Practical AI. Stronger systems. Room for the mission. The Calm Infrastructure Lab | LUR Growth Issue 001 | September 4, 2026 This week: make one report easier to finish There is a lot of AI news right now. You do not need to keep up with every release to make useful progress. For this first Calm Capacity Brief, I want to connect the updates to something familiar: the report that pulls you into three spreadsheets, two email threads, and a search for the latest numbers. The starting point is a clear process, trustworthy information, and a person who reviews the result. Three updates worth understanding 1. Claude: new models, the same need for clear boundaries On September 1, Anthropic announced Claude Fable 5.1 and Mythos 5.1. Fable 5.1 is generally available; Mythos 5.1 is restricted to trusted access programs. Anthropic also described new enterprise safeguards with a phased rollout planned for later this fall. Those safeguards should not be assumed to be active in your account today. Read Anthropic's announcement. My practical take: if you use Claude or Claude Code, test a familiar task with fictional data before changing a working process. Check the result against the source, including what the model leaves out. A stronger model still needs a defined job and a reviewer. 2. ChatGPT and Codex: more capability, staged access OpenAI's update for August 31 through September 4 describes the initial GPT-6 Astra rollout for enterprises in its Trusted Access Program, with broader access expected afterward. It is not a promise that every member can select it today. Read the official weekly update. My practical take: do not rebuild your process around an announcement. If a new model becomes available, compare it with your current tool on one repeatable task. Count review time and corrections, not just how quickly it produces a first draft. 3. Perplexity: a more explicit pause before actions

It's been a few months since I posted the welcome message for this community, and only one of you has broken the ice so far (thank you, @Catalin Tutunaru for going first). That's on me, not you. A community doesn't build itself just because people joined it. So let's actually start. If you're one of the ten people who joined and haven't introduced yourself yet, this is me asking you directly: @John Patrick @Tasia Wilson @Pedro Gomez @LeeAnn Taylor @John Zoren Gutierrez @Jameson Chua @Gretchen Thomas @Tiffany Grant @Kenya Hendrix @Marva Murray Drop a comment below with: - What you do (your role, your org's mission, or both) - The one operational thing that's eating the most time or energy right now - What "AI actually helping" would look like for you, concretely, not the hype version I'll personally reply to every single one. This is a small room right now, which means your voice here actually gets heard, not lost in a feed of thousands. That won't always be true. Use it while it is. Welcome (again) to The Calm Infrastructure Lab.

Same live build from the deck below, now the full recording: https://youtu.be/vXQWde8HlkU The live build starts at minute 37 if you want to jump straight to the part where the donor workflow gets built in real time, project, instructions, source file, scheduled task, dashboard. If you already grabbed the deck and sat with the quiet question, drop your answer in the comments over there whenever you're ready. That is where the next build starts.

We did the thing today. In the "Stop Doing This By Hand" session, we built a real donor communications workflow live: the project, the instructions, the source file, a scheduled task, and the dashboard it all flows into. The full 29-slide deck from the session is attached below, just for this community. It has the C.A.L.M. framework, the three layers of behavior change, the safety guardrails checklist, and the readiness questions we closed with. If you registered for the session, the full replay lands in your inbox tomorrow morning, along with the one-page self-assessment and the step-by-step workflow guide. If you missed it: grab the deck, then sit with your team this week and answer the quiet question. If AI made us three times faster, what would we use the hours on? Drop your answer in the comments. That is where the next build starts.

My teleprompter broke three days before this Friday's webinar. Instead of digging through code, I talked to the AI in plain English and watched it find the root cause. Full build story here, and this is exactly the kind of thing we'll do live on Friday. [YouTube link] Want to build your own? Two attachments below. The PDF walks you through it step by step. The .md file is the shortcut: fill in the three blanks at the top, hand the whole file to Claude Code or Codex, and say "build this for me."