As educators who know our students and our teaching and assessment contexts, we are uniquely positioned to know what menu items to recommend. For any one assessment (and student, for that matter), there will be ways to use AI and tools that are more suitable or less suitable. Sometimes it will be more productive and responsible for students to use particular AI tools to engage with literature and edit their text. Other times it will be more suitable for students to other AI tools to provoke reflection, draft a structure, and then provide feedback on text. Want to order five desserts? Maybe not for this assessment, but it’s a great idea for the next one.

Change happens when something starts or stops, or when something that used to happen in one way starts happening in another. It happens at a particular time, or in several stages at different times. Organizational change is structural, economic, technological, or demographic, and it can be planned and managed on a more or less rational model...

Transition, on the other hand, is a three-part psychological process that extends over a long period of time and cannot be planned or managed by the same rational formulae that work with change.

The history of science and technology is filled with cases of rejected innovations coming from outside the knowledge boundaries of a given professional community: from germ theory by Pasteur (a chemist, not a medical professional) to the theory of energy conservation by Helmholtz (a physician, not a physicist) to principles of inheritance from Mendel (an amateur biologist who relied on mathematics) (Barber, 1961; Campanario, 2009). Knowledge is embedded and invested in R&D professionals’ work processes and practices and therefore is “at stake” for those actors who have developed it (Carlile, 2002). Moreover, open innovation platforms attract individuals on the margins of knowledge boundaries either socially or knowledge-wise (Jeppesen and Lakhani, 2010), posing a threat to specialized knowledge developed in an organization or a discipline. Therefore, beyond the notorious not-invented-here syndrome (Allen, 1977; Katz and Allen, 1982), R&D professionals have unique reasons for rejecting knowledge from open innovation platforms.

Reform efforts are generally associated with large expenditures of time, effort, and money devoted to addressing three contexts of reform: the background of the teacher, the cultural and structural contexts of instruction, and teacher thinking. But, despite substantial work and investment, rarely do reform efforts at either the K–12 or the college level result in sustained,

fundamental changes in classroom practice. Why? Woodbury and Gess-Newsome (2002) propose three potential explanations for the paradox of “change without difference”:

1. Structural and cultural contexts. A variety of interconnected structural and cultural components of school systems must change to support and sustain instructional changes.

2. Purposes of reform. Many reforms are not intended to alter common pedagogies, leaving teachers’ and students’ roles untouched.

3. Teacher thinking. Teachers’ knowledge and beliefs mediate reform proposals. Because reforms seldom alter important aspects of teacher thinking, reform enactment remains remarkably traditional.

Our study (IRB#2035233-1) examines how instructors navigate the impacts of Generative AI on college writing as they adapt to technological paradigm shift. Through surveys and focus groups involving 178 faculty across 16 institutions, we document how AI changes the writing they assign and the extent to which they welcome or resist these changes. Our key finding: instructional choices are shaped by powerful, but seemingly unconscious (Turel and Kalhan, 2023), assumptions about how humans and machines will coexist. Faculty adopting a substitution paradigm (“If AI can do summaries, students won't”) make different choices from those adopting a collaborative paradigm (“AI is good at summaries; so, use AI to coach students to summarize”). While AI-pedagogy scholarship recognizes that teaching and learning are inherently relational and that interpersonal trust networks been fundamentally disrupted by the emergence of Generative AI (Blackwell-Starnes, 2025; Luo, 2024; Ryan 2025; Winthrop, 2025), it is less aware that we and our students also exist in relation to AI itself. By mapping out these relational models, we make visible the full spectrum of human-AI possibilities, while redirecting pedagogical attention toward the underlying frameworks where choices are crystallized.