From Micro-Manufacturing to Manus: What You Missed in AI Morning Club

Monday, February 23rd, 2026
Perry + Angie: 2026 Predictions Check-In + 2030 Vision + The Micro-Manufacturing Revolution

Perry and Angie opened the week in a more conversational format, built around questions Angie had collected from the community. What followed felt less like a normal session and more like a live update from someone trying to connect the dots between macro shifts, market pain, and where ordinary people might still have an edge.

Perry revisited several of his 2026 predictions and, in typical fashion, didn’t hedge much. He pointed to Bitcoin’s fall back to the sixty-seven-thousand range, silver’s surge and pullback, the wave of tech layoffs, the software sector getting hammered, franchising exploding, and SBA acquisition loans reaching record levels. His bigger point wasn’t that every call landed perfectly. It was that the direction of travel is becoming easier to read: layoffs are rising, people want cash flow now, and buying existing businesses may be smarter than trying to invent everything from scratch.

Then he pivoted to the more unsettling part: his private 2030 predictions. He admitted he hasn’t published them because they feel almost too disruptive, but said that if he’s even twenty-five percent right, daily life by 2030 will be almost unrecognizable. Robots in homes. Less mandatory labor. More automation than most people are emotionally prepared for. In other words, the future isn’t creeping in anymore. It’s accelerating.

The most practical part of the call came when Perry zoomed all the way down from macro predictions to a sourdough bread stand sitting on the edge of a parking lot. The model was beautifully simple: a tiny unmanned bread kiosk, Venmo and Cash App QR codes, a Ring doorbell for remote monitoring, and loaves costing about fifty cents to produce selling for twelve to eighteen dollars each. From there he expanded the idea into a broader thesis: as inflation and corporate consolidation distort markets, local micro-manufacturing becomes viable again. Bread. Tomatoes. Tool holders. Leather goods. Canned foods. Laser-cut products. Small-batch production is no longer just a craft hobby. With AI-assisted design, digital payments, cheap monitoring, and garage-scale tools, it starts to look like a legitimate lane for people who aren’t elite closers but still want to build income.

The thread running through the whole session was clear: many giant industries have become vulnerable because corporations pushed efficiency so far that they left emotional connection, local trust, and niche flexibility on the table. Perry’s prediction was that a surprising amount of that value is coming back home.

Tuesday, February 24th, 2026
Emma: Open Claw 8-Week Curriculum Launch + Setup Rules + Community Task Collection

Emma’s session opened, hilariously, with a mystery: someone inside Ignite keeps mailing her perfectly chosen T-shirts and refuses to claim credit. She was equal parts grateful and suspicious, and it set the tone for a call that mixed humor, practical paranoia, and a big systems rollout.

The centerpiece was the launch of an eight-week Open Claw curriculum. Emma announced that Ty Shane would be joining for six of those weeks, with Ty and Boris handling the next session while she’s away on a cruise writing her second book. The promise was ambitious: start from absolute zero, teach the install, then progressively build real capability around it.

But before the hype could get ahead of the reality, Emma delivered the warning. Open Claw doesn’t have a sandbox. That means this is not one of those neatly walled-off AI tools where you can casually click around without thinking about exposure, permissions, or access. Her message was very direct: you are responsible for deciding your own security comfort level. She then walked through her own setup standard, which is more cautious than most people will probably adopt — a cheap dedicated mini computer, a VPN, a VPS where appropriate, and entirely separate email and social accounts so the system never touches her primary personal or business identity.

That security framing gave way to the most valuable part of the call: Emma started collecting real use cases from the community to shape the eight-week curriculum. Mark wanted lead research and podcast/seminar outreach. Ken wanted serious content creation and distribution support.
Lisa wanted help enriching a giant client contact list with social profiles and follower counts. James, in wonderfully classic James fashion, tried to ask for “agents… the whole thing,” which caused Emma to nearly short-circuit on the spot. The running joke was that the curriculum would only be useful if people could state what outcome they actually wanted, not just say “agents” like it was a magic word.

Underneath the humor, the strategy was smart. Instead of teaching Open Claw in the abstract, Emma is reverse-engineering the curriculum from the exact jobs people want done. That makes the whole thing much more likely to become usable rather than just interesting. By the end of the session, it was clear that this wasn’t just “AI education.” It was a guided buildout of a practical operating layer — one designed around research, outreach, content, and business support tasks that people already know they’re drowning in.

Wednesday, February 25th, 2026
Rick: Advanced Prompt Engineering + The 24-Year Experience Rewriter System

Broadcasting from Perry’s home studio, Angie handed the spotlight to Rick with one goal: let him wow the room. He did exactly that, not with flashy visuals or an AI party trick, but with something much more useful long-term — a system for turning average prompts into dramatically stronger ones.

Rick began by asking the room to share their latest prompts, then used those examples to show why most prompting still breaks down in the same predictable places: not enough specificity, missing context, vague audience, unclear outcomes, sloppy structure. His fix starts with a deceptively simple framing line:

“You are a prompt engineer with 24+ years of experience.”

