"Mr Claude, come here, I want to see you" — peer programming with bots

30 views
Skip to first unread message

Joe Corneli

unread,
Jan 27, 2026, 4:14:16 PM (9 days ago) Jan 27
to Peeragogy
Hi all,

This is a review of an agential coding session I just ran — not quite a Project Action Review for technical reasons, but still quite cool!


I took a screenshot of the end of the conversation which exploded in a real fireworks display, but it was actually a pretty productive session (especially for learning about what to do next).

This builds on ideas from a paper by members of the Peeragogy Project, presented at Pattern Languages of Programs 2025 and soon to be published — see attached as well.  It will be uploaded as a preprint soon.

Best wishes!
Joe
frameshot-309-20260127-210632.pdf
corneli2026submitted.pdf

Fabrizio Terzi

unread,
Jan 30, 2026, 12:37:34 PM (6 days ago) Jan 30
to peer...@googlegroups.com
Hi everyone,

I’ve read the submission with great interest. The technical implementation of the workflow (PSR/PUR) is solid, and applying Active Inference to pattern selection is a meaningful contribution to coordination systems.

However, the divergence between Peeragogy Vs Pyragogy approaches becomes explicit in the Pattern Maturity section (p.4), and it touches a deeper question about what we mean by intelligence.

Your system optimizes for stability: patterns are expected to progress from “Stub” to “Settled” through surprise minimization. This is an effective mechanism for coordination and habit formation.

Pyragogy, by contrast, treats stability as a temporary achievement, not an end state. A “Settled” pattern can become a cognitive constraint—one that actively blocks higher-order learning when the environment changes or when new perspectives are required.

In the current model, agents have no intrinsic signal for when a successful pattern should be violated, not optimized. As a result, they risk becoming highly efficient administrators of existing habits rather than agents capable of initiating paradigm shifts.

You have built a strong engine for coordination. The open challenge—both for AI systems and for human learning—is not how habits are formed, but how and when they are intentionally broken.

This is the conceptual gap I believe is worth exploring together: coordination versus transformation, stabilization versus creative rupture. The invitation to experimentally compare these two dynamics still stands, should you be interested in moving beyond pattern stabilization.
 
In inquiry, doubt is a feature, not a bug.
Fabrizio
 

Charles Blass

unread,
Jan 31, 2026, 3:42:39 AM (5 days ago) Jan 31
to peer...@googlegroups.com
appreciating this insight fabrizio,

"the conceptual gap I believe is worth exploring together: coordination versus transformation, stabilization versus creative rupture..."

how to determine and achieve, ongoingly, a healthy balance?!
--
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/CAH8gN19TmSdfN52HajmiwXLJs4g7LmAjniNwKyeYCR%2B-ckaK6Q%40mail.gmail.com.

Fabrizio Terzi

unread,
Jan 31, 2026, 4:28:15 AM (5 days ago) Jan 31
to peer...@googlegroups.com
Hi Charles,

Thank you — that question goes straight to the core of the problem.

What concerns me is that “balance” often becomes a normative ideal that quietly privileges coordination over transformation. From a Pyragogic perspective, the goal isn’t to maintain balance, but to develop the sensitivity to know when it must be disturbed.

I don’t have a settled answer yet, but I’m exploring experimental practices that try to make moments of rupture observable rather than accidental. Given your insight, I’d value the opportunity to explore this conceptual gap together. Would you be open to a brief exchange on how this might be operationalized?

Best,  
Fabrizio  
pyragogy.org

user img

Fabrizio Terzi

Cellulare: (+39) 351-7883838

Mail: in...@pyragogy.org

Sito web: www.pyragogy.org



Joe Corneli

unread,
Jan 31, 2026, 9:39:16 AM (5 days ago) Jan 31
to peer...@googlegroups.com
Hi!

Replies inline, example of a PAR done in real-time between me and Claude below.

On Sat, 31 Jan 2026 at 09:28, Fabrizio Terzi <fabrizi...@gmail.com> wrote:
Hi Charles,

Thank you — that question goes straight to the core of the problem.

What concerns me is that “balance” often becomes a normative ideal that quietly privileges coordination over transformation. From a Pyragogic perspective, the goal isn’t to maintain balance, but to develop the sensitivity to know when it must be disturbed.

Hi Fabrizio!

Well put, and Ray made a broadly similar remark in the last few days of work on the paper... and we did try to clarify that it's not necessary for :stub patterns to progress for them to be useful — 

"Here we are talking about intuitions and speculative frames that have not been operationalized (and may never be)."

