On the Bun rewrite, Hazel makes a great point: the ~$160,000 spend was with a *very* wasteful / un-optimized approach to AI use. As we learn and share better techniques for agent orchestration, we can likely drive down the cost in tokens (and therefore $) significantly.

"We would never be comfortable to give a 3rd party tool access to our database" - from a dev at a mid-sized company that built its own coding harness that is used by 70%+ of all PRs They would not let any vendor run this kind of infra, so they brought it in-house. They are happy they did this

🚀 Zed v1.10 is out! Thanks to our friends at @hf.co we finally have support for llama.cpp in Zed.

just released port-file on crates.io -- if you tell a child process to bind a service to port 0, facilities for the child to tell the parent what port it actually bound to. A very simple problem solved very well

After over 3 years of layoffs, I’ve felt this year was the tipping point where it seems most people in tech are now burned out on the industry. AI is the missing ingredient which has now made working in tech simultaneously exciting and an excruciating ordeal. It looks like the feeling is widespread

social media platforms that operate to maximize engagement, nazi ideology built on hate and rage, weak regulation and enforcement... democracy is dying before our eyes and big tech is fuel powering it www.irishtimes.com/world/europe...

Talking w a larger startup whose product seems to not have degraded (vs many other products), even though they also use a AI for coding everywhere. Asked an infra lead: how? "Even before AI, we were paranoid about quality, and built this irrationally thorough E2E testing system"

If you missed it: Python Markdown has documented all of its extensions as being in maintenance mode. Good timing: we just rewrote Python Markdown in Rust, and it already powers Zensical Studio. It’s 10-20x faster and will be Open Sourced later this year. github.com/Python-Markd...

It’s here! Today, we’re announcing the launch of Blacksky-Only Posts! Now, anyone with a blacksky.app handle or anyone on the Blacksky feed can make a private post on blacksky.community. This marks a significant step forward in advancing permissioned data on the open social web.

This is a demo of Raspberry Pi running a whisper model entirely locally to do voice transcription on the edge and that's the tweet hit send.

Open source in HuggingFace! 🤗 🙌

after about ~4 hours of claude tag i'm calling the ball - anthropic's product people have officially lost the plot. 'ill-conceived' doesn't even begin to describe my feelings on this. also, they give you 25k in credit because it defaults to fable 5 and also watches _every slack message sent_

What would a rainbow look like on Tatooine? Watch the latest What If? video in collaboration with @minuteearth.bsky.social! www.youtube.com/watch?v=BO8r...

In the Bible, the punishment was to create more languages. This time around the punishment will be reversed and we’ll all end up writing rust

everywhere you look, things are goin' great

I have high hopes for @mu.social as a proving ground for democratic social media, given ample support from the atmosphere community. With the help of civilized-discourse expert @jonathanwarden.com we can have Community Notes on atproto, starting with mu.

From a Head of Engineering who spent a week in SF, meeting a bunch of AI startups: "When I was back, I wrote a memo about my impressions. The #1 was how these amazing startups I met: they mostly did not have a business. But we do. So be VERY careful in copying what they do." (cont'd)

Existing communities on Discord shouldn't have to choose between the critical mass of members they've built up on the legacy platform versus the agency afforded to them by @roomy.space Introducing Roomy's first premium feature: Discord-Roomy bridging! 🌉 (free for non-profits)

well that didnt take long www.reuters.com/world/beijin...

Gemma 4 technical report is out! lots of cool stuff, check it out! arxiv.org/abs/2607.02770

Agents-A1 🤖🔬 New agentic model from Shanghai AI Lab, InternScience team huggingface.co/collections/... - 35B MoE (built on Qwen3.5-35B-A3B) - Apache 2.0 - 256K context - Trained for long-horizon agent work - Includes quantized variants

coming in atuin 18.17 🐢✨ ctrl-r search is 78x faster 780ms -> 10ms, across >500k entries

Keith Adams thought AI would democratize software, raise the floor. His dream: a version of Doom where all the monsters have my face. But its been "more of an amplifier of differences in facility than a leveler." Expected a compressor. Got the opposite.

