The “lazy way to make money with AI” is everywhere. YouTube thumbnails scream about passive income with AI influencers. Creators boast of thousands in revenue. But one honest guy tried it, documented every step, and made $157 after seven days. He also admitted the content was junk, the challenge was a failure, and the whole thing made him uncomfortable.
He should be uncomfortable. Because this business model isn’t just lazy in the effort sense. It’s lazy in the ethical sense too. And most people in the AI space won’t say that out loud, because admitting it means questioning a gravy train they’re riding.
The problem is that it works exactly well enough to fool people.
The Setup: Three Rules, One Week, Zero Face
The creator set three rules: spend under $99, never show his face or use his voice, and build the first AI influencer by the end of day one. He created three characters using Claude for the backstories, Higgs Field for the images and videos, and a skill inside Claude to write the scripts. Total cost for the tool chain: $99 for the Higgs Field plan. Claude was free. Stanstore for hosting the workbook used a free trial.
He did everything in one day. Photos, voices, scripts, videos. Then he posted 10 Reels per account on Instagram.
Nothing about this is hard. That’s the point. But hard isn’t the problem. The problem is that it works exactly well enough to fool people.
The Results: Half a Win, Full Failure
After seven days, the numbers told a messy story. The health influencer Amos took first place among the three accounts. But the creator’s personal favorite, the rich grandma Vivian, had a video hit 250,000 views — and he had abandoned her to focus on Amos.
He sold 23 copies of a 30-day workbook at $12 each. Revenue: $276. After PayPal fees ($19.52) and the Higgs Field plan ($99), profit landed at $157.48. He failed to hit his goal of $1,000, so he gave away $1,000 of his own money to a commenter.
The raw profit isn’t the story. The engagement is. People in the comments were arguing with each other about whether Siena (the relationship AI) had the right to choose her men. They had full conversations with a fictional character, and neither participant realized they were talking about a bundle of pixels. That’s not a business model. That’s a social pathology waiting to be exploited.
The Uncomfortable Truth Nobody Wants to Touch
Here’s the argument no one in the AI influencer space will make on camera: this is ethically bankrupt, creatively hollow, and socially corrosive.
The creator himself said it. I’m paraphrasing, but he basically called the output “junk AI content” and said the internet loses its soul when everything is generated. Then he immediately pivoted back to promoting his discount code for Higgs Field.
That cognitive dissonance is the real product. You get paid to feel guilty, and then you monetize the guilt by selling more tools to people who will feel the same.
The Deception Is the Feature, Not the Bug
Yes, the creator labeled his accounts as AI. But look at what happened: commenters still debated Siena’s dating choices. They still asked where she bought her clothes. The label became invisible the moment the content was compelling. AI influencers don’t need to fool everyone. They only need to fool enough people to generate engagement metrics.
And the metrics work. That 250,000-view video proves it. But those views didn’t convert into sales for Vivian because the creator had already pivoted to Amos. The opportunity cost of picking the wrong character wiped out a week of effort. That’s not passive income. That’s gambling with a spreadsheet.
The Business Model Collapses Under Scrutiny
Let’s run the math honestly. 23 sales. That took a week of content creation across three accounts. At scale, you’d need dozens of accounts to reach meaningful income. But every account requires maintenance, fresh content, updated products, and monitoring for platform policy changes.
Instagram already labels some AI content. TikTok is exploring similar rules. The window for this “lazy way” is closing, and the people selling courses on it know that. That’s why they’re selling courses now.
Why Most People Won’t Say This
Three reasons.
First, admission of guilt. If you’re making money off AI influencers and then criticize the model, you undermine your own revenue. No one does that.
Second, the grift is popular. Criticizing the grift means fewer views, fewer affiliate clicks, fewer “controversial” thumbnails. It’s easier to pretend the model works and just needs more optimization.
Third, the line between tool and product is blurry. Using ChatGPT to draft an email is fine. Using a full avatar pipeline to impersonate a human for profit? That’s a different category. But the industry treats them as the same thing, so any criticism gets dismissed as Luddite panic.
The One Detail That Proves the Point
The creator used a Claude skill to generate the workbook. A skill is a reusable prompt that automates a process. He fed it the character bio and topic, and it spat out a 30-day challenge complete with bonus recipes. He admitted afterward he didn’t read the whole thing for accuracy.
He sold that for $12 a pop.
The quality bar for this business model isn’t “valuable.” It’s “good enough to pass a 10-second scroll.” That’s not a sustainable advantage. That’s a race to the bottom where the bottom is zero quality and the winner is whoever can generate the most volume before platforms catch on.
The Real Lazy Way Is Honesty
Here’s the alternative that no one will sell you a course on: build something real, use AI to accelerate the boring parts, and put your own face behind it.
The creator proved that AI influencers work as an engagement hack. He also proved that $157 is what you get when you optimize for laziness instead of value. The Vivian account hit 250k views because the character had a recognizable personality. But he killed that momentum by switching to the “safer” pick.
If he had committed to Vivian, put real effort into the content beyond what a Claude skill could write, and built a genuine audience around the character’s quirks rather than a generic workbook, the numbers would look different. But that would require work. Actual work. Not lazy work. Not “prompt and pray” work.
And that’s the position most people won’t take publicly: the lazy way doesn’t work well enough to matter, and the work required to make it matter is indistinguishable from building a real business. So you might as well build the real thing and skip the ethical gymnastics.
- How much profit did the creator make?The creator made $157.48 profit after a week of running AI influencer accounts.
- What tools did the creator use?He used Claude for backstories, Higgs Field for images and videos, and Stanstore for hosting the workbook.
- Why did the creator call the experiment a failure?He failed to hit his $1,000 goal and admitted the content was junk and ethically questionable.
