Thursday, 15 October 2026. Two car companies, two unveilings, one evening. In a field outside Waco, Texas, Tesla finally showed the second-generation Roadster: a car first presented in November 2017 and promised for 2020. It drove up an inverted ramp, spun in mid-air and hopped over the rest of the Tesla line-up on compressed-air thrusters.
That same Thursday, on a pier in Queens with the Manhattan skyline behind it, Porsche presented the Mission S: a concept with round headlights, physical switches and no permanent central screen.
I love cars, I believe in electric mobility, and I look at products with the eyes of a CTO. That evening, all three parts of me were busy. Because what happened on 15 October is not really a story about cars. It is a story about the difference between announcing something and shipping it.
Nine years is a long time, in any industry
The new Roadster was born as a surprise, at the end of the Tesla Semi event in November 2017.
Deliveries were promised for 2020. Customers could reserve one with a $50,000 deposit, or pay $250,000 for one of the 1,000 Founders Series cars.
Then the delays began: 2020, 2021, 2023, 2024. At the end of 2025 Musk promised a reveal by year-end, then moved it to 1 April 2026, joking that April Fools' Day would give him some deniability if it slipped again. It slipped again.
The event was finally set for 1 October in Waco, then cancelled at the last minute because of heavy rain: the show, Tesla said, could only be held outdoors. New date: 15 October.
Nearly nine years from one prototype to the next. Nine years in which reservation holders left their money parked with Tesla, interest-free.
Let's be fair. In those years Tesla went through the Model 3 "production hell", turned the Model Y into one of the best-selling cars on the planet and shipped the Cybertruck. Pushing a low-volume sports car down the backlog was a rational decision.
What is less rational is keeping the promise alive on stage, year after year, while the backlog was saying something else. In software we have a word for a product that is announced, re-announced and never delivered. We call it vaporware.
Futuristic, or theatrical?
The car Tesla showed is very different from the 2017 prototype. The lines are sharper, closer to the Lotus Esprit from The Spy Who Loved Me (Musk owns one, and it was on display) than to a modern Tesla. The company cites the SR-71 and the F-35 as inspirations.
There are three trims. Performance is the "road ready" two-seater, with active aerodynamics and a suction system that glues the car to the ground. Thrust adds rear cold-gas thrusters for a claimed 0–60 mph in under one second. Rocket is a single-seater with ten thrusters around the body, fed by 600 litres of compressed air stored in carbon-fibre pressure vessels in every usable void of the car. All three promise about 480 km of range and more than 370 km/h.
Is it futuristic? In a science-fiction sense, absolutely. In a mobility sense, much less. The suction idea is fascinating, but it is not new: the Brabham BT46B "fan car" won a Formula 1 race with it in 1978, and Gordon Murray put a fan on the T.50.
The 2017 prototype promised around 1,000 km of range; the new one states half of that. The number that matters to people who actually drive electric went down, while the number that matters on YouTube went up.
As a CTO, I have seen this pattern many times. A demo is not a product. Every engineer knows the feature that works perfectly on stage and dies in staging. Thrusters demo beautifully. Then come the non-functional requirements: pressure vessels to certify, refills, maintenance, insurance, crash behaviour. Musk himself has said that safety is not the main goal of this car, and that people who put safety first should not buy it.
That sentence would never pass a product review in any company I have worked with.
Will they actually build it?
Musk said production will start in 2028. No price was announced. Reservations reopened the same evening: $50,000 to reserve a Performance, $250,000 for a Thrust. The Rocket is "invite only".
My honest answer is: probably yes, in some form. Tesla ships late, but it ships. The Cybertruck arrived years after its reveal, and so did the Semi. I would bet on a Performance Roadster reaching customers, in small numbers.
The Rocket is another story. Nobody on stage answered the questions that matter. Who certifies a road car with ten compressed-air thrusters? Under which rules in the United States? And how would it fit into European type-approval, which already struggles with far less exotic ideas?
I don't know. Tesla itself has partly answered: its website states that the Rocket is not approved for public road use. And the car that flew in Waco had no one on board. What we saw was a test vehicle, not a product.
Even some of the most loyal reservation holders reacted with irony. A well-known tech reviewer who had reserved the car years ago publicly asked whether the whole thing was a prank. When your earliest believers start joking, the problem is no longer engineering. It is trust.
What impact will it have on the car world?
