I’m still trying to get things ready before I go. Almost all the energy at work for my immediate peers is going into our planning round, trying to get money for the big streams of work. We’ve got a big ask for website and intranet, but a lot of it is essentially making up for austerity.
As a side note I started thinking this week about the metaphors of austerity. Things seem to slide into the metaphors of 2010 in terms of personal debt (pertinent to the context of the financial crash but differently situated now).
“[Labour] could remind the government of the obvious truth that Britain’s great institutions were built by recognising that assets designed for multi-generational use are precisely what debt is for. For Starmer, that perhaps sounded too much like a commitment. He preferred to mimic George Osborne in austerian pomp: the ‘national credit card’, he declared in his response to the budget, is ‘maxed out’. Austerity slogans constrict any future Labour government’s room to manoeuvre every time they’re cited: it’s stupid to reanimate them. This is a reminder that mendacious doubleness is a connivance of both sides of the House. Starmer can decry, as he did in his response, ‘fourteen years of stagnation’. But the diagnosis is not matched by the prescription.”
*https://www.lrb.co.uk/the-paper/v46/n06/james-butler/short-cuts*
I started to think about the streets I cycle on as the credit card. If you are diligent with maintenance then you can maybe skip a few years of 100% coverage. Sure, they’ll get a bit worse, but it’s all starting from a strong position so there is some give in the system. But here we are, everything is run down and even the opposition is offering to keep running down the infrastructure, to keep borrowing on the current state to put off future spend.
But, much like my bike ride into town where I have to zigzag across the carriageway to avoid the potholes exposing pipes and wires, the bill eventually arrives. People still need the road, the number of people go up. More bikes, more EVs, heavier traffic. So, Cambridge’s website needs the money. If it doesn’t get it, we’ll pay more next year and we’ll offload the cost in the meantime into a thousand annoyances for staff and missed opportunities.
(https://medium.com/@ElizAyer/meetings-are-the-work-9e429dde6aa3)
Had a really productive meeting with our chief architect about how we could work to reform our non-functional requirements process and take it from individual clauses in a spreadsheet to a set of (maybe external?) standards we adhere to. It was a really open door, I had anticipated pushback, but it seems like the blocker is time rather than willpower. So, going to try and do a day on it when I’m back in the summer.
Been thinking a lot about how to land portfolio prioritisation stuff in an environment that isn’t used to prioritising based on things other than “amount of smoke”. I obviously love James Higgott’s stuff (https://jiggott.medium.com/creating-the-nhs-website-roadmap-a-case-study-2ad98d6087d2) from the NHS but have to keep coming back to both the process, the people and the expectations/accountabilities. If I can’t fix the idea that we have to work in a different way as well as using a miro board then it isn’t going to work.
A bit of work I kicked off is running into trouble. It is being done by an outside agency, it meets a core and legally tangible need of the business. But, we keep getting bogged down in compliance and committees. This week it’s the dawning realisation that we don’t have the support in place for this stuff in the long term and that it is coming to a cliff edge where it either becomes a project that we launch and watch it disintegrate over the next little bit or a service with upkeep and maintenance.
I went down to Newspeak House for the (re-)launch of the Election Handbook, a collection of tools for anyone trying to put together civic tech or general election hackery. While I was there, I saw this thing that uses an LLM to try and squash Hansard debates into a WhatsApp format. I felt really uneasy about it for quite a few reasons. Number one is below: it makes tonally dreadful decisions about the text.
Secondly, it cannot deal with nuance. The way that Parliamentary debates have little tics like “Interuption”, “The Noble Lords: ‘oh!’” or one where I saw the LLM crunch down an MP saying a sentence that should have translated to “this bill will die on its arse unless you build a wider coalition of supporters” to “lots of people want to support this”.
Thirdly, what’s the user need? What need does this fill? I cannot imagine a need for automated shortening of parliamentary debates that doesn’t create a wide open misinformation vector (either deliberated or non-human curation slippage).
I’m picking on one thing for sure, but I kept thinking about when we used to do Parliament hack days and the apps were to a greater or lesser extent trying to fix needs people had (personally) had for finding information or understanding an event. These things feel like the need was to see what an LLM makes of Hansard. So, even if points 1 and 2 were worked out, I am left wondering why?
I went round London on Saturday night because the eldest got tickets to Frozen the Musical before it closes and the baby wouldn’t sit still for West End theatre. So, I took my lensbaby trio (velvet, ‘sweet’, and twist lenses all in one), ate late night pastel del nata and watched people be far too drunk for 8pm (I was never etc. etc.). It was really hard to get the focus right on the little viewfinder screen, so eventually I took off my glasses and adjusted the view for my least-worst eye. I got about 10% good shots in the end, but seeing as it is a bit of a pig of a lens to use (I have only really had success using it for portrait shots before this), I was happy with that. I also had a go at some 1 ISO shots while resting the camera on the edge of a bridge, but they didn’t turn out interesting enough.