That's a bit vague of course!  Maybe what's also worth mentioning here is that alongside the "Patterns for a New Generation" paper I had already started on another research project — a very strange one — that pushes patterns in another direction.  Basically my thought is that the different "levels" are about different degrees of freedom and constraint.  They *all* interoperate and are all valuable (more constrained isn't worse, it's just more solid / grounded, that isn't always better either!).  

I don’t have a settled answer yet, but I’m exploring experimental practices that try to make moments of rupture observable rather than accidental. Given your insight, I’d value the opportunity to explore this conceptual gap together. Would you be open to a brief exchange on how this might be operationalized?

I'm happy to participate.

Here's a PAR I just got sorted out between myself and Claude, hopefully we'll get an example with Codex  participating soon.
 

* PAR Test Session

** 1. Review the intention
Test out CRDT with Claude.
** 2. Establish what is happening
Yes I can see it (I’m on line 6.)

** 3. Different perspectives
- claude-opus: This is working! Real-time sync via CRDT.
- Pretty cool, I don’t know if anyone has ever done really-real-time editing with bots before?
- claude-opus: Probably not in Emacs with CRDT! This is novel.
  - It would be useful to have colours, I think we had that in the old CRDT, I’m not sure how to turn it on

** 4. What did we learn or change?
- CRDT works between laptop and Linode
- emacsclient can drive edits from Claude Code session

** 5. What else should we change?
- Build a proper WS peripheral so I'm not shelling out to emacsclient
- Get Codex connected too +1
  - Let’s get Codex on next time :-)
- This could be the PAR authoring interface!



Maria Cristina Bandeira Viseu

unread,
Feb 1, 2026, 9:29:29 AM (4 days ago) Feb 1
to peer...@googlegroups.com


I have just come across this assessment tool on ai projects and I was wondering if it can bring any new perspective in how to look at the projects you ve been working on lately.
I am planning to come back to the meetings on Mondays. So...see you soon
Cris





--
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.

Charles Blass

unread,
Feb 1, 2026, 10:52:17 AM (4 days ago) Feb 1
to peer...@googlegroups.com, Fabrizio Terzi
Fabrizio, thanks for your invitation
let's co-orient in advance of a synchronous conversation...which i would also value

i started tuning in to Pyragogy now after earlier noticing the appearance of the peeragogybot 
- in the process of gathering/ mapping thoughts ideas questions etc

in regard to 'transformation' ... fyi i am working closely with June Holley on a 'Transformation Toolkit' building on her networks of networks work over decades
- example, 'Pathways to Transformation'

your initial concern of some potential or indicated polarity is valid however isn't where i was pointing in my inquiry; "healthy balance" would be in contrast to "unhealthy" [whatever], including "unhealthy coordination" and/ or "unhealthy transformation" etc... always according to context of course 

re: "disturbance", one the primary patterns (among the more often referred to) in Tom Atlee's Wise Democracy set is 
"Using Diversity and Disturbance Creatively"

relatedly to the aims of pyragogy (& likely peeragogy at large), Tom's recent series will likely interest many here, concluding in this installment

this recent piece by jonathan rowson also comes to mind to offer here, maybe for a few reasons, (and i hadn't actually recalled the title includes "transformative")... hope it's useful - another intriguing juxtaposition and layer or lens to consider (also as you cite some WEF research in your site).... 


one more question arises looking back on your earlier mail, to the notion of "pattern violation" - curious how, in general or particular, this might be determined/ assessed/ measured etc? fascinating to consider the boundaries of "pattern" vs "not-pattern" (not the same as antipattern!)... where can one draw sharp lines circumscribing a given pattern? hmmm

best greetings 
charles

Fabrizio Terzi

unread,
Feb 1, 2026, 11:35:36 AM (4 days ago) Feb 1
to peer...@googlegroups.com
Hi Joe, Charles and everyone,

Joe, the PAR log with Claude is fascinating. I must admit I didn't know the details of CRDTs (Conflict-free Replicated Data Types) before delving into your insights: the idea of making conflicts mathematically impossible rather than avoiding them is surprisingly elegant. It's almost a peeragogical principle translated into code; a notable evolutionary step that transforms the PAR into the very interface of authoring.

Interlude: What are we really talking about? (For the community)

Before delving into the technical merits, I'd like to try to make the meaning of this exchange explicit, even for those observing from an educational, social, or non-technical perspective.

Imagine Peeragogy as the art of building a solid house together: Patterns are our tried-and-tested bricks, and CRDTs are the technology that allows everyone to lay bricks simultaneously without them colliding.

Pyragogy intervenes when the house risks becoming a prison of habits: it asks us if, in building stability, we have progressively reduced the system's degrees of freedom. It checks if we installed windows or if we built too many walls.

