The course cost $4,000. My total earnings from everything it taught me: $0. And the part that stings most, looking back, is that I didn't fall for some shady unknown operator. I fell for the famous one.
I'm telling this story because I kept meeting people asking the same thing I typed into Google at 1am back then, some version of "can i make money with ai if i'm not technical," and the honest answer deserves to outrank the influencers selling $4,000 maps.
The math I actually live on now, up front: about $1,140 a month freelancing with AI doing the heavy lifting, $1,290 in client payments last month minus roughly $150 in tools and fees. Seven months to get there. Nine hours a week, around my actual job as a warehouse shift lead in Memphis. I am not technical. The most technical thing I do is copy and paste.
One thing you'll notice: I'm not naming the influencer or the course, for my own legal and financial safety, and honestly because the name doesn't matter. Swap in any famous make-money guy from your own feed and the story runs identically. It's not one bad course. It's a business model.
Why I bought it
Because it was everywhere, and everywhere reads as proof.
His ads were on my Instagram between stories. They were the pre-roll on YouTube videos. Friends who don't even follow this stuff had seen him. Just over 1 million followers, students posting win screenshots, press logos on the sales page. My reasoning felt airtight at the time: nobody spends that much on ads for something that doesn't work, and a million people can't all be wrong.
I now know both halves of that sentence are wrong. Ad spend isn't proof the method works, it's proof the COURSE sells, and those are different businesses. The ads weren't funded by students making money. They were funded by students like me, buying. A million followers doesn't mean he's made anyone money except himself. The audience wasn't evidence. The audience WAS the business.
What the course actually was
Four modules on making money online with AI: pick a niche, set up an AI-powered service or store, use ChatGPT for content, scale with ads. A private Discord. Weekly group calls, in theory.
What was genuinely good, because I promised myself I'd be fair
The production was slick, the dashboard was nicer than most software I use, and module one's framing of "sell a service before you build anything" is actually correct advice you'll find in every legitimate business book. The first week of the Discord had real energy, strangers cheering each other on. And in fairness, buying it got me to finally START, which sounds pathetic to pay $4,000 for, but momentum is real even when it comes from a bad purchase.
What didn't work, which was everything else
Around module two the videos got shorter and the slides took over, and the slides were unmistakably AI-generated. Generic stock phrasing, bullet points that said nothing, the same advice you'd get from a free chatbot in thirty seconds, formatted prettier. A $4,000 course about making money with AI that was itself lazy AI output. The irony took a while to stop being infuriating.
The Discord's 4,000 members thinned to about nine active ones by week three. The weekly calls became monthly, then became recordings. And then came the real product: a $10,000 "inner circle" mastermind pitch. That's when I understood. I wasn't the customer. I was the lead. The course wasn't the product, I was, and the $4,000 was my qualification fee for the $10,000 pitch.
Someone in a small-business forum put it in one sentence that reorganized my head: the influencer makes money selling the map, not walking the path. Check whether the person teaching you earns from doing the thing or from teaching the thing. A big following only proves one skill: growing a following.
What actually worked
Freelance writing and admin work, upgraded by AI. That's it. That's the unsexy answer a $4,000 course spent four modules avoiding.
My first paying work was product descriptions for a local western-wear store whose website read like a customs form, found through a friend of a coworker, not a platform. I quoted $150 for the full catalog cleanup. With Claude drafting and me fixing anything that sounded wrong for Memphis, it took two evenings. That $150 taught me more than all four modules: clients don't pay for AI, they pay for finished work. AI just lets a non-technical person finish more work than they otherwise could.
From there it compounded slowly: resume rewrites at $75, social captions for two small businesses on $200 and $250 monthly retainers, email cleanups, a menu rewrite. Nothing I couldn't theoretically have done before AI, but all of it at three times the speed, which is the entire margin that makes this work around a full-time job.
The skill that made the difference wasn't prompting tricks. It was learning to work WITH the tool like a fast junior assistant: give it real context, check its claims, never send anything I can't defend line by line. I got that from structured practice, a course on Mindwand built around actual client deliverables in 15-minute lessons. And yes, I hear it: the guy who got burned by a course bought another course. The difference is the price-to-substance ratio. Four thousand dollars for mindset slides is a scam shaped like education; thirty-ish a month for hands-on lessons is what the material is actually worth, and you cancel the moment it stops earning its keep. Skool communities and Coursera cover similar ground. Format matters more than brand. Pick one. Finish it.
And no, YouTube doesn't replace any of this anymore
I tried the free route first, three weeks of it, and I'll save you the trip. YouTube's money-making content has become the same funnel as the paid courses: half the videos are secretly selling their own $997 thing, the other half contradict each other, and the algorithm feeds you hype in exactly the proportion you'll watch it, not in the proportion it's true. YouTube is fine once you already know what you are looking for. It is a maze when you do not.
The honest timeline
Month 1 (the course months don't count, those were negative): $150, the western-wear catalog. Month 2: $75. One resume. Nearly quit. Month 3: $390. First retainer landed. Month 4: $610. Month 5: $875. Month 6: $1,020. First month over a thousand. Month 7: $1,290 gross, $1,140 after tools.
Averaged over everything including the $4,000 hole, this ran deep in the red until month three. From there it climbed to something like $28 to $32 an hour for the hours I actually work now. Nobody selling a course leads with that curve, which is exactly why you should trust curves over screenshots.
The decision that mattered most
Local and boring beats global and glamorous when you're starting. The famous course pointed everyone at crowded global platforms and dropship empires. What actually worked was businesses within twenty minutes of my house whose owners answer their own phones. No algorithm to please, no forty competing proposals, and a referral network that markets for you free. The first thousand a month is hiding in businesses whose websites embarrass them, and there are more of those than people willing to fix them.
If you're where I was at 1am: it works, slower than the ads say and faster than the cynics say. Skip the $4,000 maps from people whose real product is their follower count. The path costs about $50 a month in tools and lessons, Mindwand, Skool, or Coursera all sell a version of it, and the rest is showing up. I hope the $4,000 sentence saves you the $4,000.