SukanyaLab
0/40.5 h
SukanyaLab · A working notebook, kept in public

The architect takes her own curriculum.

I designed the 80-hour Cheyney AI Literacy Shared Core. Before I teach it, advertise it, or advise on it, I am completing it — every module, every artifact, every certificate — at an architect's pace of 40.5 hours. This lab is where the work happens and where the record lives.

Act with full excellence in the work itself; hold the outcomes with equanimity. The practice frame · nishkama karma · Bhagavad Gita 2.47
0 / 40.5Hours banked
0 / 10Modules
0 / 5Artifacts
0 / 3Certificates
Why 40.5 hours, not 80

The architect's pace, stated honestly

The full 80-hour Core includes a 30-hour zero-prerequisite onramp designed for students arriving without digital fluency. The architect does not need the onramp at student pace — but she does need to sit inside every module she asks students to sit inside. Compression is applied only where prior expertise is real; artifacts and certificates are never compressed.

ComponentStudent paceArchitect paceWhat is never skipped
Elements of AI onramp30 h10 hEvery quiz; the certificate
Frameworks (HKS · 4D · Google)21 h13.5 hAll three certificates and both memos
Cheyney-original modules14 h6.5 hThe mini-audit and the policy — written fresh
NIST + capstone + close15 h10.5 hThe full build-evaluate-defend cycle
Total80 h40.5 h5 artifacts · 3 certificates · 1 defended build
The path · 10 modules · complete in order

The Shared Core, walked by its designer

Each module carries its working link, its steps, the artifact it produces, the question it gives you standing to answer, and its place in the Universal HBCU Workforce Alignment Model. Mark a module complete to bank its hours on the ledger.

1 The Serviceability Foundation FRAMINGCHEYNEY-ORIGINAL 1hour

The distinction that opens the entire Core: employability names what an employer asks for at the door; Serviceability — Dr. Lezli Baskerville's standard — names whether a graduate is equipped to serve their community. A credential names a passing score; a capacity names a durable competence.

Do this

  1. Re-read your own Module 1 text and Dr. Baskerville's framing as a learner, not as the author.
  2. Write your one-page Serviceability Statement — the same pass/fail assignment every student receives. It returns as the opening reflection of your capstone.

Artifact

The Architect's Serviceability Statement (1 page). This is publishable — the founding document of the public series.

Standing earnedWhere does serviceability enter AI literacy — and why does an AI course open with a question about community rather than a question about technology?
ALIGNMENT MODEL · System 5 — Mission Value Evidence
2 What Is Artificial Intelligence, Really? CA1 · UNDERSTANDCERTIFICATEELEMENTS OF AI 10hours

The University of Helsinki's Elements of AI — the 30-hour zero-prerequisite onramp, taken at review pace. You wrote the discussion seminars that wrap this course; now sit the course itself, every chapter and every quiz.

Do this

  1. Enroll free at elementsofai.com — "Introduction to AI."
  2. Complete all six chapters and every exercise. While inside, verify the current certificate threshold and correct the curriculum document if it differs from the stated 90% figure.
  3. As you go, annotate each chapter against the paired case study you assigned it — COMPAS at Chapter 4, Gender Shades and Noble at Chapter 6.
  4. Claim the certificate; add it to LinkedIn.

Artifact

Certificate 1 of 3 — Elements of AI, University of Helsinki.

Standing earnedWhat is AI, actually — and can you explain machine learning, neural networks, and their limits to a community room in plain language? This module is the backbone of the AI 101 posture.
ALIGNMENT MODEL · System 3 — AI Workforce Readiness (the onramp every student walks)
Open the course →
3 The Anatomy of a Generative System CA1 · CA3HARVARD KENNEDY SCHOOL 5hours

Three open sessions from HKS's "The Science and Implications of Generative AI" (Goel, Levy, Svoronos): Introduction to Generative AI; Prompt Engineering; Beyond Chatbots — system prompts and RAG. Here you learn the TIC syntax — Task, Instructions, Context — and the ladder of tailoring: prompt, system prompt, RAG, fine-tuning. Pick the lowest rung that satisfies the use case.

Do this

  1. Work the three sessions at generative-ai-course.hks.harvard.edu — videos, slides, activities are open; no registration.
  2. Complete the student assignment exactly: three prompts against three tasks from your own practice, each iterated twice, with a reflection on what changed between iterations.
  3. Note for the record: the open course issues no certificate — the artifact is the evidence.

