Skin Marrow is submitted to the App Store. Waiting on Apple now, 48 hours to weeks. Honest version: it did not one-shot. I thought 2 days, it took 6. 350+ threads back and forth between me, Claude and Codex. I turned the recording off a couple times because I was just stuck on bugs and didn't want to bore you 😂 But I never wrote a single line of code. Not one. What surprised me most: the code was never the hard part. Ideation was. A lot of ideas that felt great died the second I ran a real market check on them. Full build, start to finish, nothing cut 👇 https://youtu.be/basupz-qRmQhttps://youtu.be/basupz-qRmQ Total cost to put an app on the App Store: $99/yr for the Apple developer program. That's it. No course, no tool stack, no agency. And you guys get TestFlight before YouTube does — I'll drop the link here the moment Apple approves 🔥 One line: what's the app you keep almost building? Tell me and I'll tell you which part Claude will actually fight you on. After 304 threads I've got a decent read on where it breaks 💪 p.s. tiny favor as always — a like or a comment on the video really helps with the anti-crickets protocol 😂 p.p.s. if you missed part 1 where we designed the whole thing in Claude Design, it's linked in the description.

I made Claude + Codex argue over my work—then give me one answer Most people use AI one tab at a time. Claude knows one conversation. Codex knows another. Then *you* become the project manager carrying context, comparing answers, and deciding what survives. I just published a walkthrough of a different setup using Buzz—the open-source workspace built by Block where humans and AI agents can work in the same channels. In the video, I: - give the same job to separate Claude and Codex agents - make them work independently instead of copying each other - have them challenge the weak parts of each other's work - make them converge on one final output while I keep the final decision - show how to create your first channel and agent from scratch The real unlock isn't “another AI assistant.” It's coordination: shared context, visible handoffs, different models with different jobs, and one human still owning the outcome. Watch it here: https://youtu.be/N7bZlTq0h5g After you watch, tell me this: **what is the first repeated job you would hand an AI teammate?** (in the youtube video) Don't give me an agent name—give me the job. I'll turn the best answers into copy-paste agent instructions for the next video. 🐝

If you don't commit, you don't get things done. At least I don't. Everything I've actually finished had someone else in the room. A deadline I couldn't move, or people who'd notice if I went quiet. Everything I started alone is still sitting in a folder somewhere 😂 We're tribe animals. That's not a weakness, it's just how it works. We grow in groups, and we need people keeping us honest. ════════════════════════════════════════ Here's what made me think about this. I went back and read every intro in this community. Three out of four of you described the same build. Life AI OS. An OS for a one-person company. Hermes + Obsidian Same build. Different names. And almost none of you have finished it. I know why, because it happens to me too. It gets to about 60%, it half works, and then it goes in the folder next to the other three you started. It was never a skill issue. Nobody was watching. Nothing was due.🙂‍↔️ Without a deadline, nobody builds anything. Nobody ships anything. So I built the only thing I know how to fix that with. A room, a deadline, and me building next to you. ════════════════════════════════════════ AI Movers+ — Season 01. Four weeks. Starts Saturday 8 August. ⭐ Week 1 — your bottleneck dies. We find the one job eating your week and you build an AI employee that owns it. Running by Sunday. ⭐ Week 2 — Building your own stack. Not someone's templates. Skills built together for how you actually work. ⭐ Week 3 — it runs without you. And you spend the time you just got back building the thing you've been putting off for a year. App, SaaS, game, whatever it is. You ship it. ⭐ Week 4 — demo day. You demo, we audit it live, give each other feedbacks, we publish it. Then talk about how to Scale. You leave every single week with something that works, and it compounds. I ship it with you, and every single person gets feedback. That's why it's capped at 50. ════════════════════════════════════════ First 50 people: $44/mo, locked in for as long as you stay. After that it's $99.

I spent my last job in venture capital reading pitch decks for due diligence, and the thing nobody tells founders is this: the reviewer is a human being before they're a rational one. Before I judged the product, before I judged the team, I'd already decided whether I wanted to keep reading. A stale deck gets closed before the strategy is ever read. So I rebuilt a real client deck — one that went on to close a high-ticket deal — and made the whole process into a video. Here's the short version, so this post is useful even if you never click. THE 3 TELLS 1. The vertical accent bar on every card. border-left: 3px solid. It's the single most recognizable thing AI makes, and once you see it you can't unsee it. 2. The grid of rounded boxes. I counted 11 on one slide. Every slide being the same card grid makes a deck read like a form — and nobody argues with a form, they fill it in and leave. 3. No texture at all. Flat white or flat dark, nothing else. THE 4 BLOCKS THAT FIX IT This is the entire prompt structure. Works in Claude Code, Codex, Gemini — it's a way of thinking, not a tool. 1. Guardrails — the don'ts and dos. Start here, not with what you want. You're not describing a target, you're fencing off the average. AI's default output IS the average of everything it's seen, so most of the work is closing doors. 2. Experience — what should the reader feel, and who is the reader? 3. Design — the concrete layer. Colour, theme, template, layout, font. 4. Assets — links, filepaths, screenshots, images. Block 4 is the one people skip and it's the one that matters most. I wasted an embarrassing amount of time typing "warm linen texture, analog, handcrafted" and getting garbage back every time. AI has never touched linen. It has no idea what that word means as a feeling. You can't adjective your way there — you show it the photo. The gradient in this deck came from a phone photo I took at a startup pitch event in Seoul. Badly, at an angle, with the back of someone's head in it. That photo was worth more than every adjective I could type.

I’m heading to Amsterdam right now! Mainly staying at Brussels, Belgium Taking flight back and maybe spend 1day at Frankfurt Any AI movers down for meetups? ☕️ Also taking recommendations for spots/ foods etc 👊