We're working on a little black magic spam scraper and we need your help. The architect for this is one of our members @Julian Zengo special hat tip. Existing rules: - Same opener sent to several DIFFERENT people (lead-fishing). - Same promotional pitch posted across DIFFERENT posts (solicitation). - Accounts that post then delete quickly (scam burst pattern). Existing list of Solicitation Keywords: commission you get a dm me i can make i can create gumroad cheap discount reach out send me a message shopify Can you think of anything else? Please let us know.

If you’re using AI to build something bigger than one image or one afternoon’s work, don’t make the chat your production system. I learned that building our 104-page graphic adaptation of Arthur Conan Doyle’s *The Lost World*. And by “our,” I mean me and ChatGPT. I made the creative decisions; ChatGPT helped execute and keep the work organized. The method was what I call **folder factories**. Each part of the job had its own saved inputs, working files, decisions, and checks. The next step could begin from those files without needing to reread a months-long conversation. Three things made the difference: 1. **Design the world before making the book.** We established character and creature designs, costumes, props, vehicles, locations, and a consistent painterly look. Every new panel had specific visual rules to answer to. 2. **Prove the process before scaling it.** We tested one finished panel, then a complete page, then a difficult ten-page run. Only after that did we build the 100 story pages and 372 panels. 3. **Treat review as production, not cleanup.** A panel can look great alone and still fail beside the next page. We checked continuity, page turns, dialogue, lettering, and eventually the print-safe area. When something failed, we could return to the part of the process that owned it without starting the whole book over. We learned why this mattered when trying to load hundreds of images crashed the chat. The work survived because the source art, approved decisions, and revisions lived in the folders. That’s the lesson I’d pass on: **generation is not completion.** A creative project becomes manageable when you can find what was decided, see what changed, and hand the next person or tool a usable file. The complete graphic novel is free to read on The Gutter: https://the-gutter.racon-gunner.chatgpt.site/comics/the-lost-world/complete-graphic-novel/?page=1

Does anybody know just how secure the above setup is (on my Daily Mac machine)? Specifically, if I give it access to ONE folder in my Dropbox. • I use Dropbox on my Daily machine... that's what's making me a bit nervous. • My sensitive files are encrypted, but I still wonder what kind of "escaping" Claude Desktop/Cowork could do when I've unencrypted and mounted sensitive volumes also "Cowork runs inside a secure Virtual Machine (VM) on your computer (using the Apple Virtualization Framework on Mac). This means it can safely "mount" your local files to read, edit, and reorganize them without risking your entire machine." Source: https://www.lumetric.ai/resources/claude-web-vs-desktop-vs-code-vs-cowork Thank youuuu!!!

This podcast says yes. I can say I disagree…

I just watched the Jake Van Cleef breakdown, and the biggest thing that stood out to me was not getting overwhelmed and starting small. It clicked because I tend to complicate things and get overwhelmed. My main takeaway was start small. The experiment at the end helped me understand how to think about the system. Going forward, I'm going to start small and utilize the system better.