December 30, 2024 – January 5, 2025
My “weeknotes” capture events, thoughts, and other items from the past week, mostly focused on work. Learn more about the weeknotes concept here.
This week was the end of 2024 and start of 2025—the two faces of the Roman god Janus rolled into what felt like 2 weeks. Being another holiday week, the list of things that happened was pretty short and disconnected. But let’s look back a little and forward a little before we plunge back into a more normal schedule.
Am I doing a 2024 retrospective on my posts or other work? Nope. I know that’s common, but I just don’t have the energy. Maybe I’ll have AI develop one in the future. But really, this is as much a grab-bag journal as anything else, so summarizing it feels… not that useful.
Flighty Passport 2024
The annual “passport” generated by the mobile app Flighty comes at a fairly high price for an occasional traveler, but it’s nonetheless a nice review of the year, as experienced through airports, anyway.
- My 8 airports for 2024: CMH, DEN, ANC, BNA, BWI, DCA, ORD, and PHX (although the app says it was 9 airports)
- 14 flights totaling more than 16,000 flight miles, all domestic
- Shortest flight: CMH to ORD
- Longest flight: DEN to ANC
This year is likely to be more home-bound. But I do start the year off flying to Florida before the end of this new month, so who knows?
As for flying itself, it’s getting more expensive and less pleasant with each passing year. I’m happy with the changes applied by the Biden administration in the last couple years, but it’s entirely possible the Trump administration wipes out those gains. The air travel industry is an example of one that actually benefits from strong and standardized regulation. There just isn’t enough competition to keep the market customer-focused.
Artificial Intelligence 2024-2025
There was a lot of sturm und drang around AI in 2024, as we all know. But I think 2025 is going to be the year where things start (just start) to get “real.” The hype-masters will keep pushing for nuclear power plants, data centers, and unlimited cash, but expectations around AI should start to come back to earth and focus on things we can actually use. A few items of note for now…
- John Gruber’s Daring Fireball linked to this tour de force review of AI developments in 2024: Things we learned about LLMs in 2024. If you’ve had your head in the sand and need to catch up, this is the place to start. Or if you’re plugged in so much you can’t remember what all went down last year, same thing: start here. Two standout sections for me:
- The VA published an AI Use Case Inventory report, a sort of 2024 review and update on how AI is getting into their operations. Their report shows 230 uses (in an Excel spreadsheet you can download), many of which fall into a few categories like transcription, data visualization, and log analysis—stuff that’s commercially available and basic. The biggest category of AI usage reporting, however is “None of the above,” and that grab bag categorization seems to be driven by self-reporting leaders in the VA noting usage of AI features picked up through commercial sofrware packages that claim to use AI, but they can’t really say it’s an innovation or deeply describe it. When I saw the publication of this report it made me think, “Oh man, this is gonna show how far behind we are,” the details in the report show me the VA isn’t that far ahead and I could generate a similar inventory if I can include all the marketing claims of our vendors. Hell, Microsoft released Copilot features into the GCC edition of M365 before the end of the year, so… pop that on the list. That said, they had some small-“i” innovative uses I liked seeing. Mostly what I appreciate here is the transparency and the details in the report. We should all publish this stuff. (I also wrote this up as a post on LinkedIn.)
- Minnesota’s in-house IT service (MNIT) hired one of their own as the new Minnesota Director of AI. This new role will try to organize AI thinking and efforts across State agencies as you might expect. More States and other government bodies need to do this—the field is moving fast, vendors are making kooky claims, and there are benefits to be had, but we have to figure out where it makes sense to innovate on our own and where it makes sense to just knit together third-party solutions. As for this kind of role, we explored asking for a similar starter role in our 2025 budget requests but ultimately chose not to ask for it. Down here at the County level it may still be too soon. But I bet we ask for it in 2026.
2025 Priority: GO
As 2025 starts, I’m looking ahead, like anyone else, thinking about priorities for the year. And one post from Michael Brennan at Civilla from last month has me thinking: The state of human-centered design. In the piece, Brennan points to challenges we’re seeing at the local level when it comes to advancing the goals of human-centered design (HCD) as it applies to government services: cultural challenges, resource challenges, and perhaps fundamentally (at least for us), challenges in understand what HCD is and why it matters. If you don’t know HCD’s value, you’ll struggle to see why a government agency or service team should bother to change and improve services with iterative improvements based on real public needs.
So how do we address this in the GX Foundry team in 2025, especially when we’re still in the early days of developing our digital service muscles? How do we expand our messaging out to our broader county audience? And how do we deal with that while there are some bigger fish to fry?
Because in 2024 I added our Delivery Services (project management and business analysis) team to my responsibilities, and we spent much of 2024 figuring out how to rearrange our priorities and approaches. I think we have the model down at this point, but now we have to operationalize it, and we have to knuckle down and get real work done. Indeed, our BA and PM teams need to become the engine by which most major organizational accomplishments are achieved.
To be brutally honest, the confused strategies and conflicted leadership around service delivery going back several years has left us with teams and individuals with spare capacity to deliver solutions and services. In the last couple years we’ve spent more time tripping over our own feet than walking or running. And what’s deeply frustrating is that we have a team actually capable of running, actually getting a lot done. We really do have a great team now. But we’re not pushing them to achieve what’s possible. We’re leaving achievements undone. Indeed, we’re leaving individual careers to languish when there’s work to do.
So the challenge for 2025 is getting our creaking machinery of service delivery oiled up and moving faster and faster. We have capacity, we have talent, and the working models are aligning rapidly. So let’s effing go.
2025: We need more whiteboarding
Speaking of making things go… I’ve always liked whiteboarding as a way to design solutions and plan projects—like, hours-long sessions of whiteboarding with colleagues to figure out what to do and how to do it. Visualization, including “bad” drafts of visualization that invite people in on co-creation of ideas, has been useful to me, but as the TEDx video below explains, it’s useful for teams, too. Visualization and the use of visual metaphors is a team sport.
So for 2025, I’m driving more meetings, more discussions, and more planning to the whiteboard.
And I will do more doodling for myself. I’ll pull out the iPad and Apple Pencil and even pen and paper to diagram what I’m thinking. And I’ll share it here if I can.
Fair warning: My word for 2025 is…
I played this little new year social media game with Merriam-Webster and, well, the results were illuminating. Consider yourself warned.
Discover more from digitalpolity.com
Subscribe to get the latest posts sent to your email.
digitalpolity.com 