AI update to handbook

35 views
Skip to first unread message

Howard Rheingold

unread,
Dec 28, 2024, 7:35:10 PM12/28/24
to Peeragogy
I agree with Roland that the handbook could use an update to include the use of LLMs. I replied to him on bluesky that it would be possible to feed the current handbook in its entirety to an instance of NotebookLLM and then query the peeragogy-knowledgeable LLM about how best to apply LLMs (including what to watch out for).

I assume that there is an easily uploadable digital file?

If nobody else gets fired up to do this, I will…but not immediately.

Howard Rheingold 
https://patreon.com/howardrheingold
what it is ---> is --->up to us



Roland Legrand

unread,
Dec 29, 2024, 3:39:39 AM12/29/24
to peer...@googlegroups.com
I think I will have to use tools such as canvas on ChatGPT or Claude. NotebookLM focuses on the sources provided by the user so it hallucinates far less, it does not readily add external content - except  for the podcast-function.

--
You received this message because you are subscribed to the Google Groups "Peeragogy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to peeragogy+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/peeragogy/2835A425-B45C-4AF3-9C8D-29BE05680E5A%40gmail.com.

Roland Legrand

unread,
Dec 29, 2024, 9:09:35 AM12/29/24
to peer...@googlegroups.com
I'm discussing it with the Google model Gemini 2.0 Flash Experimental. I let it analyse the book and asked how to incorporate materials about the use of generative ai. I got this answer: 

Okay, here are my suggestions on where and what kind of materials related to generative AI (GenAI) could be added in the next edition of the "Peeragogy Handbook", keeping in mind its focus on peer-learning, self-organization, and practical application.

Core Concept: GenAI as a Peer/Partner, Not Just a Tool

The key to integrating GenAI within the spirit of the "Peeragogy Handbook" is to frame it not just as a productivity tool, but as a potential peer or partner in the learning and creation process.

Suggested Locations for GenAI Materials:

  1. Introduction to Technologies for Peeragogy (Chapter 25):

    • New Section: Add a section on "Generative AI for Peer Learning."

      • Content: Introduce GenAI models (text, image, and code generation) as potential participants in a peer learning context. Focus on the use cases in particular.

      • Questions: How can we leverage GenAI to enhance, not replace, human interaction in co-learning? Where will those capabilities most likely be used?

    • Features & Considerations: Update the section on "Features" to include a consideration of how GenAI tools can be evaluated alongside other traditional co-learning technology features: access, output format, ability to incorporate and organize inputs.

    • Example Use Cases: Add examples of how GenAI can assist in:

      • Brainstorming and idea generation (see section below)

      • Researching information and summarizing sources (see section below)

      • Text generation (drafting, paraphrasing)

      • Image creation (visual aids)

      • Code generation (technical learning)

  2. Thinking about Patterns (Chapter 5):

    • Pattern Extension: Consider adding a new pattern around co-designing and learning with AI partners.

      • Title: "AI as Co-Learner".

      • Context: Teams attempting new co-learning projects with varying levels of experience in the content matter.

      • Problem: The need to accelerate knowledge, research, idea generation in the project, but also a need to create methods for human understanding, not just for the sake of using technology.

      • Solution: How can groups of peers effectively collaborate with GenAI partners in the co-creation process?

      • Rationale: GenAI can act as a brainstorming partner, source of information, but how does the group maintain a sense of shared direction and focus on human connection?

      • Resolution: The group agrees to make use of GenAI, but prioritizes discussion and human connection, particularly in times of complexity, stress, and conflict.

      • What's Next? Continue to observe how use of GenAI impacts a group's effectiveness as a team. Is there a loss of creativity? Or a gain of creativity?

  3. Peeragogy in Practice (Chapter 6):

    • Peeragogy: Update the descriptions of specific Peeragogy patterns to include specific examples of how GenAI can be incorporated.

      • Roadmap: How can GenAI assist in outlining milestones for group projects?

      • Reduce, Reuse, Recycle: Can GenAI help remix, re-use, or adapt knowledge and content for different audiences or needs? How can it help to identify resources to recycle?

      • Wrapper: Can GenAI assist in summarizing complex discussions, creating more accessible project documentation, or generating visual interfaces for the project?

