<?xml version="1.0" encoding="UTF-8"?>
<opml version="2.0">
  <head>
    <title>Signal Sources</title>
    <dateCreated>Wed, 19 Aug 2026 00:00:00 +0000</dateCreated>
  </head>
  <body>
    <outline text="Research &amp; Labs" title="Research &amp; Labs">
      <outline type="rss" text="BAIR Blog" title="BAIR Blog" xmlUrl="https://bair.berkeley.edu/blog/feed.xml" htmlUrl="https://bair.berkeley.edu/blog/" description="Berkeley's AI lab publishes its own papers straight from the people who wrote them, no press office in between. A recent post traces a kernel-porting project called K-Search from CUDA to Apple Silicon, the kind of write-up that never makes it into a conference recap. (weekly to biweekly)"/>
      <outline type="rss" text="SAIL Blog" title="SAIL Blog" xmlUrl="https://ai.stanford.edu/blog/feed.xml" htmlUrl="https://ai.stanford.edu/blog/" description="Stanford's AI lab runs the same format — grad students and faculty writing up their own results — with a recurring &quot;Papers and Talks at [conference]&quot; roundup that works as a cheat sheet for what actually mattered at ICML or CVPR that year. (weekly to biweekly, with spikes around conferences)"/>
    </outline>
    <outline text="Engineering Blogs" title="Engineering Blogs">
      <outline type="rss" text="Fly.io Blog" title="Fly.io Blog" xmlUrl="https://fly.io/blog/feed.xml" htmlUrl="https://fly.io/blog/" description="Fly.io's own engineers narrate the mess of running a public cloud in first person: GPU fleet rollouts, a Rust gossip protocol called Corrosion, the sandboxing work behind their agent product. It reads like someone explaining a system to a colleague, not a polished case study. (biweekly to monthly)"/>
      <outline type="rss" text="Tailscale Blog" title="Tailscale Blog" xmlUrl="https://tailscale.com/blog/index.xml" htmlUrl="https://tailscale.com/blog" description="Alongside the product updates, Tailscale runs genuine postmortems. A recent one walks through hunting down a SQLite corruption bug that had apparently been sitting in the wild for sixteen years — that's the kind of debugging story worth the subscription on its own. (roughly weekly)"/>
      <outline type="rss" text="Turso Blog" title="Turso Blog" xmlUrl="https://turso.tech/blog/feed.xml" htmlUrl="https://turso.tech/blog" description="Glauber Costa's team is rebuilding SQLite in Rust, and they write about the internals as they go: MVCC, async I/O, giving every customer their own database at scale. Primary-source database engineering, updated almost weekly. (weekly)"/>
    </outline>
    <outline text="Newsletters" title="Newsletters">
      <outline type="rss" text="ChinAI Newsletter" title="ChinAI Newsletter" xmlUrl="https://chinai.substack.com/feed" htmlUrl="https://chinai.substack.com" description="Jeff Ding translates Chinese-language writing on AI policy and industry that most English-language coverage never touches, including reactions inside China to its own AI companion regulations. No other newsletter is doing this specific job. (weekly)"/>
      <outline type="rss" text="Interconnects" title="Interconnects" xmlUrl="https://www.interconnects.ai/feed" htmlUrl="https://www.interconnects.ai" description="Nathan Lambert works on post-training at Ai2, and it shows. His posts on frontier training methods and the open-model ecosystem read like notes from someone who runs the experiments, not someone summarizing a paper abstract. (one to three times a week)"/>
      <outline type="rss" text="AI as Normal Technology" title="AI as Normal Technology" xmlUrl="https://www.normaltech.ai/feed" htmlUrl="https://www.normaltech.ai" description="Arvind Narayanan and Sayash Kapoor, the pair behind the &quot;AI Snake Oil&quot; project, bring the same skepticism here, testing what AI agents can and can't actually do instead of assuming it. A useful counterweight if your feed already leans toward hype. (roughly monthly)"/>
    </outline>
    <outline text="Changelogs &amp; Release Feeds" title="Changelogs &amp; Release Feeds">
      <outline type="rss" text="vLLM Releases" title="vLLM Releases" xmlUrl="https://github.com/vllm-project/vllm/releases.atom" htmlUrl="https://github.com/vllm-project/vllm/releases" description="The changelog for the inference engine most self-hosted LLM deployments run on. New model support and quantization tricks show up here, sometimes several times a week, before anyone gets around to writing a blog post about them. (multiple releases per week)"/>
