AI Side Hustles Hype — the math doesn't survive a calculator

By the Income Reality Check Editorial Team · Updated 2026-11-14 · 7 min read

Codie Sanchez’s AI bots turned $100 into $1,000? Here’s what the cash actually says

Verdict: Hype — the math doesn’t survive a calculator. The $1,000 finish line is only crossed by counting leads as “value,” not dollars; the real cash across five bots was about $87.

Codie Sanchez set herself a clean challenge in “I Tried The ‘Easiest’ Ways To Make Money With AI”: take $100, spend ten days, and let an AI agent she calls “Grokbot” turn it into $1,000. Three rules — the bots do the work, she spends no more than $100, and “the money only counts if it’s cold hard cash or saves real money.” By the end she declares victory. But if you hold her to her own third rule and add up only the cash, the total lands nowhere near four figures. That gap is the story.

What the video actually claims

Sanchez runs five experiments with the same AI agent. A trading bot named Wolfie pulls stock ideas from Reddit, X, and news and trades them through Robinhood. A three-bot team (Alvin, Simon, and Theodore) sources free items, arranges couriers, and re-lists them on Facebook Marketplace. A clipping bot hunts for editing gigs on Fiverr and Upwork. A bot named Finn builds a Shopify landing page and fires re-engagement emails through Klaviyo to 3,000 people who attended her book launch. And a cold-outreach bot named Dora scrapes business owners off Google, LinkedIn, and Apollo to fill seats at an Austin workshop.

Her scorecard, as she reads it out: the Facebook reselling made about $88 net, the book-email bot netted roughly $14 in a two-hour test, the clipping bot made $0 in outside revenue, Wolfie the trading bot lost about $15, and the cold-lead bot produced “about 2K right there in value.” Add the big number and she’s over $1,000. “Can Grokbot turn $100 into $1,000? Yeah,” she concludes, then invites viewers to comment “bot” for the templates.

Worth saying plainly: this is a promotional format. The exact workflows sit behind a link in the description, and Sanchez runs a $100M-plus portfolio of businesses. That matters for whether you can copy the result.

The number that doesn’t hold

Here’s where her own rule bites. Rule three said money counts only as cash or real savings. Look at what the cold-lead bot actually produced: ten interested replies and two booked calls. Zero sales. Sanchez values it anyway — “$100 each” per response and “$500 bucks” per booked call, because that’s what those actions cost her on Meta. That’s how $0 in revenue becomes “$1,942 net.”

Imputed pipeline value is a real concept in sales. It is not cash, and it is not a saving — it’s a bet that some of those calls will close later. Strip it out and tally only what hit a bank account: about $88 from reselling, $14.48 from the book emails, $0 from clipping, minus roughly $15 on the trading bot. That’s around $87. The $100-to-$1,000 headline survives only because a sales forecast got laundered into the win column.

What the method actually requires

Start with the agent itself. Sanchez shows bots that “click, scroll, browse the internet” — the computer-use category of AI. Independent testing is blunt about where that technology sits. In the Online-Mind2Web study of 300 everyday tasks across 136 live websites, the strongest agent completed only 61.3%, and most agents finished closer to 30% (arXiv, 2025). The video quietly confirms this: the clipping bot stalled on an “am I a human?” check, and Wolfie had “no idea” what it was buying. “AI does it all for you” is true at the drafting step and shaky at the doing step.

Then there’s the cold-email engine, which is the segment generating most of the claimed value. Blasting business owners you scraped is regulated. In the U.S., the FTC’s CAN-SPAM Act requires every commercial email to carry accurate headers, a truthful subject line, a valid physical postal address, and a working opt-out you honor within 10 business days — and it holds you liable even when a contractor (or a bot) sends on your behalf. Penalties run up to $53,088 per email (FTC). A bot firing 200 messages a day across four inboxes is 200 chances a day to get that wrong. None of that appears in the video. (Readers in the U.K., EU, Canada, and Australia face stricter consent regimes still — many require opt-in before the first message, not opt-out.)

The trading bot deserves its own caution. Wolfie’s job was to trade on social-media hype, which is precisely the behavior U.S. regulators keep flagging. The SEC has charged operators who lured retail investors with “AI-generated investment tips” pushed through social feeds and chat groups, misappropriating over $14 million in one 2025 case (SEC). Sanchez is clearly not running that scam — she loses her own money on camera and says “please do not do this.” Good. The point stands: pointing automation at a bad process just loses money faster, which is exactly what Wolfie did.

The quieter costs add up too. The book-email bot runs on a Shopify store (plans start around $39/month before card fees) plus Klaviyo (Shopify). The Facebook flips needed paid couriers, product photography, and — crucially — items to sell. She was clearing junk from her own office, including a set of stools and a free bowling ball.

Who actually wins this game?

The honest answer sits in plain sight: the person who already owns the business the bots plug into. The re-engagement bot only “worked” because Sanchez had a 3,000-person list from a book launch she’d “beat the hell out of” with prior marketing. The cold-lead bot checked HubSpot before emailing and handed replies to a human sales team. The clipping bot’s real payoff was becoming her existing editor’s assistant, trimming a cost she was already paying ($50–$250 per short).

See the pattern? Every genuine win depended on infrastructure a beginner with $100 doesn’t have — an audience, a CRM, a sales team, a product, and existing paid-marketing benchmarks to value leads against. The AI amplified distribution that was already there. It didn’t create any.

What you’d realistically earn

Say you’re starting cold, with $100 and no list. The only segment that produces cash without a pre-built business is the reselling flip, and even that made $88 largely by liquidating free office furniture. A realistic beginner running Marketplace flips might clear $50–$200 a month once courier fees, gas, and dud listings are netted out — real money, but a grind, and nothing like $1,000 in ten days. We’ve walked through that specific math before in our look at the $1,000-in-7-days Facebook Marketplace claim.

The email and cold-outreach bots are where the big numbers live, and where a beginner has nothing to point them at. Two sales from 3,000 emails to a warm, already-marketed list is roughly a 0.07% conversion — below the 1–4% that re-engagement campaigns typically manage, because that list was already squeezed. The “$2K in value” from cold leads converts to exactly $0 until a call closes. For a fuller tour of which AI methods pay and which stall, see our 5 ways to make money in 2026 with AI breakdown.

Who this is (and isn’t) for

This workflow makes sense if you already run a business with a customer list, dead quotes, abandoned carts, or recurring supplier costs — the “competing supplier quote” idea she floats (negotiate 8% off a $200k parts bill) is genuinely useful, and so is cutting an editor’s busywork. If that’s you, and you have a few hours a week to supervise flaky agents and a budget to absorb the Shopify, Klaviyo, Hunter.io, and API fees, the tools can trim real cost. It is not for someone with $100, no audience, and no business, hoping the bot manufactures income from nothing. The video’s own ledger shows that version earns about $87 and a lost $15 on stocks.

What to remember

Sanchez is more candid than most — she shows the losing bots and labels the lead number as “value.” But the challenge was framed around cash, and on cash it falls short. The methods that worked were leverage on assets she already owned, and the one that supposedly cleared the bar never collected a dollar. Treat the video as a demo of what agents can assist with inside a running business, not a blueprint for turning a hundred bucks into a thousand.