Weeknotes

Weeknote #19: Wrap up

Rob Cawston

— 3 min read
Share
Pink blossom covering a path.
Blosom on The Meadows.

More of a mini-note this time around, ironically due to a very busy schedule. I'm writing this at the end of Friday, trying to leave things in a reasonable state before heading off to London for the Hackney Half.

What happened this week?

Lots of chasing suppliers and resellers as we continued the migration to a new virtual environment due to large price rises. Watching progress bars and shifting approaches has been an interesting lesson for the everyone in reacting to change, staying calm and problem-solving. Hoing to wrap things up next week, although the project has surfaced some clear areas of improvement and backlogs across our digital storage.

Much of my week was spent in tender evaluation for a CRM solution and partner (perhaps my least favourite activity). Plenty of admin to navigate and compliance hoops to jump through. I always want to skip to the end, but there's been some decent submissions.

A lot of time was spent preparing a full report for the next Board meeting. There's always a balance between championing the team's achievement and flagging the risks and need for further investment.

A refreshing catch-up on Wednesday with our UKRI intern, Elena, who is investigating the sector's application of AI policies, with a particular focus on Library chat services. It was great to see her present with knowledge and passion to a wider group of staff.

We said goodbye this week to our Director of Engagement, Jackie Cromarty, after 37 years at the Library! She will be much missed and the sadness was compounded with news of another colleague set to leave in the Summer.

Interesting things

"That's a good way to judge a presentation. Decisions per slide. If every slide represents a bunch of good decisions then it's probably a good presentation. Presenting is deciding."

Getting the most out of AI

Lots of posts spotted this week pondering how to make effective use of AI.

  • Your AI use is Breaking my Brain: we are becoming the AI police.
  • The dangers of software brain: "Not everything about our lives can be measured and automated and optimized, and it shouldn’t be." (h/t Neil Williams)
  • Why AI adoption is a full organisational endeavour: "you cannot procure your way out of hard organisational change."
  • A much-shared piece by Chris Coyeir on AI alignment and why the ability for AI agents to write code at pace doesn't translate directly to productivity gains. In short, there's still value in moving slowly.
  • Similarly, the short-term impact of coding agents can come with long-term penalties: "If you’re producing twice as much code, you need code that costs half as much to maintain."
  • You can probably ignore all of the above and just read this short piece, from Matthew Sheret at Public Digital: How do you automate with AI in a way that people can actually trust?. I'm comforted that these are the approaches we are adopting at the Library - some by choice, some by necessity:
    • Automate the work people find boring
    • Be honest about your risk appetite
    • Break problems into pieces
    • Build with users

Watching, listening, reading, doing

📖 All list are inevitably wrong, but I'm enjoying the Guardian's 100 best novels of all time. I've read about a quarter, less than I thought.

📺 I've saved up a couple of episodes of Half Man for the train. We're also half way through Babies, but taking it slowly 😔 With the boy almost free of obligations for the whole Summer, a classic film list is being compiled. Suggestions, please...

🏃 The less said about my half marathon prep the better. I'm now in the zone of "turn-up-and-see-what-happens". Rain is promised so maybe that will be a great leveller, like a muddy pitch for an FA cup tie. Straws are being clutched.

Read more