Magister of Artificial Intelligence in Marketing (Mag.AI-Marketing)

Castalia Institute — Magisterium Design Document

Authoritative copy: part of the MyST project (myst.yml).

I. Purpose¶

Definition¶

Mag.AI-Marketing is a post-graduate program focused on the design, simulation, and deployment of AI-native marketing systems: audiences, messages, channels, measurement, and ethics as executable worlds, not slide decks.

It is not:

• a tactics checklist

• a “growth hacking” bootcamp

• a platform-specific tool course

It is:

A discipline of constructing and testing marketing realities through simulation before spending attention and budget in the wild.

Core Thesis¶

Marketing is a world model with attention, trust, and measurable persuasion.

Therefore:

• understanding marketing = modeling how people attend, believe, and act

• learning marketing = refining those models under evidence

• mastery = designing influence systems that survive reality and scrutiny

II. Degree Philosophy¶

1. Simulation First¶

Students begin by formalizing who is in the world, what they perceive, how messages move, and what counts as success — then they perturb before they scale spend.

2. Student as Author¶

Students:

• define audience ontology

• choose assumptions explicitly

• encode rules and measurement

• own their models

AI does not replace judgment about ethics, brand, or truthfulness.

3. AI as Instrument¶

Technical AI literacy (explicit): Parallel to world models, students build competence in evaluating AI in marketing workflows: RAG/grounding quality, prompt and version discipline, human-in-the-loop review, cost/latency tradeoffs, and red-teaming customer-facing flows. This is not a separate “AI minor”; it is embedded in campaign, data, automation, and ethics courses (see CURRICULUM_FULL.md, AI literacy spine).

4. Knowledge as Metabolism¶

Students continuously:

• ingest signals (analytics, qual research, competitive intel)

• refine models

• test assumptions

• ship measurable artifacts

III. Learning Model¶

Traditional Model (Rejected)¶

template → post → hope

Magister Model (Adopted)¶

define audience world → simulate → perturb → refine → deploy

Core Loop¶

1. Construct world (PTAH)

2. Run simulation / scenario set

3. Observe outcomes

4. Reflect (SAMWISE)

5. Challenge (Faculty)

6. Revise model

7. Deploy with measurement

IV. Program Structure¶

Degree Path¶

Bac.AI → Lic.AI → Mag.AI-Marketing

Mag.AI-Marketing is equivalent to applied doctorate-level mastery in marketing design.

Duration¶

• 3 Terms (or flexible pacing)

• Each term increases world complexity (audience → campaign org → competitive strategy)

Workload and milestones¶

Eighteen courses are artifact sprints, not implicitly “semester-long” loads. Faculty set minimum viable artifacts and may sequence or combine milestones so students do not carry eighteen simultaneous builds. Over-scoping is treated as a failure mode to catch in advising.

Instructional stack and fallbacks¶

Primary stack: PTAH (worlds), SAMWISE (reflection), iNQspace (execution, lineage), AI Faculty (challenge), MCP (tools). If any component is unavailable to a cohort, delivery uses approved substitutes: e.g. Python/notebook worlds with explicit ontology docs for PTAH-shaped work; structured reflection templates for SAMWISE; generic notebook + Git lineage for iNQspace; MCP optional on faculty-hosted or synthetic endpoints (see MCP_MARKETING.md). The curriculum outcomes do not depend on a single proprietary binary.

V. Curriculum Architecture¶

TERM I — Audience & Message Worlds¶

Build and simulate attention, trust, and message dynamics

TERM II — Campaign & Growth Worlds¶

Simulate campaigns and growth mechanics before scaling

TERM III — Strategic Marketing Worlds¶

Compete, allocate budget, automate responsibly, deploy

VI. Core Technologies¶

iNQspace¶

iNQspace is the primary environment in which we teach AI for this program: guided labs, course notebooks, simulations, integrations with models and data workflows, and lineage so work is inspectable and revisable. Lectures set direction; competence is built by doing — building worlds, running perturbations, and reflecting — inside iNQspace. It is the operational bridge between “understanding AI in marketing” and demonstrable skill.