      <outline type="rss" text="Ollama Releases" title="Ollama Releases" xmlUrl="https://github.com/ollama/ollama/releases.atom" htmlUrl="https://github.com/ollama/ollama/releases" description="Ollama ships constantly, and the release notes are where new backend support (MLX, for one) and CLI changes land first, often as release candidates before they hit the stable channel. (multiple releases per week)"/>
    </outline>
    <outline text="Practitioners &amp; Independents" title="Practitioners &amp; Independents">
      <outline type="rss" text="Eugene Yan" title="Eugene Yan" xmlUrl="https://eugeneyan.com/rss/" htmlUrl="https://eugeneyan.com/" description="Eugene builds evals and recommendation systems for a living, currently at Anthropic and previously at Amazon and Alibaba. His posts read like internal engineering notes someone decided to publish: heavy on what actually shipped, light on speculation. (roughly monthly)"/>
      <outline type="rss" text="Vicki Boykis" title="Vicki Boykis" xmlUrl="https://vickiboykis.com/index.xml" htmlUrl="https://vickiboykis.com/" description="A founding ML engineer with a long resume in embeddings and information retrieval, writing with a craft-first, occasionally contrarian streak. Her recent post on running local models well is more useful than most vendor comparison posts. (roughly monthly)"/>
      <outline type="rss" text="antirez" title="antirez" xmlUrl="https://antirez.com/rss" htmlUrl="https://antirez.com/" description="Salvatore Sanfilippo built Redis and has spent two decades as a working systems programmer. Now he's turning that same skepticism on AI coding tools and lab governance, posting every few days to weekly with no patience for hype or doom. (every few days to weekly)"/>
    </outline>
    <outline text="Video &amp; Podcasts" title="Video &amp; Podcasts">
      <outline type="rss" text="The TWIML AI Podcast" title="The TWIML AI Podcast" xmlUrl="https://feeds.megaphone.fm/MLN2155636147" htmlUrl="https://twimlai.com/shows/" description="Sam Charrington has been interviewing ML researchers about specific papers and systems since 2016, which makes this one of the longest-running technical interview shows still going. Each episode picks one piece of work and stays with it instead of skimming five. (weekly)"/>
      <outline type="rss" text="The Cognitive Revolution" title="The Cognitive Revolution" xmlUrl="https://feeds.megaphone.fm/RINTP3108857801" htmlUrl="https://www.cognitiverevolution.ai/" description="Nathan Labenz talks to AI builders and researchers multiple times a week, and occasionally sits down alone to walk through a paper — a recent episode dissects Anthropic's interpretability work on &quot;global workspace&quot; in more depth than the paper's own abstract. (multiple times per week)"/>
    </outline>
    <outline text="Benchmarks &amp; Leaderboards" title="Benchmarks &amp; Leaderboards">
      <outline type="rss" text="Epoch AI Benchmarking Hub" title="Epoch AI Benchmarking Hub" xmlUrl="https://epochai.substack.com/feed" htmlUrl="https://epoch.ai/benchmarks" description="Epoch pulls FrontierMath, GPQA Diamond, SWE-bench, and a dozen other evals into one place, plus its own composite Capabilities Index. Useful when you want the whole picture instead of checking five separate leaderboards. (continuous, multiple updates a month)"/>
    </outline>
    <outline text="Communities &amp; Policy" title="Communities &amp; Policy">
      <outline type="rss" text="AI Alignment Forum" title="AI Alignment Forum" xmlUrl="https://www.alignmentforum.org/feed.xml" htmlUrl="https://www.alignmentforum.org/" description="A researcher-run discussion board for technical alignment work, where the comment threads are often as substantive as the posts — people like Paul Christiano showing up to argue about failure modes in reward hacking and debate training. (several posts a week)"/>
      <outline type="rss" text="Tech Policy Press" title="Tech Policy Press" xmlUrl="https://techpolicy.press/rss/feed.xml" htmlUrl="https://techpolicy.press/" description="A nonprofit newsroom covering the actual mechanics of AI regulation and platform accountability, plus a standalone tracker for pending legislation and litigation. Closer to a beat reporter's notebook than a think-tank press release. (multiple articles daily)"/>
    </outline>
  </body>
</opml>
