13.6 · Áp dụng chuyên môn ngành vào AI Fluency — Chuẩn bị sinh viên để không thể thay thế

Mở đầu: Điều bất ngờ về AI Fluency thực sự

Nhưng đây là điều bất ngờ: Mọi thứ bạn sắp học trong bài này có thể xảy ra mà không cần chạm vào AI hệ thống nào.

Điều bạn xây dựng ở đây là human capacity — giúp sinh viên (và chính mình) articulate cái mình biết, biết thế nào, và tại sao matter.

Công việc foundational này là điều khiến chúng ta irreplaceable. Nó giữ con người trong loop, theo nghĩa đen và nghĩa bóng.

Nghĩ về nó như thế này: Bạn không thể cộng tác với AI trong ngành em effectively cho đến khi em có thể articulate:

• Quality trông thế nào trong ngành em
• Experts giao tiếp thế nào trong ngành em
• Vấn đề nào matter trong ngành em
• Standards nào apply trong ngành em

Trước khi em teach AI anything, em phải teach it yourself what your field values.

Và khi em do that, em tạo ra something quan trọng hơn cả AI Fluency: sinh viên có khả năng direct AI — hoặc bất kỳ tool nào — một cách thoughtfully trong ngành của họ.

Bài này về cái đó.

Starting Point: Discernment — Biết "Good" trông thế nào

Khác với 3 bài trước (dạy Delegation, Description, etc. theo thứ tự framework gốc), bài này bắt đầu với Discernment. Tại sao?

Discernment, khả năng evaluate quality, luôn matter trong giáo dục. Bây giờ, với AI generating endless content, nó trở thành essential.

Chỉ con người có thể judge cái gì truly phục vụ goals và values của chúng ta. Trong education, chúng ta phải define goals và values này explicitly trong disciplinary context.

Chiến thuật 1: Build quality criteria together

Với colleagues và sinh viên, articulate excellence trong ngành.

Move beyond vague terms:

• "Well-argued"
• "Creative"
• "Original"

Build detailed rubrics capturing deep quality markers. Ví dụ:

Triết học paper — "what makes a philosophy paper compelling?"

• Precise định nghĩa terms, đặc biệt những từ everyday use conflates
• Strongest possible version of opposing views (steelmanning)
• Counterargument từ internal logic của opposing view, không chỉ external criticism
• Examples connecting abstract principles tới concrete cases
• Clear "so what" — implications nếu argument correct

Engineering solution — "what makes a solution elegant?"

• Minimal parts cho required function
• Graceful failure modes
• Extensibility không major redesign
• Readability to fellow engineers
• Performance ratio to complexity

Document những criteria này theo cách sinh viên có thể internalize và apply.

Chiến thuật 2: Systematically analyze exemplars

Collect outstanding work từ ngành — published papers, professional portfolios, breakthrough solutions.

Đừng chỉ show chúng. Systematically analyze với sinh viên:

• Mark up texts
• Diagram structures
• Decode decisions

Help students see what YOU see khi em recognize excellence.

Create annotation guides make expert thinking visible.

Ví dụ annotation cho 1 excellent opening paragraph essay:

"The question of whether AI can be conscious 
 [1] has haunted philosophers for decades, 
[2] yet the question itself may be malformed.
[3] In this essay I argue that..."

[1] 🔑 Stakes set: why reader should care
[2] 🔑 Tension signalled: not "here are facts" 
     but "we think we know, but..."
[3] 🔑 Claim explicit: thesis moves from 
     speculation to argument

Chiến thuật 3: Diagnose failures together

Equally important: Study gì doesn't work, gì excellence isn't.

Collect flawed examples:

• Failed arguments
• Buggy code
• Ineffective designs
• Weak proofs
• Confused writing

Examine forensically:

• Where EXACTLY does this proof break down?
• Why doesn't this solution scale?
• What makes this composition fall flat?
• How does this fail to meet industry standards?

Understanding failure modes builds evaluative skills sinh viên cần — whether working alone hoặc với AI.

