For Faculty — JV briefing
A cross-disciplinary graduate program designed around how faculty actually teach — not how interdisciplinary programs have historically asked them to.
Twelve ASU schools, three per class, a Lead + Two Rotating Guest faculty model that fits inside standard course-load economics, and an AI substrate that handles the cohort-scale knowledge compounding between sessions — so faculty spend their time on the part of teaching only they can do, and the cohort produces work no single-discipline program can match.
Why most cross-disciplinary programs die in semester three
Three failure modes — and what this program does about each.
Failure mode
Faculty headbutting
Why it kills programs: Three professors each prepping their own slice end up competing for class time. The cohort gets three mini-courses stapled together, not one cross-disciplinary working session.
How we fix it: The Architect / Critic / Synthesizer triad. Three faculty playing three different roles against the same open question — rotating weekly. No professor competes for "the right answer" because no professor is supposed to drive to one.
Failure mode
Unsustainable faculty load
Why it kills programs: Asking three faculty to fully co-teach 15 sessions is three full course loads on one module. Provosts kill these programs by Year 2 on economics alone.
How we fix it: Lead + Two Rotating Guest. One Lead (one standard course load); two Guests at half-load each. Total faculty cost per module: two course loads spread across three faculty — manageable inside standard appointments.
Failure mode
Knowledge that evaporates
Why it kills programs: Cross-disciplinary insights from Week 1 are gone by Week 8 because no substrate captures them across the cohort. Each student rediscovers their own slice; nothing compounds.
How we fix it: AI substrate as the connective tissue. Cohort knowledge graph, REFLEXION journals, just-in-time card retrieval, verifier-loops on artifacts. The cohort's thinking compounds; the substrate makes that compounding tractable at depth.
Faculty load that pencils out
Two course-loads of faculty time. Spread across three faculty. One module.
Cadence: 2 days per week × 2 hours 45 minutes per session × 7.5 weeks = 15 sessions, roughly 41 hours of class time per module. The Lead attends every session; each Guest attends roughly seven sessions in their discipline window plus the triad bookends.
| Role | Sessions / module | Class hours | Equivalent |
|---|---|---|---|
| Lead Professor | 15 | 41h | One standard graduate course |
| Guest Professor A | 7 | 19h | Half a standard course load |
| Guest Professor B | 7 | 19h | Half a standard course load |
| Total / module | 29 | ~80h | Two course loads, distributed |
Year-2 upgrade path:Once the program has FTE budget committed, the Guest load can scale to all 15 sessions for both guests — the aspirational “triad core” model. The starting model is designed to fit standard appointments; the upgrade is optional and economics-driven.
Why now — and what the operator's PhD studies
Two democratizations. Both shift who builds and who decides.
Capability democratization. Commercial Earth observation, ground-station-as-a-service, open-source ML stacks, cloud geospatial processing — capabilities that were exclusive to states and well-resourced corporations a decade ago are now a credit-card purchase. A hedge fund counts cars in a parking lot from orbit. Bellingcat geolocates a war crime on Twitter. The bottleneck shifted from access to judgment.
Platform-building democratization. Five years ago, building an aerospace decision-intelligence platform or a 50-state compliance system required ten engineers and venture funding. Now it is months of work by one person directing AI. SpaceForge is exhibit A. ASIS is the second. Cottage Launchpad is the third. The means of production is becoming conversational. The practice is sometimes called vibe coding; the academic phrase is AI-mediated platform development. Either name, it is a population existing technology-assessment frameworks were never designed to govern — a single person and a model can now ship what an institutional engineering team used to.
The cohort program is the consumption-sideresponse: graduates of a Masters or Graduate Certificate prepared to lead when capability is broadly accessible and cross-disciplinary judgment is the binding constraint. The operator's PhD is the production-side response: an empirical study of who governs the single-person AI builder when there is no institution to relocate that governance into.
Target home: HSD (Human and Social Dimensions of Science and Technology) at SFIS, Fall 2027 entry. The PhD this proposes is one— the operator's. The cohort earns a Masters or Certificate. Those scopes do not get conflated.
