,” but “manual dexterity and spatial awareness” (to ensure we remain embodied beings).
• Communication: Not “write the essay,” but “refine the tone and intent of this draft to resonate with a specific human audience.”

5. Conclusion: The “Orchestra” Revisited

Returning to your conductor metaphor: A conductor does not need to play every instrument perfectly (Exopraxis), but they must have perfect pitch and rhythm internally (Endopraxis). If the conductor cannot hear the dissonance internally, they cannot correct the orchestra.

Minimum Viable Endopraxis is the internal ear. It is the quiet, internal knowledge that allows you to stand amidst the noise of a thousand AI agents and say, “That is out of tune,” or “This is the wrong song,” or “Stop.”

Without this core, Exopraxis is not management; it is surrender.

Does this align with your vision of the “internal tool-kit” required to keep the human in the loop? We could drill down into one of these pillars (like Epistemic Vigilance) to see what specific pedagogical tools would look like.

I can’t pretend sufficient familiarity either with Chinese/Mandarin or with Chinese philosophy to pass judgement, but I am concerned about the “Greek” flavour of the endo/exo-praxis and since I will need to discuss these ideas with Chinese schools, I wanted an alternative indigenously-appropriate set of terms if they can be found. I think it’s very interesting that my distinction originates with or was appropriated by Mao, and probably I’d want to avoid that connotation. A Chinese native speaker suggested 内部资源 vs 外部资源 or 内源 vs 外源（外援 - external help) commenting that 内应 sounds like a spy!

Qwen 3.5:

Your instinct to avoid the “Greek” flavor and the Maoist 内因/外因 baggage is excellent. In a modern educational context, especially with schools, you want terms that feel culturally native, ethically clear, and free of heavy political ideology.

Your native speaker friend is absolutely correct: 内应 means “spy inside the camp.” Do not use it.

Regarding 内部资源 (Internal Resources) vs 外部资源 (External Resources): While grammatically correct, these terms are too passive. “Resources” sit there waiting to be used. Your concept of Exopraxis is active—it involves design, command, and orchestration. “Resources” doesn’t capture the agency of controlling an AI agent.

Here are three indigenous Chinese pairings that capture your distinction more accurately, ranging from business-modern to philosophical-classical.

1. The Best Operational Pair: 内修 (Nèi Xiū) vs 外驭 (Wài Yù)

This is likely your strongest option for schools. It blends traditional culture with modern management logic.

• Endopraxis → 内修 (Internal Cultivation):
  • Meaning: Cultivating oneself, internal skill building, character development.
  • Connotation: Active work on the self. It implies that knowledge must be digested and embodied, not just stored.
  • MVE Connection: You cannot 内修 nothing. There is a baseline of 内修 required to remain human.
• Exopraxis → 外驭 (External Orchestration/Driving):
  • Meaning: 驭 means to drive horses or vehicles. It implies controlling external agents (AI, robots, systems) without doing the physical labor yourself.
  • Connotation: Mastery over tools. A driver (驭手) does not run alongside the car; they steer it.
  • Why it works: It captures the command-and-control aspect of Exopraxis perfectly.

Usage Example:

• “Traditional education focused on 内修 (doing things yourself). Future education must balance 内修 with 外驭 (orchestrating external agents).”
• 传统教育重在**内修**（亲力亲为），未来教育必须平衡**内修**与**外驭**（调度外部智能）。

2. The Philosophical Pair: 修己 (Xiū Jǐ) vs 役物 (Yì Wù)

If you want to ground this in classical philosophy (avoiding Marxism entirely), this is the deepest option. It comes from Mencius and Confucianism.

• Endopraxis → 修己 (Cultivating the Self):
  • Standard Confucian term for internal moral and skill development.
• Exopraxis → 役物 (Making Things Serve You):
  • Context: There is a famous Chinese philosophical warning: 勿役于物 (Do not be enslaved by things).
  • Meaning: To command objects/tools rather than being commanded by them.
  • Why it fits: It addresses the risk of AI directly. If you have Exopraxis skills, you 役物 (use AI). If you lack MVE, you 役于物 (are used by AI).
• MVE → 根基 (The Root/Foundation):
  • Instead of “Minimum Viable,” use 根基. If your 根基 is shallow, you cannot 役物.