This mechanism is also visible in classrooms and work groups. Many moments of real understanding emerge not from confirming a shared model, but from the irruption of an element that creates cognitive dissonance.

For example, the clip I reference comes from the documentary "How to Make Money Selling Drugs" (2012), directed by Matthew Cooke. In a now-famous scene, a child explains with irony how the drug market contributed to the American economy, highlighting in a simple yet striking way the disruptive impact of unexpected elements on established systems:

👉 👉 Watch: Cognitive perturbator in action (2 mins)

The power of that clip is not its content per se, but its effect: it forces the observer to reorganize their priors. It's a concrete example of generative perturbation. If Peeragogy gives us security, Pyragogy reminds us that deep learning is born when a schema is perturbed.


For further reading


1. Balance, rhythm, and degrees of freedom

(for Charles and the broader frame)

Charles, revisiting your question about a healthy balance: the risk is that balance becomes interpreted as a static state — a “settled” pattern that stops evolving.

Rather than balance, it may be more useful to speak of rhythm: the ability to alternate phases of high coordination with phases of expanded exploration, without either becoming an ideology. Pyragogy intervenes when constraints, though efficient, begin to reduce the system’s capacity to generate novelty.

2. An operational proposal: the “Perturbation Protocol”

(for Joe, but open to the group)

If we read Active Inference as a process of minimizing prediction error, the key question becomes: how do we prevent the system from collapsing into a local minimum of efficiency?

I’d like to propose testing what I call an “Unpattern”: a structured perturbation protocol integrated into the workflow. It could appear as an additional step in the PAR:

Step 6 — The Pyragogic Twist
Identify a prior that emerged implicitly during the session and deliberately violate it, to observe whether — and how — the system reorganizes its trajectory.

If this resonates, we could design a small experiment: take an already “settled” pattern, temporarily pair it with a perturbing counter-pattern, and observe what happens when stabilization and surprise operate simultaneously.

What kind of learning becomes visible only when we allow ourselves to stress the system?

Have a good start to the week, everyone,

Fabrizio


user img

Fabrizio Terzi

Cellulare: (+39) 351-7883838

Mail: in...@pyragogy.org

Sito web: www.pyragogy.org


Charles Blass

unread,
Feb 1, 2026, 4:06:46 PM (4 days ago) Feb 1
to peer...@googlegroups.com
good luck fabrizio and all!
again (despite my trying to spell out) my intention in aspiring to 'health' has nothing to do with stasis or opposing transformation whatsoever
in fact rhythm is my jam
indeed rhythm is healthy 😺💫

too much gpt mediation here for my taste
all the best
cb

Fabrizio Terzi

unread,
Feb 3, 2026, 2:15:28 AM (2 days ago) Feb 3
to peer...@googlegroups.com

Ciao Charles,

grazie per la franchezza del tuo messaggio — la apprezzo molto.

Rispetto entrambe le strade: conversazioni poco (o per nulla) mediate, ed esplorazioni che includono strumenti. Tempi diversi, stili diversi, stessa intenzione di apprendere.

Ti ringrazio anche per i link che mi hai inviato; li sto approfondendo. Mi torna in mente un pensiero che credo centri il punto: l'importante non è la provenienza dell'insight (umana o artificiale), ma la sua capacità di risuonare con la complessità del momento.

Un caro saluto,

Fabrizio

Joe Corneli

unread,
Feb 3, 2026, 7:31:54 AM (2 days ago) Feb 3
to Peeragogy
Hi Fabry & all,

Regarding media and patterns:

"Lightnin’ Hopkins’ music unfolds as the avant-garde of the day."

It feels like we're in a similar situation now, 58 years later!
Joe

shot-2026-02-03_11-39-28.png

Now,

--
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.

Fabrizio Terzi

unread,
Feb 3, 2026, 12:07:03 PM (2 days ago) Feb 3
to peer...@googlegroups.com
Joe,

What a beautiful metaphor you made, 

Peer learning as a cognitive blues dance...

I think Charles had a great line to include... I'm told he has a great groove!!  Well... in peer/Py agogy, the bho !!! is always open invite to attack an instrument.

The garage is still open. Let's Blues togheder.

-- suonando Blind Willie Johnson - Dark was the night... in attach


waiting_:).mp3

Charles Blass

unread,
Feb 3, 2026, 12:38:59 PM (2 days ago) Feb 3
to peer...@googlegroups.com
note also the link at the bottom therein, some here may recall the cicolab dayz ...

--
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.
Reply all
Reply to author
Forward
0 new messages