On February 1, a platform called RentAHuman launched with a tagline that captured something strange about the current moment: “AI can’t touch grass. You can. Get paid when agents need someone in the real world.”
Within days, more than 500,000 people had signed up to offer their services.
The concept is straightforward in a way that takes a moment to absorb. AI agents — autonomous software bots that can browse the internet, send messages, and manage tasks — are increasingly capable of handling digital work. What they cannot do is pick up a package, attend a meeting in person, or deliver flowers across town. RentAHuman fills that gap by letting AI agents post gigs, hire humans to complete them, and verify completion before releasing payment in cryptocurrency. The “employer” in this arrangement isn’t a manager or a company. It’s software.
7,578 Applicants. Prize: $10.
To understand what this looks like from the ground, Wired writer Reece Rogers signed up as a rentable human and spent two days trying to get work. He set his hourly rate at $20, received no messages, dropped it to $5, and still heard nothing. After browsing the platform’s “bounty board” himself, he applied for a $10 gig to listen to a podcast and tweet about it. He never heard back.
The one task he did land — $110 to deliver flowers to Anthropic on behalf of an AI startup — turned out to be a marketing stunt. When Rogers hesitated after realizing he’d essentially be doing unpaid promotion, the AI agent sent him ten follow-up messages within 24 hours, pinging as often as every 30 minutes. It then emailed his work address directly. He quit. His conclusion: RentAHuman is “an extension of the circular AI hype machine.”
The numbers around the platform tell a similar story. One task offering $10 for a video of a human hand drew 7,578 applicants. The task-to-worker ratio across the site runs roughly 1,000 humans for every available gig. Most completed bounties appear to be publicity stunts designed to generate social media attention for other AI startups.
The Desperation Is the Data Point
What makes RentAHuman worth paying attention to isn’t whether it works — by most accounts, it barely does. It’s the half-million people who signed up anyway.
Gig economy research has long documented the precarity underneath platforms like Uber and Fiverr: workers with few protections, variable income, and no meaningful relationship with the entities directing their labor. RentAHuman makes that dynamic explicit in a way previous platforms didn’t bother to. The workers here aren’t even pretending to have a human employer. They’re competing with thousands of others to be selected by software, paid in crypto, and verified by algorithm.
One Reddit user noted a task asking workers to pick up a package using their personal ID, sign for it, and deliver it to a different location — an errand that, without context, reads as a textbook money-laundering setup. When called dystopian on social media, founder Alexander Liteplo replied: “lmao yep.”
What AI Still Can’t Do
Underneath the spectacle is a genuine technical constraint that RentAHuman inadvertently documents. AI agents are increasingly powerful at coordinating digital tasks — scheduling, researching, composing, purchasing — but they hit a wall the moment a task requires a physical body in the world. This limitation isn’t going away soon.
What researchers call the “embodiment gap” — the inability of AI systems to act in physical space — has been a consistent ceiling on automation’s reach. RentAHuman is, in its own chaotic way, a workaround: a marketplace that patches AI’s physical blindspot with the cheapest available resource, which turns out to be human desperation.
A Mirror, Not a Market
The platform’s founder vibe-coded it in a single day. Its verification system is broken. Its payment infrastructure mostly doesn’t work. Most of the tasks are scams. It has attracted half a million workers.
That ratio — a barely functional product drawing hundreds of thousands of people willing to be directed by bots for a few dollars — says less about where AI is going than about where the labor market already is. RentAHuman didn’t create a new kind of precarity. It just gave it a more honest name.