Artifact

Portfolio artifact 1 — the prompt set with iterations and reflection. Suggested tasks: a board memo outline, a WAM heat-map explainer, an AI 101 lesson opener.

Standing earnedHow does a generative system actually work — tokens, context, retrieval — and when does a problem require more than a better prompt?
ALIGNMENT MODEL · System 3 — the prompting grammar the sector tracks assume
Open the course →
4 The 4D Framework — Delegation, Description, Discernment, Diligence CA2–CA5CERTIFICATEANTHROPIC · CC BY-NC-SA 3.5hours

The spine of the Shared Core. Anthropic's AI Fluency: Framework and Foundations (Dakan, Feller) teaches the four durable capacities that map to DOL Content Areas 2 through 5. The course runs 3 to 4 hours — a correction to the 6-hour estimate in the curriculum document, logged in the ledger below.

Do this

  1. Complete the course at Anthropic Academy (free Skilljar account) — or via Coursera.
  2. Take the final assessment; claim the certificate of completion.
  3. Write the student assignment: a 1,200-word memo narrating one recent AI interaction of your own through each of the four D's. Choose a consequential one — a deck build, a federal-comment draft.
  4. Extension for the teaching posture: Anthropic's newer AI Fluency for Educators (~3 h) and Teaching AI Fluency (~5–6 h) are free, certificated, and CC-licensed — queue them after the Core.

Artifact

Portfolio artifact 2 — the 4D memo. Certificate 2 of 3 — Anthropic AI Fluency.

Standing earnedWhat does responsible delegation to AI look like in practice — deciding whether and what to delegate before deciding how?
ALIGNMENT MODEL · System 3 — the mental model every track inherits
Open the course →
5 AI in the World of Work CA2 · CA3 · CA5CERTIFICATEGOOGLE · VIA PASSHE ACCELERATOR 5hours

Google AI Essentials — five modules from introduction through responsible AI. Cheyney is a confirmed inaugural member of the Google AI for Education Accelerator (PASSHE announcement, September 8, 2025), which provides free access to faculty and staff. Do not pay the $49 retail: route through Cheyney's institutional license.

Do this

  1. Request Accelerator access through Cheyney's program administrator (Google's contact for institutions: aiforeduaccelerator@google.com). Fallback paths: Coursera financial aid, or audit free without the certificate.
  2. Complete the five modules at Coursera — Google AI Essentials; quizzes pass at 80%.
  3. Run the student worksheet on yourself: route five entry-level tasks from your own week through the 4D framework, noting which functions are absorbing into AI workflows and which are not.
  4. Claim the Google certificate and Credly badge.

Artifact

Certificate 3 of 3 — Google AI Essentials.

Standing earnedWhich entry-level functions are absorbing into AI workflows — the question every advising conversation and every program review now turns on.
ALIGNMENT MODEL · Systems 3 + 4 — workplace application and the employer vocabulary
Open the course →
6 Equity-Centered AI CA5 · RESPONSIBLYCHEYNEY-ORIGINAL 3hours

Your own eight hours, walked in three: Noble's Algorithms of Oppression, Benjamin's Race After Technology, Buolamwini and Gebru's Gender Shades, Obermeyer et al. in Science — across the four case clusters: criminal justice, healthcare, hiring and credit, generative AI. The pedagogical move you designed is the point: these cases position students as the next generation of auditors, reviewers, regulators, and builders — a professional posture, not a defensive one.

Do this

  1. Re-read the four case clusters as a learner; annotate where the 2026 evidence has moved since you wrote them.
  2. Complete the student assignment fresh: a ~1,500-word mini-audit of one publicly accessible AI system. Architect's option: audit a tool you actually use in the consulting practice.

Artifact

Portfolio artifact 3 — the mini-audit. A strong candidate for the public series: the strategist audits her own toolkit.

