How I used AI to get out of $92,488.28 of debt in under 90 days — without ever showing my face.

There's a specific feeling I want to describe, because if you know it, the rest of this letter is written for you.

It's the feeling of opening an app you already know the number in. Mine was Credit Karma. You're not checking. You're confirming. And you do it anyway, four or five times a day, like pressing on a bruise.

Mine said $92,488.28. Total debt, every account, one number. Not "about ninety grand" — I knew it to the cent, the way you know a number you've looked at every day for a year.

I hadn't spent my way there. I'd built my way there. I started a car rental business. The cars were financed on loans, the day-to-day ran on credit cards, and my savings filled every gap in between. The business lost money — and I kept covering the losses, month after month, because I was sure it was about to turn around. It never turned around.

What "failed" looks like on paper: the cars went back on tow trucks. The loans behind them defaulted. The credit cards were maxed, then closed. About $20,000 was owed to the IRS, in letters I'd stopped opening. And here's the part that separates my story from the usual one — I wasn't behind on the minimum payments. I wasn't making them at all. There was no clever budgeting strategy available to me, because the amount of money I had was zero. My credit score sat on the same screen, right above the damage: it had been 762 — a number I'd been quietly proud of — and now it read 491.

When it was over I moved back into my mom's house at 28, which is a sentence that still costs me something to type.

I want to be precise about what that does to you, because most people writing pages like this skip it. It isn't dramatic. It's flat. I stopped seeing my friends — not officially, nothing announced. I just started making excuses, one invitation at a time, because I couldn't afford to be wherever they were going and I was too embarrassed to say why. It takes maybe two months of that before people stop asking, and the worst part is the relief when they do. It doesn't feel like a crisis. A crisis is loud, and this is silent — you're just a person at zero, in a childhood bedroom, and nobody around you knows, because you've made sure of it.

So I did the responsible thing. I looked for a job.

Every single day. Applications in the morning, follow-ups in the afternoon, DoorDash at night to keep some kind of money coming in, tweaking the resume after that like the font was the problem. I moved home in July of 2024. The system you're about to read about didn't show up until October of 2025. That's fifteen months — and in fifteen months of daily applications, I didn't get a single interview. Not a rejection with a name on it — nothing. I was a guy with a 491 credit score and a failed business on his hands, delivering food and refreshing an inbox that had decided I didn't exist.

Here's the thing about applying to jobs from your childhood bedroom while $92,000 in the hole: every application is hours of your life spent asking someone else for permission to start climbing out. And they weren't even saying no. They were saying nothing.

And while my inbox stayed empty, my feed didn't. The algorithm — with what I can only assume was a sense of humor — kept showing me nineteen-year-olds posting watches that cost more than my entire debt. Kids a decade younger than me, casually wearing the number that was ruining my life. I'd love to tell you I scrolled past with dignity. I didn't. I watched every one, doing the same math each time: that watch is $92,000. That watch is my whole problem, on a wrist.

That's what finally made me look into the internet money stuff — not ambition, honestly. Something closer to how? And here's where I lost the most time: every single "make money online" path I found required me to become a person on camera. Post daily. Build a following. Document your life. Show your morning routine, your workspace, your wins — and then, in eighteen to twenty-four months, maybe monetize it.

Document what, exactly? I was a 28-year-old sleeping on a twin bed in his mom's house with a 491 credit score. I didn't have an apartment to show. I didn't have a routine worth filming. My whole life at that point was a job board, a twin bed, and a number I was ashamed of. The camera path doesn't just ask you to work — it asks you to already have a life worth watching. I didn't.

And I didn't have twenty-four months to go build one. I needed a path that didn't care who I was.

It turned out that one existed. I just didn't know its name yet.

From the TikTok video to having everything built took four days.

Not because I'm fast — because there was almost nothing to build and almost nothing to pay for. The AI tools have free plans. The web hosting has a free plan. The only money the whole setup cost was the domain name — less than $11, paid for out of a night of DoorDash deliveries. Signed up for an affiliate network — the middleman that connects companies paying commissions with people like me — found an offer paying $280 a sale, had AI write the page, had it write the ads for that page. The actual building — page live, ads set up and ready — took me under four hours. A guy who'd never done any of this, prompting his way through it in an afternoon.

And then everything sat there, finished, for two weeks. Because the one thing on the list that isn't free is the ads themselves — and DoorDash money was survival money. It kept me fed; it didn't leave anything over. So for two weeks I took every shift I could get and ate as little of it as possible, until I'd scraped together a few hundred dollars that could be spent on something other than staying alive — knowing the whole time that the machine was already built, sitting in a browser tab, waiting for fuel.

On October 14, I turned the ads on. Here's that first week, day by day, exactly as it happened — because this is the week that kills most people who try this, and I want you to see what it actually looks like:

Tuesday the 14th: spent $18. No sales.
Wednesday: spent $78. No sales. Two days in, nearly $100 of DoorDash money gone, nothing back. This is the moment. Every instinct I had — fifteen months of trained losing — said of course it doesn't work, nothing works, turn it off.
Thursday the 16th: spent $118. One sale. $280. One stranger, somewhere, bought.
Friday: spent $136. Another sale.
Saturday: spent $93. Nothing. (Saturdays, I'd learn, are just like that sometimes.)
Sunday: spent $108. Third sale.

The network pays out weekly. So on Monday, October 20, the first check landed: $840. Three sales at $280 each, against $551 in ad spend. Around $290 in profit, in week one, run by a guy who was still doing DoorDash shifts between all of it.