Discernment summary

┌──────────────────────────────────────────────────────────┐
│                                                          │
│   DISCERNMENT — Define Quality trong Your Discipline     │
│                                                          │
│   1. Build detailed rubrics (beyond vague terms)         │
│   2. Systematically analyze exemplars                    │
│   3. Diagnose failures forensically                      │
│                                                          │
│   Result: Sinh viên có robust vocabulary để describe     │
│   "good" cho AI assistants và cho nhau.                  │
│                                                          │
└──────────────────────────────────────────────────────────┘

Moving to Description — Articulating How Experts Communicate

Mỗi ngành có ways communicate embody values và methods. Making tacit knowledge explicit qua practices below naturally builds strong description skills.

Chiến thuật 1: Map your discipline's products

Document key products trong ngành with precision.

Đừng chỉ say "lab reports". Specify exact sections, conventions, why they matter.

Create templates revealing underlying logic:

• Why methods section come before results?
• What work does passive voice do in scientific writing?
• Why does the Abstract have exactly the structure it does?
• What makes a good literature review vs. annotated bibliography?

Have students reverse-engineer professional outputs to understand deep structure.

Ví dụ — Research paper sections decomposed:

ABSTRACT (150-250 từ):
├── Background (2-3 câu): context + gap
├── Purpose (1 câu): question/hypothesis
├── Methods (2-3 câu): design + participants
├── Results (2-3 câu): key findings with numbers
└── Conclusions (1-2 câu): implications

INTRODUCTION (10-15% tổng):
├── Hook: broad problem
├── Narrow to specific gap
├── Statement of purpose
└── Outline of paper

[...etc cho mỗi section]

Sinh viên thấy: mỗi section có purpose functional, không phải arbitrary convention.

Chiến thuật 2: Reveal how experts think

Make the processes visible that experts trong ngành use to approach problems.

• How does a historian evaluate sources? Trace each micro-decision.
• How does a designer move from concept to prototype? Document each iteration.
• How does a doctor build differential diagnosis? Show reasoning chain.
• How does a lawyer read a contract? Reveal the "red flag" scanning pattern.

Have students:

• Interview practitioners
• Create flowcharts of expert thinking
• Narrate own thinking as they try to replicate

This metacognitive work builds awareness em need cho any collaboration — human hoặc AI.

Chiến thuật 3: Name your norms

Surface behaviors defining your field.

• Scientists don't just research — họ follow specific protocols for skepticism, replication, peer review.
• Artists don't just create — họ experiment, critique, revise.
• Lawyers don't just argue — họ steelman counterparties, cite precedent carefully.
• Engineers don't just build — họ design for failure, document decisions.

What does it mean "to think like a mathematician" in concrete, observable terms?

• Look for patterns
• State assumptions explicitly
• Test edge cases
• Prefer general over specific
• Accept elegant over clever

Build these behavioral modes of performance với sinh viên.

Description summary

┌──────────────────────────────────────────────────────────┐
│                                                          │
│   DESCRIPTION — Master Your Discipline's Communication   │
│                                                          │
│   1. Map products with precision (specific sections,     │
│      conventions, why they matter)                       │
│   2. Reveal expert thinking processes (trace decisions)  │
│   3. Name norms (concrete behaviors, not abstract)       │
│                                                          │
│   Result: Sinh viên có language để articulate work       │
│   of the field — to AI, to peers, to themselves.         │
│                                                          │
└──────────────────────────────────────────────────────────┘

Moving to Delegation — Understanding What Work Happens

Trước khi sinh viên work effectively với AI — hoặc với anyone — họ cần understand cái gì xảy ra trong ngành và tại sao matter.

Đây là nơi sinh viên develop ability to decompose complex work thành understandable components. Điều này cho phép em understand khi nào delegate work tới AI và khi nào không.

Chiến thuật 1: Reveal the anatomy of problems

Teach students to break down challenges trong ngành thành component parts.

Take typical task như writing a literature review:

LITERATURE REVIEW — Anatomy

├── Finding sources
│   └── Is this automatable? [Mostly YES — AI search, databases]
│
├── Evaluating relevance
│   └── How can AI augment? [Partial — AI suggests; human judges]
│
├── Synthesizing arguments
│   └── Requires human judgment? [Heavily YES]
│
├── Identifying gaps
│   └── Human insight needed [Yes — requires domain intuition]
│
└── Crafting final narrative
    └── Collaborative with AI? [Yes — AI drafts, human polishes]

Create problem anatomy diagrams cho every major task type trong ngành em.

