FeedCity

Most people's mental model of Claude Code is that "it's just a TUI" but it should really be closer to "a small game engine".

For each frame our pipeline constructs a scene graph with React then

-> layouts elements
-> rasterizes them to a 2d screen
-> diffs that against the previous screen
-> finally uses the diff to generate ANSI sequences to draw

We have a ~16ms frame budget so we have roughly ~5ms to go from the React scene graph to ANSI written.

— Thariq Shihipar

Tags: react, claude-code

I still code as an engineering manager. Not to stay busy, but to show my team what good looks like.

Something actually new? There’s finally been a big leap forward in coding tech unlocked by AI — not just “it’s doing some work for me”, but “we couldn’t do this before”. What’s new are a few smart systems that let coders control fleets of dozens of coding bots, all working i...

what's a few systemd services in initramfs between friends?

Background and bytecode design The ZJIT compiler compiles Ruby bytecode (YARV) to machine code. It starts by transforming the stack machine bytecode into a high-level graph-based intermediate representation called HIR. We use a more or less typical1 control-flow graph (CFG...

Claude's new constitution Late last year Richard Weiss found something interesting while poking around with the just-released Claude Opus 4.5: he was able to talk the model into regurgitating a document which was not part of the system prompt but appeared instead to be baked...

Dinsdag 27 januari om 18:30 is er in Den Haag in de Tweede Kamer een rondetafeloverleg over de consequenties van de (beoogde) overname van Solvinity voor DigiD. Er komen diverse sprekers, onder andere namens Bits of Freedom, de Dutch Cloud Community, Clingendael, NLDigital (big tech), Stichting Digitale Infrastructuur Nederland. Ook zijn Brenno de Winter en Paul Timmers van de partij. En ik ben er ook. En u kunt ook komen! Het zou weleens een leuk signaal zijn als de tribune vol zat - soms komen er maar drie mensen opdraven.

Today's links Google's AI pricing plan: From each according to their ability (to pay). Hey look at this: Delights to delectate. Object permanence: Disney buys Pixar; Bruces on "vernacular video"; Hickey paralysis; RIP David Hartwell; Personalized pricing; They were warned. Up...

After 16 years on the LAMP stack, I finished migrating this website from Drupal to Hugo a few weeks ago.

What's old is new, as this blog was originally built with Thingamablog, a Java-based Static Site Generator (SSG) I ran on my Mac to generate HTML and FTP it up to my first webserver (over 20 years ago!).

The main reason I moved from an SSG to Drupal was to add comments. I wanted my blog to have the same level of interactivity I had pre-Thingamablog, when I was (briefly) on Xanga.com.

Electricity use of AI coding agents

Previous work estimating the energy and water cost of LLMs has generally focused on the cost per prompt using a consumer-level system such as ChatGPT.

Simon P. Couch notes that coding agents such as Claude Code use way more tokens in response to tasks, often burning through many thousands of tokens of many tool calls.

As a heavy Claude Code user, Simon estimates his own usage at the equivalent of 4,400 "typical queries" to an LLM, for an equivalent of around $15-$20 in daily API token spend. He figures that to be about the same as running a dishwasher once or the daily energy used by a domestic refrigerator.

Via Hacker News

Tags: ai, generative-ai, llms, ai-ethics, ai-energy-usage, coding-agents, claude-code

read on, queen

Giving University Exams in the Age of Chatbots

Detailed and thoughtful description of an open-book and open-chatbot exam run by Ploum at École Polytechnique de Louvain for an "Open Source Strategies" class.

Students were told they could use chatbots during the exam but they had to announce their intention to do so in advance, share their prompts and take full accountability for any mistakes they made.

Only 3 out of 60 students chose to use chatbots. Ploum surveyed half of the class to help understand their motivations.

Via lobste.rs

Tags: education, ai, generative-ai, llms, ai-ethics

I redesigned my www website. Why? The end of year / holiday break is a great time to work on such things. I wanted to scratch an itch. Websites are a worry stone [gestures at current state of the world] Do I really need a reason? Nope. I read something along the lines of “...

A list of tools and items that JustWork (tm)

Today's links AI is how bosses wage war on "professions": The only people *required* to tell their bosses to fuck off. Hey look at this: Delights to delectate. Object permanence: Enshittification v VC; Byrne v DRM; Forger does it for the exposure; Beach trash hotel; Mafia nic...

jordanhubbard/nanolang Plenty of people have mused about what a new programming language specifically designed to be used by LLMs might look like. Jordan Hubbard (co-founder of FreeBSD, with serious stints at Apple and NVIDIA) just released exactly that. A minimal, LLM-frie...

Today's links Social media without socializing: For Zuck, your friends are a problem to be solved. Hey look at this: Delights to delectate. Object permanence: "Among Others"; Vagina tea-balls; MLK was a socialist; Planet Money's free Gatsby audiobook; Bounty for wiretapping l...

On escaping burnout in The New Age.