EdTech Connect | Episode Summary

Episode Title:

Dr. Cabrini Pak Breaks Down the Future of AI in Academia

Date:

July 11, 2025

Host:

Jeff Dillon

Guest:

Dr. Cabrini Pak, Professor, Bush School of Business at Catholic University of America

Brief Overview

In this episode, Dr. Cabrini Pak joins EdTech Connect to discuss the transformative potential of artificial intelligence (AI) in higher education. Drawing on her extensive, interdisciplinary expertise—spanning biology, business, theology, and culture—Dr. Pak explores practical applications, emerging challenges, and philosophical considerations of AI. The conversation ranges from classroom-tested use cases and error rates, to the promise and perils of agentic AI bots, ending with her vision for a radically more adaptive and responsive university ecosystem.

Key Discussion Points & Insights

1. Dr. Pak's Interdisciplinary Background and Approach (02:23–05:03)

• Dr. Pak holds degrees in biology, management information systems, theology, and religion & culture.
• Her passion: Connecting insights across disciplines to fully understand the human experience and improve educational and business processes.
  • “Technology has a secret infatuation with nature...just look at how we've named things in technology, a lot is inspired by nature.” — Dr. Pak (03:43)
• She draws from fields as diverse as embryology, prenatal psychology, and philosophy to inform her perspective.

2. Explaining Stigmergy and Its Classroom Application (05:03–07:08)

• Stigmergy: A coordination system inspired by social insects (ants, bees), involving action-trace feedback loops, as adopted from cybernetics.

  • Action leaves a trace; traces trigger further actions, forming a feedback loop that fosters complex collaboration (05:03).

• She integrates stigmergic feedback loops in her classes to enhance teamwork and adaptive learning.

  “You have action, trace, medium and agent—those four elements can be used to design really powerful coordination mechanisms.” — Dr. Pak (06:05)

3. AI in Higher Ed: The Living Lab Mindset (07:08–09:49)

• Dr. Pak advocates for a "Living Lab" model: treating classrooms as experimental spaces to test and iterate technologies directly with students.
  • Example: Her students tested the Zoom AI assistant’s meeting summaries for quality and accuracy, uncovering real-world limitations (misgendering, focus errors).
  • AI is positioned as “a very fancy hammer”—a practical tool rather than an end in itself (10:30).

4. Faculty Adoption, Literacy, and Resistance (09:49–12:38)

• Faculty attitudes towards AI vary widely.
  • Anxiety, lack of digital literacy, and privacy concerns are significant barriers.
  • Practical workaround: Combine written and oral exams to discourage AI-enabled cheating.

5. The Advent of Agentic AI in Academia (12:38–15:21)

• Agentic AI (bots that can independently act on tasks) is already entering university life—virtual TAs, code helpers, advising assistants.

  • Application examples: Essay feedback, coding help, course planning, workflow automation.

• Cautions about hasty implementation:

  • System compatibility, data security, and operational risk need careful review.

  “I have a little bit of a healthy suspicion right now of even using [agentic AI] on a native platform to my environment.” — Dr. Pak (14:56)

6. Automation and Practical AI Opportunities (16:50–19:19)

• AI bots can relieve mundane campus tasks:
  • Example: Website maintenance (cleaning dead links, updating content) likened to “automatic vacuums” for digital spaces (17:01).
• Cautions on jargon, context, and verification—AI suggestions should have human review before being implemented.
• Importance of unbiased bot assistance for improved efficiency and relevance.

7. Tracking and Addressing AI Error Rates (19:40–23:15)

• Dr. Pak underscores the critical need to monitor and analyze AI error rates, citing real-world algorithmic bias cases (MIT, facial recognition).

  “Unless you’re actively testing the error rate, you’re not going to find ways to improve the training protocols.” — Dr. Pak (20:23)

• Acceptable error rates vary by application; critical tasks require near-perfect accuracy, while lower-stakes workflows (like course advising) can tolerate more mistakes if humans retain oversight.