Chiến thuật 2: Map AI possibilities explicitly

Create detailed maps của cái gì có thể:

• Automated (AI done on its own)
• Augmented (Human + AI together)
• Delegated to AI agents (AI acts on human's behalf)

Specific questions to ask:

• What routine tasks could AI automate? (Data cleaning, initial drafts, pattern recognition)
• Where does human-AI collaboration add value? (Design iteration, hypothesis generation, literature synthesis)
• What might an AI agent act independently on your behalf? (Monitoring experiments, filtering information, giving initial feedback on student work)

Have students debate these boundaries using real examples from ngành em.

Chiến thuật 3: Design decision trees

Create decision frameworks for when và how to involve AI or any tool or collaborator.

• Analyzing qualitative data: Where does software help vs. hinder?
• Solving engineering problems: When does simulation replace prototyping?
• Medical diagnosis: When does AI augment vs. when does human intuition override?
• Creative writing: When does AI inspire vs. when does it homogenize voice?

Build these trees through case studies. Make delegation decisions explicit và defendable.

DECISION TREE: Should I use AI for this research task?

Task type = ?
├── Information retrieval from public sources → AI YES (with verification)
├── Synthesis of literature → AI AUGMENT (AI drafts, human judges)
├── Novel methodology design → AI LIMITED (AI suggests; human decides)
├── Ethics review → AI NO (human judgment central)
└── Data contains PII/sensitive? → AI NO (or only PII-safe AI tier)

Delegation summary

┌──────────────────────────────────────────────────────────┐
│                                                          │
│   DELEGATION — Understand Work of Your Discipline         │
│                                                          │
│   1. Reveal problem anatomy (decompose tasks)            │
│   2. Map AI possibilities (automate/augment/agent)       │
│   3. Design decision trees (explicit delegate logic)     │
│                                                          │
│   Result: Sinh viên có judgment về when to delegate,     │
│   independent of specific AI capabilities.               │
│                                                          │
└──────────────────────────────────────────────────────────┘

Finishing with Diligence — Embodying Discipline's Values

Mỗi ngành có ethical standards, professional expectations, integrity requirements. Making these explicit prepares sinh viên cho responsible practice — with or without AI.

Chiến thuật 1: Codify ethical frameworks

Move beyond general academic integrity. What does "do no harm" mean trong ngành em?

• AI analyze patient data? → How ensure privacy?
• AI assist legal research? → How maintain client confidentiality?
• AI evaluate job applications? → How prevent systemic bias amplification?
• AI tutor students? → How ensure equitable access?

Build case studies exploring these edges. Create ethical decision matrices sinh viên có thể apply cho novel situations em sẽ encounter trong professional futures.

Ví dụ — Ethics matrix cho Journalism:

Chiến thuật 2: Clarify transparency

Research và document transparency norms với sinh viên.

• When must methods be fully disclosed? (Scientific research)
• When is process intentionally obscured? (Trade secrets, national security)
• When is "AI-assisted" acceptable attribution? (Journalism, academic)
• When does AI use require specific methodology section? (Research papers)

Create disclosure templates cho common scenarios trong ngành em based on existing practice now applied to AI.

• How would you cite AI's assistance in a research paper vs. creative portfolio vs. business report?

Chiến thuật 3: Co-create accountability

What does responsible practice look like trong ngành em's AI-enhanced future? Why do these practices matter?

Draft classroom policies together. Create honor codes addressing AI collaboration. Develop peer-review protocols checking appropriate human oversight.

Co-creation process builds buy-in và deeper understanding về tại sao accountability matters.

Ví dụ — Collectively drafted class policy:

In this class, AI is a collaboration partner — not a substitute for thinking. Here's our shared understanding: 1. AI use is expected. Not using AI doesn't earn 'moral high ground'. 2. Transparency is mandatory. Every assignment asks how AI contributed. 3. Verification is your responsibility. AI hallucinates. You sign off on claims. 4. Your voice, your judgment. AI doesn't replace your analysis — it supports it. 5. We help each other. If you see a peer over-relying on AI, start a conversation. — Signed by each student and the instructor at semester start.

Diligence summary

┌──────────────────────────────────────────────────────────┐
│                                                          │
│   DILIGENCE — Embody Your Field's Values                 │
│                                                          │
│   1. Codify ethical frameworks (field-specific)          │
│   2. Clarify transparency norms (what, when, how)        │
│   3. Co-create accountability (collective ownership)     │
│                                                          │
│   Result: Sinh viên internalize ethics, so they can      │
│   navigate any collaboration responsibly — with AI or    │
│   with humans.                                           │
│                                                          │
└──────────────────────────────────────────────────────────┘

The Feedback Loop — Why This Deep Discipline Work Matters

Deep disciplinary work này creates a powerful feedback loop.

