bookmarks

links to articles, websites, and resources I find interesting or useful

🔖

Sophia French on microblogging being the next big art movement:

As mega media players and their algorithms burrow themselves further into our lives, a return to the tactics of the early days of microblogging signals more than just nostalgic yearning. It could be the way forward for how artists share their work, process, and life online – on their own terms.

🔖

Andros Fenollosa on why we want small systems instead of decentralized ones when talking about online communities.

There are no commercial interests and people know each other. It’s like a shoe shop in a small town. The shop can only serve a few customers because there is little space, you know the owner’s name (maybe even his wife’s) and the shelves don’t offer much variety; and yet you know you will be well served and the owner wants you to leave with what is best for you.

🔖

Doug MacDowell writes about drawing data visualisations completely by hand.

“What are you working on these days?”
“Data visualizations.” I told him.
“Ah, you using algorithms, machine learning, cloud computing, things like that?”
“No.” I said. “I’m just trying to draw a line graph.”

My neighbor thought I was getting into some complex sh**. But what’s been more interesting to me lately than using […] is learning to draw data by hand. 50 Hours to Draw Some Lines is about spending more than a week on something that software can accomplish in 20 minutes - and a catalog of resources and methods acquired along the way.

🔖

Amanda Shendruk writes about the work that’s done after Burning Man, where 150 people line up, side by side and arms apart, and slow walk the entire playa looking for Moop (Matter Out of Place).

This forensic-style sweep takes weeks; everything they find is removed and logged. At the end, they’re left with a remarkable accounting of what 70,000 people left behind: The MOOP Map. And I’m obsessed.

The Moop Map is a pretty remarkable view of what is left behind:

Moop Map from Bunring Man 2025
🔖

Mario Zechner writes about the current usage of LLM coding agents in software development and how we just need to slow down our release cycles.

You have zero fucking idea what’s going on because you delegated all your agency to your agents. You let them run free, and they are merchants of complexity. They have seen many bad architectural decisions in their training data and throughout their RL training. You have told them to architect your application. Guess what the result is?

An immense amount of complexity, an amalgam of terrible cargo cult “industry best practices”, that you didn’t rein in before it was too late. But it’s worse than that.

🔖

Jacob Filipp writes about a moment in the 1980s when the Ontario Ministry of Education commissioned a computer designed specifically for students.

A time when public institutions believed they could build their own tools, and did.

1984 was an alternative reality where we could do things. Like build a computer in Ontario.

🔖

Ladybird (a new web browser), recently migrated part of their code base to a new programming language with the help of Claude Code and Codex (OpenAI’s code model). I thought this idea of human-directed LLM tasks is really great:

I used Claude Code and Codex for the translation. This was human-directed, not autonomous code generation. I decided what to port, in what order, and what the Rust code should look like. It was hundreds of small prompts, steering the agents where things needed to go. After the initial translation, I ran multiple passes of adversarial review, asking different models to analyze the code for mistakes and bad patterns.