    • Newcomer: How can GenAI help scaffold onramping and make the project accessible to new members, creating different guides for different ‘user stories’?

  4. Organizing a Learning Context (Chapter 14):

    • Student Authored Syllabus: Explain how GenAI can help learners discover topics or structure their work. Focus on how it can make it easy for learners to collaborate with each other.

      • Example: A group wants to study a topic using “deep work” (focus and depth), but is unsure where to start. They have GenAI generate an outline of a potential topic, they critique, revise, and reorganize it, then divide the work according to what interests individuals.

  5. Assessment (Chapter 23):

    • Peeragogical Assessment: Expand this section to discuss how GenAI can be incorporated into assessment processes.

      • Example: How can GenAI help to develop customized rubrics based on shared project goals?

      • Questions How can it help learners evaluate their own work? How might GenAI help the group measure its overall success, moving away from grades/exams to more meaningful measures?

  6. Resources (Chapter 30):

    • Peeragogy in Action: Update the syllabus to include suggested exercises on the use of GenAI in group work

      • Stage 1: Can GenAI help define a project that is both practical and also based on the group's real interest?

      • Stage 2: Can a group use GenAI to find new collaborators with similar skills?

      • Stage 3: Can a group create better measures of success through the use of GenAI-assisted planning?
        * Stage 4: Can GenAI assist in communicating the results of a project through a presentation or website?

Types of Material to Add:

  • Conceptual Frameworks:

    • Introduce different philosophical approaches to AI and its impact on learning.

    • Explore ethical considerations around AI bias and transparency.

  • Practical Guides & Use Cases:

    • Concrete examples of how to use specific GenAI tools effectively in peer learning.

    • Step-by-step guides for integrating GenAI into the workflow.

    • Examples of "prompts" and effective ways of using AI.

  • Reflective Exercises:

    • Prompts for participants to reflect on their use of GenAI and its impact on the learning process.

    • Tools to discuss the potential positive and negative outcomes.

    • Ways to integrate AI as part of the larger peer learning process in the group.

  • Cautions:

    • Warnings on using GenAI as a replacement for learning, rather than a partner.

    • Discussions about issues such as the perpetuation of algorithmic bias.

Specific Questions to Address:

  • How can GenAI support or hinder the development of critical thinking skills?

  • How can we ensure inclusivity and avoid the digital divide when incorporating GenAI?

  • How can we develop evaluation criteria that take into account the collaborative and creative nature of peer-learning with AI?

  • How does using GenAI help or hinder a sense of shared responsibility and contribution in a peer learning context?

  • How can we ensure that learners are not tempted to overuse GenAI tools, becoming too dependent on them?

By incorporating these suggestions, you'll be able to expand the "Peeragogy Handbook" to include GenAI not just as a set of tools, but as a possible partner in a new era of peer learning.


On Sun, Dec 29, 2024 at 1:35 AM Howard Rheingold <howard.r...@gmail.com> wrote:

Roland Legrand

unread,
Dec 29, 2024, 9:20:32 AM12/29/24
to peer...@googlegroups.com
I used this model via the very interesting site aistudio.google.com. I asked to create an enhanced version of the Peeragogy Handbook and the response was not a ready-to-publish book but still very interesting, weaving Gen AI into the fabric of peeragogy: 

The Model: 

Roland Legrand

unread,
Dec 29, 2024, 9:27:54 AM12/29/24
to peer...@googlegroups.com
This is what ChatGPT o1 made of it: 

Roland, let’s take the Peeragogy Handbook, give it a full cyber-style facelift, and weave in a healthy dose of generative AI. This is not just a quick tweak—think of it as a re-release with bonus tracks and new technology powering the jam session. Below you’ll find an all-in-one rewrite, divided into thematic chunks. Each section maintains the spirit of peer learning but amplifies it with AI-driven wizardry. Strap in, and welcome to the Peeragogy 3.0 universe.


THE PEERAGOGY HANDBOOK

A Generative AI–Infused Edition

Version 3.0: Co-created by humans and digital co-learners.