Cohort knowledge graph
Every artifact, journaled reflection, primary source, and cohort discussion gets indexed into a graph that links across modules. By Week 8, students can pull every thread the cohort has touched on a question — including from Q1 they did six months ago.
Just-in-time evidence retrieval
When a class discussion surfaces a question, the system retrieves the relevant cards — primary sources, prior cohort journal entries, faculty-uploaded cases — in real time. The faculty don't have to pre-load every possible angle; the substrate finds the angle the cohort actually asked for.
Pre-defense stress test on student artifacts
A drafted memo or model gets stress-tested against the named stakeholder persona before class — the substrate plays the persona's likely first three questions, the student iterates, and only the iterated draft enters the classroom. Class time is spent on real defense, not first-draft cleanup.
Counterfactual provocation
When a cohort risks consensus-by-fatigue on a question, the substrate generates the strongest counter-argument from the literature it has indexed. The Critic professor can deploy it directly. This is the substrate as a tool, not a teacher.
Between-session synthesis pass
Between classes, the substrate consolidates the cohort's threads into a synthesis brief — what the three professors meet to read at their 20-minute mid-week sync. The next session's open question is informed by what the cohort actually surfaced, not what the syllabus pre-planned.
Research substrate for the operator's PhD
The same cohort knowledge graph becomes the operator's PhD research substrate. The PhD candidate's work studies how Masters-level cross-disciplinary outcomes compound when an AI substrate handles the cross-referencing load — a research question the field hasn't been able to answer at depth because the substrate didn't exist.
Substrate provenance
The AI substrate that runs this is a working AI orchestration platform Enkari developed for high-stakes analytical work in other domains, repurposed here for cohort-scale teaching. It is not a teaching-product wrapper. That provenance is what makes the depth tractable: no other program teaches this way because no other program has the substrate underneath.
The operator's PhD proposal — three possible arcs
Three research arcs for one PhD. Not for the cohort — the cohort earns a Masters or Certificate.
The cohort produces Masters-level cross-disciplinary work, defended against stakeholder personas — not dissertations. The PhD this proposes is the operator's single research candidacy, taken alongside building and running the program. The three arcs below are candidate framings for that research; the actual PhD will be one of them (or a hybrid), shaped by the chair and the IRB-approved data the program generates.
Governing the Single-Person AI Builder
When AI-mediated development collapses an enterprise- or aerospace-grade build from a team-plus-a-year to one person over months, who governs the result, and how? This is the spine of the dissertation.
Three solo-built platforms — SpaceForge (aerospace workforce education), ASIS (aerospace supply-chain decision intelligence with cascade simulation and digital twin variants), and Cottage Launchpad (50-state cottage-food compliance) — are the empirical base. Methodology: participant-builder inquiry (autoethnography + participatory-action-research lineage) with adversarial coding and explicit ENKARI LLC conflict disclosure. The dissertation treats these platforms as evidence, not products under promotion.
Anticipatory Governance When the Builder Is One Person
Anticipatory governance assumes there is an institution to build steering capacity into. When the builder is one person and a model, that assumption breaks. Where does governance have to relocate?
Theoretical home: Guston's anticipatory governance + boundary organizations + responsible innovation, applied to a limit case the framework was not designed for. Maynard's Advanced Technology Transitions and human-agency-under-AI work supplies the agency-side framing. Three regulated publics — aerospace decision-making, a regulated food economy, and graduate education — each get something different from one builder. The PhD proposes what each is owed and where the boundary work has to happen when the organization has disappeared.
What One Person Can Govern, and What Exceeds Anyone
What can a single AI-directed builder govern well, and what exceeds any one person's span of control no matter how capable the tools? Where is the line between capability and overreach?