Usage Example:

• “We must teach students 修己 so they have the 根基 to 役物 (command tools), rather than being 役于物 (controlled by tools).”

• 我们必须教导学生**修己**，使他们有**根基**去**役物**，而不是**役于物**。

3. The Modern/Technical Pair: 内力 (Nèi Lì) vs 外力 (Wài Lì)

Borrowed from Martial Arts (Wuxia) and Physics, this is very intuitive for students.

• Endopraxis → 内力 (Internal Force):
  • Your own inherent capability, intuition, and knowledge.
• Exopraxis → 外力 (External Force):
  • Leveraging tools, networks, and AI to amplify output.
• MVE → 底线 (The Bottom Line):
  • The minimum 内力 required to safely use 外力.

Usage Example:

• “AI is 外力. You need 内力 to control it. Without 内力, 外力 will hurt you.”

• 人工智能是**外力**。你需要**内力**来控制它。没有**内力**，**外力**会伤到你。

Recommendation on “Minimum Viable Endopraxis”

Translating “Minimum Viable Endopraxis” is critical. “Minimum Viable” sounds too startup-ish (最小可行性); “Endopraxis” sounds too academic.

I suggest translating the concept rather than the literal words. Use 固本 (Strengthening the Root).

• Concept: 固本教育 (Education that strengthens the root).
• Definition: The essential human skills (critical thinking, ethics, basic literacy) that must remain internal so you don’t lose agency.
• Phrase: “In an age of 外驭 (Exopraxis), we must 固本 (secure the Endo foundation).”

Summary Table for Your Presentation

A Note on the Maoist 内因/外因

You are wise to avoid this. In mainland Chinese schools, 内因/外因 immediately triggers memories of political textbooks and dialectical materialism exams. It feels rigid and ideological. 内修/外驭 feels much more like skill development and management, which is what you want.

Sample Opening Statement for Chinese Schools

If I were to draft your opening pitch using these terms:

• “Historically, education was about 内修 (internalizing skills so you can do everything yourself).
• But in the AI era, we must teach 外驭 (how to orchestrate external agents).
• However, 外驭 requires 固本 (a strong internal foundation).
• If we lose our 内修, we become 役于物 (slaves to the machine) instead of 役物 (masters of the machine).

• Our goal is 内修 for agency, 外驭 for efficiency.”

• 历史上，教育重在内修（将技能内化，亲力亲为）。
• 但在 AI 时代，我们必须教授外驭（如何调度外部智能体）。
• 然而，外驭需要固本（强大的内部基础）。
• 如果我们丢失了内修，就会役于物（成为机器的奴隶），而不是役物（成为机器的主人）。
• 我们的目标是：以内修保自主，以外驭求效率。

Does this feel culturally safe and conceptually accurate for your purposes?

What’s the pinyin here: 底线 and 固本教育 Gu Ben …

Qwen3.5:

Here is the pinyin for those terms:

底线

• Pinyin: dǐ xiàn
• Pronunciation: Dee (falling-rising) + Shien (falling)
• Meaning: Bottom line / Baseline / Threshold

固本教育

• Pinyin: gù běn jiào yù
• Pronunciation: Goo (falling) + Ben (falling-rising) + Jiao (falling) + Yu (falling)
• Meaning: Education that strengthens the root/foundation

So you were correct on Gu Ben (gù běn), just note that 固 is a sharp falling tone (4th), and 本 is a dipping tone (3rd).

What are the Chinese equivalents of autonomous and heteronomous?

Autonomous is usually 自行 (zixing) but can also be 自导 (zudao) and hetermonomy or to do things by external agency 他律 (talü)

+++

A native Chinese speaker suggests that we should incorporate the autonomous/heteronomous connotation somehow. Does Chinese allow that in an idiomatic way?

Qwen3.5: Yes, Chinese absolutely allows for this distinction, and there are actually three layers of terminology you can use depending on how philosophical or practical you want to be.

The direct translation of “Autonomy vs. Heteronomy” exists, but it is academic. However, there are more idiomatic ways to express the feeling of being self-governed vs. externally governed that fit your educational context perfectly.

