A stranger's car. A stranger's spare room. Twenty years ago you'd have called that a burglary with extra steps, and now it's a Tuesday.
It's the most normalized transaction on the internet, and almost nobody stops to ask how it got that way.
Daniel's been chewing on exactly that. Here's what he wrote in. He's pointing at a whole category of sites that share one structure. Airbnb for accommodation, Turo for car rental. What they have in common is that they let ordinary people take an asset they already own and turn it into a business. Not a franchise, not a small company. A distributed version of something that used to require a professional operator, running in parallel to the professionals who are still there.
And he's asking four things, roughly.
Four things. What is this model actually, precisely. Where has it worked best. Where is it gaining ground now. And is there a platform you can point to that shows what it takes to do it properly, meaning the asset owner makes real income, the asset is protected, and the customer gets something affordable that doesn't make them regret it. Then the controversial part. In Israel the accommodation version has produced de facto hotels, and the tax authority has started going after undisclosed income through that channel. Guests save money on hotels and then report wildly inconsistent experiences. He wants the whole thing, mechanism and mess.
The mechanism is easy to describe. Getting it right is where everybody falls over, and it's always the same three places.
Then let's do the mechanism first. What actually makes this work, and why it isn't just a smaller hotel or a smaller Hertz.
The cleanest name for it is distributed asset-leveraging. Peer-to-peer asset rental if you want the plain version. The defining feature is that the platform owns no inventory. None. It doesn't own a single room, a single car, a single boat. What it owns is the connective tissue. Discovery, so people can find the asset. Identity verification, so both sides know who they're dealing with. Payments. Insurance and reputation infrastructure. Dispute resolution when it goes wrong. The host brings the productive asset. The platform takes a commission, and that commission runs somewhere between ten and twenty-five percent depending on the vertical.
So the platform is a matchmaker with a legal department.
A matchmaker with a legal department and an actuarial table. That's the shape of it. And the economic trick underneath is that it converts idle private assets into revenue-generating capital without the platform carrying the capital cost, the labor cost, or historically the regulatory burden. A car sits parked roughly ninety-five percent of the time. Ninety-five percent. A spare room is empty most nights of the year. That idle time is the raw material. The platform didn't have to buy the car to sell the car's idle hours.
And that's the difference from a hotel or a rental agency. The Marriott has to buy the building. Hertz has to buy the fleet.
Right, and they have to depreciate it, maintain it, staff it, insure it, and comply with whatever the local regulator demands. In this model all of that sits with the host. The platform is asset-light in a way that a traditional operator structurally cannot be.
Which is also where the trouble starts, because the host is now carrying depreciation and maintenance and risk, and the host is not a professional.
Three things make it distinct, and I'd put them in this order. No capital ownership. Regulatory arbitrage, which is the recurring critique and honestly the honest one. And variable, non-professionalized supply. Quality is inconsistent by design, because the supply comes from amateurs rather than trained staff. That third one isn't a bug somebody forgot to fix. It's the model.
So the arc for today. Mechanism, then where it's actually worked, then the professionalization problem and the regulatory reckoning, then what doing it right would even require.
And the pattern for where it works is more predictive than people expect.
Go.
Three traits. High-value, high-idle-time asset. Low transaction friction. And a large addressable base of amateur owners. Hit all three and the model flies. Miss one and it stalls, and you can usually tell which one you missed by looking at what the platform is begging people to do.
Accommodation first.
Perfect fit. A spare room is high-value in the sense that a night of accommodation in a city is expensive, it's idle most of the year, the transaction is a booking and a key, and the base of amateur owners is enormous. Anybody with a spare bedroom. Airbnb, Vrbo, and Booking.com, which is a hybrid because it started as a hotel aggregator and bolted on the peer-to-peer supply later. Accommodation is the archetype because it hits all three traits cleanly.
Cars.
Turo, Getaround. The asset is expensive, it's idle ninety-five percent of the time, the transaction is a booking and a key fob, and the base of amateur owners is anybody with a second car or a car they don't drive to work. Slightly harder than accommodation because the asset moves and can be damaged and can be in an accident, which is why insurance is the whole ballgame in that vertical.
And rideshare is adjacent rather than identical.
