Weeknote 59
Reflection as an outcome, question banking, making an 80s style game that pits pits an applicant AI vs an Assessment AI. usual interesting things.
What I did
Monday I spent time with Liz and the team from Lloyds Bank Foundation as we had a reflection/wrap up session for the Organisational Resilience programme we ran for the last 6 months. It's very easy to just move on at the end of pieces of work and miss out on the opportunity to go "what worked? what didn't? what would you change? what next?" so it was good to do that. We got to see some of the interim evaluation of the programme which was validating in that it was excellent, and that there were some things that could be improved.
We had intentionally built in reflection as a muscle throughout the programme. We did it in every session and had whole sessions on it. It felt a little alien to some people at first, but by the end everyone was naturally doing it, using that muscle. So it was especially nice to see this in the evaluation
Reflection was an outcome in itself
The what next question was a bit trickier. Although the programme evaluation shows it was excellent, with people and places really valuing it, the delivery and content really good, I get the feeling it doesn't really seem to fit the new LBF strategy and approach, or at least it's not seen to within the leadership team. The champion of the programme Jill has retired, and this work has sat a bit on the periphery of the new leadership team and approach. I hope I'm wrong, because, well, it works. But it's not flashy. It focuses on conditions and adaptation, not growth.
Ran a question banking session for a team thinking about data and relationships. As always, once people stop thinking just about data, and just dump all their questions down for everyone to see, and you cluster them, and prioritise them, and ask what they mean, you get a whole different sense of what people/organisations really want and value. Another none flashy thing that just works.
Caught up with Doug while he was in Portugal - discussed some potential work, with an interview next week and a pitch for something else. Also discussed the first FIELD STATION experiment a bit more, the new September TechFreedom cohort and checking in with the first cohort to see how things are going.
This is the last week of school for my daughter, so I'm winding down work for the summer now. I'll still do some bits, but mainly will be spending time in europe in the van(which van? yes good question) I was reflecting that the last 12 months have been quite tough work wise. Data For Action was sort of winding down as Tom F looked for security or just something else to get himself into, finding work for DFA had been tricky in 2025 despite people saying we were doing excellent stuff. I had sort of neglected The Good Ship a bit and so although I had some work, it never felt like I had much security. I had to spend probably more time than I wanted to ramping that all back up. Feel like I'm sort of getting there again now, so taking time off over summer hopefully won't knock that.
Random things
Put together the most random and ridiculous of things on Friday. It's an 80's style game that takes a satirical look at AI in grants. It pits an applicant AI vs an Assessment AI. The applicant tries to get a higher score based on some obscure percentage score and vague feedback. The funder AI introduces biases that it never describes. Who will win?
Tom Watson
Underneath the ridiculous nature of this, there is probably a more serious note. There has been a rise in platforms for AI applications and for AI assessments. Some even do both. The game highlights that biases can be introduced but never actually surfaced. It also highlights that 'scores' with no feedback, to either applicant or grant funder is a very real thing happening in real life situations right now.
This is a game. A rather crude one done in not much time. But the underlying concept, well, that is potentially real. Is it possible to game the AI funder assistant of a real funder - a less ethical person may well do that. I hope the game also highlights that if you are going to put AI checkers, at any point in the process, you should have evals. And if you have evals you should be able to give honest and open feedback to all parties about how and why scores are made. If you're not, then you need to take a long look at what you are doing.
Campervan and running
I went a bit backwards with the campervan this week, refitting the subfloor. It was 'almost right' first time, but almost isn't going to work long term, especially for the floor, so I redid it and I'm much happier. Made some decisions about walls and electrics. Decisions are the hardest bit. Should i? Could I? What if I f*ck it up etc. Progress and feel much more confident to just go with it now. Also accepted it's probably not going to be ready for this summer. Which is a shame, and a load off the mind.
Running has taken a hit this week with working on the van. Got out for forest friday though and felt pretty decent for most of it, was able to push. There's a section at the end which is about 3km of undulating trail which when feeling good I'm able to push and enjoy it, which I did. Not the running year I wanted, but it's the running year I've got.
Interesting things
The State of Open Source AI — V1.0 · July 2026 - open source AI is an option people
AI 2040: Plan A - A detailed forecast and recommendation for how the US, China and the rest of the world should navigate superintelligence.
Our Future Is Now - Will we be owners or tenants? It’s time to decide.
> Our Future Is Now - by Raffi Krikorian - Owners Not Renters
The paradoxes of the knife - The Cynefin Co - If you want good roses, sharpen your knife and harden your heart. - I enjoyed this one. Choose efficiency or effectiveness and be clear which one you are choosing
The portal trap - No one likes portals, one-stop shops, single front doors or gateways, so why are they the go-to solution for so many large organisations?
curious patterns #37 - Architecture as relational infrastructure, the glitch in the public arts funding matrix, and the ecosystem turn in the creative industries. - I was included in this for an older piece ( Field Notes: Designing Entropic Systems) which was lovely to see older pieces picked up - but also it was nice to just see all thr other things included! Lots to read in here!
Football Data Portraits - who doesn't love a bit of data viz
medium.com
> A story from Jihoon Jeong on Medium
Read “The Year of the Loop“ by Jihoon Jeong on Medium:
UK’s largest seagrass restoration project underway - The £1.8 million project will restore seagrass and native oyster habitats in Cornwall, to support marine life like seahorses and sharks
> UK’s largest seagrass restoration project underway
(W)HOLE : Home - Clusterduck’s cooperative world-building exercise aimed at revealing the hidden world of digital infrastructures through multiple media, languages, and realms, while providing the means to act upon it.