AI Side Hustles Half-true — works only if you do the unspoken work

By the Income Reality Check Editorial Team · Updated 2026-10-23 · 7 min read

AI agents saved him $50,000 on a pool — can they make you money?

Verdict: Half-true — works only if you do the unspoken work. The agents really do handle the boring steps; the part that produces income is still a business you have to build.

Chris Koerner’s video “I Let AI Make and Save Me Money for 6 Weeks (Here’s What Worked),” on The Koerner Office, has pulled in more than 57,000 views with a simple hook: he handed AI agents his credit card and they went to work — buying, negotiating, filling out “82 contact forms at 2:00 in the morning.” The headline moment is a pool project where he says agents saved him $50,000. Is any of this real? Mostly, yes — the tools exist and they do what he says. The catch is that almost nothing he shows is actual income. It’s savings, and a handful of unsolicited leads, dressed up as six ready-made businesses.

What the video actually claims

Koerner lays out “six businesses you can start this week,” all run through consumer AI agents he names as Muse (Meta’s), Grockbot (from X’s Grok), and a startup called Instinct. The pool story is the anchor: after two weeks of in-person meetings produced zero quotes, he says one prompt to Muse found 82 pool builders, filled out their contact forms, and logged eight quotes in 24 hours — ranging from $120,000 to $240,000 for the same build. Getting more quotes, he argues, is how he’ll avoid overpaying by “50 or even $100,000.”

From there the pitches widen. He flips garage items on Facebook Marketplace and sources a spray-foam repair for $1,200 that first quoted at $12,000. He describes “agentic drop shipping” — listing fruit trees you don’t own, buying from the cheapest nursery after a customer pays, pocketing a 30–60% markup. He shows $11 domains (a crane-rental site in Laredo, a shower-door site in Brownsville) that he says generate leads he can rent to local contractors for $800–$1,200 a month. And he walks through an automated outreach machine for landing AI-consulting clients.

Two things sit underneath all of it. Koerner runs a paid program, playmakersai.com, where he says he’s “teaching over 300 people to do AI consulting,” and the video carries a GoHighLevel affiliate offer. The “six ways” are genuinely interesting. They’re also the top of a funnel.

What the method actually requires

Start with the tools, because here the video is accurate. Muse is real: Meta announced it on September 8, 2026 as a personal AI agent that “doesn’t just answer questions, it actually does the work” — completing forms, negotiating, and making purchases through a secured checkout, with a Sentinel step that approves sensitive actions. So the production step Koerner describes — agents doing the clicking — isn’t hype. That part survives scrutiny.

The money step is where the gaps open. Take the pool. Saving $50,000 by collecting eight quotes instead of two is plausible, but it’s a one-time personal saving on a quarter-million-dollar purchase — not a repeatable income stream. The “business” he spins out of it (charging homeowners 4–6% to gather quotes, like a pool consultant) is a real service model, but it requires you to find clients, build and maintain a productized tool, and deliver results reliably. None of that is the agent’s job.

Agentic drop shipping is the clearest example of skipped work. Buying a tree from a nursery after a customer orders is retail arbitrage with extra steps, and the margins are nothing like the “fugazi, it’s all pennies” framing. Shopify’s own 2026 analysis puts dropshipping margins at 20% to 50% gross — before ads, hosting and fees — and notes customer acquisition costs running from about $21 in some categories to $377 in electronics. Beginners, Shopify says, typically take home $1,000–$2,000 a month, and only about a third of US businesses survive a decade. Koerner’s model also depends on suppliers who don’t know they’re your supplier, which means no inventory feed, no guaranteed stock, and chargeback risk when a tree arrives dead or never ships. His own fix — the agent refunds the customer — is a cost, not a feature.

The $11 lead-gen sites have a timing problem the video glosses. Koerner shows a site ranking #3 for a Spanish-language search within weeks. Google’s own guidance is blunter: indexing a new page takes “anywhere from a day or two to a few weeks,” and ranking is a separate, slower process — “some changes might take effect in a few hours, others could take several months.” Rank-and-rent operators generally plan for three to six months of SEO work before a site produces steady leads. One early lead in an uncompetitive city is possible; a portfolio of sites paying $800/month each is a project, not an afternoon.