┌────────────────────────────────────────────────────────────┐
│                                                            │
│   FEEDBACK LOOP                                            │
│                                                            │
│   Students who can articulate                              │
│   QUALITY STANDARDS                                        │
│        │                                                   │
│        ▼                                                   │
│   → better evaluate ANY output (human OR AI)               │
│                                                            │
│   Those who understand                                     │
│   DISCIPLINARY METHODS                                     │
│        │                                                   │
│        ▼                                                   │
│   → can guide ANY process more effectively                 │
│     (với humans AND AI)                                    │
│                                                            │
│   Those who internalize                                    │
│   PROFESSIONAL ETHICS                                      │
│        │                                                   │
│        ▼                                                   │
│   → can navigate ANY collaboration responsibly             │
│     (với humans OR AI)                                     │
│                                                            │
└────────────────────────────────────────────────────────────┘

And we benefit too. Thinking through lens of 4Ds forces us make tacit knowledge explicit. Considering AI capabilities helps us value uniquely human contributions. Preparing cho AI-enhanced future makes us better teachers và practitioners TODAY.

4Ds as framework để develop discipline-specific expertise

┌──────────────────────────────────────────────────────────┐
│                                                          │
│  DISCERNMENT: Define và recognize quality                │
│  → Collaborative rubrics                                 │
│  → Exemplar analysis                                     │
│  → Failure diagnosis                                     │
│                                                          │
│  DESCRIPTION: Master field's communication               │
│  → Mapping outputs                                       │
│  → Decoding processes                                    │
│  → Naming behaviors                                      │
│                                                          │
│  DELEGATION: Understand work processes                   │
│  → Problem decomposition                                 │
│  → Automation mapping                                    │
│  → Decision frameworks                                   │
│                                                          │
│  DILIGENCE: Embody field's values                        │
│  → Ethical frameworks                                    │
│  → Transparency norms                                    │
│  → Co-created policies                                   │
│                                                          │
└──────────────────────────────────────────────────────────┘

These competencies prepare students cho whatever comes next. Không chỉ qua understanding how they can best work với AI, mà cũng provide them metacognition skills necessary to apply this process tới any change họ may encounter trong future — technological or otherwise.

Case studies theo chuyên ngành — Deep dive

English Literature — Applying 4Ds to discipline

Discernment — defining literary quality:

• Not: "well-written", "creative"
• Instead: specific markers — voice consistency, metaphorical precision, structural rigor, emotional authenticity
• Exemplar practice: annotate 1 Márquez paragraph showing how 7 words create 3 layers of meaning
• Failure practice: compare Hemingway draft vs. final — what Hemingway cut

Description — how literary scholars communicate:

• Close reading format: quote + explication + connection
• "Hedging" language: "suggests" vs. "demonstrates" vs. "implies"
• Citation norms: block quote vs. inline vs. paraphrase

Delegation — work anatomy of a literary essay:

• Source search: AI-friendly
• Text analysis: human-AI collaborative (AI finds patterns; human judges significance)
• Thesis development: human-led with AI rubber-ducking
• Final argument: primarily human

Diligence — ethics of literary scholarship:

• Respect living authors' agency
• Don't AI-generate close readings (foundational human skill)
• Cite AI assistance in methodology
• Don't claim AI analysis as own insight

Chemistry — Applying 4Ds

Discernment — what "elegant mechanism" means:

• Fewer steps > more steps
• Predictable intermediates > exotic intermediates
• Literature-supported > speculative
• Scalable > impractical

Description — communication norms:

• Reactions drawn with arrow conventions (curved, electron flow)
• Reagents above/below arrow standard positions
• Product selectivity noted with %

Delegation — work anatomy:

• Literature search: AI-heavy
• Mechanism prediction: AI-augment (AI suggests; chemist checks feasibility)
• Actual synthesis: human only (physical)
• Characterization analysis: AI-augment

Diligence — lab ethics:

• Reproducibility (real vs. convenient reporting)
• Hazard disclosure
• Data integrity (no fabrication)
• AI-generated mechanisms: always verify in literature