8. Student Advising: AI as an Accelerator, Not Replacement (23:22–24:45)

• AI bots can automate repetitive advising tasks (degree requirements, scheduling), enabling human advisors to spend more time on meaningful guidance.
  • “A machine never loses its patience...once they come back to me...now we can have a deeper conversation.” — Dr. Pak (24:30)
• The aim: AI handles 85% routine work, humans focus on nuanced, supportive engagement.

9. Lessons from Corporate Practice & the Living Lab Mindset (25:06–26:16)

• Higher ed can gain from the private sector’s agility by adopting a living lab, design-thinking approach to technology and process improvement.
  • Promotes a “circular knowledge economy” engaging all campus stakeholders.

10. Underrated AI Use Case: Unlocking Institutional Tacit Knowledge (26:24–27:26)

• AI could bridge silos by connecting university data hidden across isolated drives and teams, unlocking valuable, underutilized “tacit knowledge.”
• Big assumption: Data quality and access must be addressed for this to succeed.

11. A Vision for AI in Higher Ed Bureaucracy (27:36–29:06)

• Dr. Pak’s dream: Deploy AI to break up entrenched, inefficient bureaucracy—what she calls the “constipated dinosaur”—making university operations agile, adaptive, and user-focused.
  • Professional development for staff should not be neglected, as they are “the glue that holds everything together.”

  “If I had a magic wand...it would be something to transform those constipated dinosaurs into agile, whip smart adaptive service providers that not only anticipate the road ahead, but helps its users get to their destination the most effective way.” — Dr. Pak (28:00)

Notable Quotes & Memorable Moments

• On Nature and Technology:
  “Technology has a secret infatuation with nature.” — Dr. Pak (03:43)
• On AI in the Classroom:
  “For us, AI is like a hammer. It's a very fancy hammer, but it's a hammer. It's a tool.” — Dr. Pak (10:30)
• On Agentic AI Risks:
  “You run the risk of some system failures, potential breaches, poisoning your data.” — Dr. Pak (14:40)
• On Error Rates:
  “Unless you’re actively testing the error rate, you’re not going to find ways to improve the training protocols.” — Dr. Pak (20:23)
• On Bureaucracy:
  “I have two words for a typical university bureaucracy—a constipated dinosaur.” — Dr. Pak (27:36)
• On Her AI Magic Wand:
  “Turn those constipated dinosaurs into agile, whip smart adaptive service providers...” — Dr. Pak (28:00)

Timestamps for Key Segments

| Segment | Topic | Timestamp | |---|---|---| | Dr. Pak’s Background | Interdisciplinary mindset | 02:23–05:03 | | Stigmergy | Coordination & learning loops | 05:03–07:08 | | Living Lab Model | Experimental approaches to AI | 07:08–09:49 | | AI Adoption Challenges | Faculty motivation, digital literacy | 09:49–12:38 | | Agentic AI | Virtual TAs, code bots, risks | 12:38–15:21 | | Automation | Site maintenance, practical uses | 16:50–19:19 | | AI Error Rates | Importance & impact | 19:40–23:15 | | Advising Bots | Use case, human/AI collaboration | 23:22–24:45 | | Lessons from Corporate | Living labs, agility | 25:06–26:16 | | Underrated Use Case | Institutional tacit knowledge | 26:24–27:26 | | AI-Driven Bureaucracy | Future vision, magic wand | 27:36–29:06 |

Summary Takeaways

• AI should be approached as a flexible tool, not a panacea, with careful focus on error rates, privacy, and literacy.
• Agentic AI is already impacting higher ed, but trust, oversight, and systemic readiness remain issues.
• Living Lab approaches—treating institutions as spaces for experimentation and iterative learning—can unlock innovation and rapid adaptation.
• Streamlining tedious administrative processes with AI (e.g., advising, web maintenance) frees up human expertise for more complex, value-add tasks.
• Universities must break down data silos and attend to staff development to truly realize AI’s promise.
• There is both playfulness and urgency in Dr. Pak’s vision: higher education is at an inflection point, and those who embrace agile experimentation will thrive.