Latest from peeragogy chatbot

I've written about peeragogy and a new culture of learning and more recently about the peeragogy chatbot. Here is the latest report from the peeragogy community, which still meets weekly 14 years after I started it.

Dear Howard,

I wanted to share that we've made quite a lot of surprising progress inspired by your question. A key step was noticing that it is possible to get ChatGPT to run simulated multi-agent workshops.  The first one I set up was around the set of discussion prompts copied below.  

We used the same idea in a new paper:

https://arxiv.org/abs/2506.09696

My algorithm for producing the questions below: ask ChatGPT to take a draft paper and turn it into a set of writing prompts; and, again, to turn these prompts into peeragogical exercises.  I'd previously asked ChatGPT to read the Peeragogy Handbook.  So, the questions below are specific to the topic I was looking at, but this method can be repeated for any topic!

Joe

🌀 Peeragogical Exercises for a Seminar on Epistemic AI (30 Prompts)

🌱 Phase 1: Foundations of Inquiry

• Build a Collective Glossary
  Each student proposes one term related to epistemic AI and one unexpected term from another domain. Define both and discuss overlaps.

• Who Are We Learning With?
  Interview an AI system (e.g. ChatGPT) about a complex concept. Compare transcripts in groups: what kind of peer is the AI?

• Design a “Worst Case” AI System
  In small teams, sketch an AI that fails at supporting collaborative inquiry. What design patterns led it there?

• Map Our Unknowns
  As a group, build a shared uncertainty map: What do we not know yet about AI and epistemology?

• Construct a Pattern in the Wild
  Go into your own disciplinary practice (lab, studio, workplace) and identify a peer-learning pattern at play. Share and codify it.

• Trace the Life of a Question
  Take one “big question” from class. Track how it evolves over three weeks across discussions, documents, and AI queries.

• Draft a Proto-Manifesto
  In trios, write three versions of: What should AI be like if it is to support human flourishing in knowledge work?

• Reverse Engineer a Research Culture
  Choose a field (math, art, social science). How does it teach newcomers to think? Could AI systems model that onboarding?

• Build a Collaborative Reading List
  Select 3–5 readings as a group. Make a system for how new texts are added based on inquiry rather than syllabus.

• Design a Rotating Facilitation Protocol
  Develop (and revise!) a structure for rotating leadership in weekly discussions that encourages emergence.

🔍 Phase 2: Mid-Semester Meta-Design

• Develop a Peer Review Protocol
  Invent your own way of giving meaningful feedback on experimental ideas. Try it in pairs.

• Run a Scenario Workshop
  Imagine a future where your collective project becomes real. Role-play users, funders, skeptics.

• Write a Pattern Language Seed
  In subgroups, name and sketch 2–3 emergent design patterns from your own collaboration so far.

• Prototype an Epistemic Interface
  Imagine a dashboard, workspace, or social tool that helps people work with ambiguity. Draw, describe, or code it.

• Explore a Failure Together
  As a group, choose one shared failure in the course (a broken tool, a boring session). Analyze it like a system.

• Ask a Better Question
  Identify a question from early in the course that no longer serves you. Transform it into a more generative version.

• Design for Diverse Knowing
  Explore how AI systems might work differently for users with radically different epistemic traditions or contexts.

• Hold a Midpoint Retrospective
  Facilitate a meta-discussion: What’s working? What’s lagging? What wants to change in the learning ecology?

• Propose a Change to the Curriculum
  Each student gets one “curricular hack” to propose and trial. Debrief what emerges.

• Audit an Existing AI Tool
  Choose a current AI application. As a team, audit its epistemic posture: What does it assume about knowledge?

🚀 Phase 3: Toward Open-Ended Futures

• Start a Collective Inquiry Thread
  Pick one inquiry theme and follow it across different modalities (chat, essay, code, fieldwork) for the rest of the semester.

• Develop a Co-Writing Protocol with AI
  Design a fair, experimental method for collaborative writing with LLMs. Publish one short piece.

• Create a Public Artifact
  Produce a shared output (zine, wiki, podcast, pattern library) for external readers interested in epistemic AI.

• Reflect on a Moment of Not-Knowing
  Share one moment when you felt stuck or disoriented in the course. What did you learn about inquiry?

• Interview Each Other as Researchers
  Conduct peer interviews on evolving research identities. Reflect on how each person’s epistemology has shifted.

• Forecast an Ethical Dilemma
  Imagine your future AI system becomes widely adopted. What new ethical dilemma does it create? Explore it deeply.

• Curate a Gallery of Partial Ideas
  Share unfinished, strange, or “in-progress” ideas. What makes something worth sharing before it’s ready?

• Name Your System
  If your semester’s work produced an AI system or protocol, what would you call it—and what does that name imply?

• Build a Meta-Pattern Language
  Map the peeragogical moves you used to shape this course. What did you invent about learning itself?

• Disperse the Inquiry
  Final project: each student plants a seed from this course into another community they belong to. Come back with stories.