Deliberate Struggle
I had the honour to attend a Cynefin Exploratory in Dublin last week. It had been 10 years since my last conference, which I usually ended by writing up a little something so I decided to summarise my experience in this journal.
Triopticon
The facilitation format the participants were guided through goes by the name of Triopticon. To paraphrase Dave Snowden: "something between a Conference and an Un-Conference".
Let me summarise it in case you don’t want to read up on the format in the official Cynefin Wiki
Day 1
The facilitator makes sure the participants are grouped in trios, made up of people who do not know each other.
You get to hear a formal presentation; the other speakers get the opportunity to critique the content while the presenter is not allowed to answer.
You gather in your trio, discuss the presentation, determine a speaker and the speaker (3 presentations, speaker 1-2-3) gets to discuss with all the other designated speakers from the other trios.
That is repeated for each presentation
Day 2
On day 2 all designated speakers were grouped together (all speaker1, all speaker 2 etc.) to work out observations, questions and possible action items which was followed by an individual choice to assign their name to a cluster and engage in a working group around one cluster.
What worked well
Entangled trios are a common theme in anthro-complexity methods so being deliberately paired in one after having used them is an interesting experience. Obviously the connections made in the trio have been way stronger than to other participants. I got to know 3 other awesome people sharing their stories about personal and work life while reflecting on the presentation (my table was the only entangled quartet)
The gathering of the designated speakers to me honestly was the pinnacle because by design if you are not a speaker you are not allowed to interfere or ask questions so while I was burning to yell "CONCENTRATION OF POWER" and to give some little explanations about how LLMs technically work I was only allowed to take notes and preserve what was important for me for a later stage.
On day 2 I actually got the incredible honor of being led through a little estuary mapping exercise because we orbited around critical thinking or the fear to lose it in the dawn of ai providing us with new shortcuts.
While it is not news to Dave Snowden what I experienced when talking about Wardley mapping is that people think: oh, it's about how to move everything into commodity. No, it's not. It's about having a common understanding of the landscape and making a deliberate decision.
Estuary Mapping is about mapping the energy gradient and while people more involved with the method understand that it is also about creating a shared understanding and making a deliberate choice I do think there could be the underlying assumption in private and public organisations that we should generally remove the energy cost of everything.
Estuary Mapping deliberately introduces a volatile zone where items are placed that can actually change with low time and energy investment but in the context of this format, Human and AI Reasoning I was envisioning an "AI can do this" line that is positioned higher than the volatility zone which allows to create a shared understanding of what under no circumstances should be automated and so to speak should maintain an amount of deliberate Struggle
I do have used friction as a deliberate design element in software but to go a level deeper and realise that we do not really have tools and methods or even the room to talk about 'where do we actually want to increase friction' gave me some to-dos for my list.
What I missed
Another element in an Estuary map is the counterfactual: the area where we move things whose energy and time cost of change is so high it isn't even worth approaching them.
Here and there you'd catch phrases like 'this is the manifestation of capitalism' or 'we are building capabilities on top of two companies,' but given the amount of expertise and knowledge in the room, I longed for some discussions around:
- Data centers
- Social media + deepfakes
- Democratic election cycles in the context of AI
- Open-source models vs. private models
- EU sovereignty
I'm not sure how I personally would adjust the event to make sure bigger topics are in focus, but I think it was partially due to each phase having a high focus on reaching consensus.
Optional extension, Day 3: a James Joyce–inspired walk through Dublin
Did I extend my hotel and reschedule my flight when I heard there was a walk happening after the official event? Absolutely.
Did I start to read Ulysses only to give up because I deemed the English too foreign for me to grasp the meaning without an intensive time investment, and have no understanding of the cultural relevance of James Joyce's work? Also true.
I do love a good walk though, and the conversations while circling around Dublin were deeper by magnitudes than the quick in-between chats during the event.
The walk ended with a visit to the James Joyce Centre, which left me with one deeply sad realisation:
With AI, quantity has lost meaning.
Ulysses is a book that runs between 640 and 1,000 pages depending on the edition.
I do assume AI will forever be part of humanity.
Never again will someone be able to write an ~800-page book over 7 years that becomes a cornerstone of a city's cultural literature, because with AI there will always be the obligation to disclose how much — and in what form or shape — has been written, edited, or audited by AI.
Closing
I want to thank each individual person at the event. In the event it was complex endeavour, a sum of its parts.
Head over to https://thecynefin.co/ if you'd like to attend an event like this