I know $290 doesn't sound like a headline. But the number I couldn't stop looking at wasn't the $840 — it was the ratio. I put $551 in and got $840 out. That's not a payday. That's an exchange rate. And I'd been feeding the machine roughly $90 a day in DoorDash money, which meant the only question that mattered anymore was: what does this look like at $200 a day? At $300?

Which is how I hit the problem nobody warns you about: I finally had something that worked, and it was starving. Every card I owned was maxed out and closed — the banks had written me off a long time ago. The machine could clearly handle more fuel than DoorDash shifts could supply, and I had no other tank to draw from.

So I made the most uncomfortable phone call of that entire year. I called a friend — someone who'd known me long enough to trust me even though the evidence didn't look great — and asked to borrow their credit card. Not their money. Their card. I'd run ads on it, and when the commissions paid out, I'd clear the balance before they saw a statement.

Here's how the next three weeks went, straight off the payout reports:

October 28 — $1,680. Six sales. Cleared what I'd put on the card and ran it again.
November 5 — $840. Three sales. Slower week. Cleared the card. Ran it again.
November 10 — $4,480. Sixteen sales. Paid the friend's card to zero, and for the first time in a year I was holding money that didn't already belong to somebody else.

By that point the math had settled into something almost boring: for every dollar I put into ads, I was getting that dollar back plus another one in profit. Spend a dollar, get two. The only question left was how many dollars I could put in.

That's the whole method in one sentence: front the ad spend, get paid on the sale, cover the spend, keep the difference, do it again. From there the only thing that changed was the size of the numbers.

The same day that $4,480 landed, I did the thing that actually got me out. I stopped treating this as one lucky campaign and started treating it as a system — which meant the next question wasn't "how do I squeeze more out of this offer" but "what's the second one?"

I went back to the same process that found the first offer and ran it again. Found one that paid $300 per sale. Built the page the same way. Wrote the ads the same way. Launched it alongside the first offer instead of replacing it, because the first one was still working and there was no reason to touch it.

And here's why I could launch it with a straight face instead of white knuckles: it wasn't a new bet. Both offers were in the same market — weight loss, specifically the GLP-1 medications everyone and their doctor was talking about that year. Same customer, same demand I'd already proven I could reach, just a second company paying me to reach it. I wasn't asking "does this work?" twice. I'd answered that. I was asking "does this work again?" — which is a much cheaper question.

Two payouts landed the same day.

$14,100 from the second offer — its first payout, forty-seven sales.
$8,400 from the first offer, thirty more.

Twenty-two and a half thousand dollars in one afternoon. After ad spend, a bit over $12,000 in profit — thirty-six days after the first $840.

I want to be clear about what that number was and wasn't. It wasn't a windfall. It was two boring systems, each doing what it had done every week since it launched, stacked on top of each other. The system didn't get better. There was just more of it.

The next morning I did three things, in this order.

I cleared my friend's card one final time — ads had kept running on it through November, so there was a fresh balance — and told him he was officially retired from the lending business. Then I paid off my Amex — $11,328.19, in full, the first card I took to zero. And then I called Amex, got a human on the phone, and asked them — begged them, honestly — to reactivate the card they'd closed on me.

Not because I wanted the credit. Because I wanted to stop running my business on someone else's card. There's a version of this story where that phone call is embarrassing. It didn't feel that way. It felt like the moment I stopped being someone the debt was happening to.

December was November with more fuel in it. Same two offers. Same pages. Same ads. The only variable that changed was how much I could put in each day, and now that I was funding it from my own account instead of a friend's, the answer was: more.

Across the month, the two offers paid out $138,000. After ad spend, $81,420 of that was profit.

I'll say the obvious thing so you don't have to: that's most of a $92,000 debt, cleared in one month. I'm not telling you that to make you think it'll happen to you in your second month. I'm telling you because it's what the same boring loop looks like once you're funding it properly and running it twice. Nothing about the method changed between the $840 week and this one.

I don't remember the exact day the last card hit zero. I've tried, and it's gone — I think because by then it had stopped being an event. I do remember December 17, because that's the day I sent the IRS $20,000 and closed out the back taxes I'd spent a year not opening the letters about. Two months earlier I couldn't cover a minimum payment. Now I was voluntarily paying the one creditor nobody voluntarily pays.

By the end of December: every credit card at zero. The IRS at zero. The car loans, gone. All of it, gone.

The last domino was a car loan, and I couldn't tell you the exact day in December it fell — by then, payments had stopped feeling like events. One of those weekly payouts simply became the final payment on the final loan. The number that had been $92,488.28 read $0.00 — under eighty days from the afternoon I turned the ads on. (January, for the record, cleared another $103,000 in profit. The debt just wasn't there to take any of it.)

I still open Credit Karma sometimes. Out of habit, mostly. But I have not opened it out of dread since. That's the actual product. Not a lifestyle — just the absence of a specific, daily, low-grade fear.

It's August now, so I'll tell you where it went, and then I'll stop, because this isn't that kind of page.

Since October: $1,277,513 in payouts, $613,206 of it profit. I moved out of my mom's house — the same one I told you about moving back into — into an apartment I actually like, up high enough that the view is the thing I notice first when I get home. I bought a car. A 2025 Camry — and I'm naming the model on purpose, because you've seen the version of this story where the guy is leaning on a Lamborghini, and I need you to understand that this is not that story and I'm not that guy. And I've put close to $100,000 into a brokerage account, which is a sentence I'd have laughed at a year ago.

None of that is the point. The point is that the loop I found in October is the loop I still run today — across more offers and more markets now, but the same four steps, and still no face on anything.