It’s like if Rosetta Stone were a rap song - open.spotify.com/track/5gpyyq...

From "not everything is a chat" category for LLM apps. A couple of interesting experiments from Sean Goedecke: www.endlesswiki.com/wiki/main_page (an infinite, AI-generated wiki) www.autodeck.pro (AI driven spaced repetition app)

Frenchies took over MS's docs department - www.typescriptlang.org/docs/handboo...

A bitter lesson about the bitter lesson - www.dbreunig.com/2025/08/01/d... via @dbreunig.bsky.social. Excellent as always

1k

How to Setup an Ergonomic Workstation - More productivity, less back / neck / wrist pain hamy.xyz/blog/2025-09...

Claude code agentic loop breakdown blog.promptlayer.com/claude-code-... h/t @timkellogg.me

The Litany Against Fear

Yup. I've been referring to it as an "A-word"

"have you tried turning it off and on again" (IT crowd reference ) spotted in gemini-cli. is this a sign that engineers are in charge in gemini department?

This looks awesome. And the tune is 👌

Not here do defend Mr Armstrong but > and nothing about quality, maintainability, correctness, lack of duplication - right? it's in the second "paragraph" of tweet. "... needs to be reviewed" etc To borrow, Balaji's "0 AI is SLOW, 100% AI is SLOP" - companies are trying to find a sweet spot

But of course, Italians are behind the word "extravaganza". 🤌

Some of the most remarkable lost artefacts from the ancient world were the titanic wrecks of the Nemi ships, built for a Roman emperor. In their first-century heyday they held gardens, palaces and baths in a floating wonderland. But barely a decade after their recovery, they were lost forever.

Wow. I think about this a lot. What a great way to put it!

Finished this just in time to make a print for Rose City! It's a colorized and repurposed version of my Porco Rosso drawing from my old Ghibli-themed inktober set back in 2016!

What a great example of partial autonomy llm feature plugged into existing UI. With some additional affordances this could be a beast of a feature

This is very good ⤵️

"Machinery is aggressive. The weaver becomes a web, the machinist a machine. If you do not use the tools, they use you. All tools are in one sense edge-tools, and dangerous." - Ralph Waldo Emerson

alias claudius="claude --dangerously-skip-permissions"

Interesting typographical choices here (from en.wikipedia.org/wiki/COBOL)

Blueberry is new strawberry

[NOTES] How Kimi K2 post trained on tool usage + some general considerations regarding tools usage in LLMs. Another excellent piece from @dbreunig.bsky.social - www.dbreunig.com/2025/07/30/h...

TIL: all those whimsical gerunds in Claude code get generated on the fly via Haiku. I am lost for words. Should have asked Haiku for an appropriate verbal concoction to use

A lot of interesting observations here - balajis.com/p/ai-is-poly... - there many AIs not 1 AI - AI is middle-to-middle, not end to end - amplified intelligence - AI lets you do any job (ok'ish) - AI takes prev AI's job - AI is better for visuals than verbals - killer AI: drones [...]

Ok. This book is insane! Kudos to @javierblas.bsky.social for making this such a page turner.

1) AI researchers have often tried to build knowledge into their agents 2) this helps in the short term 3) [but] plateaus and inhibits further progress long-term 4) breakthrough progress eventually arrives [via] scaling computation by search and learning www.incompleteideas.net/IncIdeas/Bit...

🤌 mai una gioia

I think i am going to give this a try as well

[Notes] Mini SWE agent - a radically simple agentic loop mini-swe-agent.com

They: your product is just a Chat GPT wrapper You: genius.com/Chase-and-st...

🤌 non c'è due senza tre

Trying to get Claude to give me a nice ASCII art style Winnie-the-Pooh. I mentioned Claude Monet (half jokingly) and things escalated.

Recently, I suggested a way forward may be AI systems that maintain their cognitive and knowledge systems separately. Now Karpathy - one of AI's heavy hitters - is getting on the train. Choo choo, folks - we're on our way to ASI (artificial simple intelligence) 🚂

Marcel chauffe! 😂 love it. `marcel grève` is a chef's kiss github.com/brouberol/ma...

Sharing one more time for folks who did not see the announcement during the weekend: I'm building a box of building blocks to customise constraints/commands in terminal agent chats. There's some neat tricks you can do already, details in vid: youtu.be/XLbi00RI0V8

This is wild!

This looks awesome!

Mr Knuth must be delighted

This thread is 🔥. I am still wrapping my head around @badlogic.bsky.social's original post.

Here is the link to the GitHub repository: github.com/Shopify/roast There are quite a few examples available in the repo, see here -> github.com/Shopify/roas...

Watched “High noon” last night. Here is the old marshal explaining Cooper why no one is willing to help him defend their town: “People gotta talk themselves into law and order before they do anything about it. Maybe because down deep they don't care. They just don't care.”

[Notes] How to Fix Your Context - www.dbreunig.com/2025/06/26/h... via @dbreunig.bsky.social

It took me a while to finally get my hands on uv. Sounds like there's no more pip for me - github.com/astral-sh/uv

I've just integrated Claude code into my blog github.com/callmephilip... to help converting my bsky threads into blog posts. Here's the first post created using this new pipeline - callmephilip.com/posts/notes-...

Starting a list of practicing AI optimists on bsky. Who did I miss? go.bsky.app/LSdjm8s

[Notes] How Long Contexts Fail - www.dbreunig.com/2025/06/22/h... via @dbreunig.bsky.social

Did traditional programming condition humans to act more like machines and aren't LLMs liberating us (developers) from this conditioning then? Instead of "coding" as identity, let's do "building stuff people actually want". from corecursive.com/coding-agent... h/t @simonwillison.net

I confess my first thought was about missing fingers despite my general positive outlook on generative AI. Might be my sense of humour.