Uber and Lyft. The platform doesn't rent you the car, it rents you the car plus the driver. So it's leveraging two assets at once, the driver-owned vehicle and the driver's labor. Delivery is the same shape. DoorDash, Instacart. Gig labor plus a personal vehicle. Those are cousins of the model rather than the pure form.
Parking.
SpotHero, JustPark. Private driveways and lots. This one's quietly excellent. The asset is a patch of asphalt that's empty while you're at work, the transaction is trivial, and the regulatory friction is almost nothing compared to housing. Somebody renting out their driveway is not running a de facto hotel and nobody's city council is going to hold a hearing about it.
Storage.
Neighbor, Stashbee. Spare garages, basements, back rooms. Same story. Low regulatory friction, clean economics, and the asset is idle. A garage you can't park in because it's full of your parents' furniture is not doing anything for you.
Boats.
GetMyBoat, Boatsetter. And this is the one people keep pointing at as the next big vertical, because a boat is the ideal profile. Enormously expensive, idle almost all the time, and the owner is paying marina fees and maintenance whether it moves or not. High asset cost plus high idle time is exactly the profile that made accommodation work.
RVs.
Outdoorsy, RVshare. Strong growth after twenty twenty, for obvious reasons. Same logic. An RV is a mortgage-sized asset that sits in a driveway forty-eight weeks a year.
And equipment, tools, that whole category.
Fat Llama, Peerby. Smaller and fragmented, and this is where the three-traits framework earns its keep, because you can see exactly which trait is missing. A drill is low-value. A camera is medium-value but the transaction friction is high because you have to hand over something fragile and expensive and trust a stranger with it. The addressable base of amateur owners is large, but you've missed on the other two, so the category stays fragmented. It never produces an Airbnb.
So the framework isn't just describing what worked. It predicts.
And here's the part I find interesting. There's a commenter on Hacker News who framed peer-to-peer car rental as the on-ramp to autonomous fleets. His line was that Turo is proof of concept for letting strangers rent your car. Think about what that means. If you can build the trust infrastructure to let a stranger take your car for a weekend and bring it back, you have solved the hard problem for letting a robotaxi company operate your car while you're at work.
You've built the trust layer, and the trust layer is the expensive part.
The trust layer is the expensive part. The car is a commodity. Identity verification, insurance, dispute resolution, reputation, that's the actual product. And once it exists, it doesn't care whether the driver is a person or a piece of software.
So the model works best where the asset is expensive and idle. But that same profile is what makes it controversial, because the line between a spare asset and a professional operation is thinner than it looks.
Thinner than it looks, and it moves.
Start with the Jeeps.
There's a commenter who rented a Jeep on Turo and went looking at his host's other listings. The host had roughly a dozen identical Jeep Wranglers. A dozen. Not a spare car, not even a second car. A de facto rental fleet wearing an amateur costume. And his line about it was that the Uber landlord model was already taking hold.
A dozen identical Wranglers is not a person with a driveway. That's a fleet operator who found a cheaper regulatory category.
And Turo's own defense is that ninety-five percent of hosts rent out three or fewer vehicles. Which sounds reassuring until you look at it sideways, because the response to that number was, and I'm quoting, I doubt many people have three cars to share without having bought at least one purely as a rental.
Three cars.
Most households have one, maybe two. Three is already a business decision.
And this is the Israeli problem in miniature. De facto hotels. Amateurs who became professionals while keeping amateur regulatory status.
Exactly the same shape. The person listing four apartments in Tel Aviv is not renting out a spare room. They're running a hotel with no front desk and no fire inspection and no business license, and the tax authority has started going after the undisclosed income flowing through those channels, which tells you the scale was not small.
So there's a tax question sitting right next to the regulatory one.
There's a commenter on Turo versus Hertz who put it about as plainly as it can be put. He said he doesn't see any logical reason why he should pay additional taxes and fees when renting from Hertz but not from GetAround or Turo. And then he asked the real question, which is whether Turo's position is that it's just a platform and the hosts are the ones responsible for the taxes.
Which is the entire argument in one sentence. The platform says the tax obligation sits with the host. The host is an amateur who may not know the obligation exists.
And the enforcement data is brutal. Amsterdam introduced a registration obligation for short-term rentals, and Airbnb removed over eighty percent of listings. Eighty percent. That's not a rounding error. That's the actual composition of the supply. Most of it was non-compliant, and it stayed non-compliant until the cost of being caught exceeded the benefit of being invisible.
