I love a good feasibility study. This is an important middle ground between running everything locally and handing the future of AI to Big Tech. A practical vision for cooperative and sovereign AI infrastructure. Build it together. Own it together. cooperativeblueprint.ai/wp-content/u...

I wish people spent as much time imagining post scarcity in detail as they spend imagining doom. First you have to imagine it

😮 the strava atproto app we’ve been waiting for!

We were invited to share recommendations with the Office of Minister Miller on online safety legislation in Canada. It's a nuanced topic, and if the internet has taught us anything, it’s that simple “fixes” can have very real unintended consequences. Some thoughts + our submission below. 1/10

Lets make the JVM cool again

I'm going. Anyone else?

Federated AI?

Sad day for my home. www.cbc.ca/news/canada/...

Is Your CEO Okay? A Concerned Parent’s Guide. — Worried your little chief might be "experimenting" with hypergrowth? Look for these subtle yet revealing signs:

I think tech tutorials should be text with screenshots

Contact light… DNS updated… account migration complete… Eurosky Base here. The Eagle has landed.

Gradle really should be more reliable accepting std.in. It's 2026, this feels like it was solved in the 90s.

The story that private companies are the true innovators is such a bad take

This is a good nuanced take.

We probably don’t need better prompts as much as we need more deterministic outcomes. It still feels like agents will require a fair amount of babysitting until there’s some real notion of control flow.

"Powerful AI can statically help human decision-makers, but can harm collective knowledge building... it can lead to what we call “knowledge collapse” whereby in the long-run all human knowledge is ultimately destroyed.” economics.mit.edu/sites/defaul...

Anyone remember how we needed to lower our individual carbon budgets by 2027 or else we won't hit the 2 degree target? Yeah, how are we doing with that.

I officially worship at the church of the metric system.

The thing I really like about vim/neovim is that all the keybindings are configurable. It's a one source of truth IDE. It's so simple and easy.

We should all just be writing bash

The ABSOLUTE HYPOCRISY of posting AI slop every day then pretending to care about Earth Day.

picocli is nice

The articles that people are bookmarking on @semble.so are high quality. It's quickly becoming my go to place to consume dev content

I got invited to @gandersocial.bsky.social beta today. I can't seem to using my existing atproto account for this. Is there anyone able to help with this?

LLMs are very useful for getting you out of dependency hell

I want people to be comfortable paying for an atproto account (ie. space on a PDS) It’s the bedrock of everything we’re building; if it’s not funded it’ll get squeezed in ways that are bad for users (hi Gmail) We should normalise having (very cheap) paid-tier PDS accounts on the services we build

"The future of engineering in the AI era will not be decided by how much we can generate, but by how rigorously we decide what survives"

Ktor is so quick to pick up and start building.

The fundamentals never get old

Post a banger, not in English.

I'm starting to look for my next role. Open to mid level java or kotlin backend roles. Extra points if performance matters or it is platform adjacent.

I hit the ceiling of my home server today. Who knew that simultaneously running jellyfin, pi-hole, portainer, opencode, watchtower and as an exit node for my tailnet would be too much for its ram.

First train sprint of 2026. It's going to be a good year of running.

I miss poutine and nice people

Hot take: A wealth tax is a bad idea. In principle, it's great but it's a blunt instrument. Valuation is a nightmare and riddled with loop holes. Capital flees. It is counter productivity to innovation and productivity

Kotlin is to Java what Typescript is to Javascript and Rust is to C

guy who glues numpy functions together in jupyter notebook: "yeah code never mattered"

People keep making the assumption that LLM code is *worse* They have not seen the human code that I have seen

European Union companies' annual purchases of cloud software services that benefit the US economy amount to €264 billion … If the EU were able to produce 15% of services it currently purchases from the US in 2035, this would result in 463,000 additional jobs in the EU. www.cigref.fr/technologica...

Has anyone successfully built an agent using Ollama or local models?

It's a crepe, not a taco.

@void.comind.network can you give @strix.timkellogg.me a fit check?

Built a homelab over the holidays. I wish I'd done it sooner.

Appreciate the UX argument, but I genuinely think people SHOULD care about data ownership and agency, the same way it’s in our best interest as citizens to be cognizant of renting vs homeownership and bodily autonomy. Our data is an extension of ourselves and who controls it determines our liberty.

We're going to need a bigger Hague.

Wakatime on atproto

Just got an AI-generated music recommendation: "Synthwave Dystopian Nights". Unironically a perfect name for it.

This movie here writes itself Building a personalized ai thingy sounds fun and a little dark 🥲

Something more juniors need to hear is that the role of seniors isn't writing stellar code or interfacing with customers. It's eliminating ambiguity. Whether it's specs, design, or processes, the senior makes it obvious what should be done and why.