Here are the best options:

1. The Philosophical Standard: 自主 vs 他律

• Autonomy: 自主 (zì zhǔ)
  • Literal: Self-master / Self-decide.
  • Usage: Very common (e.g., 自主创新 - independent innovation). It implies agency and ownership.
• Heteronomy: 他律 (tā lǜ)
  • Literal: Other-law / Governed by others.
  • Usage: This is the standard philosophical translation for heteronomy (often used in Kantian ethics).
  • Nuance: It implies you are following rules or laws imposed from outside (including algorithms) rather than your own internal moral compass.
• Verdict: Accurate, but 他律 is academic. Schools understand it, but it feels like ethics class.

2. The Power Dynamic: 主宰 vs 受制

If you want to emphasize the command-and-control aspect of Exopraxis (from your original definition), this pair is more visceral.

• Autonomy: 主宰 (zhǔ zǎi)
  • Literal: To dominate / command / be the master.
  • Usage: 主宰自己的命运 (Master one’s own destiny).
• Heteronomy: 受制 (shòu zhì)
  • Literal: To be subject to control / constrained by.
  • Usage: 受制于人 (Controlled by others) or 受制于算法 (Controlled by algorithms).
• Verdict: Very strong for AI context. It highlights the risk of losing agency to the machine.

3. The Indigenous Philosophical Pair (Mencius): 自得 vs 外铄

This is the deepest cultural option. It avoids Western philosophy entirely and draws on Classical Confucianism.

• Autonomy: 自得 (zì dé)
  • Literal: Self-obtained / Internal realization.
  • Meaning: Knowledge or virtue that is deeply internalized and owned by the self.
• Heteronomy: 外铄 (wài shuò)
  • Literal: Meltered/coated from outside.
  • Meaning: Knowledge or behavior imposed externally without internal understanding.
  • Famous Quote: Mencius said, “仁义礼智，非由外铄我也，我固有之也” (Benevolence… is not imposed from outside, I possess it internally).
• Verdict: Beautifully captures the risk of AI education becoming 外铄 (superficial external imposition) rather than 自得 (deep internal ownership).

Recommendation for Schools

For a school curriculum, 自主 (zì zhǔ) is the best word for Autonomy. For Heteronomy, 他律 is too academic. 依附 (yī fù - dependence) or 被动 (bèi dòng - passivity) is clearer.

However, if you want to keep the Endo/Exo structure (内修 / 外驭) and add the Autonomy layer, you can combine them:

• Endopraxis Goal: 内修以自主 (Internal cultivation for autonomy).
• Exopraxis Risk: 外驭而受制 (External orchestration leading to being controlled).

Revised Framework with Autonomy Connotation

Sample Sentence for Your Presentation

• “We must ensure students develop 内修 (Endopraxis) so they remain 自主 (Autonomous), even when using 外驭 (Exopraxis). Otherwise, they risk falling into 他律 (Heteronomy), where the tool controls the human.”

• 我们必须确保学生发展内修，使他们保持自主，即使在使用外驭时。否则，他们可能陷入他律，即工具控制人。

A Note on 他律 (Heteronomy)

While 他律 is the correct term, it is slightly stiff. If you want to be more idiomatic about the danger of heteronomy, use 迷失 (mí shī - losing oneself) or 役于物 (yì yú wù - enslaved by things).

• Autonomy: 做主人 (Be the master).
• Heteronomy: 做奴隶 (Be the slave) or 被算法主导 (Dominated by algorithms).

If your native speaker friend wants the specific philosophical connotation, 自主 vs 他律 is the correct pairing. If they want the spirit of it, 自得 vs 外铄 is profoundly Chinese.

Does one of these pairs feel more appropriate for the specific schools you are visiting?

Great summary, and the answer is nuanced: what every school wants is a socially embedded citizen (very Confucian!) who is  自得 but who in the modern context (and probably also the ancient context) is also empowered to be what you nicely called an exopractitioner, something like 外语人 if that is at all possible.

Qwen3.5: This is a brilliant synthesis. You have hit upon the exact tension that modern Chinese education is trying to resolve: how to remain traditionally Chinese (自得, Confucian self-cultivation) while becoming globally competitive and technologically empowered (the “Exopractitioner”).