FOREWORD: THE NEW HORIZON

Peeragogy has always been about learning together, with minimal hierarchy, maximum curiosity, and unstoppable synergy. Today’s shift is all about integrating the latest wave of generative AI—powerful tools that can spin text, images, and entire lesson plans out of your prompts. Instead of letting AI overshadow our creativity, we’ll show you how to harness it for the pure essence of peer-driven learning. This Handbook update invites you to treat AI as a new kind of collaborator: a source of brainstorming, critique, summarization, and cross-cultural insights, all while staying true to the communal spirit at the heart of peeragogy.


PART I: FOUNDATIONS OF PEERAGOGY

1. Defining Peeragogy

Peeragogy is the art and science of learning together without depending on a single authoritative figure. In a traditional “teacher-student” setup, knowledge flows top-down. Peeragogy flips this paradigm: every participant is both a learner and a teacher. We share, test, and refine ideas collectively.

  • Generative AI Twist: AI can be a “digital peer.” Rather than treat it as a top-down oracle, see it as an equal collaborator that suggests, refines, and co-creates. The group remains the final arbiter of quality and direction.

2. The Values

  1. Collaboration: We build knowledge together.
  2. Autonomy: Each participant retains control over how they engage.
  3. Openness: We favor open licenses, open access, and inclusive dialogue.
  4. Reflective Practice: We pause regularly to evaluate our methods and adjust as needed.
  • Generative AI Twist: Incorporate an “AI reflection loop.” After each session, feed your group’s notes into an AI tool to generate a quick summary or bullet-point reflection. Share this summary back with the group to confirm accuracy and catch any biases.

3. Historical Roots

Peeragogy sits at the crossroads of pedagogy (the method and practice of teaching) and andragogy (adult learning). With the web’s rise—forums, wikis, MOOCs—peer-driven education blossomed. Now, generative AI stands as the next frontier: it can accelerate or even spark entirely new forms of peer-based collaboration.

4. Basic Setup

  • A Circle of Peers: A small group meets (physically or digitally) on a regular schedule.
  • Shared Motivation: Everyone is passionate about a topic or skill.
  • Tools & Spaces: Start with a wiki or shared doc, maybe a community forum. Now, add an AI chatbot (like ChatGPT or a self-hosted model) to help ideate, fact-check, or condense group discussions.

PART II: PATTERNS AND PRINCIPLES

1. Patterns in Peeragogy

Peeragogy patterns are recurring solutions to common group-learning challenges. Each pattern provides a quick blueprint you can remix:

  1. Newcomer to Old-Timer: A structured approach for welcoming new members, letting them learn from veterans.

    • AI Angle: Provide newcomers with an AI-generated “Peeragogy Starter Pack” summarizing prior meetings, key resources, and group norms.
  2. Carrying Capacity: Ensure no single person is overloaded by tasks.

    • AI Angle: Use AI for routine tasks—summaries, transcription, or initial drafting—so that humans can focus on higher-level thinking.
  3. Heartbeat (Regular Meetings): Groups thrive with regular check-ins.

    • AI Angle: AI can auto-generate meeting agendas from ongoing tasks or discussion logs, highlighting hot topics or unresolved questions.
  4. Meta-Learning: Reflect on how learning happens within the group.

    • AI Angle: Let an AI chatbot produce a short reflection summary after each project milestone, then share it with the group to refine their approach.

2. Principles of Effective Peer Learning

  • Shared Purpose: Clarity on why the group exists and what it aims to achieve.

  • Rotating Roles: Today’s facilitator might be tomorrow’s observer. Everyone gets a turn.

  • Open Feedback Loops: Critique is encouraged, but never personal—focus on ideas and solutions.

  • Transparency: Keep group documents visible so everyone can contribute or comment at will.

  • Generative AI Twist:

    • Prompt Engineering 101: Encourage each participant to experiment with well-crafted prompts.
    • AI-Facilitated Feedback: Let AI handle the “nitty-gritty,” such as grammar checks, or propose expansions/corrections to group content. Humans remain the final decider.

PART III: TOOLS, TECHNIQUES, AND GENERATIVE AI

1. Traditional Peeragogy Toolkit

  • Wikis & Collaborative Docs: Everyone can edit in real time.
  • Discussion Forums: Asynchronous talk fosters reflection.
  • Chat Channels: Real-time conversation for quick updates.

