YouTube Shorts — Batch 01
**Format rules (genre-proven):**
- Hook lands in the first 3 seconds. No intro, no "hey guys."
- One idea per Short. One concrete payoff.
- ≤60 seconds = ~140–155 spoken words max. These are written to that.
- End on a CTA that captures attention into something you own (newsletter / DM keyword / free resource).
- On-screen text reinforces the hook and the payoff — not a transcript.
**HeyGen notes:**
- Each script = one avatar take. Scene breaks below map to where you cut to B-roll or screen-recording overlays.
- Voiceover column = what the avatar says. On-screen text = burned-in captions/titles.
- Keep avatar energy high on the hook line; HeyGen reads flat by default, so mark stress with caps in your script copy if needed.
---
1 — UAE AI House / sovereign routing
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–3s | Two Dubai schools. Almost four thousand students. And a rule: their data can't leave the country. | "3,900 students. Data can't leave the UAE." |
| 3–10s | So I built a three-layer AI system that routes every request by how sensitive it is. | "3 layers. Routed by sensitivity." |
| 10–35s | Public stuff — homework help, general questions — goes to a free model. Internal work goes to a paid model I control. And anything with student data? It never leaves a sovereign model hosted in the region. One question, three possible destinations, decided automatically. | [screen-rec of routing diagram] |
| 35–50s | Most people think "use AI" means "send everything to ChatGPT." For regulated orgs that's a lawsuit waiting to happen. The trick isn't the model. It's the routing. | "The trick isn't the model. It's the routing." |
| 50–58s | I'm breaking down the full architecture this week — link in bio to get it. | "Full architecture → link in bio" |
---
2 — MakeLive
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–3s | You don't need to learn to code to put a page on the internet anymore. Watch. | "No code. Watch." |
| 3–30s | I type what I want in plain English. "A landing page for a gold trading signal service, dark theme, one signup form." That's it. It builds the actual HTML, live, in front of me. | [screen-rec: typing prompt → page renders] |
| 30–45s | No templates to fight. No drag-and-drop that breaks. I describe it, it ships, and I can keep editing by just talking to it. | "Describe → ship → edit by talking." |
| 45–58s | I'm opening a few free spots to try it. Comment "LIVE" and I'll send you the link. | 'Comment "LIVE" for access' |
---
3 — ZTAB / teaching AI
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | Everyone says AI levels the playing field. They've got it backwards. | "AI levels the playing field? Wrong." |
| 4–25s | AI is a multiplier. It multiplies whatever judgment you bring to it. A beginner with weak judgment, times AI, still gets weak output — faster. An expert with sharp judgment, times AI, pulls further ahead than ever. | "AI multiplies judgment — not output." |
| 25–45s | So the gap between someone who knows what "good" looks like and someone who doesn't isn't shrinking. It's exploding. The skill that matters now isn't using AI. It's judging what it gives you. | "The new skill: judging the output." |
| 45–58s | That's what I teach in Zero to AI Builder. Link in bio if you want to be on the right side of that gap. | "Zero to AI Builder → link in bio" |
---
4 — goldsilversignals (engine logic, no financial advice)
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | I stopped trying to predict where gold goes next. Here's what I do instead. | "Stop predicting. Start matching." |
| 4–30s | Every day, my system takes the current shape of the market across six linked assets — gold, silver, the dollar, treasury yields — and asks one question: when has it looked exactly like this before? | "6 assets. One question." |
| 30–48s | It finds the closest historical matches, normalized so old prices and new prices compare fairly, and looks at what tended to happen next. It's pattern-matching at scale, not fortune-telling. | "Pattern-matching, not fortune-telling." |
| 48–58s | I break down how the engine works — not signals, the system. Link in bio. (Not financial advice.) | "How the engine works → link in bio" |
---
5 — UAE AI House / positioning
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | I have almost no marketing budget and I'm landing AI deals in Dubai. Here's the actual lever. | "No budget. Still landing deals." |