Is the automated outreach even allowed?

This is the question the video never asks. The outreach machine fills out 50 contact forms with “a number that you control,” calls businesses after hours to check for voicemail, and — in the drop-shipping and Marketplace sections — automates activity on platforms whose terms restrict exactly that kind of scripted behavior. Mass automated cold outreach brushes up against spam and telemarketing rules (in the US, that’s the CAN-SPAM Act and TCPA), and automated posting or buying can get Facebook and Marketplace accounts banned regardless of which agent is driving.

There’s a second regulatory angle that matters more for the course than the agents. US readers should know the Federal Trade Commission has been actively prosecuting “AI will make you money” pitches. In its Operation AI Comply sweep, the FTC alleged Ascend Ecom defrauded consumers of at least $25 million with AI-powered storefront promises, charged Ecommerce Empire Builders over AI stores sold for up to $35,000, and targeted the FBA Machine scheme over more than $15.9 million in losses. To be clear, Koerner isn’t among those cases, and gathering quotes for your own pool breaks no rules. But under the FTC’s Business Opportunity Rule, anyone selling a money-making program faces a hard line: it’s illegal to make an earnings claim without written proof on hand that typical buyers get similar results. “One of these pays your mortgage” is the kind of claim that rule exists to police.

Who actually wins this game

Look at who’s positioned to profit from exactly these six plays, and a pattern shows up. Koerner has a podcast audience, 2.5 years of documenting AI, a paid community of 300-plus people, and businesses in tree trimming and sport courts that already give him contractor instincts. When he says “any real business owner can ballpark a quote,” that’s domain expertise talking — expertise he tells beginners not to worry about.

The people who turn agent workflows into real money tend to be operators who already have one of three things: an audience to sell the course or tool to, prior expertise in the niche they’re automating, or the patience and budget to grind SEO and outreach for months. The agent compresses the labor. It doesn’t hand you the audience, the trust, or the fulfillment. That’s why the honest business in this video isn’t pool quotes or fruit trees — it’s teaching other people to try, which is what playmakersai.com and the GoHighLevel link are for.

What you’d realistically earn

Set the video’s numbers next to the data. The $50,000 is savings on a purchase, not income — treat it as what good comparison shopping is worth, not a paycheck. The $3,000, $9,000 and $23,000 figures are the values of leads, not money he’s banked; he says he’s giving those leads away free for now. The drop-shipping margins are real but modest once you count ads and refunds, and Shopify’s own range for beginners is $1,000–$2,000 a month after a learning curve, with most stores failing.

A realistic path for someone starting today, part-time, with no audience: expect roughly $0 for the first few months while you build a site or a client pipeline, then — if you stick with it and the niche cooperates — maybe a few hundred dollars a month from a first rank-and-rent site or a first consulting client after six months to a year. The agent fees really are cheap (Koerner’s “tens of dollars a month” is fair). The time and persistence are not. If you want a fuller picture of how these AI-agent claims tend to shake out, our breakdowns of running a Claude dropshipping store and whether AI has really changed how people make money cover similar ground.

Who this is (and isn’t) for

This makes sense if you already have a service skill or a local network, 5–10 hours a week to spend on outreach and SEO, a small tolerance for platform bans and dead ends, and the discipline to treat the agent as a tireless assistant rather than a money printer. If that’s you, the outreach and quote-gathering workflows can genuinely save real hours. It does not make sense if you’re expecting the “I did nothing, I just sat there and watched” experience to produce rent money in week one, or if you’d be buying a course on the strength of lead values that haven’t converted to cash. The agents are cheap enough to test. The businesses are not passive.

What to remember

Koerner isn’t selling snake oil — Muse exists, the agents do automate the tedious parts, and $4 in tokens to collect pool quotes is a smart move. But “saved me $50,000” and “a $23,000 lead” are not the same as “made me $50,000,” and every one of the six businesses still needs the distribution, trust and fulfillment work the video treats as an afterthought. The agent does the clicking. You still have to build the company.