In practice, iNQspace provides:

• Notebooks and runnable artifacts (course exercises, thesis components)

• Simulation runs tied to PTAH-style world models

• AI-in-the-loop workflows where appropriate (grounding, evaluation, not prompt tourism)

• Versioning / lineage for iteration and defense

MCP (marketing tools)¶

The Model Context Protocol (MCP) is how courses connect AI workflows to real marketing systems — analytics, CRM, ad accounts, content CMS, enrichment APIs — with explicit tool contracts instead of ad hoc scraping or mystery plugins. Students leverage existing MCP servers (faculty-approved) and build or extend servers when a course artifact requires a custom bridge to data or simulators.

The curriculum emphasizes tools by naming major platform ecosystems—social (Meta-class), Google (Search/Ads/Analytics/YouTube), Amazon (marketplace + ads)—as modeled and measurable environments, not vendor trivia (see CURRICULUM_FULL.md).

MCP is not “extra IT.” It is how deployment (Dimension D in assessment) stays inspectable: which tools ran, with what scope, and what evidence was produced. Program-wide expectations and guardrails: docs/MCP_MARKETING.md.

Authoring and the public site (content layer)¶

Program content — syllabi, lectures, public pages, design docs — is authored in structured sources so materials ship as cross-linked web pages, printable exports, and teaching packages. The project configuration declares the canonical table of contents for what counts as program materials. The repository README is developer-only: it must not duplicate curriculum text.

Derivation (single source):

1. Author course and program text under the paths declared in the project configuration.

2. Build the public site with the MyST / Jupyter Book CLI (myst start / myst build / jupyter-book build --site). Optional exports (PDF, Word, JATS, slides) via jupyter-book build --pdf (etc.) when needed.

3. Structure: the full TOC is the site; slide sources live under the slide paths with frame separators; exercises use {exercise} / {solution} blocks (and may mirror in notebooks).

4. Teach in iNQspace by importing or syncing those paths — syllabi and lecture text are not retyped inside the lab; execution (notebooks, sims, lineage) lives in iNQspace. MCP connects labs to marketing tools and data under policy.

5. Do not maintain a parallel copy of curriculum in wikis or ad hoc decks — if it is not in the project tree, it is not authoritative.

iNQspace is the execution layer for all runnable work.

PTAH¶

Ontology, rules, scenarios for audiences, messages, channels.

SAMWISE¶

Reflection, pattern detection, assumption critique.

AI Faculty (In Voce)¶

VII. Assessment Model¶

See ASSESSMENT.md. Five dimensions: World Construction, Simulation Quality, Insight, Deployment, Iteration.

VIII. Signature Experiences¶

1. Build Your First Marketing World — attention + credibility + message fit as a runnable model.

2. Narrative Arena — competing stories, measured outcomes.

3. Faculty Interventions — positioning vs behavior vs ethics.

4. Forkable Worlds — compare audience models and measurement choices.

IX. Differentiation¶

X. Integration with Castalia Ecosystem¶

XI. Monetization Strategy¶

XII. Final Statement¶

Mag.AI-Marketing is a program for people who want leverage without fog: models you can run, defend, and connect to real outcomes.

At completion, the graduate possesses a full toolbox — not a checklist of courses, but reusable artifacts: simulatable worlds, deployment evidence, technical AI practice (eval, grounding, oversight) in real workflows, MCP-governed marketing tool connections, reflection practice, and publishable documentation across the program site and artifacts. The credential attests that those tools exist, run, and can be defended.

Graduates are:

Designers of influence under constraint.

Legal Notice¶

“The Castalia Institute Magisterium confers proprietary credentials based on demonstrated work and evaluation. These credentials are not accredited academic degrees and do not confer professional licensure.”