After last week, have a couple more additions to this list. But before I do… Do you all have any suggestions for additional tips? www.dbreunig.com/2026/05/04/1...

there's infinite talk about ai taking jobs but i hardly see people talking about it disrupting the company that employs them we're tossing out products we've used forever it's not a price thing, we just get a better experience by hooking ai up to raw data

I had a chat with Lori Lorusso and Niko Matsakis about the Rust Foundation Maintainers Fund Funding open source is a huge topic right now, the Rust Foundation has some great ideas. It will be exciting to watch this one grow and evolve #OpenSourceSecurity #rust #RustFoundation

💫 GM Tanglers! Its release time 💫 Here's whats new: - Full text code-search - Focus mode - Tangled XRPC API - A new homepage - MicroVM powered CI engine

🙋‍♂️ want a zero-config backup of your github repos and think @tangled.org is pretty cool? then synchub.to is for you! 👉 it's a github app that automatically syncs your public github repos to tangled ✨ let me know what you think! 🙏

Let me know if this is just me: Noticed someone I know who is very "AI-pilled" and uses agents 24/7 to... start to talk IRL noticeably more like these LLMs write. Eg more heavily using adjectives like "geniune", frequently terms like "the shape of" and many more examples

Big companies aren't hiring juniors anymore, which sucks for young people. This year we hired Ian, a CompSci student, as our summer intern at @transistor.fm He's been making a podcast about his experience, and just interviewed me: share.transistor.fm/s/abfc686a?c...

A creepy story: an old Fedora Linux contributor’s account was apparently hijacked, and the attacker then started faking activity with an LLM. Likely to build trust before launching a supply chain attack. lwn.net/Articles/107...

@sifa.id is awesome 😎 Imported my profile from LinkedIn and 🚀 sifa.id/p/jehrhardt....

So what the hell is a loop and why won't people shut up about them? https://www.linkedin.com/pulse/what-hell-loop-anyway-laurie-voss-ldmdc/

DawrfStar with DeepSeek 4 Flash 4 bit, PRO 2 bit, and GLM 5.2 4 bit. M3 Ultra 512GB. Results on hard programming tasks with GPT 5.5 as a judge. So the GLM 5.2 branch is going to be merged and supported for CUDA + Strix Halo as well.

it takes effort to read something so I want to read things that have had thought put into them with llms it's possible (though not inevitable!) to create writing without adequate thought so when people use llm-generated text it makes it very difficult to tell

Łukasz Niemier is doing cool stuff. See their Sifa profile 🛠️ @hauleth.dev

It fills me with joy that so many of you are organizing book club and group reads of The Psychology of Software Teams! What a culture champion move. To support, I am putting together a BOOK CLUB DISCUSSION GUIDE for the Psychology of Software Teams 🥰🔥

wait this was ON PURPOSE? and they're making it opt-in "for now"? github.com/anthropics/c... unsafe at any speed (their product decisions)

I've been thinking about my previous life in data science/applied computational social science and how many people have always been cut off from that. I hear a lot of examples of "I can't believe I can do this" from friends about LLMs that are along the lines of finally being able to look at data

Did you know? Renewables have just overtaken coal as the world's largest source of electricity 34% – global renewable share, up 2x in 20yrs 48% – EU 37% – China 26% – US 24% – India

AI startup Twelve Labs is raising $100M from Amazon, NEA, Naver etc to accelerate AI models for video The Seoul & Silicon Valley-based startup's models can search & understand vast archives of Hollywood studios, sports franchises etc, making them monetizable www.bloomberg.com/news/article... #AI

There are few people who have impacted the software engineering industry like @kentbeck.com has. He'd never before told his career story from start to today in one sitting - until now. What a treat: • YouTube: www.youtube.com/watch?v=ddHQ... • Spotify: open.spotify.com/episode/3M6h...

Maintainers spend hours every week closing duplicate issues. 😮‍💨 That changes with duplicate detection. Surface up to 3 matching issues as you create one, before you hit submit.

NEW BLOG FOR YOU Google's energy consumption numbers in their new climate report are mind-blowing. 2 years ago they flipped from linear to exponential growth, and their climate impact is blowing out, too. A WILD testament to the obscene bloat and waste of GenAI: ketanjoshi.co/2026/07/01/g...

looks like gpt-5.6 has Pro versions for all Luna, Terra & Sol, and the smallest (Luna) is still better than got-5.5-Pro presumably 6x single-model size, so Luna-Pro is likely: - $6 input / $0.60 cached - $36 output

Godot’s new AI policy is here. the TL;DR: all code must human authored.

💗 NanoBanana 2 Lite is my go-to now for image generation and editing! Very, *very* high quality, at only a small fraction of the price of our other image editing models. 🙌 Bonus: Omni is now available via the API! Create and edit *even videos* with natural language: 🔗 blog.google/innovation-a...

for what it’s worth, the prevailing beliefs about bluesky on threads are that it’s dry, boring, and everyone here just likes to be angry all day. so show them interesting and fun things they’re missing.

Happy to share that we have 8 papers accepted to #ICML2026! 🎉 Most of the authors will be traveling to Seoul, Korea 🇰🇷 — feel free to reach out and chat with them if you’re interested in the work. See the full thread below for our papers 👇

Who could have possibly seen it coming that the problem will not be AI taking over the world but instead assholes ruining everything themselves again, without any AGI necessary.