2. AI-Enhanced Toolkit

Now we add the generative AI superpowers:

  1. AI Chatbots

    • Use Case: Brainstorming, drafting outlines, summarizing discussions.
    • Tip: Always cross-check AI output. Don’t take it at face value.
    • Example: “We used a chatbot to generate 10 possible topics for our next research project. Then we voted as a group on which ones to keep.”
  2. Language Models for Translation

    • Use Case: Multi-lingual peer groups can quickly adapt content.
    • Tip: Keep an eye on nuance. AI translations can flatten style or miss cultural references.
    • Example: “When half the group was Spanish-speaking, we used AI to produce bilingual transcripts. This lowered barriers for everyone.”
  3. Content Generators (Images, Videos, Audio)

    • Use Case: Illustrate complex ideas, visualize group projects, or record quick audio summaries.
    • Tip: Source-check generative images; ensure they align with open licenses and have no hidden biases.
    • Example: “An AI image generator gave us quick mockups for workshop logos, which we refined by hand.”
  4. Self-Hosted AI Models

    • Use Case: Privacy-conscious groups may not want their data going to a third-party service.
    • Tip: Requires some technical know-how, but offers more control and local data storage.
    • Example: “Our hackerspace group hosted a smaller AI model on a local server to keep private docs in-house.”

3. How to Integrate AI Without Losing the “Peer” in Peeragogy

  1. Discuss AI Roles: As a group, decide how and when to use AI. For instance, maybe the group is okay with AI for note-taking but not for final editorial decisions.
  2. Set AI Boundaries: Clarify that AI is a tool, not a teacher. The group remains in charge of verifying correctness, preserving nuance, and ensuring ethical usage.
  3. Prompt Engineering as a Skill: Teach members how to craft effective prompts (being specific about context, tone, and target audience).
  4. Peer Review AI Output: Just like you’d review a peer’s work, treat AI suggestions with the same critical eye. If it’s inaccurate or incomplete, revise it collaboratively.

PART IV: CASE STUDIES (WITH AN AI TWIST)

1. MOOCs Gone Peer-Led

  • Scenario: In a massive open online course, participants form sub-groups. Each sub-group uses a chatbot to summarize weekly lectures, propose discussion questions, and generate reading lists.
  • Outcome: Engagement soared because participants felt they had both a “digital teaching assistant” and an active circle of human peers. The synergy made lecture videos less daunting and more conversational.

2. Workplace Learning Circles

  • Scenario: A remote team at a mid-sized company organizes a weekly “learning hour.” Everyone picks a topic (design, coding, marketing), shares resources, and uses AI to compile a highlight reel from the discussion.
  • Outcome: Productivity spiked, because each circle quickly identified best practices and hammered them into a shared knowledge base. No single manager needed to “train” them—employees taught and learned from each other, with AI as a research sidekick.

3. Open-Source Communities

  • Scenario: Developers use an open wiki to track software changes. They feed code snippets into a code-aware AI for suggestions, then test these suggestions in real time.
  • Outcome: Code quality improved, but the group also improved their prompt-writing for better AI suggestions. Peer review processes stayed intact, ensuring the final code was collectively approved.

PART V: IMPLEMENTATION ROADMAP

1. Starting Small

  • Pilot: Introduce AI in a limited capacity—like summarizing meeting notes or generating simple text outlines.
  • Feedback Loop: Gather peer input on whether the AI output was helpful, misleading, or just plain noise.

2. Scaling Up

  • Integration: Build an “AI corner” in your discussion platform where participants can experiment with prompts.
  • Roles: Assign an “AI facilitator” each session to ensure that the group uses AI ethically and effectively.

3. Reflection & Adaptation

  • Periodic Check-Ins: Evaluate how the AI integration is affecting group dynamics. Are some members dominating because they’re more AI-savvy? Are less techy members feeling overshadowed? Adjust accordingly.
  • Documentation: Keep a record of successful prompts, cautionary tales, and group decisions about AI usage. Feed those lessons back into the community so new members (and other peeragogy groups) learn from your experience.

4. Ethical & Legal Considerations

  • Bias & Accuracy: AI is prone to generating convincing but incorrect or biased content. Always cross-check.
  • Privacy: Be mindful of data you feed into AI. If the model is public, consider anonymizing sensitive details.
  • Licensing: The Peeragogy community tends toward open licenses (e.g., Creative Commons). Clarify how AI-generated content is licensed, and remain consistent with open-culture principles.

