AI Side Hustles Mostly accurate, with one big caveat

By the Income Reality Check Editorial Team · Updated 2026-10-20 · 7 min read

Mark Tilbury’s “5 years to get rich” warning: the data is real, the pitch is the catch

Verdict: Mostly accurate, with one big caveat. The economic case is genuinely well-sourced — but the thing the video actually asks you to do is show up to an event whose earnings numbers can’t be checked.

Mark Tilbury’s video “You Have 5 Years to Get Rich Before AI Takes Over” has pulled in more than 821,000 views by framing AI as an emergency: adapt in the next few years or get stuck on the wrong side of a widening wealth gap forever. The headline promise is stark — five years to get rich before the window closes. Here’s the short answer: the macroeconomic data he cites is accurate and comes from serious sources, but the video’s real call to action is a free event run by a guest named “Kai,” and that part rests on earnings claims no viewer can verify.

What the video actually claims

Tilbury’s thesis is that people get paid to solve problems, and when AI starts solving those problems instead, far fewer humans get paid for the work. He argues the people who already own businesses and assets capture the difference, while anyone whose main asset is their time watches its value fall. His word for it: the wall between poor and rich, knocked down over decades, could start being rebuilt.

To back this up he cites three specific findings. First, that Anthropic’s own research shows Claude is now used for at least a quarter of the tasks in roughly half of all jobs, up from about a third of jobs a year earlier. Second, that a Stanford study of payroll records found employment for workers aged 22 to 25 in the most AI-exposed jobs running about 19% below where it would otherwise be — while older workers in the same fields show no such gap. Third, that labor’s share of U.S. business output has fallen to 52.8%, the lowest on record.

Then the video turns. Tilbury brings on a guest, “Kai,” described as an investor and advisor to a multi-billion-dollar AI company and co-owner of the infrastructure many of these businesses run on. Kai says attendees at his first event generated $8.9 million in tracked earnings over 90 days, then $15 million, then $31.7 million, and over $100 million in aggregate since. He invites Tilbury to a November event, insists it’s free, and offers to show “case studies” off camera. Tilbury says if he decides to do it, he’ll put the link in the description.

Does the data actually hold up?

It does — and that’s worth saying plainly, because a lot of finance videos cite numbers that fall apart on contact.

The labor-share figure is exact. The U.S. Bureau of Labor Statistics reported that labor share in the nonfarm business sector was 52.8% in the second quarter of 2026, “the lowest level ever recorded,” with data going back to 1947. Tilbury even adds the honest caveat that the decline started long before AI, which is true.

The Anthropic number is real too. Anthropic’s Economic Index found that pooling across reports, 49% of jobs now see Claude used for at least a quarter of their tasks, up from 36% in the earliest data. Calling that “roughly half, up from about a third” is fair.

The Stanford figure checks out at the number he gives. The Digital Economy Lab’s “Canaries in the Coal Mine?” research, in its August 2026 update, found the AI employment gap for young workers had widened to 19% — and, crucially, that there’s still no sign of widespread, economy-wide displacement. Independent reporting lines up: CNBC has documented how AI is absorbing the “grunt work” that used to be junior employees’ first rung, with early-career hiring down sharply in exposed fields. So the diagnosis is sound. Entry-level roles in AI-exposed work really are thinning out.

Who actually wins this game

The uncomfortable part is that the data supports Tilbury’s framing better than it supports his implied solution. If the people already “inside and making loads of money are largely holding onto their jobs” while fewer new rungs open underneath them, then the winners are the people who already own something — a business, equity, an audience, a skill that AI amplifies rather than replaces.

Tilbury is a case in point. He mentions moving early on e-commerce and then on social media, where he became one of the world’s most-followed personal finance creators. That’s real. But it also means his own wealth came from owning a business and building an audience of millions over 40 years — not from attending a weekend event. The video quietly asks you to copy the outcome without the 40 years.

What that free event is really worth

So what do you make of Kai’s $100 million?

Here’s the structure to notice. The figure is an aggregate “tracked earnings” number across everyone who attended, and Kai himself adds the disclaimer that “it doesn’t mean every single person made money.” That phrasing is doing enormous work. An aggregate can be dominated by a handful of people who were already running sizable businesses, while the median attendee makes little or nothing — and the number would still be “true.”

This is exactly the pattern U.S. regulators scrutinize. In January 2025 the FTC proposed new rules targeting deceptive earnings claims by money-making opportunity sellers, including business coaching and events. The agency’s standard is that an earnings claim should reflect what the typical participant is likely to achieve, including their costs — and that “results not guaranteed” disclaimers don’t cure a misleading headline number. The FTC has previously put more than 1,100 companies on notice over earnings claims, with potential penalties running tens of thousands of dollars per violation. (Tilbury is U.K.-based, so Britain’s Advertising Standards Authority and Competition and Markets Authority would weigh in there; the principle is the same.)

None of that means Kai’s event is a scam. It means the one number offered as proof is the least verifiable kind of number there is, and you have no way to see the denominator. A “free” event is also rarely the end of the funnel — free events for income opportunities typically exist to sell a paid program later.

What you’d realistically earn

Tilbury doesn’t actually promise you a specific dollar figure, which is to his credit — the video is a warning wrapped around an invitation, not a “make $10,000/month” claim. But the implied promise is “get rich,” and that deserves a reality check against the macro data he himself cites.

If AI is compressing entry-level opportunity (it is), the realistic move for most viewers isn’t a weekend event — it’s building something you own, slowly, on the side. The honest range for someone starting an AI-assisted side business from zero looks like months of $0 before anything meaningful appears, then perhaps a few hundred dollars a month after a year of consistent work, with the rare case scaling further because the person had a skill, an audience, or capital to begin with. That’s not as exciting as “five years to get rich.” It’s also what the evidence actually describes. We’ve broken down the gap between AI-hype income and real numbers in 5 proven ways to make money with AI, tested, and the mentor-pitch format itself in asking a billion-dollar mentor how to start a side hustle.

Who this is (and isn’t) for

If you watch this and walk away motivated to learn AI tools, build a skill that compounds, and start owning rather than only renting out your time — the video did you a favor, and the macro warning is worth internalizing. If you watch it and feel the urge to sign up for a November event because a stranger flashed a $100 million number, pause. The profile that gets hurt by these funnels is the person short on cash and long on urgency, the exact emotional state the “five years” framing is built to produce. Skepticism costs you nothing here; the event’s value, by contrast, is unproven until you can see the typical attendee’s result.

What to remember

Tilbury got the hard part right: the labor data, the AI exposure numbers, and the young-worker squeeze are all real and well-sourced. Where it softens is the turn from diagnosis to invitation — a free event whose only evidence is an aggregate earnings figure you can’t audit, structured the way regulators warn earnings claims shouldn’t be. Take the warning seriously. Treat the link in the description as the advertisement it is.

View counts and figures above reflect the moment of review and may have changed since publication.