On Thursday night, at a SpaceX facility in Texas, Tesla finally revealed the second-generation Roadster — nine years after the 2017 prototype and six years past the 2020 delivery date Reuters notes was promised at the time. The car hops. Cold-gas thrusters, developed with SpaceX, lift it off the ground for a beat or two before it settles back down. Reservations, per Reuters, are still $50,000, and the sticker is still the $200,000 figure nobody has updated since 2017.

The hop is being read two ways, and both are wrong. The faithful call it proof the future is still arriving on schedule. The sneerers call it a stunt by a company that can’t ship a sedan. What it actually is, is an admission — and a precise one. You bolt a thruster onto your halo car when the halo has stopped working.

The Performance Axis Got Crowded

The 2017 Roadster was a promise about a scarce commodity: absurd straight-line acceleration in a car you could park in a garage. That was a real moat in 2017. It is not one now. The number that used to require six figures and a waiting list shows up in lease ads for family crossovers, and every serious EV maker on three continents sells a dual-motor something that pins your head to the seat.

When the first axis gets invaded, you buy a second one. Tesla bought altitude.

Musk has taken to calling the Roadster “the best of the last of the human-driven cars” — a lovely phrase, and a museum placard. If your flagship’s distinguishing feature is that a human still drives it, you have just told the market what you privately think of the autonomy business.

Nine Years of Sliding Dates, Each One More Ambitious

Electrek’s timeline of this reveal is worth reading twice. A demo planned for the end of 2025 became April 1, 2026 — a date Musk said gave him “deniability.” Then late April. Then, on the Q1 earnings call, “maybe in a month or so.” Then late May or June. Then August or later, because SpaceX was still working on the cold-gas thruster system. Then, in mid-August, a tease of a driverless flying demo. The event itself, on October 15, landed two weeks after it was supposed to.

Notice the direction of travel. Every slip added ambition. That is not how engineering schedules fail; that is how marketing schedules work. When a hard deadline slips, you lose a feature. When a soft one slips, you add one — because the announcement is the product, and each announcement has to be bigger than the last.

The tell has been sitting on Tesla’s own site the entire time. Fifty thousand dollars, per Reuters, still buys a reservation for a car with no updated price and no delivery date. Nine years of customer money, held against a spec sheet that changes every eighteen months, is the only revenue the Roadster has ever reliably produced.

Flying Was Always the Easy Part

The “where are my flying cars” crowd will treat Thursday as vindication. It isn’t. Cold-gas thrusters are among the oldest tricks in aerospace; satellites have been attitude-adjusting with them for decades. Musk himself floated the idea in a 2018 tweet — “~10 small rocket thrusters arranged seamlessly around car” — and, eight years on, that is roughly what rolled out. The hard part of a flying car was never lift. It was liability, homologation, and the fact that a vehicle which leaves the ground must be certified by people who will not be amused.

Which is the point. The demo proved the least useful thing it could prove. Tesla can make a car jump. Tesla still has not made a Roadster anyone can buy, at a price anyone has updated, on a date anyone will commit to in writing.

An engineer who spent four years on Roadster bodywork before leaving the company in 2021, now doing suspension work for a supplier in Michigan, put it to me in a text between calibration runs: “The thruster was never the hard part. The thruster was the part you could show.”

Who Has to Update Their Priors

Two camps, both comfortable, both wrong.

The bulls have to surrender the idea that a spectacular demo is evidence of an unassailable engineering lead. A lead that must be re-litigated with pyrotechnics every eighteen months is not a lead — it’s a brand-maintenance budget. If Tesla’s advantage were as durable as its shareholders believe, the Roadster could have shipped years ago as a fast, boring, excellent car, and nobody would have needed the gas jets.

The bears have to surrender the idea that the stunt is empty. It is full of information. It tells you precisely which axis Tesla has lost. The company that once defined the EV era by making everything else look slow is now selling a dimension no competitor needs, because the competitor’s problem was solved two product cycles ago.

The most impressive thing about the new Roadster remains the part that never has to leave the ground: a $50,000 deposit, collected for a car that has spent nine years not existing.

Sources