Reads across the three platforms for the point past which a solo builder is shipping systems they cannot responsibly oversee. SpaceForge's recursive case — built by the method that the program also teaches, with the program itself becoming a data site — adds a layer the other two cannot. The cohort-program question (whether the Masters cohort actually produces cross-disciplinary leaders for a post-democratization workforce) lives here as a subordinate strand: the headline of the PhD is governance of the builder, not the program.
The twelve schools
Every school's involvement, by module.
No school appears as a guest only. Backbone schools (W.P. Carey, SESE, Fulton, SFIS, Law) appear in two or three modules; specialist schools appear in one with deep ownership. Every module is a triad.
| ASU School | Modules where involved | Count |
|---|---|---|
| Ira A. Fulton Schools of Engineering | Q1, Q5 | 2 |
| W.P. Carey School of Business | Q1, Q3, Q5 | 3 |
| School of Computing and Augmented Intelligence (SCAI) | Q2 | 1 |
| School of Earth and Space Exploration (SESE) | Q2, Q4 | 2 |
| Walter Cronkite School of Journalism and Mass Communication | Q2 | 1 |
| Sandra Day O'Connor College of Law | Q3, Q6 | 2 |
| Thunderbird School of Global Management | Q3 | 1 |
| School of Sustainability | Q4 | 1 |
| School of Politics & Global Studies | Q4 | 1 |
| College of Health Solutions | Q5 | 1 |
| Herberger Institute for Design and the Arts | Q6 | 1 |
| School for the Future of Innovation in Society (SFIS) | Q1, Q6 | 2 |
The six classes
Six modules. Three ASU schools per class. One reason each trio is in the room.
Q1
How Space Programs Actually Get Built
- Lead — W.P. Carey School of Business
- Guest — Ira A. Fulton Schools of Engineering
- Guest — School for the Future of Innovation in Society (SFIS)
Why this trio
Engineers design what's technically possible. Business school sets what's financially viable. SFIS reveals what's organizationally survivable. Programs die at the seams between all three.
Q2
Sensor to Signal: Turning Data Into Decision
- Lead — School of Computing and Augmented Intelligence (SCAI)
- Guest — School of Earth and Space Exploration (SESE)
- Guest — Walter Cronkite School of Journalism and Mass Communication
Why this trio
Scientists know what the sensor measures. AI scales it. Cronkite knows how to translate it into a decision-maker's language — the gap that kills 80% of intelligence products before they reach a desk.
Q3
New-Space Ventures: Building, Funding, Surviving
- Lead — W.P. Carey School of Business
- Guest — Sandra Day O'Connor College of Law
- Guest — Thunderbird School of Global Management
Why this trio
Founders die from three causes: capital structure, regulatory misstep, or losing the international market. Each school is one of those failure modes. Putting all three in the room is the only way the student sees all three coming.
Q4
What's Worth Doing in Space?
- Lead — School of Earth and Space Exploration (SESE)
- Guest — School of Sustainability
- Guest — School of Politics & Global Studies
Why this trio
"Worth doing" requires scientific merit + long-horizon ethics + the political coalition that makes it actually fundable. Remove any one and the question collapses into the wrong answer.
Q5
When Things Go Wrong: Operations Under Pressure
- Lead — Ira A. Fulton Schools of Engineering
- Guest — W.P. Carey School of Business — Supply Chain Management
- Guest — College of Health Solutions
Why this trio
When a mission anomaly cascades, the cause is almost never purely technical. It's a supply-chain delay → a hardware compromise → a crew-fatigue decision. All three are co-causal — and no single-discipline program puts all three in the room.
Q6
Designing the Rulebook: Policy as Engineering
- Lead — School for the Future of Innovation in Society (SFIS)
- Guest — Herberger Institute for Design and the Arts
- Guest — Sandra Day O'Connor College of Law
Why this trio
Policy is engineering — but most policy schools don't prototype. Herberger brings design-thinking and simulation. SFIS brings the systems frame. Law writes the statute that has to survive the first commercial dispute.
Want the working artifacts?
The classroom protocol, the card-deck templates, the REFLEXION-journal schema, the AI-substrate access for faculty preparation — all available on request as part of the JV briefing.