Tesla Cybercab: What Everyone Gets Wrong

Farzad records this one from inside a Cybercab with Dave Lee. No steering wheel. No pedals. The argument is not another cabin tour. It is what people are getting wrong about the soft Austin launch, about Tesla quitting passenger cars, and about whether Cybercab needs human controls to scale.

Dave opens by saying he will try to convince Farzad that the wheel-and-pedals volume plan will not happen. Farzad’s answer on tape is zero percent. He calls what is on the street now the setup for the real product in about two years.

Soft launch, then public rides

On tape they describe a quiet invite event: no livestream, Elon not in the room, a few of them got to ride. Then Cybercab showed up on the Robotaxi app for public hails shortly after. Dave remembers that public window as the next day. Farzad says a couple of days. Both say the mix surprised them: low-key event, then faster public access than they expected. Treat the sequence as their recollection of the September 2026 Austin window, not a Tesla press timeline.

If you fly into Austin you download the Robotaxi app and hail a Cybercab the way you would hail Uber. Point A to point B. No steering wheel. No pedals. That is already live inside the geofence, not citywide. Dave’s open question on this tape is not whether the first cars work. It is how many more Cybercabs are on the road in six months.

People think Tesla quit consumer cars

Farzad’s first correction is demand framing. A lot of commentary treats Robotaxi as proof Tesla stopped caring about cars people buy. He says that read is wrong. On tape he puts Tesla’s long-term car volume in the millions, and floats perhaps 10-plus million units a year as the kind of scale he still thinks Elon is aiming at. Most of that, in his framing, is still consumer cars. Robotaxi and Cybercab-style network cars are a smaller slice. He soft-guesses that slice in the two-to-three-million range, not the whole company.

Treat those volume numbers as Farzad’s illustration on this tape, not a Tesla guidance print.

Why he killed his own wheel-and-pedals idea

Earlier, Farzad had floated adding steering and pedals later so excess factory capacity and supply contracts could sell into a driveable variant while the geofence and safety case matured. Dave presses that line hard on this ride.

Farzad walks it back. The strongest regulatory path, he argues, is a vehicle with no driver controls at all. If there is no human driving, mirrors, pedals, and the rest of the driver-control rulebook stop being the right frame. He names NHTSA as the agency whose driver-control system rules Tesla is trying not to trigger. His forecast on tape is that a lot of that compliance question clears in roughly six to nine months, maybe a year. That is Farzad’s opinion, not a NHTSA docket outcome. Dave pushes back: liability for crashes, Uber disruption, legacy auto, injury lawyers, job fear, and, on this tape, that Uber is already partnering with the driverless union. Farzad does not pretend those fights vanish. He separates them from the narrow NHTSA driver-control stack on a no-controls car.

Capacity is not make-or-break

Dave asks whether Tesla can run the Cybercab line full bore in the next few years. Farzad’s answer is flexibility. Whether next year looks like 50,000, 100,000, or 500,000 Cybercabs, Tesla as a company is fine. He uses Cybertruck as the analogy on tape: they were hoping to do at least 250 thousand a year, nowhere close, and it still was not the end of Tesla. That 250 is Farzad’s hope-figure on this ride, restated as 250,000 when Dave comes back to it, not a current Tesla filing. Cybercab does not have to max the line to prove the product.

Stage one map, stage two cabin catalog

Farzad splits the story into two stages. Stage one is this Cybercab: expand the operating area over the next couple of years. His Austin example is a fence that eventually reaches Georgetown and Lakeway, not just the first pocket. Stage two is the next lineup of wheelless vehicles built on the next generation of the stack: four-seat, six-seat, seven-seat shapes that are cheaper and more useful than a two-seater, including for ownership rather than hail-only trips.

That is why he keeps saying the current two-seater is the setup, not the final catalog.

Dave’s $2,000 controls objection

Dave’s counter is practical. Say controls add about $2,000. Build them in, run unsupervised where allowed, then yank wheel, pedals, and mirrors later so the same metal falls under a different rule set. Farzad’s pushback is that the “just bolt them on and pull them off” mental model understates how the vehicle and the regulation are designed together. The no-controls car is the argument. Adding a human control suite to make volume easier, then removing it later, is the plan 99 percent of CEOs would buy. Dave calls that the underrated-Elon point: Elon is optimizing for a different shape of product than the practical hybrid Farzad floated months ago.

What this tape is not

This is not a new Domain-to-UT fare check. It is not a claim that Uber is already dead. It is not Tesla announcing a four-seater Cybercab. It is Farzad and Dave arguing, in a live Cybercab, about which story the soft launch actually supports: a temporary robotaxi demo, or stage one of a no-driver vehicle line that also leaves room for consumer volume.

This Exclusive is from the long-form at https://www.youtube.com/watch?v=gJb3A7RCu_g.