5.1: Building Faculty Capacity

Generative AI did not arrive at College of the Siskiyous as a planned initiative. It arrived in screenshots and links, in hallway conversations and worried emails. One moment caught the range of reactions. During an Academic Senate discussion, a senior colleague turned to me and said, “On the moral scale, I rank AI below Satan.” The hyperbole made me laugh. The conviction underneath was serious.

At the same time, I was watching students use AI as a lifeline. Usually late at night, often alone, trying to make sense of assignments they did not feel ready for. That tension, between a genuine threat and a genuine opportunity, shaped how our professional development had to work.

What Did Not Work: Self-Paced Learning Alone

Our first design borrowed from a familiar playbook: self-paced learning plus a tangible product. Faculty would finish an online course over the summer, build AI-aware materials in the fall, and share at the end of the semester. The appeal was obvious. Flexibility, concrete outputs, alignment with patterns we already used.

Three things pushed us to redesign. Faculty time and role overload, since at a small rural college people wear many hats, and a substantial online course read less like an opportunity than an extra unpaid job. Uneven comfort and strong moral positions, since telling a colleague who sees any AI use as compromised to go complete an “AI integration” course alone was never going to move the needle. And rural isolation.

What Worked: Community of Practice

We shifted toward a community of practice. Self-paced learning stayed useful as background, but the center of gravity moved to shared, synchronous, relationship-rich work.

• Anchor the year in shared sessions. Live sessions with outside speakers gave faculty a shared experience and concrete examples rooted in our context.
• Build a standing group, not a workshop series. Regular meetings focused on what happened when I tried this, not abstract best practices.
• Ensure cross-disciplinary participation. We recruited from English, Biology, Paramedic, Fire Technology, Administration of Justice, Library, Early Childhood Education, and more.
• Tie incentives to engagement, not perfection.

Ground Rules That Mattered

Psychological safety. People brought half-formed ideas, failed experiments, and open questions to the table. The portfolio-grading script that posted a zero to every student when an API call failed. The assignment redesign that worked in theory and confused everyone in practice. No one got shamed for not knowing, for changing their mind, or for deciding partway through that AI did not belong in their course after all. That permission to fail in front of colleagues is most of what made people willing to try anything.

Honoring skeptics. The colleague who ranked AI below Satan kept coming to meetings. So did others with serious reservations, and we treated their objections as contributions rather than obstacles to manage. They asked the questions the enthusiasts skipped. What happens to a student who never learns to struggle through a hard reading? Who owns the data we are feeding these tools? Is the environmental cost defensible in a region shaped by fire and drought? Faculty who did not plan to use AI at all turned out to be essential to the conversations about policy, equity, and academic integrity. The skeptics made the rest of us more honest.

Shared responsibility. The work was collective. No single instructor or discipline was on the hook for covering AI, and no one was the designated expert. That mattered at a college where most of us were learning this in the margins of an already full week.