| 4–28s | It's not ads. It's not a slick pitch deck. It's one aggregator partnership — a chamber, a free zone authority, an industry body — that already has the trust and the relationships I'd spend years building. | "One partner. Their trust. My system." |
| 28–48s | When a respected body opens the door, the deal is half closed before I walk in. In this region, who introduces you matters more than what you say. | "Who opens the door > what you pitch." |
| 48–58s | I'm documenting the whole go-to-market as I build it. Follow if you're building B2B in the Gulf. | "Building B2B in the Gulf? Follow." |
---
6 — MakeLive / objection
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | Most AI website tools fail for one boring reason. They let you generate, but not iterate. | "Generate ≠ iterate." |
| 4–28s | You type a prompt, get a page, and the second you want to change one thing, you're stuck — start over, or fight a clunky editor. The magic isn't the first generation. It's being able to say "make the header smaller, change this to teal" and have it just happen. | "The magic is the second edit." |
| 28–48s | That's the whole bet behind what I'm building. Unlimited edits, conversational, on a page that's already live. First draft is easy. The loop is everything. | "First draft is easy. The loop wins." |
| 48–58s | Comment "LIVE" if you want early access. | 'Comment "LIVE"' |
---
7 — ZTAB / mindset
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | If you're learning a new AI tool every week, you're wasting your time. Here's why. | "Chasing tools = wasting time." |
| 4–26s | Tools change monthly. The tool you mastered today is deprecated by next quarter. What doesn't change is the thinking: how to break a problem into steps, how to spot when the AI is confidently wrong, how to verify instead of trust. | "Tools expire. Thinking compounds." |
| 26–46s | I've shipped real systems — sovereign AI for schools, trading engines, live web apps. None of it came from knowing the hottest tool. It came from knowing how to think with whatever tool's in front of me. | "Skills > tools." |
| 46–58s | That's the entire premise of my masterclass. Link in bio. | "Link in bio" |
---
8 — UAE AI House / proof
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | My entire sales strategy is one sentence: never pitch. Build it first. | "Never pitch. Build first." |
| 4–28s | Decks describe what you might do. A working prototype shows what you already did. When I meet a client, I don't open slides — I open the actual thing, running, with their problem already half-solved. | "Slides describe. Prototypes prove." |
| 28–48s | First impression is the last impression. And nothing makes a first impression like "I already built a version of what you need — want to see it?" | "I already built it. Want to see?" |
| 48–58s | Follow for how I turn prototypes into closed deals. | "Follow for more" |
---
9 — goldsilversignals / discipline angle
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | My trading system is allowed to talk to me exactly six times a day. On purpose. | "6 times a day. On purpose." |
| 4–28s | Markets scream noise every second. The fastest way to lose is to react to all of it. So I built fixed report times — aligned to my actual trading day — and outside those windows, silence. | "Noise is the enemy. Silence is the edge." |
| 28–48s | Six structured reads beat six hundred panic checks. The system enforces the discipline I'd never hold on my own. | "Structure beats willpower." |
| 48–58s | I break down how I designed it. Link in bio. Not financial advice. | "Link in bio" |
---
10 — Cross-cutting / personal brand
| Time | Voiceover (avatar) | On-screen text |
|------|--------------------|----------------|
| 0–4s | I run multiple AI products, mostly solo. People ask how. Here's the honest answer. | "Solo. Multiple products. How?" |
| 4–30s | I don't do more work than a team. I let AI do the parts I can specify precisely — the building, the boilerplate, the first drafts — and I spend my time on the part it can't do: judgment, taste, knowing what's actually worth building. | "Specify the work. AI does it. I judge it." |
| 30–48s | That's the whole game now. Not "AI replaces you." It's "AI amplifies the one person who knows exactly what they want." | "AI amplifies clarity." |
| 48–58s | I share the systems behind it all. Follow along. | "Follow for the systems" |
---