Modularity · the architecture's stability seen from inside

The main document's first curriculum-design principle, modularity, names the structural property that lets independent course modules be reorganized, replaced, and recombined as the surrounding technology and audience change. This section elaborates the principle from the practitioner-pioneer's reflexive position. What I add is the autoethnographic record of what stayed the same across four iterations and eight contexts when the surface changed, and the analytic naming of the underlying pattern as compression-as-curriculum-maturation.

D.2.1 The architecture as the structural form of modularity

The four-theme architecture (Education, Industry, Ethics, Accessibility) is the structural form modularity takes in this curriculum. I built the themes into the course from Iteration 1 onward (DK-1.FG slides 5-9; FP-1's nine reflection questions distributed across the four themes; see also §A.4). Each theme operates as an independent module: any one theme could be taught without the others; the themes can be reordered; the tool-and-content list within each theme can be replaced as the field changes. Across the four iterations the architecture stayed intact while the surrounding curriculum shifted.

This is the autoethnographic observation that supplements the DBR principle. The DBR analysis says modularity emerged as a useful curriculum-design property; the autoethnographic supplement adds that the specific architecture that operationalized modularity (the four themes) is what carried the iterations through tool turnover and format compression.

D.2.2 What the corpus shows about architectural stability

The architecture held across four configurations:

Beyond the four iterations, the architecture appears in the HCI summer 2024 guest series (HC corpus: ten decks named by theme — HC-EDU, HC-INDUSTRY, HC-ACCESS, HC-AUDIO with Ethics, etc.) and in the public-facing channels (Keep Up Newsletter episodes organized by theme cluster; CU RMACC webinar organized around the four themes; see §D.3 for the multi-context elaboration).

D.2.3 The sub-claim · compression-as-curriculum-maturation

The autoethnographic analysis surfaces a pattern within modularity that I name compression-as-curriculum-maturation. The pattern is that the same modular architecture compressed approximately six-fold in delivery time (from 15 weeks to 5 days) while preserving its structural form. The compression is not content loss; it is content distillation that became possible because the architecture itself was modular and stable.

The diagnostic for the pattern is architectural stability across compression in delivery time. Where curriculum compression in technology fields is sometimes described as expansion (growing into multi-course sequences) or loss (shedding content under time pressure), my iterations exhibit a third pattern: the architecture persisted at higher pedagogical density while the surrounding scaffolding adapted to the new format.

D.2.4 The format confound I take seriously

Iterations 3 and 4 were not only compressed versions of the semester course. They were also a distinct delivery format: a free public online workshop hosted on Luma and partnered with GenAI Works, delivered through YouTube to an audience that registered for a five-day commitment. The compression from 15 weeks to 5 days is therefore the joint product of two forces: (a) what I had learned to identify as essential through two full-length iterations, and (b) what the online-workshop format itself requires (online learners commit to short, intense windows; the platform supports 5-day rather than 15-week structures; the GenAI Works partnership was framed as a workshop from the outset).

A reader who attributes the compression entirely to instructor maturation would overstate what (a) accounts for; a reader who attributes it entirely to format constraints would understate what the two semester iterations made possible. My claim is the joint one: the format made compression possible in a way it would not have been without external constraints, and the two full-length iterations made the compression successful (the architecture held, the topics persisted, the satisfaction ratings were high) in a way it would not have been from a workshop-first starting point.

D.2.5 Why this pattern emerges in pioneer practice

The pattern emerges because pioneer practice has a learning curve that non-pioneer practice does not. A pioneer instructor in a fast-moving technology field cannot consult a literature on what is essential to teach, because the literature does not yet exist. The pioneer learns what is essential through full-length iterations that surface what students engage with, what tools persist beyond a season, and what conceptual frameworks travel across topics. Once the pioneer has accumulated that knowledge, compression to a short format becomes possible because the pioneer knows what to keep.

This is why my workshops succeeded the semester courses in the timeline rather than preceding them. Iteration 1 was the discovery phase; I did not know in advance which themes, tools, and assignments would land. Iteration 2 was the consolidation phase; I tested the framework against a new student population and refined the module placement, the tool list, and the guest-speaker roster. Iterations 3 and 4 were the distillation phase; with the framework stabilized, I compressed to the workshop format without losing what mattered.

A non-pioneer instructor adopting an established curriculum would not need to do this. The accumulated literature would tell them what to keep and what to drop. The pioneer's compression is therefore not just a smaller version of the semester course; it is the outcome of the pioneer's two-semester learning process.

D.2.6 Theoretical contribution

The literature on iterative curriculum design (McKenney and Reeves 2018) treats iteration as a method for refining design conjectures through structured cycles. The pattern this elaboration identifies is consistent with that literature in the sense that iteration was the mechanism through which the curriculum matured. The elaboration adds two specifics that the iterative-design literature does not commonly name.

First, the maturation pattern is compression rather than expansion. The pioneer's iterative learning produced a shorter, denser delivery, not a longer or more elaborated one. This contrasts with how curriculum maturation is sometimes described in technology fields, where successful curricula are said to grow into multi-course sequences. Both patterns are possible; mine is the compression pattern.

Second, the compression preserves architecture rather than content alone. The themes were stable across all four iterations; the tools changed across the iterations; the duration compressed; the architecture survived. This suggests that what pioneer instructors learn through iteration is the architecture, not the content list. The content list is replaceable, and is replaced as new tools emerge. The architecture is the durable pedagogical contribution.

The elaboration contributes the claim that compression-as-curriculum-maturation is a documentable pattern in pioneer instructor practice at fast-moving technology fields, and that the diagnostic for the pattern is architectural stability across compression in delivery time. The claim sits as a sub-claim within the modularity principle the main document develops.

D.2.7 Guidelines for instructors entering fast-moving technology fields

The compression pattern documented in this elaboration is not a recipe; it is a posture that other pioneer instructors of emerging-technology curricula may find useful. Four guidelines drawn from the cross-iteration record:

• Plan to run the curriculum at full length before compressing it. The pioneer's first iteration is the discovery phase, and Iteration 2 is the consolidation phase (§C.3.8). Compression to a workshop format becomes possible only after architectural stability emerges through full-length delivery. Starting with a compressed format risks omitting what the longer iterations would have surfaced as essential.
• Treat the conceptual architecture as durable; treat the tool list as disposable. Themes stay. Tools turn over. The Midjourney Self-Portrait Assignment recurred across all four iterations under three generations of image-generation tools. The architecture absorbed the turnover; an architecture organized around specific tools would not have.
• When compressing, drop the lecture-discussion-lab-assignment loop, not the topics. The topics are what learners came for. The pacing structure is what semester length supports and workshop length does not. In the workshops, live demonstration takes the explanatory load that lecture-discussion cycles carried in the semesters.
• Expect roughly a six-to-one compression ratio. Iterations 3 and 4 documented this ratio from the fifteen-week semester to the five-day workshop. Pioneer instructors planning similar moves can use this as a rough planning anchor, recognizing that the ratio is empirical to my case rather than universal.