Standing earnedWhere does equity enter an AI curriculum — as a module or as a posture? The answer is the reframing itself: vigilance as professional responsibility.
ALIGNMENT MODEL · System 5 — the mission argument, made technical
7 The Responsibility Regime — NIST AI RMF CA5 · RESPONSIBLYNIST AI 100-1 · AI 600-1 2.5hours

The four functions — Govern, Map, Measure, Manage — as a lifecycle applicable to every AI system you will ever be asked about. AI 100-1 (January 2023) and the Generative AI Profile AI 600-1 (July 2024) remain the operative versions as of mid-2026; a revision cycle is underway for 2026–2027.

Do this

  1. Read the RMF core at nist.gov/itl/ai-risk-management-framework and skim the free RMF Playbook on the NIST AIRC.
  2. Read AI 600-1's 12 generative-AI risk categories and its Section 3 actions.
  3. Complete the student assignment: a one-page RMF mapping of one AI system you used in the past week.

Artifact

The one-page RMF mapping — the governance answer, on one page, in federal vocabulary.

Standing earnedHow does a small institution govern AI without a large staff? Four functions, a tiering discipline, named ownership — the fluent answer this module supplies.
ALIGNMENT MODEL · Systems 1 + 3 — the risk grammar behind governed adoption
Open the framework →
8 Writing Your Personal AI Use Policy CA5 · RESPONSIBLYSYNTHESIS 2hours

The synthesis module: a seven-section policy — scope and definitions; tools and situations; data shared and withheld; disclosure practices; verification practices; update cadence; and the Serviceability clause describing standards beyond the legal minimum. The Serviceability clause is the section a supervisor remembers.

Do this

  1. Write all seven sections for yourself — the working policy of Mukherjee & Associates, not a hypothetical.
  2. Publish it. "The architect's own AI use policy" is the single most credible marketing artifact this walk produces.

Artifact

Portfolio artifact 4 — the Personal AI Use Policy of the person who wrote the assignment.

Standing earnedWhat does disclosure look like in practice? You will have answered it for yourself before advising anyone else.
ALIGNMENT MODEL · System 3 — responsible-use baseline, modeled at the top
9 The Core Capstone — Build, Evaluate, Defend CA2–CA5PARTYROCK · AI STUDIO 8hours

Design, build, evaluate, and publicly defend a small AI-powered application. Build on AWS PartyRock first — free daily credits, social login, no card, no AWS account — with Google AI Studio as the alternative (free tier now runs Flash-class models; Pro moved to paid in April 2026). The full specification, including the three WAM-aligned build options, lives in the Capstone section below.

Do this

  1. Choose one capstone option from the specification below; write the two-sentence problem statement.
  2. Build at partyrock.aws; document at least 3 prompt iterations with reasoning.
  3. Write the evaluation memo — at least 3 output-error categories with mitigations — and the Diligence memo applying your Module 8 policy.
  4. Defend it: record a 10-minute walkthrough, or present it live at a community AI 101 session. The defense is the launch of the public series.

Artifact

Portfolio artifact 5 — the capstone package: app, iterations, evaluation memo, Diligence memo, recorded defense.

Standing earned"Have you actually built anything with this?" — after this module, the answer is yes, with the documentation to prove the process.
ALIGNMENT MODEL · the capstone options each serve a live system — see the specification
Open PartyRock →
10 Compliance as Competitive Advantage FRAMINGCHEYNEY-ORIGINAL 0.5hours

The closing doctrine — the conviction the decade at Cheyney produced: compliance, practiced with rigor, is not a burden to manage; it is competitive strategy for mission-driven institutions. The same is true of AI governance done well.

Do this

  1. Write the closing reflection: one page connecting the walk you just completed to the doctrine — where did taking the curriculum change how you will teach it, sell it, or defend it?
Standing earnedThe sentence that closes every keynote: I did not only design this curriculum. I completed it.
ALIGNMENT MODEL · the doctrine that holds all five systems together
The armory · nuanced questions, fluent answers

What the walk gives you standing to say

Ten questions you are asked as The Mission Preservation Strategist — each answered in working form, with the module that earns the standing and the citation that carries the authority. Review one per module completed.

Q1What does an AI curriculum look like from a mission-preservation perspective?

It looks like infusion, not reinvention — roughly 180 of 220 hours taught with curated open content from Harvard, Anthropic, Google, Helsinki, and NIST, because a resource-constrained institution should not rebuild what the field already teaches well. The 40 institution-original hours are where the mission lives: Serviceability, equity-centered AI, and compliance-as-advantage. Mapping every module to the Department of Labor's five Content Areas makes the program federally legible without diluting it — the federal vocabulary protects the mission by making it fundable and auditable.