Law — Applying 4Ds

Discernment — quality legal writing:

• IRAC clarity
• Case citations precise (jurisdiction, year, page)
• Counter-arguments steelmanned
• Statutes quoted verbatim
• Authority graded (primary > secondary)

Description — legal communication:

• Formal register
• Hedged conclusions with confidence ("likely", "arguably")
• Cite-as-you-go convention
• IRAC structure baseline

Delegation — work anatomy:

• Legal research: AI-friendly (but verify!)
• Brief drafting: AI-augment with caution (hallucination risk)
• Client counseling: human only (judgment, empathy)
• Court appearance: human only (procedure, live reasoning)

Diligence — legal ethics:

• Attorney-client privilege (NEVER feed to public AI)
• Verification of every citation
• Competence duty (knowing AI limits)
• Court disclosure requirements (varies by jurisdiction)

Design — Applying 4Ds

Discernment — design excellence:

• Functional fit (form follows function)
• Aesthetic coherence (consistent visual language)
• User respect (clarity over cleverness)
• Craft (detail precision)
• Context-appropriate (luxury vs. utility)

Description — design communication:

• Concept statements (1-2 sentence essence)
• Design rationale (why these choices)
• Spec sheets (measurements, materials, colors)
• Critique vocabulary (formal elements: line, shape, color, texture)

Delegation — work anatomy:

• Concept divergence: AI-generous (idea generation)
• Curation: human-led
• Technical execution: human + AI iterative
• Final refinement: human (taste)

Diligence — design ethics:

• Attribution of visual influences
• Avoid cultural appropriation
• Accessibility (WCAG compliance for digital)
• AI-generated imagery: disclosure in portfolio

Medicine — Applying 4Ds

Discernment — clinical judgment:

• Occam's razor (simpler diagnoses first)
• Don't anchor early
• Consider worst-case always (won't miss life-threatening)
• Confidence calibration (Bayesian thinking)

Description — medical communication:

• SOAP notes structure
• ICD/CPT coding precision
• Patient communication (plain language, empathy)
• Clinical reasoning documentation (chain)

Delegation — work anatomy:

• Literature review: AI-heavy
• Differential diagnosis: AI-augment (AI broadens; physician narrows)
• Physical exam: human only (can't automate)
• Treatment decision: physician-led (liability)

Diligence — medical ethics:

• HIPAA/PHI absolute
• Informed consent (including AI involvement)
• Beneficence + non-maleficence
• Never reassign clinical authority to AI

Business — Applying 4Ds

Discernment — strategic excellence:

• Clear thesis
• Actionable recommendations (vs. platitudes)
• Quantified assumptions
• Scenarios considered
• Stakeholder analysis

Description — business communication:

• Executive summary first
• BLUF (Bottom Line Up Front)
• Data viz norms (chart types for data types)
• Memo vs. deck vs. email register

Delegation — work anatomy:

• Market research: AI-heavy
• Financial modeling: AI-augment
• Customer discovery: human only (relational)
• Strategic decisions: human-led

Diligence — business ethics:

• NDAs and trade secrets (not in public AI)
• Sourcing of claims
• Disclosure of AI use in proposals
• Conflict of interest awareness

K-12 Teaching — Applying 4Ds

Discernment — pedagogical quality:

• Learning objectives specific, measurable
• Differentiation for diverse learners
• Assessment valid (measures what claims)
• Cultural responsiveness

Description — pedagogical communication:

• Lesson plan format
• Rubric clarity for students
• Parent communication register
• Curriculum alignment language

Delegation — work anatomy:

• Content preparation: AI-augment
• Lesson design: human-led
• Assessment creation: AI-augment
• Student interaction: human only (relational)

Diligence — teaching ethics:

• Student data privacy (FERPA)
• Equity (AI access gap)
• Modeling ethical AI use
• Accountability to parents

Prompt mẫu: Co-develop discipline-specific 4Ds

1. Discernment quality rubric

Tôi dạy [ngành]. Help me develop DETAILED rubric for "excellent
work" — beyond vague terms. For each of 5 quality markers:
- Concrete description (what it looks like)
- Specific example from field
- Common failure mode (what NOT it)
- How to teach recognition of this marker
Markers: [5 markers em identify]. Output: rubric markdown.

