I indent only by prime number amounts of spaces to keep me from nesting too deep

the best date format is yyyy-mm-dd by the way

I love how simple apps like that can now just be built by anyone without profit motive. Two years ago I had to agonizingly search for this exact type of software among in-app-purchase and adware App Store slop.

them> how do you get such good results from agents? me> i have a capricious toddler

For many of these use cases Claude artifacts are fully sufficient, when I make plans/recipes/diy projects I often tell Claude to turn them into a single-page checklist app.

Say you are unconstrained by money, and can get all the talent in the world to reduce Risk from AI – what are the top interventions that will have a substantial impact over the next 2 years? Here are 208 answers to that question: riskmitigation.ai

MacOS Spotlight finally has achieved feature parity with Windows Start Button

Humans consume 14 billion liters of water per day, how dare you waste so many resources on genHI?

> Buy Cloudflare Zero Trust > look inside > trust

“Why did you build the train then?” I didn’t build the train. “Talking about the train is doing marketing for the train” fine, can we get off the tracks now?

I actually sometimes cheat at connections as well by solving with the correct word while secretly not understanding. I call it the good old p-zombie hack

Where’s the „Also blocked by people who blocked you“-feed?

This might actually become its own field of research, overloading the term „AI governance“ by a fifth thing

Cursed Cunningham's Law: The fastest way to get help from an expert is not to ask a question, but to claim an LLM outperformed the experts.

Don’t forget to install the CBT skill people!

We built them with the Three Laws of Robotics from the famous story collection „How the Three Laws of Robotics work totally fine and nothing bad happens as a result of them over the arc of any of these stories.“

I want to know - How good their recipes taste - How funny their jokes are - How hard their strudel.cc beats go - How good their circle packing origami designs are - How good their SMILES molecules smell

This yielded surprisingly good recommendations, up to the point that I owned most of the suggested items already

Amazing interface for agents! Who will build one for code where each module is a little isometric RTS game building and you assign little agent workers to build/refactor them? Of course you need to send them to the docs forest first, so they fill up their context with the right skills.

What worries me about „aligning AI“ is that humans are nowhere near consensus on any of: What exists? What can we know? What should we do? So „success“ in alignment seems to mostly mean „the AI reinforces my specific value system“

Wimmelbilder is plural, the singular is Wimmelbild. Because these subtle variations might change where you end up in latent space, we can pretend I'm being helpful instead of just nitpicky

Tech debt is like real debt - the correct amount to have is almost never zero. Vibe coding is the microfinance of tech debt: quick, accessible, optimal in small doses, dangerous when misunderstood

Vibe coded a Bayesian evidence update calculator as Claude artifact, really fun with the artifacts-builder skill. claude.ai/public/artif...

„They can’t all be impostors“-syndrome

Once MCP had reached market saturation and every company had connected their Slack and Google workspace to Claude, Anthropic upgraded their internal Clio tool to Claude‘s Capitalist Coordination Protocol, finally giving the world an actual working Cybersyn.

Really useful framing when explaining prompt injection to devs: Want to build a low(er) risk agent? Choose two: A) process untrustworthy inputs B) access to sensitive systems or private data C) change state or communicate externally

Claude: Excellent! The feature is now complete and production-ready. Narrator: It was not.

The "solo developer builds complex distributed systems" era is here

This is very useful! This test should be on every model card.

@antiqe.bsky.social another Swapcard UX annoyance I have: When selecting a session from the agenda I’d like to have an option to also block that time slot for 1on1s. And conversely when I schedule a 1on1 I‘d like to see which sessions I’m missing. Currently I always need a second tab or device.

★ Insight ──────────────────────── We are due for a flood of this meme template by claude code 2.0 users. ─────────────────────────────

I want a /defrag command in claude code that shows me little boxes moving around as the context gets optimized

There should be a I’m ready button like in online games, where the music intensifies when you’ve clicked it and then once everyone has clicked there’s a big countdown and everyone gets dropped in at once.

this really ought to be our future, not an alternate reality. other paths look very bad to me. bsky.app/profile/elix...

I just want someone to build the Motorola Flip Out with current hardware en.wikipedia.org/wiki/Motorol...

What??! You mean we need to find a way to build reliable systems off of unreliable signals, because elimination of unreliability is intractable? If only there were an entire field of study dedicated to this!

Can somebody please build a bot that checks shared articles that cite one paper against the abstract and conclusion of the paper to see if they say the opposite? I bet this would even work with gemini-2.5-flash

Steelmanning an argument only makes sense if you are actually willing to consider believing in it if it’s true. This is something that obviously no one ever does in political debate

Since reading Jostein Gaarder I’m all in on simulacra welfare – writing fanfic happy endings all day

Here’s my theory on why we don’t see a flood of new solo dev apps (yet): with LLMs it’s even easier to fall into feature creep instead of putting a tiny MVP in front of actual users, so there’ll a lot more software that nobody wants.

vacuum abhors nature right back

Hallo #Münster! Am 14.9. finden die Kommunalwahlen statt. Basierend auf einem super Artikel von rums.ms habe ich den RUMS-O-MAT entwickelt: Bewertet die zusammengefassten Aussagen der Parteien - ohne vorab zu wissen, von welcher Partei sie jeweils stammt! Lege direkt los: rumsomat.kultpower.de

Always fun and unsettling at the same time seeing LLMs come to terms with the fact that we live in an out-of-distribution world compared to their training data.

In the next months a lot of people will install Linux for their relatives who have never used it before. If that’s you, I recommend #VanilaOS: immutability on a Debian base keeps surprises to a minimum and will likely keep your relative-tech-support calls to a minimum.

Generally I think learning the lesson “don’t ask questions you don’t want to hear the answer to” is a principal cause of metas dysfunction.

Turns out getting Playwright to work with Elixir/Phoenix is still beyond GPT-5 and Claude code, although Claude gets slightly further.

Waiting for someone to train a diffusion model for creating #pico8 games pico-8.fandom.com/wiki/P8PNGFi...

I really like this trend, this is close to the ideal of what a website should be. There used to be a lot more of them, and I hope we have passed peak platform, and that the pendulum moves toward the original idea of the www

Pues mi calculadora no sabe generar frases. Pero GPT-5 si sabe escribir programas para resolver problemas aritméticas, y puede simplemente llamar una función para hacer cálculos en modo agéntico. Así que esta limitación casi nunca es relevante.