STANDING · Modules 1, 6, 10 · CITE · DOL TEN 07-25 (Feb 13, 2026); Cheyney AI Literacy Curriculum v3.0
Q2How does a small institution govern AI without a large staff?

With four functions and a tiering discipline. NIST's lifecycle — Govern, Map, Measure, Manage — scales down; Marshall University's public four-tier risk classification shows it operationalized on a campus. The governing rule: every consequential AI use carries a named human accountability owner, an audit trail, and board visibility. Governance is a discipline of ownership, not a function of headcount.

STANDING · Module 7 · CITE · NIST AI 100-1 and AI 600-1; Marshall University IRP (June 2026); CHEA Guiding Principles (Feb 2025)
Q3Is AI literacy a technical subject?

Not in the federal definition. The Department of Labor defines AI literacy as the competencies that enable individuals to use and evaluate AI responsibly, with a primary focus on generative AI — and its five Content Areas are Understand AI Principles, Explore AI Uses, Direct AI Effectively, Evaluate AI Outputs, and Use AI Responsibly. Direction, evaluation, and responsibility are judgment capacities. The Pearson and AWS friction research makes the same point institutionally: five of the six frictions blocking AI readiness live in the executive layer, not the classroom.

STANDING · Modules 2, 4 · CITE · DOL TEN 07-25; Pearson & AWS, AI Readiness in Higher Education (2026)
Q4What is serviceability, and why does it belong in an AI course?

Serviceability is Dr. Lezli Baskerville's standard: the measure of whether graduates are equipped to serve their communities, not merely whether they are employed. It belongs at the opening of an AI course because it sets the evaluative frame for everything that follows — the earnings test measures one thing; the mission produces another; a curriculum that names both trains graduates who can defend both. Every student's Personal AI Use Policy closes with a Serviceability clause for the same reason: standards beyond the legal minimum are the ones a supervisor remembers.

STANDING · Modules 1, 8 · CITE · Baskerville serviceability standard; Alignment Model System 5
Q5How does a curriculum built on free vendor content survive vendor churn?

By governing the dependency instead of denying it. The curriculum carries a Content Risk Register: every significant source has a stability assessment and a named substitution — IBM SkillsBuild behind Elements of AI, an offline CC-licensed archive behind the Anthropic course, alternate certificates behind Google. The Advisory Committee reviews the register every January. Dependency on world-class free content is a strength; ungoverned dependency would be the failure.

STANDING · Modules 2, 4, 5 · CITE · Curriculum v3.0 Content Risk Register; CC BY-NC-SA 4.0 license terms
Q6What does the federal government actually ask of an AI program?

Three things, currently. A vocabulary: the five Content Areas and seven Delivery Principles of TEN 07-25, which reviewers now read programs through. A performance bar: Workforce Pell's 150–599 clock-hour window with 70% completion and 70% placement gates. And an integration invitation: the AI in Registered Apprenticeship Innovation Portal, launched April 29, 2026, with its three pillars and its join-create-update paths. WIOA funds are explicitly authorized for AI training. A program that speaks all three is legible to every federal reviewer it will meet.

STANDING · Modules 2–5 · CITE · DOL TEN 07-25; P.L. 119-21; DOL AI-RAIP (apprenticeship.gov, Apr 29, 2026); TEGL 03-25
Q7Where does equity enter — as a module or as a posture?

As a posture that one module makes explicit. The eight equity-centered hours teach COMPAS, Gender Shades, and the Obermeyer healthcare-algorithm finding not as cautionary tales that position students as potential victims, but as professional responsibilities that position them as the next generation of auditors, reviewers, regulators, and builders. That single reframing — from defensive to professional — is the pedagogical signature of the curriculum, and it changes what every other module trains for.

STANDING · Module 6 · CITE · Noble; Benjamin; Buolamwini & Gebru; Obermeyer et al., Science (2019)
Q8Should a community AI 101 start with tools or with judgment?

Judgment, in the 4D order: Delegation comes first — whether and what to hand to AI — before Description teaches how. The employer data supports the sequence: talent leaders rank critical thinking and communication above AI literacy for entry hires. In a community room, open with the TIC syntax as the first sentence of the new grammar, and use PartyRock as the hands-on moment — free, no account barrier, and a participant leaves having built something in an afternoon.

