Three years ago I was in Morocco with my 16-year-old brother and $500 a month... and today I run an AI film studio that works with billion-dollar companies, and governments are putting our films into their national education systems.
This is the exact blueprint, all of it, for free.
Everything here applies outside of film. The system is transferable, the specifics are mine.
The opposite direction
When ChatGPT dropped, the entire internet decided the winners of AI would be the most technical people in the room... fine-tuning models, training LoRAs, learning Python.
I went the other way.
If AI handles execution, the only thing left is what you tell it to do, and the quality of your instructions depends entirely on what you actually know about the thing you're making.
Someone typing "cinematic shot of a man walking" gets back stock footage, while someone who understands focal lengths, motivated lighting, and emotional blocking gets back something that stops people mid-scroll. Same tool, same cost... the gap is knowledge, not code.
Technical skills aren't the moat, and they might actually be the anchor dragging you down. While you're grinding through Python tutorials, someone who understands the craft is giving better instructions and getting better output, and the tools will deprecate next quarter while the knowledge won't.
The real moat is your character, the ideas only you can have because of the specific life you've lived. So I made it my job to have the most unusual life possible.
Five countries, zero code
Holland to Morocco, took my 16-year-old brother with me, $500 a month. Everyone thought I was insane... "You're 20, you don't even know how to take care of yourself." That's exactly why I went. Six months of survival on nothing rewires your brain in ways that six years of comfort never will.
Decided on Friday, moved on Monday.
Morocco to Serbia, alone this time... no brother, no family, no friends, new phone number, new everything. Six months in complete solitary where nobody knew where I was, and I rebuilt how I think from scratch.
Decided on Thursday, moved on Sunday.
Serbia to Dubai, then Bali. Started selling AI services... funnels, websites, sales systems. Met founders, got into rooms, learned how business actually works when real money is on the table.
Got scammed, lost everything, moved back to zero.
Poured all of it into generative AI and went all in on film. Six months later ARQ launched, and six months after that we made a film for the government of El Salvador.
Morocco, Serbia, Dubai, Bali, El Salvador. None of it required a CS degree, but all of it gave me something no model can generate on its own... a way of seeing things that can't be replicated with compute.
That's the foundation, and here's the framework.
Build a brain that thinks for you
No matter what niche you end up in, you need this first, because this is the operational foundation that makes everything else compound.
Grab a laptop, install Claude Code, and create a specific folder that only this terminal has access to. Inside that folder, create a file called CLAUDE.md... this becomes your operational brain.
Fill it with everything you've got. Your goals, where you want to be in six months, your current situation, income, expenses, team, resources, how you make decisions, what you prioritize and what you avoid, who you're working with and who you want to reach, everything you've learned so far.
90,000+ characters. Just dump it all in... your outreach strategy, your content plan, your monthly numbers, your email templates, all of it.
When it gets too long (and it will), you compress. Move the specifics into separate knowledge files, replace 300 lines with 2 lines that say "when this topic comes up, reference this file." Your main file stays lean while your knowledge base grows forever.
I built knowledge files by getting AI to watch 50+ commercials and extract every visual pattern, and I got Gemini to absorb a million tokens of Christopher Nolan, Tarantino, Asimov... storytellers in one file, cinematographers in another. One of my system prompts:
You are a cinematography specialist trained on Christopher Nolan's complete filmography. For any scene, break down: camera angle, focal length, lighting motivation, color temperature, actor blocking, and emotional intent. Reference specific Nolan scenes as proof for every decision you make.
That file sits in my knowledge folder, and whenever I need a reference for how to shoot something, I have an expert that never sleeps and never forgets a frame.
Then you build agents, where each one gets its own system prompt, its own knowledge scope, its own job.
An X Agent that understands your entire brand, your trajectory, everything you've done, and where you're going. A Planner that takes your progress and maps out what content comes from it... you shipped a project? The planner knows that means a tweet, a breakdown, a behind-the-scenes clip, and a follow-up thread next week. A Trend Tracker that monitors your niche on X, tracks specific accounts, and flags when formats go viral so you can adapt fast.
Now connect the data. The first thing I did after building my X Agent was connect it to the Twitter API and run a full forensic analysis on everything I'd ever posted, and here's the type of prompt that produces real insight:
Here is my complete X data export. Every tweet, every reply, every DM, every engagement metric over [timeframe].
Phase 1: Behavioral analysis. What percentage of my activity is original posts vs replies vs DMs? What does this ratio say about my actual growth strategy vs what I think my strategy is? Map my output by month. What happened in the high months? What happened in the quiet ones?
Phase 2: Performance forensics. Sort every original post by engagement. What do the top 5 have in common? What do the bottom 5 have in common? Find every spike moment where engagement was 5x my baseline. What triggered each one? Is there a pattern in timing, format, or content type? Does reply volume correlate with engagement in the same period?
Phase 3: Blind spots. What patterns am I repeating that the data says don't work? What's the gap between what I think my audience wants and what they actually engage with? Where am I leaving relationships on the table by not following up?
