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

By the Income Reality Check Editorial Team · Updated 2026-07-07 · 7 min read

Wholesale Ted’s 5 AI side hustles: the SEO work behind the $6 can cooler

Verdict: Half-true — works only if you do the unspoken work. The tools are cheap and real; the traffic that turns them into income is neither.

Wholesale Ted’s “The Best AI Side Hustle Ideas NO ONE Is Talking About” has pulled in more than 56,000 views by promising five AI businesses “making their owners rich” while everyone else chases Claude Code. The five: faceless AI music channels, personalized Etsy can coolers, printable junk-journal supplies, an AI research agency on Fiverr, and local “micro lead sites.” Every tool she names is real and mostly cheap. The catch is that four of the five run entirely on search ranking — and ranking is the part the video treats as a footnote.

What the video actually claims

The pitch is a workflow tour, not a hard income figure — which is smart, because it’s harder to fact-check a vibe than a number. Still, the numbers are there if you listen. She points to an AI music channel that’s nine months old with 58 million long-form views (24 million in the last 90 days) and flashes an estimated ad-revenue figure she calls a “really nice” monthly check. She shows a two-month-old channel already doing 50,000 to 70,000 views a day.

Then come the products. An Etsy can cooler selling for $6 that “66 people bought in the past 24 hours,” which you can print on demand through Printify for $3.28 a unit at a “25 to 30%” margin. Junk-journal shops that hit 1,700 sales in three months, 3,500 in six, 53,000 in eleven. Fiverr research agencies with “five orders in the queue.” And local lead-generation sites — she says she owned some in New Zealand and sold them — that rank on Google for terms like “pool service in Anaheim” and sell each lead to a local contractor.

The production stack is genuinely accessible: Suno for music, Google Flow and Nano Banana for images, Kling for animation, CapCut to assemble, Canva and Printify for products, and Base44 to vibe-code a lead site. Nothing here is vaporware.

What the method actually requires

Here’s where the video gets quiet. Making the asset is the easy 10%. Getting it found is the other 90%, and it isn’t free.

Start with the music channel, because it’s the one that hides the most. YouTube doesn’t pay a cent until you’re in the Partner Program, which requires 1,000 subscribers plus 4,000 valid public watch hours in the past 12 months (or 1,000 subs and 10 million Shorts views in 90 days), per YouTube’s own eligibility page. For hour-long ambient music, 4,000 watch hours is real runtime you have to accumulate before your first payout — and the channels she shows as “proof” are survivors, not the median. She even flags the risk herself: YouTube treats AI music as high-scrutiny content, and Content ID claims or demonetization can end a channel. The tooling is cheap (Suno’s Pro plan is $8–$10 a month for up to 500 songs, exactly as she says). The audience is not something you buy for $10.

The Etsy math deserves a calculator, because “25 to 30% margin” quietly ignores Etsy’s stack of fees. Per Etsy’s fee policy, every sale carries a $0.20 listing fee, a 6.5% transaction fee (charged on item price plus shipping), and payment processing of 3% + $0.25 in the U.S. NerdWallet totals that at roughly 9.5% plus $0.45 per order. Run the can cooler through it:

That’s about 28% — her number holds if shipping is covered by the buyer and if you never pay for an ad. But once a shop clears $10,000 in a rolling year, Etsy auto-enrolls it in Offsite Ads and skims another 12–15% on ad-driven sales. And NerdWallet’s own guide warns that print-on-demand goods are among the most saturated categories on Etsy — the exact niche she’s recommending. Selling ten cheap coolers nets you around $17. The business only exists at volume, and volume needs search ranking.

Which is the thread running through all of it. The research-agency and lead-site ideas aren’t even AI products — they’re SEO services with an AI assistant bolted on. The lead site only earns if it ranks for “landscaping Chandler,” and ranking a brand-new domain in Google takes months of content, backlinks, and patience that a $1 domain and a Base44 build don’t buy you.

Who actually wins this game?

The same people who win every “just upload it” pitch: those who already have distribution. The music channels she cites as proof are the handful that broke out of a niche now crowded with identical Suno playlists — for every one at 58 million views, there are thousands stuck under the monetization threshold. The 53,000-sale junk-journal shop is an eleven-month head start over anyone starting today, in a niche Etsy publicly named as fast-growing (which is another way of saying fast-flooding with competitors).

The people who reliably profit are operators who understand keyword research, can grind out listings or videos for months without income, and treat this as a marketing job that happens to use AI. AI collapsed the cost of making the product. It did nothing to the cost of ranking it — arguably it made ranking harder, because everyone now has the same cheap tools.

What you’d realistically earn

Honest ranges, not the highlight reel. A new AI music channel earns exactly $0 until it clears the Partner Program bar, which can take months of daily uploads — and plenty never get there. Once monetized, long-form ambient music runs a modest RPM; the viral examples are outliers, not the base case.

An Etsy shop is a coin flip weighted against you. Etsy has millions of active sellers, and independent estimates consistently put the median seller’s income at a few hundred dollars a year, not a month — a small top tier captures most of the sales. A realistic first year selling POD can coolers or printable journals looks like $0–$100 a month for six months, and maybe a few hundred a month after a year if your SEO lands. The Fiverr research gig can earn per-order money quickly but caps at your own hours — it’s freelancing, not passive income. Lead sites can genuinely pay, but the payout comes after the months it takes to rank.

Nothing here is a scam. But if a U.S. seller repackaged these clips into a paid course promising specific earnings, the FTC’s proposed Business Opportunity Rule changes would require them to substantiate those numbers in writing on request. A free YouTube tour showing other people’s wins clears that bar easily — which is exactly why the “proof” is always someone else’s dashboard.

Who this is (and isn’t) for

This works for someone with 10–15 hours a week, near-zero startup cash, no fear of learning SEO, and the patience to earn nothing for a season. If you like the marketing puzzle — keyword research, competitor teardowns, iterating on listings — the AI tools genuinely lower the barrier to entry, and the lead-site model can pay well once ranked. It’s a poor fit for anyone expecting passive income by month two, anyone who won’t touch keyword tools, or anyone counting on the product itself to sell without promotion. For the wider AI money landscape, compare this against our breakdowns of what actually makes money with AI and the easiest beginner AI side hustle.

What to remember

Wholesale Ted didn’t lie. Suno costs what she says, Printify’s coolers cost what she says, the fees pencil out to roughly her margin, and the niches really are trending. What she skipped — across all five ideas — is that AI builds the asset in an afternoon and the market pays you only after months of ranking work she compressed into a sentence. The half-truth isn’t the tools. It’s the timeline.