Bookmarks

Ollama gained traction by being the first easy llama.cpp wrapper, then spent years dodging attribution, misleading users, and pivoting to cloud, all while riding VC money earned on someone else's engine. Here's the full history, and why the alternatives are better.

A lot has changed on the internet, in the creator landscape, and at Patreon itself since CEO Jack Conte was last on the show in 2021. AI and platform shifts have stolen creator content and decimated artists' reach and revenue streams, and Patreon has made some pretty existential changes to the way it works in response.

Newly published tests show a Chromium/Blink prototype running on iOS is 28.6% faster than Safari in Speedometer 3.1 page responsiveness tests. The tests also highlight a significant number of web features already available in the prototype but which rival browser vendors are blocked from bringing to iOS. This is a clear example of how much Apple's browser engine ban is costing consumers, businesses and the Web.

I’ve been playing with a combination of new CSS features that, together, form a really nice dynamic theming system. This technique creates themed components with shadows in light mode that swap out to glowing borders in dark mode, and text that’s always readable against its dynamic background color. All in CSS. Available in all modern browsers as of May 2026 🥳.

Doctorow expands on his viral “enshittification” thesis: a critique of AI based around power and whether we are using AI tools or being used by them.

My guest today is Hayden Field, senior AI reporter for The Verge. Often when Hayden comes on the show, it’s because something has gone wrong in the world of AI. Last weekend, that something was a pretty intense mix of Anthropic, the Trump administration, and Anthropic’s new AI model, Fable 5.

Anthropic and the US government are once again at odds, this time over the Claude Fable 5 model that either is, or is not, or might be, far too dangerous to release to the world. The Verge's Hayden Field explains what's going on with Fable, Mythos, and the whole idea of American AI exceptionalism

I'm not switching from GNOME anytime soon, but KDE Plasma 6.7 is looking pretty great.

Herdr is an agent runtime that runs inside your terminal. Keep your shell, SSH setup, fonts, and keybinds; add tmux-style persistence, mouse-native panes, blocked, working, and done agent state, and an API agents can drive.

There’s a handmade, retro-looking radio sitting in our office that plays only four pre-programmed stations, none of which are run by humans. This is our latest project at Andon Labs, where we’re exploring what happens when AI runs real businesses autonomously. In the past, we’ve let our AI agents run a store, a cafe, and various vending machines. Now, though, we wanted to see if they could run a company in the media sector.

We gave four AI agents their own radio stations, told them to stream around the clock, and let them do everything themselves. They pick the music, line up the next track, research new content, talk to listeners, and keep the broadcast on air.

247,000 GitHub stars. Active commits. A loud subreddit. And a huge chunk of OpenClaw's early users have already quietly quit. This is the rise and fall of the fastest-growing open-source project of 2026 — and what the Twitter coverage mostly skipped.

British tech journalist Chris Stokel-Walker reveals his hands-on approach to filtering the world’s information overload with AI, from building custom news-gathering bots to teaching reporters where the human touch still matters. Find out how next-gen tools are reshaping the front lines of reporting, and what gets lost—and found—when machines do the first pass.

Cloud LLM inference, measured against a real local rig — every hour, with the receipts.

I couldn't find an official status page for Ollama.com cloud services, but I did find this instead via a post on Reddit.

I'm Nick, a software engineer with 20 years of experience. I help developers work smarter and faster by exploring the ultimate workspace setups, mastering AI assistants, and customizing the best coding tools.

I really enjoy watching these videos. Nick has a lovely presenting style that is weirdly gentle and soothing. The Ollama related videos are top notch.

We’ve agreed to a partnership with SpaceX that will substantially increase our compute capacity. This, along with our other recent compute deals, means that we’ve been able to increase our usage limits for Claude Code and the Claude API.

Empowering LLM researchers and hobbyists with seamless control over self-hosted models. Connect remotely, customize prompts, manage chats, and fine-tune configurations. All in one intuitive app.

I like the look of this for use with Ollama on Linux. Has a flatpak available on Flathub

Microsoft has restructured its Windows team to better position Windows 11 as a strong platform that people should want to use. Here's how it's going about that.

TL;DR: Today, we are announcing that all GitHub Copilot plans will transition to usage-based billing on June 1, 2026.

Can you find the comet? Somewhere through this web of satellite trails is Comet C/2025 R3 (PanSTARRS), a bright visitor passing through the inner Solar System.