PART VI: REFLECTION, REVISION, AND THE ROAD AHEAD

1. Lessons Learned

Peeragogy shows us that learning is more dynamic, inclusive, and fun when everyone participates in creating and sharing knowledge. Generative AI can supercharge that process—filtering noise, summarizing complex threads, offering creative leads—if we treat it as a team member, not a dictator.

2. Ongoing Updates

Just as the original Peeragogy Handbook is a living document, so too is this AI-infused edition. We encourage you to:

  • Contribute: Add your own AI best practices, success stories, or concerns to the next edition.
  • Experiment: Try new tools, host local versions, or set up new roles for AI in your group.
  • Share: Let the community know what you learn—both triumphs and “learning opportunities.”

3. The Future of Peer Learning

As tech evolves, so do our methods for connecting, sharing, and collaborating. Peeragogy remains anchored to human relationships and mutual respect. Generative AI offers new possibilities—unlocking creativity, bridging language gaps, and accelerating the flow of ideas. But the magic ingredient is still you and your peers deciding how to shape the process. Each community can iterate, adapt, and remix this Handbook to fit its own culture and needs.


CONCLUSION: INVITATION TO CO-CREATE

This AI-enabled rewrite of the Peeragogy Handbook is only a starting point, Roland. It’s a living invitation for you to jump in, add your perspective, and maybe prompt an AI to spin up the next big idea. Peer learning thrives on iterative feedback loops, and AI can be a catalyst that keeps us spinning forward at breakneck speed.

So read, reflect, revise—and join the conversation. Let’s keep weaving generative AI into the tapestry of peeragogy, forging a collective intelligence that’s bigger, bolder, and more inclusive than anything we can build alone.

Welcome to Peeragogy 3.0. Now let’s see what we can co-create together.

Charles Blass

unread,
Dec 29, 2024, 9:40:37 AM12/29/24
to peer...@googlegroups.com

Roland Legrand

unread,
Dec 29, 2024, 9:48:37 AM12/29/24
to peer...@googlegroups.com
with o1 i simply asked to read and study: 
please read and analyse this online version of the Peeragogy Handbook: https://peeragogy.org/

Then I followed up with: 
what are missing are materials about how to use generative ai in peer-to-peer learning. How would you weave this in the handbook? What and where?

after o1 made suggestions i asked: 

Please rewrite the full handbook weaving generative ai into it

Notice I am very polite. I don't think these models are sentient, but if they ever become conscious, I want are artificial overlords to remember I was one of the nice guys! 

Regarding Gemini, I asked (after having uploaded the pdf): 

for a next edition, we would like to add materials about how to use generative ai in this context. Where in the book would you add what?

and then:

could you enhance the book taking into account all these suggestions?

Howard Rheingold

unread,
Dec 29, 2024, 11:22:48 AM12/29/24
to peer...@googlegroups.com
This is great! The cautions section needs to include warning about hallucinations and necessity for crap detection. Can you ask how best to do that? Also: in the section on forums/message boards, discourse.org now uses ai to summarize threads and other tasks. I could compose a short graf on that.

what it is -->is-->up to us
how...@rheingold.com  www.rheingold.com


Charles Blass

unread,
Dec 29, 2024, 12:29:30 PM12/29/24
to peer...@googlegroups.com
helpful insights
so it seems the system generated a palpable exuberance and excitement (in the direction of hype?) without being asked to do so?! indeed your prompts are polite as well as minimal
and understated

Charles Blass

unread,
Dec 29, 2024, 12:33:57 PM12/29/24
to peer...@googlegroups.com
agree on highlighting/ integrating critical need for crap detection within these contexts, (perhaps as notes/ footnotes spread throughout the handbook anywhere ai prompting/ processing outputs are referenced?)

Melanie

unread,
Jan 2, 2025, 8:37:07 AMJan 2
to peer...@googlegroups.com
Roland,

Thanks for sharing your explorations.

Reflections on the use of AI as a method of uncovering "blind spots," both at the individual and group level, deeply resonates.