Use specific numbers. Reference specific posts. Tell me things about my own behavior that I would never notice from inside the feed.
What came back was uncomfortable and invaluable. The system found that my growth was entirely spike-driven, where big proof-of-work moments drove the majority of new followers while my baseline between spikes barely moved.
Reply volume directly correlated with engagement... my biggest month had the most replies and 5x the engagement of my quietest month.
And it caught a pattern I was completely blind to: I was consistently failing to follow up on promising conversations, having incredible calls, getting excited, then letting them die two or three days later... not because the other person lost interest, but because I did.
None of this was visible from inside the feed. The system saw all of it in a week.
Every week, run a board meeting with one prompt that brings all your agents into the same room:
Board meeting. Agents present: X Agent, Planner, Trend Tracker.
Round 1: X Agent presents the top performing post this week and the worst performing post. Explain why each hit or missed using historical patterns as comparison.
Round 2: Trend Tracker identifies one format or topic gaining momentum in [niche] right now. Shows 3 specific examples with engagement numbers.
Round 3: Planner cross-references X Agent's performance data with the content pipeline. Which planned piece should be accelerated based on what's working? Which should be killed based on what's not?
Round 4: All agents. Find one contradiction in the current strategy. Something we say we're doing that the data says we're not. Find one blind spot. One thing we should be doing that we haven't started.
Round 5: Agree on the single highest-leverage action for next week. Not three priorities. One. Defend it with numbers.
You read the output and you'll find things about your own blind spots, your repeated mistakes, your wasted effort... things you'd never catch alone. I run this every week and it's the single best exercise I've found for staying honest about what's actually working versus what just feels productive.
Build the brain first, then feed it.
Pick your niche, feed the brain
One niche, not two, not "a little bit of everything."
AI filmmaking, AI music production, AI UGC and personal brands, AI tools and automation for specific industries... all of these are producing real money right now for non-technical people.
But you need to master the actual craft, not just the tools, because AI executes what you describe and you can only describe what you understand. If you don't understand cinematography, your AI film looks like a tech demo, and if you don't understand sales psychology, your AI funnel converts at zero.
The tools change every month while the knowledge compounds, and the people who know the craft will always outperform the people who only know the tool... because the tool is the easy part to learn and the first part to become obsolete.
If you're in AI filmmaking, watch Dune, pick a scene, and remake it with AI. It will look terrible... the colors will be wrong, the composition will be off, the lighting will make no sense. Good, now figure out why. Then do Fight Club, then Blade Runner, study every frame, read the reviews, because that process of failing and fixing is where the real learning happens, and it's the process that gets skipped because it doesn't feel like progress.
If you're in AI tools, start with yourself... save 20 minutes of your day by automating something, then 30 minutes, then an hour, then solve a friend's problem, then a friend of a friend's, and keep going until you have a suite of solutions someone would pay for.
Everything you learn goes into the brain. Every technique, every pattern, every mistake feeds back in, the knowledge files grow, the agents get smarter, the board meetings get sharper... the brain compounds as you compound.
Become the person who gets referred
You don't know enough yet, and nobody does at the start. The ones who make it aren't smarter, they just push longer than everyone else.
Here's the move that changes everything: find someone who is already successful and add value for free.
Not someone at your level... someone above you. 500K followers, a $150M company, a government contract. Figure out what they need that you can provide and do it before they ask.
The higher you climb in any industry, the less money matters and the more people matter, because at the top what everyone needs is reliable, high-quality people who execute without being managed. Rich people don't need your money, they need someone they can trust to deliver, and they will always refer reliable people to their other rich friends because reliability at the top is genuinely scarce.
That's the entire game... become the most reliable person in the room for your specific thing.
I am the guy for AI videos in my network, and when anyone in my circle needs one, I get referred. Not because I'm the cheapest, not because I'm the most technical, but because I'm extremely good at the craft and extremely reliable.
Craft plus reliability... that combination is the entire engine, and everything else is noise.
Turn the network into a business
Every achievement, post about it, every lesson, post about it. On X you grind and build publicly, on Instagram you show results, and before you've made real money or closed a real client, it's X... that's where your first fans, your first relationships, your first lessons come from.
Gather a community, a group, a channel, somewhere people can follow the journey and access the value you're creating. Give them everything... the breakdowns, the prompts, the process, the failures. The people who stick around through the building phase become your most loyal audience when you start winning, and they become it because they were there before the wins made it easy to show up.
The deals come from the network, not from cold outreach, not from ads, but from people who watched you build, saw you deliver, and know you're the person for the job. Every client you close leads to the next one, every relationship compounds... $5K to $8K to $10K to $20K, client to client to client.
You don't chase the money, you let the work chase it for you.
Build the brain, master the craft, become the referral, turn the network into a business.
There's no hack and there's no shortcut, but there is a cheat code... 99% of people will read this and do nothing, and the bar is just doing the work.
Your move.
(i also have a tg channel where i share secrets and stuff like that - ask for it and i will provide)