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

By the Income Reality Check Editorial Team · Updated 2026-08-15 · 7 min read

5 Claude AI side hustles from Shane Hummus: the traffic problem he skips

Verdict: Half-true — works only if you do the unspoken work. Every named success in this video is real, and every one of them was built on an audience the pitch treats as an afterthought.

Shane Hummus opens his video “5 Claude AI Side Hustles You Can Run While Working Full-Time” with a big number: the thing he built on his pharmacy lunch break “has made me over $10 million.” Then he runs through five ways to use Claude on the side — building apps, teaching what you know, guiding people’s decisions, selling AI agents, and explaining the system your job lives inside. The hustles are legitimate. But there’s a thread running under all five that he mentions once, near the end, almost in passing — and it’s the part that actually decides whether you make a dollar.

What the video actually claims

The pitch is that Claude does “the heavy part” so these fit the hours you have. You don’t code — you describe, and Claude writes the software. You don’t need to be an expert — you record what you already explain for free. Each hustle comes with a named person and what Shane calls “a real receipt.”

The receipts are specific. Max shipped 28 apps in 8 months and went from $200 to $10,000 a month while keeping his engineering job, spending about $300 a month on costs. A teacher named Lisa Collum turned a classroom writing curriculum into a business now doing $1.9 million a year in profit. English with Lucy pulls “$10 to $29,000 a month from AdSense alone.” Isaiah Harris gained 8,000 subscribers and made $32,000 in two months. Two friends, Sal Aiello and Monica Powers, spent $185 building an AI research tool and sold it for $150,000 — a figure CNBC Make It did in fact verify with documents in 2023.

So the numbers aren’t invented. That’s not the problem. The problem is what sits between you and them.

What the method actually requires

Watch how three of the five hustles are described. Teaching what you know: “get paid every time somebody watches.” Guiding decisions: Isaiah “gained 8,000 subscribers.” Explaining your system: Nurse Jen “was at a full-time income” in under two months. Notice the common ingredient — views. Subscribers. Watching. These aren’t Claude outputs. They’re distribution, and Claude doesn’t do distribution.

Shane says it himself, once: “all of these side hustles work, but they all need one thing and that is traffic. They all need distribution and YouTube is the best form of traffic.” That single sentence is the whole ballgame, and it arrives right before the pitch for his coaching program.

Here’s what building that traffic costs in 2026. To earn ad revenue at all, you have to clear the YouTube Partner Program bar: currently 1,000 subscribers plus 4,000 valid public watch hours in the past year (or 10 million Shorts views in 90 days). And that bar is rising. TechCrunch reported that starting February 1, 2027, the watch-hour threshold doubles to 8,000 hours a year. Eight thousand hours of watch time is not a lunch-break milestone. For a new channel it can mean months of consistent uploads before the first cent of AdSense.

Then there’s what a view is worth. RPM — revenue per mille, your earnings per 1,000 views after YouTube’s 45% cut — swings hard by niche and country. Finance content runs $9–$11; gaming and comedy sit at $2–$4. In the U.S. the blended average lands around $6, lower in many other markets. Do the arithmetic on English with Lucy’s “$10 to $29,000 a month from AdSense”: at a $6 RPM, the top of that range implies roughly 4–5 million monthly views. Achievable — Lucy has spent years getting there — but not a side effect of pressing record once.

The app hustle has its own gap. Max’s 28-app portfolio is real, and so is the survivorship math behind it. RevenueCat’s 2025 developer survey found the median indie iOS app earns under $1,000 a month, with roughly four in five never crossing that line within two years; the top 1% capture the lion’s share of store revenue. Max didn’t get lucky on one app — he shipped 28 because he expected most to fail. That’s the actual strategy, and it’s a very different pitch from “build an app in 20 minutes.”

Is any of this genuinely passive?

No — not in year one, and the video doesn’t really claim otherwise, which is to its credit. But the framing invites the misread. “Get paid every time somebody watches” sounds like income that keeps arriving while you sleep. What it leaves out is the front-loaded, unpaid stretch: the 20, 50, 100 videos that build enough watch history for YouTube to open the monetization door, plus the thumbnails, titles, and SEO that get anyone to click. Claude can draft a script. It can’t make strangers subscribe.

There’s also a regulatory backdrop worth knowing if you’re the skeptical type. In the U.S., the FTC has proposed rules specifically to curb deceptive earnings claims by coaching and money-making programs — including the tactic of dangling big specific numbers and leaning on “results not typical” disclaimers. Shane’s video isn’t an FTC action, and stacking real named case studies is a defensible way to pitch. But when a video’s throughline is “here are people who made millions, now book a call,” the reason a regulator built a rule for exactly that shape is worth holding in mind.

Who actually wins this game

The winners cluster into recognizable types. Lisa Collum had a decade of classroom expertise and a finished curriculum before she photocopied a single binder. Lucy spent years compounding search rank. Isaiah and Nurse Jen went from “zero” fast — but both did it inside a paid coaching program with a niche-selection system, scripting help, and a feedback loop, which is precisely what Shane is selling. Max was a professional iOS engineer who could ship and debug at speed.

Strip it down and the people who clear real money share one of four things: existing domain expertise, an existing audience, paid help or capital, or early-mover timing. Claude lowers the production cost of the content. It does nothing to the distribution cost, and distribution is where beginners actually stall.

What you’d realistically earn

Set expectations against the survivorship data, not the highlight reel. A committed beginner posting consistently should plan on roughly $0 for the first several months while building toward the Partner Program threshold. After a year of steady work in a decent niche, a few hundred to a couple thousand dollars a month from AdSense is a fair target — with wide variance and no guarantee. The app path is more binary: most attempts earn under $1,000/month, a few break out, and the winners usually come from a portfolio, not a single 20-minute build. The $10 million and $1.9 million figures are ceilings from multi-year operators, not forecasts for your first year.

Could you outrun those ranges? Sure — with a strong existing skill, a real audience, or a niche that happens to command a high RPM. That’s the honest version of the upside.

Who this is (and isn’t) for

This makes sense if you already have deep, specific knowledge people search for, can commit 5–10 hours a week for six to twelve months before meaningful money appears, and treat the first stretch as unpaid audience-building rather than a paycheck. It’s a poor fit if you’re counting on income within weeks, have no topic you can speak on with authority, or expect Claude to handle the part where strangers find and trust you. The tool is real. The audience is the job.

What to remember

Shane Hummus is describing legitimate hustles with real people behind them, and he’s not hiding the ball — he names traffic as the thing they all need. The catch is that the traffic is 90% of the work and gets one sentence. Claude writes the script; you still have to earn the audience, clear a monetization bar that’s about to get steeper, and outlast the months before the numbers move. For more on how these Claude pitches hold up, see our looks at 8 Claude AI side hustles that “replace” a full-time job and 7 realistic Claude AI side hustles for people with full-time jobs.