Vision

For the most part, it did. Electricity, communication, mobility, a hot shower and a manufactured bed. Each of those started out as a privilege and turned into something many got to enjoy. That is the pattern of the last hundred years and it is the best reason I know to be optimistic about the next hundred. But housing never joined the list.

So the question I keep coming back to is this: whose lives have not been getting better? It is the best filter I know for deciding what is worth building.

At SpaceX I spent five years building Starlink terminals and led the development of Starlink Mini. Inventing the prototype is the easy part. Making millions of them is the hard part. It is also the useful part. It feels like housing is still in the prototype phase, doesn’t it?

I left in May 2025 to go find the thing that had been left out of the pattern. I read for a while and one thing led to another and my next mission became very clear.

A year after the Palisades and Eaton fires, roughly seven in ten survivors were still not back home. FEMA needed five months to stand up temporary housing after the Louisiana floods and more than a year in Maui. We can get you water and food in seventy-two hours, but a place to live takes months.

25.2 million adults under 35 live in a childhood bedroom, and among those aged 25 to 34, seven in ten have jobs.

On a single night in January 2025, HUD counted 745,652 people with nowhere to sleep.

And somewhere this morning a crew building critical infrastructure drove two hours to their job site. Not out of choice, but of necessity.

None of it is new. These problems have looked the same for forty years, and by most measures they are worse now than when I was born. Every other hard thing in this country eventually gets a chart that bends the right way. This one never has.

Uplift exists to bend it.

Uplift is building an entirely new type of housing. One that scales like a product, not one shepherded like a housing project. We are calling it the Uplift Home.

A home you can put anywhere, immediately. Solar on the roof, battery in the floor, water on board, hot shower, a full bed, a room that holds its temperature from twenty below to a hundred above. It arrives whole, powers itself, and is ready the second it lands.

It is designed around one principle: it can be made in the millions. One design, built on a line, measured in homes per hour rather than months per project. We are applying the same design and manufacturing discipline that produced Model Y and Starlink globally. Anything less is just a gesture.

It is also how the price comes down, because a large amount of what a home costs today never reaches the person actually living in it. The general contractor, the trades, the site work, the financing, and the waiting, all add up. So we delete what a home currently demands from the person who needs it. There is no electrician to schedule. No plumber to call. No foundation to pour. No construction crew at all. Every dependency we delete could be months somebody does not spend waiting.

And it has to be obvious. No manual, no training, nobody sent out to show you how it works. This is true housing on demand, ready the second it touches the ground.

In 1776 Adam Smith laid out a hierarchy of tradeable goods in The Wealth of Nations. At the top, precious metals, whose market “extends to the whole world.” At the bottom, housing. The materials of lodging, he wrote, “cannot always be transported to so great a distance as those of clothing, and do not so readily become an object of foreign commerce.” Timber in the Scottish Highlands was left to rot on the ground because it could not reach a market.

For 250 years he was right. You could ship gold across an ocean. You could not ship a home.

Three moves change that.

Sever every dependency. Power, water, waste, connectivity, leveling. All of it sits inside the unit, so nothing is waiting on a utility hookup or a trench or a permit queue.

Fit the network that already exists. A forklift moves it. So does a flatbed, a container ship, and every loading dock built in the last sixty years.

Improve in the field. An integrated ecosystem of connected Homes around the world.

Then lodging itself becomes the fungible unit. Not the timber and not the siding. The home. Housing can finally, as Smith said of metals, “bear the expense of a very long land, and of the most distant sea carriage.”

Roughly 117 billion people have ever existed. About seven percent are alive right now. Nearly all of the rest died without heated water, without a manufactured bed, without a room that held its own temperature. Pause on that. Those comforts are very new. They should also be inalienable. Right now they are moving further out of reach, not closer.

Costs for now common goods have fallen because they moved into a factory. Optimize the methods and materials, then build a production system that gets cheaper with every product you complete. None of this is a housing insight. It is a manufacturing one, and housing is close to the last big thing nobody has really done it to.

What changes with it is the product itself. Not price per square foot, not price per door, but cost per person per night. Housing goes from a fixed cost welded to a parcel of land to a variable one you can add, move, and switch off. No home has ever been able to do that at scale.

So what changes when housing becomes a product instead of a project?

A twenty-seven-year-old moves out. A crew sleeps three minutes from the job instead of ninety. A county puts families back on their own land in a week instead of six months. And rent finally moves, because for the first time new supply can show up without entitling land, stacking capital, and waiting two years for a certificate of occupancy.

And a kid looks up at their parent and asks, wait, people used to just sleep outside? That is what we are after. A generation that finds the old way hard to believe.

Five years ago the parts did not exist at a price or a quality worth building around. Now they do: satellite internet that works anywhere on earth, cells cheap enough to run a home for days, heat pumps that finally got small and efficient at the same time, actuators that level a house in minutes, solar that pays for itself in a season.

Off-grid was always a misnomer. Nothing is truly off the grid unless every unit is self-contained, and until now nothing was. Ours will work as well in Asheville as in San Francisco. As well in the Mojave as on Mars. Well, maybe not Mars yet. One day.

It does not have to feel like a compromise, either. Off-grid used to mean a dim, cold imitation of a home, and people accepted that because there was no alternative. Ours is finished to the standard of the nicest short-term rental you have stayed in, and it costs less than anything else you could live in.

What is left is integration. Every dependency we delete has to be genuinely replaced, not apologized for. The power has to hold on a January morning. The water has to be there in August.

And the country is asking for it out loud. On July 11 the 21st Century ROAD to Housing Act became law, which is the first major bipartisan housing package to become law since the 1990s and is calling for the exact thing we are building.

Housing companies die building houses. Car companies win building factories.

Manufactured housing has been tried for eighty years and it keeps failing the same way. A factory is enormous capital, and enormous capital needs enormous demand to pay for it. So every attempt reached for luxury, or novelty, or a niche small enough to feel safe.

We are doing the opposite. One home, one configuration, for the largest number of people possible. That is the only demand deep enough to justify a factory, and a factory is the only thing that makes the price fall. Cost per unit does not drop because somebody got clever. It drops because you built a line, ran it, and learned, which is the whole history of the automobile and why you probably have one parked out front right now.

We have paying customers now. The demand people are asking us about is in the thousands of units. This is a business case, not a cause.

Scale changes a country.

Trevor O’Leary and I founded Uplift. a16z and Valor backed us. We have paying customers and a factory in Austin.

Nobody walks out of our prototype without a use case we had not thought of. Fire camps. A clinic that needs somewhere for the nurse to sleep. A visiting doctor on tribal land. Mine sites, film sets, national parks, a block that burned last week, somebody’s backyard. We have stopped being surprised by it. Once a home is genuinely self-contained and can be set down anywhere in an afternoon, demand quits arriving from one market and starts arriving from all of them. If you are reading this and you can see a use we are missing, tell us.

The part nobody asks about yet is the software. Every one of these is connected, which means housing is about to get a control layer it has never had. Remote diagnostics. Improvements pushed over the air. Nobody has ever tried to build that, because until now there was nothing to build it on.

People look at a problem this size and assume the answer has to be one sweeping system that fixes every part of it at once. Let me be abundantly clear: no. We are a species of tools, and combined they make wonders. We cannot build every tool this problem demands. But we can build one. A useful one. An easy one. Then we earn the chance to build another.

We have the customers, the cash, the building, and the founding team.

In a world racing to build another application, let’s make something real. Something helpful. Something immediate.

Whose lives have not been getting better? I know the answer now. This is what I am doing about it.