2. Exemplar analysis guide

Take 1 masterwork từ [ngành]: [specific work]. Create annotation
guide revealing:
- 3 specific moments where craft excellence shows
- 3 structural decisions that matter
- 3 less-visible choices that separate good from great
Format: quote + annotation + lesson for students.

3. Failure diagnosis collection

Identify 5 FAILURE MODES specific to [ngành]:
- Name failure mode (specific, memorable)
- Example text/case showing it
- Why it fails (technical reason)
- How to recognize early
- How to recover
Output: "Field guide to [ngành] failures" 1500 words.

4. Expert thinking decomposition

Interview-style: walk me through HOW an expert [profession] would:
[specific task]. Go step-by-step, include:
- Micro-decisions (every small choice)
- Why each choice matters
- Alternatives considered and rejected
- Internal dialogue
Make tacit knowledge explicit.

5. Problem anatomy diagram

Take typical task in [ngành]: [task]. Create problem anatomy:
- All component sub-tasks
- For each: human-only / AI-augment / AI-automate
- Reasoning for classification
- Cross-dependencies
- Current state of automation (what already exists)
- Projected state 2 years out

6. Discipline-specific ethics matrix

For [ngành], build ethics matrix for AI use:
Columns: [Situation types unique to ngành]
Rows: Allowed / Prohibited / Conditional
Include: AI tools, data sensitivity, transparency requirements,
accountability lines.
Output: 1-page reference for students.

7. Co-create policy statement

Draft class policy statement for [ngành] course incorporating:
- Expected AI use (not banned, but structured)
- Transparency requirements
- Student accountability standards
- Instructor commitments
- Community norms (peer oversight)
- Revision process
Tone: collaborative, not prescriptive. 500-800 words.

8. Colleague discussion prep

Prepare me for 60-min department meeting on "Discipline-specific
4Ds". Include:
- Opening question surfacing tacit knowledge
- 1 activity (15 min) collective exemplar analysis
- 1 activity (20 min) failure diagnosis
- 1 activity (15 min) problem anatomy
- Closing: shared document draft
- Follow-up plan
Output: facilitator guide.

Anti-patterns — Sai lầm khi apply expertise

Skip "discipline-specific" — dạy generic

Biểu hiện

"4Ds apply same way regardless of field."

Tại sao là sai

Miss the very point — disciplinary expertise is what makes AI Fluency meaningful.

Cách đúng

Every example, case, rubric specific to your field's context, vocabulary, values.

Keep tacit knowledge tacit

Biểu hiện

"Students will figure it out by reading good examples."

Tại sao là sai

Without explicit decoding, most students can't extract lessons from exemplars alone.

Cách đúng

Explicit articulation of what makes good good. Name it. Annotate it. Rubric it.

Work solo

Biểu hiện

You alone develop rubrics, case studies, policies.

Tại sao là sai

Limited perspective. Colleagues have blind spots you don't (and vice versa).

Cách đúng

Co-develop with colleagues. Messier process, richer output.

Skip failure diagnosis

Biểu hiện

Only show exemplars of excellence.

Tại sao là sai

Students learn "what good looks like" but not "what good is NOT". Miss discriminative power.

Cách đúng

Equal time on failures. Forensic analysis builds evaluative muscles.

Treat ethical frameworks as abstract

Biểu hiện

"Be ethical. Don't cheat."

Tại sao là sai

Not actionable. Students can't apply to specific situations.

Cách đúng

Specific cases, decision matrices, concrete examples. "In situation X, here's what responsible looks like."

Force all 4Ds equally

Biểu hiện

Equal time/weight on each D regardless of field.

Tại sao là sai

Different disciplines naturally emphasize different Ds. Medicine-heavy Diligence; Design-heavy Description; Research-heavy Discernment.

Cách đúng

Weight by field. Acknowledge and teach to strengths.

Reinvent everything from scratch

Biểu hiện

Ignore existing disciplinary norms in favor of "AI-new" everything.

Tại sao là sai

Disciplines evolved their norms over centuries. Most are still valid.

Cách đúng

Extend existing norms. "How does our field's citation practice adapt to AI?" > "Let's redo citation from scratch."

Forget students as co-creators

Biểu hiện

You develop rubrics. Students receive them.

Tại sao là sai

Miss buy-in and co-learning opportunities.

Cách đúng

Drafts with students. Let them challenge. Revise together.