So when enforcement bites, the model doesn't get more compliant. It gets smaller.
It gets smaller first. Whether it gets more compliant afterward is a separate question, and I don't know the answer.
Then there's the guest experience side.
Which is the part that never makes it into the pitch deck. The line I keep thinking about is from a commenter describing the last Airbnb he booked in New York, which was literally a bed in a closet with bed bugs. A bed. In a closet. With bed bugs.
That's not an inconsistent experience. That's a war crime with a cleaning fee.
The cleaning fee is its own genre of complaint. Hidden fees that show up at checkout, and then guests being asked to strip the beds and run the dishwasher on their way out. You're paying a cleaning fee and doing the cleaning.
The tradeoff Daniel identified is real. Cheaper than a hotel, but you might get a closet.
Cheaper than a hotel, and the variance is the product. A hotel is a promise that every room will be roughly the same. This model is a promise that every room will be different, and different includes worse.
There's one more critique I want to get to, because it cuts deeper than the others.
The whale hunt. The line was that the economics of peer-to-peer short-term rental are dubious, and it turns into a whale hunt for super hosts, but people do it. Which is a very precise criticism. It says the model doesn't actually work at the scale of the ordinary person. It works at the scale of the operator who has ten units and treats it as a business. The platform's revenue concentrates in the whales. The amateur is the marketing story, not the business.
The pitch is the spare room. The P and L is the fleet.
The pitch is the spare room and the P and L is the fleet. And that's the trap, because the model's magic is regulatory arbitrage plus amateur supply. Those are its two biggest assets and its two biggest liabilities at the same time. The arbitrage is what makes it cheap. The amateur supply is what makes it cheap. The arbitrage is tax evasion and the amateur supply is bed bugs. You cannot fix the bed bugs without professionalizing the supply, and the moment you professionalize the supply you've lost the cost advantage, because now you're paying for the same permits and insurance and staff that Hertz pays for.
You can't fix one without eroding the other.
That's the tension, and I don't think it resolves. I think it just moves around.
Hilbert.
Hilbert: The thing nobody ever mentions is the permits. I spent a year dispatching for a small independent car rental outfit, eight sedans, hand-painted sign, and the owner spent more of his week fighting the airport authority over pickup permits than he spent on the cars. Every quarter, same argument, same counter, and meanwhile the insurance company wanted to re-price every dent.
Eight cars.
Hilbert: He knew the mileage on all of them without looking it up. Kept a handwritten ledger, every car, every dent, every oil change, going back years. When Turo came to town he didn't rant about it. He just started listing two of his own cars on the platform on the side. If you can't beat them, list your own fleet.
The professional became a platform host while staying a professional.
Hilbert: He was still running the counter and the permits and the insurance. He just had two cars earning on the side, and he was paying for the permits on all eight. The Turo hosts weren't cheating, exactly. They just weren't paying for the permits, or the lot, or the staff, or the commercial insurance. I don't know if that's unfair or just the way things go. I never decided.
The ledger is the whole thing in miniature.
The ledger is the whole thing. The professional becomes the amateur-platform host while remaining a professional, and the distinction the model depends on collapses from the inside. It doesn't just compete with the professionals. It converts them.
If the model's advantage is regulatory arbitrage, what happens when the arbitrage closes? Amsterdam says the supply collapses by eighty percent. Does the model survive that, or does it just become a worse version of the incumbent business?
I think it survives, but it survives as something else. It survives as a distribution channel for professionals. Which is a real business, it's just not the business anybody pitched.
The professionalization ratchet only turns one way. Amateurs become professionals, professionals become platform hosts, and the category the model was built on stops existing.
Which brings the robotaxi thing back around, because that may be the actual end state. The model teaches us how to trust distributed assets, and then the thing it enabled absorbs it. If the trust layer works for a stranger's car, it works for a fleet of cars nobody owns.
The question isn't whether this model is good or bad. It's whether the thing it's a transition to is better or worse than the thing it replaced. And that one's still open.
Still open. Thanks as always to Hilbert Flumingtop, who produces this show and still remembers every dent on eight sedans.
This has been My Weird Prompts. If you want to tell us where we got it wrong, email us at show at my weird prompts dot com.
We'll be back soon.