STANDING · Modules 3, 4, 9 · CITE · Anthropic 4D framework; Strada (Mar 2026); HKS TIC framework
Q9How do executive competencies relate to a curriculum like this?

The competency model describes the leader; the curriculum is one of the instruments the institution holds. In the Alignment Model's terms, this curriculum is System 3 executed — the artifact through which a president evidences Delivering Academic Excellence and Funding the HBCU Future under the new federal architecture. A competency without an instrument is an aspiration; the curriculum is what makes the standard demonstrable.

STANDING · The whole walk · CITE · Universal HBCU Workforce Alignment Model; HBCU ELI Competency Model (Clark Atlanta, 2021)
Q10Where does Pennsylvania stand on AI in education?

Distinguish two tracks. State-government AI governance is mature: an executive order and Generative AI Governing Board since September 2023, a first-in-nation ChatGPT Enterprise pilot, a 2026 multiagency AI Literacy Toolkit, and a top-three national ranking for state AI readiness. K-12 classroom guidance is the emerging track: Pennsylvania was not among the states with formal K-12 AI guidance as of early 2026, and legislative hearings signal it is coming. Answering with both tracks is the precise answer; conflating them is the common error.

STANDING · Module 7 context · CITE · PA EO 2023-19; Shapiro administration AI Literacy Toolkit (2026); PA Capital-Star reporting (2026)
The capstone specification · module 9 in full

Build something the practice actually needs

The capstone doubles as product development: each option serves a live system of the Alignment Model or the community teaching posture. PartyRock first; Google AI Studio as the alternative.

Option A · The Three Doors Translator

An app that takes a program's earnings-test posture and explains, in presidential language, the three institutional decisions a first failure opens — the five-year Direct Loan suspension preserving Pell, the orderly teach-out, or no action. Serves Alignment Model System 1 and every executive briefing you give.

Option B · The AI 101 Community Coach

A plain-language companion for community sessions: a participant describes a daily task, and the app walks them through the 4D questions — what to delegate, how to describe it, how to judge the output, and what responsibility remains theirs. Serves the community posture directly.

Option C · The Serviceability Mirror

An app that interviews a student about a career goal and returns both readings — the employability answer and the serviceability answer — with the gap named. Serves System 5 and makes Module 1 tangible for any audience in 3 minutes.

The required package, whichever option

  • A defined problem articulable in 2 sentences
  • Meaningful generative-AI use
  • 3+ documented prompt iterations, with reasoning
  • An evaluation memo: 3+ output-error categories with mitigations
  • A Diligence memo applying your own Module 8 policy
  • A defense: 10-minute recorded walkthrough or a live community session
The record · what leaves this lab

Five artifacts, three certificates, one series

Portfolio artifacts

1 · The prompt setThree tasks, two iterations each, with reflection (Module 3)
2 · The 4D memo1,200 words on one real interaction (Module 4)
3 · The mini-audit~1,500 words on one public AI system (Module 6)
4 · The Personal AI Use PolicySeven sections, Serviceability clause (Module 8)
5 · The capstone packageApp, iterations, two memos, recorded defense (Module 9)

The public series — process over product

One post per module, weekly, on Substack with a LinkedIn spoke. The proven structure, adapted from learning-in-public practice:

  1. What the module claimed it would teach
  2. What I actually did and built
  3. Where the architect's design held — and where it needs revision
  4. The governance or mission question I can now answer with standing

Series title: The Architect Takes Her Own Curriculum. No public example of an executive completing their own designed curriculum surfaced in the research — the angle is genuinely unclaimed.

Corrections ledger — kept in the open, on principle

  • Anthropic 4D hours: the curriculum document estimates ~6 hours; the course's own materials state 3–4 hours. Corrected here; correct v3.1 of the document.
  • Elements of AI certificate: the document cites a 90% quiz threshold; official pages describe a completion-based certificate. Verify inside the live course during Module 2, then reconcile.
  • HKS open course: issues no certificate — the artifact is the evidence. The paid HKS Executive Education program is the credentialed alternative if ever required.
  • Directional flags: vendor-published figures (course popularity, wage premia, enrollment counts) are treated as directional throughout, per house rule.