No. 115 — Under-95-cm applications ⫶ Rules of Resonance ⫶ How LLMS work
Thinking about that small fragment
My name is Linda. I write a bi-weekly newsletter about computer science, childhood, and culture.
This is the first year I felt Christmas season in Paris was too short: so much we didn’t get to do. The silver lining: in France, holiday cards double as New Year’s cards, so I’m still on time for that. Now I’m in Switzerland, typing away at my in-laws’ kitchen table, trundling towards the end of 2025.
This is a short one, as I’m editing together the last of the Year in Review, to be published after the holidays.
Joyeuses fêtes de fin d’année, as they say.
1.
European social platforms need a rebuild. That was the message Ditte, Thomas, and their team at Rebuild shared last week in Copenhagen. Messaging, groups, communities, creator tools, photo sharing.. it is time to build European alternatives made for people. There is real energy in rethinking scale beyond the hyper-stretched, one-size-fits-all model we have been offered.
I’m watching the children’s layer with special interest. What would a web for users under 95 cm look like? How could Europe-built products offer something new for parents and communities on the big questions around our children and screens? What would app development modeled like a city play-catalogue look like?
If this sparks something, or you know someone in Europe building or dreaming of this, tell me, and point them to apply for the Rebuild cohort.
2.
In similar spirit: Resonant Computing Manifesto which outlines some of the design principles to move away from the monolithic, magnetic, and all-encompassing tech. Build products that are private, dedicated, plural, adaptable and prosocial. Here, here! It’s always hard to imagine alternatives, and yet they exist.
3.
Sam Rose explains how LLMs work with a visual essay. I love all the small interactive explorables here: the exploding matrices and the attention heatmaps. Sam has great essays on memory allocation, load balancing and big O notation.
4.
I really liked this post by Mason Currey on where ideas come from. He strings together a lovely chain of quotes from André 3000, Elizabeth Gilbert, John Cage, Elif Batuman, Terry Riley, Solvej Balle, to Lynda Barry around the notion of acting like ideas are hovering just above our heads is a productive creative stance.
Paul Holdengräber: You know, there’s a line I’ve always loved of Leonard Cohen. He said, “If I knew where the good songs came from, I would go there more often.”
David Lynch: Absolutely. People, we don’t do anything without an idea. So, they’re beautiful gifts. And I always say: Desiring an idea is like a bait on a hook. It can pull them in. And if you catch an idea that you love, that’s a beautiful, beautiful day—and you write that idea down so you won’t forget it. And that idea that you caught might just be a fragment of the whole whatever-it-is you’re working on. But now you have even more bait. Thinking about that small fragment, that little fish, will bring in more, and they’ll come in and they’ll hook on. And more and more come in, and pretty soon you might have a script, or a chair, or a painting, or an idea for a painting.
PG: But they come as—
DL: More often than not, small fragments. I like to think of it as: In the other room, the puzzle is all together. But they keep flipping in just one piece at a time.
PG: In the other room?
DL: Over there.
PG: In a sense, David, there’s always another room somewhere.
DL: Mm-hmm. That’s a beautiful thing to think about.
PG: Let’s think about it a bit.
DL: No, you think about it.
For me, the tell is an itch. I’m holding a single puzzle piece (a shard of a sentence/style/image). The rest is work.
I like the idea that ideas float in the universe, because it maps neatly onto large language models. The ideas are there, but the connections still have to be found.