Mẹo nâng cao

Mẹo 1: "Exemplar archaeology"

Each assignment, collect 1-2 outstanding student exemplars (with permission). After course, annotate them for future classes. Living repository.

Mẹo 2: Alumni network as validation

When developing rubrics, share with 3-5 alumni professionals. Ask: "Does this match what excellence looks like in practice?" Valuable reality check.

Mẹo 3: Cross-disciplinary comparison

Occasionally share your discipline's quality rubric with colleague from different field. Surprising insights emerge — you see assumptions you didn't know you made.

Mẹo 4: Student-led co-creation session

Dedicate 1 class period to: "Let's build the rubric for this assignment together." Messy but transformative buy-in.

Mẹo 5: Failure showcases

Once per semester: "failure gallery". Students submit (anonymously) their worst piece of writing and what they learned. Normalizes failure as learning.

Mẹo 6: Ethical dilemmas library

Start library of real ethical dilemmas students faced. Use for case discussion next semester. Content accumulates naturally.

Áp dụng ngay — Final exercise: Talk to humans

Cho final exercise of this course, chúng ta đang give AI partners a break và just talk to human colleagues!

However, having all participants work through the exercise từ previous lesson individually sẽ facilitate these conversations.

In whatever manner makes sense cho your context, schedule some organized time to work through 4D framework với colleagues.

Suggested topics to guide discussion

Cho Discernment — "What does quality look like in our field?"

• Work together articulate excellence trong discipline beyond vague terms
• Identify specific features distinguishing outstanding work từ mediocre work
• Discuss how teach students recognize these quality markers
• Document criteria helping students evaluate both human và AI-generated work

Cho Description — "How do we communicate in our discipline?"

• Map key products trong field với precision (not just "reports" but specific artifacts và why they matter)
• Document thought processes experts use khi approaching problems trong field
• Identify behavioral norms và conventions defining professional practice
• Explore how make these communication patterns explicit cho students

Cho Delegation — "What work happens in our field?"

• Break down typical tasks trong discipline into component parts
• Identify which elements require human judgment, creativity, expertise
• Discuss where AI could automate, augment, hoặc act as agent
• Create decision frameworks for when và how involve AI trong disciplinary work

Cho Diligence — "What are our field's values và standards?"

• Codify ethical frameworks specific to discipline
• Clarify transparency norms và disclosure expectations
• Discuss accountability standards và professional responsibilities
• Consider how these apply khi AI involved trong work

Building a shared document

• Compile discipline-specific interpretations of each D
• Include concrete examples từ field cho each competency
• Add teaching strategies colleagues suggest cho developing these competencies
• Identify areas where em have consensus và where perspectives differ
• Consider how share work với students để make AI Fluency concrete
• Discuss how integrate discipline-specific 4Ds into curriculum
• Consider how assess whether these frameworks helping students
• Agree on next steps để continue discussion và implement new initiatives

Ghi lại kết quả:

• Colleague tôi sẽ meet với: ___________
• Date/time scheduled: ___________
• Biggest question tôi muốn tackle collectively: ___________
• Document tôi sẽ draft based on conversation: ___________

Tóm tắt bài học

The bigger picture — A closing note

Throughout this course, chúng ta đã:

• Explored teaching và assessing qua AI Fluency framework
• Engaged meaningfully với AI's impact on teaching và learning
• Discussed how make 4Ds discipline-specific
• Built foundation for teaching meaningful human-AI collaboration

Your expertise — your ability to recognize quality, communicate precisely, decompose problems, uphold values — is what makes you essential.

By making this expertise explicit và teachable, em giving students foundation they need to thrive.

Future needs humans who can:

• Think critically
• Communicate clearly
• Collaborate wisely
• Act responsibly

You're not preparing your students to be replaced by AI.

You're preparing them to be irreplaceable.

We appreciate em taking this journey với us.

Tài liệu tham khảo

• Video bài gốc: https://www.youtube.com/watch?v=BUj8mjy6oxI
• Bài 13.5 — Tác động của AI lên ngành (context cho bài này)
• AI Fluency Framework v1.5 (PDF tài liệu gốc)
• Dakan, R. & Feller, J. — The AI Fluency Framework

Bản quyền gốc 2025 Rick Dakan, Joseph Feller, và Anthropic (CC BY-NC-SA 4.0). Bản tinh chế tiếng Việt 2026.