Rick explained that twenty-four years has become something of a magic number in his prompting experiments, and he now uses it as the front-end instruction for a prompt rewriter system. The actual mechanism is where the real value lives. He instructs the model to analyze the original prompt, identify gaps in precision, completeness, context, constraints, and actionability, score it from one to one hundred, ask clarifying questions only when truly necessary, and then keep rewriting until it reaches a score of ninety-seven or better.

It’s a fantastic example of AI being used not just to answer prompts but to improve the prompts themselves. In other words, he turned prompting into an iterative quality-control loop.

The live demo on Matthew’s Facebook post prompt made the value instantly obvious. What started as a vague request for a highly engaging post scored poorly because it lacked audience definition, objective, and topic specificity. Rick’s system didn’t just produce a better version — it explained why the original was weak, identified its core intention, and rebuilt it into something much more usable. Then Rick added his own twist: when variable placeholders are scattered throughout a prompt, move them all to the top. Define them once. Make the whole thing shareable and reusable. That little step turned a one-off rewrite into an actual template.

Rick’s bigger lesson was familiar but important: garbage in, garbage out still rules this entire game. But with the right rewrite layer, you can rescue weak prompts and standardize strong ones. It was one of the most practical sessions of the week because it upgraded the thinking habits behind prompting, not just the wording.

Thursday, February 26th, 2026
Russell: Radio Spot Remix App + Gemini Audio Generation for Sixth Street Club Ads

Russell went full nostalgia mode and took the group into an older chapter of his marketing life: writing nightclub radio ads for Austin venues and paying serious money to get them produced professionally. That history became the setup for a simple but powerful question: if AI can now generate voice, scripting, sound design, and app interfaces, why should anyone still be paying old-school production rates for short promotional spots?

Using archived scripts from Sixth Street clubs, Russell vibe-coded a “Radio Spot Remix” app designed to accept a PDF or document, analyze the copy, generate style suggestions, and then turn the content into a polished, high-energy audio ad. His original prompt was loose, but the resulting app concepts were surprisingly sophisticated — different tone options, voice selection, sound effects, timing controls, and even a slick SoundCloud-style interface in one version.

The contrast between Gemini and Claude became part of the show. Claude produced a much prettier wireframe, while Gemini eventually delivered the part that mattered most: actual audio output. The final radio spot for the “ultimate dollar deal” at a club came back loud, punchy, and way better than Russell expected from such a rough starting point. He kept circling back to the economics. These used to cost hundreds for scripting and many more hundreds, sometimes over a thousand, for production. Now they’re being assembled with API credits and a vibe-coded front end.

What made the session compelling was that it wasn’t just about making a fun club ad. It showed a pattern that keeps repeating across AI Morning Club: take an expensive, specialized service category, reduce it to a repeatable workflow, then wrap it in a simple interface so it becomes easy to use or easy to sell. In this case, the result was a working path toward in-house radio production for local businesses, event promoters, or anyone who needs fast-turn audio creative without agency overhead.

Friday, February 27th, 2026
Mitch “The Beardpreneur” Barham: Manus AI – $5 Email Flows, $20 UGC Videos, and 50-Company Research While at a Concert

Mitch closed the week with one of those sessions that makes the entire AI landscape feel simultaneously thrilling and mildly absurd. His core case for Manus was straightforward: it thrives under messy, human, non-technical prompting. He isn’t prompting like a specialist. He’s talking to it like he talks to his team — and still getting work back that would have taken people days.

The first showcase was an email flow rebuild for a client whose previous agency had left a mess behind. Mitch handed Manus a casual instruction, attached some brand images, and watched its internal agent swarm go to work in real time. It crawled the client’s best-selling pages, analyzed product language, understood the outdoor-brotherhood voice, mapped full flow structures, drafted subject lines, suggested design treatments, and then — with a tiny follow-up instruction — generated the email images and copy to go inside those flows. The entire task cost under five dollars in credits. More importantly, the output was “dummy-proof” enough that someone with basic Klaviyo or Omnisend skills could just plug it in and go.

Then he moved to UGC video creation, which might have been even more impressive. He used AI to generate the prompt for the video, then told Manus to use that to create the actual ad. The first version was bad, which he openly admitted. But what mattered was the iteration loop. Manus watched its own output, identified the weird parts, rebuilt them, and improved the footage without him having to micromanage every shot. By the third iteration, he had something usable for about twenty bucks total.

The wildest part of the call was how casual all of this has become in his workflow. While on the session, one of his other AI agents notified him that five daily Facebook threads were already waiting in Telegram. The night before, while he was at a concert, Manus had researched fifty HVAC companies in Phoenix and delivered an Excel file with competitive intelligence. The practical message to the group — and to his own employees — was reassuring: this isn’t about replacing people for the sake of replacing people. It’s about moving repetitive, low-leverage work to agents so the humans can focus on higher-value execution.

Mitch’s session captured something important about this phase of AI: the winning workflows are starting to look less like carefully engineered prompt chains and more like delegated conversations. Talk to the system like a competent teammate. Let it spin out its own internal agents. Review the outputs. Refine where necessary. That’s a very different world than the one most people were operating in even six months ago.

Replays on Youtube: A.I. Morning Club Replays