The New Product Tesla Is Betting Its Entire Company On

There is a new product that companies like Tesla and Waymo are betting their entire companies on. You can already see pieces of it on the road in places like Austin, Miami, Orlando, and Vegas. Self-driving cars will not only change the car. They will change cities, roads, homes, businesses, and the whole stack around getting to work, meeting people, shopping, or having things come to you.

I am already living a slice of that in 2026 inside my own vehicle. That is the point of this upload.

The Cybertruck that works for me

I have a Tesla Cybertruck that just got a Grok update. Grok is the chatbot in the same category as ChatGPT, Gemini, or Claude. It is the xAI stack, publicly branded SpaceXAI inside SpaceX, and it is now hooked up to agents that can do computer work for me while I am in the car. Because the car runs on Full Self-Driving, I can talk to it instead of staring at a keyboard. It uploads videos. It updates my website. It clips footage. It stores a video idea and builds an outline.

That is one concrete example of what happens when the vehicle stops being a machine you operate and starts being a place you inhabit, hands free, for the length of the trip.

Elevators, then cars

The clean analogy on the tape is elevators. In the late 1800s and early 1900s, getting to the fifth floor meant stairs. Then elevators arrived with a human operator inside. Over time the operator left. Today in the United States you expect the box to run itself. You press a button. You are not hunting for a magical operator.

Transportation is heading the same way, only horizontal. Point A to point B instead of floor one to floor ten. The difference is time. You freak out if an elevator ride lasts more than a minute. In a future transport pod you will spend a lot longer sitting there. That time becomes yours: work, sleep, calls, a movie, looking out the window, or spinning up agents the way I already do.

Our cities, our ownership model, Uber, and a huge chunk of the supply chain are still priced around one equation: how much do you have to pay a human for their time to move the pod from A to B.

Cost: ownership, Uber, and driverless

On tape I put three cost numbers on the table. Treat them as my framing from this video, not a Tesla fare table or a government series.

If you own a vehicle in the United States, the average all-in cost is somewhere around 80 cents per mile across new and used cars. Uber-style rides land much closer to about $1.80 per mile once you count the driver's time, maintenance, and the rest, with Uber taking margin above that. Strip out the driver, keep electricity, maintenance, insurance, and the rest of running an autonomous system, and I put the long-run cost closer to about 20 cents per mile.

Two things move at once. Cost falls. People get their time back. Cheaper trips mean more trips: jobs you would not take, errands you would skip, longer drives you used to avoid because of fuel. Time back means the cabin becomes a second workplace or a rest place. Cities then have to adapt to people who arrive having already worked, slept, or met, instead of people who burned the commute just operating the car.

Why this hits ride-hail and ownership

This disrupts ride-sharing and ownership at the same time.

On the ride-hail side, once autonomous trips are cheaper, more consistent, and safer than a random human driver in a random Civic, why take the human version. Driving as a paid profession does not survive that transition if you look far enough out. I cite the United States number of about 40,000 deaths a year tied to human error on the road. Drunk, phone, showing off. That is my framing on tape for why a camera-covered, AI-driven fleet changes the safety baseline. It is not a Cybercab safety certificate.

On ownership, people start asking why a car should sit in the driveway doing nothing when a cheaper purpose-built self-driving unit can show up, drop you off, and go charge while you are at the stadium. No parking hunt. No steering wheel tax if the product never needed one.

The agent layer is the second product

The part that is easy to miss is that the car and the agents land together. The idea for this video showed up while FSD was already driving. I pressed the Grok control, talked it through, sent it to my agent team, and they researched, fact-checked, and outlined. I can pick that thread back up on my phone, in a browser, or again in the car. That is what "time back" means in practice for someone who already runs a media business on digital workers.

If you want the deeper robotaxi economics and second-car math, I already covered those in earlier Exclusives. This piece is the product bet itself: autonomous transport as the thing the company is built around, plus what you do inside the cabin once you are not the driver.

Watch the upload if you want the elevator story and the Cybertruck demo in my voice. The transcript and this write-up are the same argument without the sponsor break.