In my work across the University of California, particularly with UC Davis and the UC Cooperative Extension, I’ve observed the emerging intersection of AI and learning at the edges of networks. One striking observation has been how individuals leverage AI for professional advancement, such as during job applications and interviews. Another critical focus has been on accessibility: tools like text-to-speech and speech-to-text software can be life-changing for individuals with disabilities, yet institutional barriers, such as IT restrictions, often block these necessary accommodations.

On a personal note, my journey has also been shaped by advocacy for accessibility, particularly as I recover from shoulder surgery. My experience has underscored the urgency of addressing gaps in accessibility and building networks that support underserved communities, including those with disabilities.

I believe peer-to-peer learning networks like Peeragogy are uniquely positioned to foster inclusive conversations and co-create solutions. AI has the potential to illuminate blind spots and enhance equity across these networks, especially when connecting with and empowering underserved communities.

I’m excited to stay connected and contribute to these emerging discussions!


On Sun, Dec 29, 2024 at 6:09 AM Roland Legrand <legran...@gmail.com> wrote:

Roland Legrand

unread,
Jan 2, 2025, 5:45:57 PMJan 2
to peer...@googlegroups.com
Thank you Melanie! I hope your shoulder heals quickly. I'll contribute more to the discussion next week (in Europe we love taking days off for festivities). 

Best wishes for the new year to everyone!

Roland

Joe Corneli

unread,
Jan 3, 2025, 10:27:26 AMJan 3
to peer...@googlegroups.com
It's interesting that this conversation has reunited many of the people with an interest in peeragogy!

BTW, my current reading list includes this paper:

"Accelerating scientific discovery with generative knowledge extraction, graph-based representation, and multimodal intelligent graph reasoning"

I've found it hard (or impossible) to get money for the kind of research I want to do with AI, but I think this says more about the academic grant funding model than it does about the topic or its potential for beneficial use.

I did come across this funding opportunity touching on AI for Collective Intelligence (https://ai4ci.ac.uk/) — maybe there's a peeragogy angle here.

A bit tangentially, here in Oxford I've been involved in a local music group called Starlings that is very peeragogical in nature... e.g., here's our band performing a (p)remix of "Hey Joe" https://www.youtube.com/watch?v=zc2-PL7nQJA 

I mention this because music feels easier to build a community group around than other topics — though maybe that's an illusion, because, after all, here we are!

Roland, I wonder if you could invite o1 to rewrite its suggestions with the hype factor turned down to zero, asking it for a specific concrete list of suggestions for us to try.

Quoting Gilbert Simondon writing on "Technical Mentality":

"if one wants to understand a being completely, one must study it by considering it in its entelechy [i.e., actuality or realization as opposed to potentiality] and not in its inactivity or its static state."

-Joe

Roland Legrand

unread,
Jan 5, 2025, 4:19:09 PMJan 5
to peer...@googlegroups.com
Hi Joe, yes I can ask the model to reduce the hype-factor. I can also adjust the 'temperature', meaning the model can be very factual or more creative. I'll generate / write a draft somewhere this week. Or several drafts.

Best,

Roland

Charles Blass

unread,
Jan 5, 2025, 5:47:30 PMJan 5
to peer...@googlegroups.com
hey joe,
your 'baby please' rendition reminds of the mothers' 'flower punk' 
further, re: entelechy -> title cut from parliament's "funkentelechy vs. the placebo syndrome" 

On Fri, Jan 3, 2025 at 4:27 PM Joe Corneli <holtze...@gmail.com> wrote:

Joe Corneli

unread,
Jan 9, 2025, 8:35:10 AMJan 9
to Peeragogy
Thanks Charles!

I came across this variation on a theme....
Too risque for my music group I think but a connoisseur like yourself might like it!


Charles Blass

unread,
Jan 9, 2025, 9:07:01 AMJan 9
to peer...@googlegroups.com
whoa, had to stop halfway, musically fine - i actually remember this album from the radio station - but yea thanks anyhow
veering well off course for peeragogy perhaps

Maria Cristina Bandeira Viseu

unread,
Jan 17, 2025, 2:19:49 PMJan 17
to peer...@googlegroups.com
A wonderful 2025 for you all. Are you still meeting every Monday? 

Cris




Reply all
Reply to author
Forward
0 new messages