Monthnotes November-December 2025
Every month or so, I share a quick digest of what I've been working on and reading. Here's the latest. More in the series here.
I sent my October notes right at the end of October, so when it came around to mid-November I didn't have very much in the newsletter tank, and figured one less email in everyone's lives is probably welcome at this time of year. So here's November and December in one big bundle instead.
A few bits of news to kick off with today. First, Elevate - the creative data visualization learning community that I run with Alli Torban, Gabriele Merite and Will Chase - is winding down as a business. We've been running it since February 2022 (plus a beta period before that), but our attention has increasingly been diverted elsewhere in the last 18 months and it's not getting as much love as it deserves. So we're going to stop charging money for it.
But that doesn't mean it's going away. There's a wonderful, vibrant Slack community with loads of people sharing questions and learnings and joys and disappointments, and we're going to leave that open indefinitely, transferring ownership to the community itself. While we're stepping back from our job as paid moderators, I certainly plan to remain an enthusiastic participant.
We did a final Elevate Show about what we've learned over the last four years, and wrote an announcement blog post. Personally, I hope that the community will now self-organise and find a way of taking things forward. But even if not, I'm still really proud of all of the amazing people who've been a part of Elevate over the last few years, and I can't wait to see what everyone does next.
Elevate
Second announcement: You might remember that last year I helped set up the Data Sonification Awards. Well, we've just opened submissions for 2026.
Any sonification works created in the last two years are eligible. To enter, all you need to do is fill out the form linked at sonificationawards.org before 31 Dec. We have three categories (art, communication, analysis) with criteria defined by the community. We're also recruiting people to judge the awards.
The coolest thing about these awards, IMHO, is that they're not competitive - everyone whose entry meets the criteria gets an award. I wish more awards worked that way, but hey - all we can do is create the world we want to live in in the small parts of it that we have influence over.
Third announcement: In 2026 we'll be running some online Loud Numbers workshops! We'll start with one on making your first sonification, and then follow up later in the year with deeper dives on audio storytelling with data, sound design for sonification, and how music theory can be useful in sonification. If that sounds very exciting, then sign up for the Loud Numbers newsletter.
0
I went to a great workshop on AI last month, run by the folks at Kulturarbetare Förena Er (which translates roughly to "Cultural Workers Unite"). It was a wild mix of attendees, which was great - loads of different perspectives. After hearing everyone's experiences, I think what really stood out for me is the extent to which so many people have internalised the claims of the AI companies , even if they're opposed to those companies. I lost count of the number of times people said "well, AI isn't very good at X right now, but in a couple of years..."
This, I think, is a common misconception that people have about technology - that progress in a given area moves forward predictably and inexorably over time. But in my 25-odd years of following tech, I have almost never seen that to be the case. What I have seen is technology moving forward in fits and starts, driven by attention and investment. One bit of the tech world gets cool and interesting, progress is made in huge jumps, but then it fairly quickly meets diminishing returns. What gives the perception of constant progress is that another kind of technology then gets cool and moves forward (perhaps enabled by the progress made in the previous bit). From a distant perspective, there's a lot of forward movement going on, but when you focus in on one part it mostly just sits there without progressing. For examples, look at VR, or voice assistants, or smartphones, or wireless networking. Major leaps are always followed by long periods of stagnation.
That's why I'm very skeptical of the narrative that AI is going to keep accelerating. ChatGPT 3.5 was astonishing, sure, but 4.0 was a smaller leap, and there's not been particularly meaningful progress since then (from a value-to-the-user perspective, as opposed to a technical benchmarking perspective - benchmarks can always be juiced). To me, it feels like the industry is in that place where Wile E. Coyote has run off a cliff, but has not yet started falling.
I think we'll see the AI bubble burst in the next year or so. What we don't really know is how that'll affect the rest of the economy - whether it'll be a 2000 dot-com style crash where life outside the industry broadly goes on unaffected, or whether it'll be a 2008 housing-bubble style crash which spurred a decade of austerity and laid the foundations of the Western political and economic omnicrisis we now find ourselves in.
However, another, more fun, question is what cool and useful things people will find to do with all of the zillions of datacentres that have been built off the back of the AI boom. The 1840s Railway Mania bubble created the foundations for the extensive British railway system today. The infrastructure built during the telecom mania of the 1990s paved the way for the widespread growth of the Internet. Crashes tend to leave behind interesting things in the wreckage, and it's likely that the next leaps forward in technology will be enabled by widespread access to very cheap computing power in datacentres that were built for AI. See Tim Harford's "Are Bubbles Good, Actually?" FT column for more nuance on this.
For now, though, my advice is that whenever you hear anyone make a claim about what AI can do or will soon be able to do, be very, very skeptical.
OK that's enough seriousness. Here are my 25 top songs released in 2025, in playlist order. Core themes: politics, regional accents, girls.
- illuminati hotties - 777
- Girl Group - Yay! Saturday
- jasmine.4.t - Tall Girl
- CMAT - Take a Sexy Picture of Me
- Adwaith - Miliwn
- Lambrini Girls - No Homo
- Panic Shack - Girl Band Starter Pack
- The Hives - Enough is Enough
- Kneecap & Paul Hartnoll - Sayōnara
- Coach Party - Girls!
- Snõõper - Worldwide
- Lava La Rue - Easy Come, Easy Go
- Kae Tempest - I Stand on the Line
- Princess Nokia - Drop Dead Gorgeous
- Four Tet - Into Dust (Still Falling)
- Tiberius b - IMMACULATE
- Wolf Alice - Just Two Girls
- Divorce - Antarctica
- Tautumeitas - Bur Man Laimi
- Water From Your Eyes - Playing Classics
- Antony Szmierek - Angie's Wedding
- Getdown Services - Eat Quiche, Sleep, Repeat
- Sports Team - Sensible
- They Are Gutting a Body of Water - The Chase
- Car Seat Headrest - Gethsemane
You can get that as a playlist on Apple Music or YouTube, and I very much recommend that you do. Send me your best-of-2025 playlists in return!
I've been slowly but surely getting into zine-making, courtesy of the riso-printing workshop I attended in October. To me, "making and selling a physical zine" is to "writing on the internet" as "playing a concert at a small local venue" is to "uploading a song to Spotify". It's so much easier to get people to connect with what you've made, and way more rewarding as a result.
Top of my to-do list is a zine version of the "Doing Creative Work in a Climate Crisis" workshop that I gave in April this year (and will be giving again in March next year). It's been fun figuring out ways to compress two hours of chat and questions down into 24 A6 pages. But I think I've got a good layout going, and now I just need to add some pics. I've found this guide from ANEMONE tremendously helpful, and their new app Layout Department is perfect for getting the whole thing into the right size and shape for printing.
Other things I want to make little zines about:
- Making electronic music sustainably
- Low-tech approaches to sonification
- My favourite Norns scripts
Speaking of playing concerts at small local venues, I'm a bit proud of the fact that I played 15 shows this year. Well, 14 so far. I've got one planned for 20 Dec.
Such a density of playing shows has put a bit of a damper on learning new instruments and techniques, but the whole point of learning those things is to use them, so I'm not too worried about that. I'm working on a few recordings, too, but I want to listen to them a few times to see if they're worth releasing or not.
Finally, I'll leave you with a somewhat niche Christmas present - this article on how to use toner transfer to put graphics on basically anything.
Have a restful holiday season and see you in mid-January.
Duncan