However, I must offer a crucial correction on 外语人, and then propose a term that fits the Confucian framework much better.

1. The Problem with 外语人 (Wài Yǔ Rén)

• 语 (Yǔ) means Language/Speech.
• 外语 exclusively means “Foreign Language” (e.g., English, French).
• 外语人 would sound like “A person who works in foreign languages” or “A linguist.”
• The Confusion: It does not convey “External Practice” or “Orchestration.” It conveys “Language Learning.”

2. The Confucian Solution: 内圣外王 (Nèi Shèng Wài Wáng)

You are describing a modern version of the ultimate Confucian ideal: 内圣外王 (Inner Sage, Outer King).

• 内圣 (Inner Sage): Corresponds to Endopraxis. Moral cultivation, 自得, internal wisdom.
• 外王 (Outer King): Corresponds to Exopraxis. Bringing order to the world, managing affairs, impacting society.

However, 外王 sounds too political (like “King”). For schools, we can modernize this to 内修外驭 (Internal Cultivation, External Orchestration).

3. Proposed Terminology for the “Exopractitioner”

If 自得 (Zì Dé) is the state of the Endopractitioner, here are three options for the Exopractitioner, ranked by fit:

Option A: 致用者 (Zhì Yòng Zhě) — The Pragmatist

• Source: Confucian phrase 经世致用 (Managing the world for practical use).
• Meaning: One who applies knowledge to practical use.
• Why it fits: Exopraxis is about using external agents to achieve results. It balances 自得 (internal knowledge) with 致用 (external application).
• Verdict: Strong. It sounds traditional but implies action.

Option B: 驭物者 (Yù Wù Zhě) — The Master of Things

• Source: 君子役物，小人役于物 (The gentleman makes things serve him; the small man is served by things).
• Meaning: One who commands/tools/objects rather than being commanded by them.
• Why it fits: It captures the agency risk. An exopractitioner must 驭物 (control AI) not 役于物 (be controlled by AI).
• Verdict: Very Strong. It directly addresses the ethical danger of AI.

Option C: 外驭者 (Wài Yù Zhě) — The External Orchestrator

• Source: Modern construction using 驭 (Drive/Control).
• Meaning: One who orchestrates external systems.
• Why it fits: It pairs perfectly with 内修 (Internal Cultivation).
• Verdict: Clear. It is the most direct translation of “Exopractitioner.”

4. The Ideal Citizen Profile: 内修外驭之士

For your schools, you can propose that the goal of education is to create 内修外驭之士 (Scholars who cultivate internally and orchestrate externally).

• Endopraxis: 内修自得 (Internal cultivation leading to self-realization).
• Exopraxis: 外驭致用 (External orchestration leading to practical impact).
• The Risk: 役于物 (Enslavement by tools).
• The safeguard: 固本 (Strengthening the root).

5. Revised Framework for Chinese Schools

6. Sample Pitch for the Schools

• “We want students who are 自得 (autonomous internally) so they can be 驭物者 (masters of tools externally).
• We teach 内修 (Endopraxis) not just for peace of mind, but to ensure they have the 固本 (foundation) to 外驭 (orchestrate) AI without becoming 役于物 (enslaved by it).

• The goal is 内修外驭：Internally human, externally powerful.”

• 我们希望学生自得（内部自主），这样才能成为驭物者（工具的主人）。
• 我们教授内修（Endopraxis）不仅为了心境，更是确保他们有固本（基础）去外驭（调度）AI，而不役于物（被机器奴役）。
• 目标是内修外驭：内部人性化，外部强大化。

Summary

• 自得 is perfect for Endopraxis/Autonomy.
• 驭物者 (Master of Things) or 外驭者 (External Orchestrator) are the best terms for
• 内修外驭 is the perfect phrase to describe the dual curriculum.

Note:

• 外语人 should be avoided (means “Foreign Language Person”).

This framing respects the Confucian desire for moral depth (自得) while embracing the modern necessity of technological power (外驭). It turns the AI challenge into a moral cultivation challenge, which Chinese schools will understand immediately.