Directly, very little. The Roadster is a halo car, and Musk admitted it will not move the needle much on revenue. Tesla's real business today is the Model Y, and its real bet is autonomy: it started offering limited Cybercab rides in Austin only weeks before this event.
There is a beautiful irony here. Musk called the Roadster "the best of the last of the human-driven cars". The company's most emotional product celebrates the human driver, while its strategy is built on removing him. The Roadster is less a vision of the future than a farewell letter to the past.
The indirect impact is about narrative, and I am not sure it is positive. The idea that electric cars are faster than anything else no longer needs proving: brutal acceleration has become a commodity. What the EV world still needs is credibility: honest range figures, real delivery dates, reliable charging. Every promise that slips by nine years makes the next promise, from anyone in the sector, a little harder to believe.
As someone who advocates for electric mobility, that is the part that bothers me most.
Porsche, same day: a meme?
The Mission S was unveiled at the Queens pier where, in 1953, the photos of the Porsche 550 Spyder were taken: the mid-engined car that defined the brand and built its success in America. The choice of place is not accidental. Neither is the message.
The Mission S is a concept, but not a one-off dream car. Porsche says it previews the design language of its next products and a new mid-engined supercar platform positioned permanently above the 911. The powertrain has not been officially disclosed; the specialised press expects some form of hybrid.
Was the date chosen to steal Tesla's thunder? Australian press noticed the match while both teasers were still running. But there is a detail I find more interesting. Tesla only moved to 15 October at the end of September, after cancelling for rain. An event on a historic pier in New York is not organised in two weeks. So either Porsche reacted with incredible speed, or it was Tesla that walked onto Porsche's stage.
I don't know which version is true. And it doesn't matter much: the juxtaposition writes itself. A car that flies against a car with an analog rev counter. It works as a meme even if nobody planned it.
And Porsche has done this before. In 2015 it showed the Mission E, a concept everyone read as an answer to the Model S, while insisting Tesla was not its benchmark. Four years later the Mission E went on sale as the Taycan.
A marketing move? Of course. But look at what each one sells
Every concept car is marketing. The interesting question is what is being sold.
Tesla sells spectacle. The future as a stunt, the car as content, the launch as an event to stream. It is a formidable machine for attention, and it has worked for years.
Porsche sells continuity. Its press release uses three words: authentic, stable, timeless.
Inside the Mission S there is no permanent central display and no passenger screen. The rev counter has a real needle. There are physical switches you push and pull. Your smartphone, on a retractable MagSafe holder, becomes the navigation screen only when you need it. The whole concept is part of a strategic reorientation Porsche calls "Sportwagenschmiede '35".
This is classic counter-positioning. When your competitor gets louder, you get quieter. On the day Tesla showcased nine years of delays, Porsche showcased seventy-three years of history.
Porsche is no saint, to be clear. It has been revising its electrification timeline and has its own problems to solve. But on 15 October it read the room better.
What this has to do with software
If you replace "car" with "product", 15 October looks a lot like a normal week in the tech industry.
Keynote-driven development
Tesla runs the Roadster like a startup on demo day: announce, collect pre-orders, build later. A reservation is non-dilutive funding, paid by your most enthusiastic users. I have written before that traction is worth more than a thousand pitch decks. A reservation is not traction. It is a pitch deck with a deposit attached.
The risk is the same one I see in many software companies: when the announcement becomes the product, the roadmap starts serving the stage instead of the user.
Concepts as public RFCs
Porsche works more like a team publishing a design system or an RFC. First it shows the language and the architecture, the platform, then it lets products adopt them over time. The Mission E in 2015 was, in practice, a public RFC. The Taycan in 2019 was version 1.0. That is semantic versioning applied to a brand, and it builds trust because the releases actually arrive.
Deleting features is the hardest decision
Removing the central screen is the choice I respect most. In software, adding a feature is easy and removing one is political. Delegating the display to the user's smartphone is also smart engineering: a car lives fifteen years, a phone gets replaced every three. Anyone who has maintained a ten-year-old embedded interface knows exactly what Porsche is avoiding. It has trade-offs, of course: you depend on someone else's platform. But it is a deliberate decision, not a default.
Counter-programming is not new
Every product manager knows the trick of publishing on the day a competitor holds its keynote. It costs little and steals the conversation. The difference is what you put next to the other headline. Porsche put a coherent story next to a spectacle.
Works on my stage
Finally, the oldest lesson of all. Thrusters are a demo-driven feature: perfect for a livestream, terrible for a backlog. The features that kill products are never the ones you show. They are the non-functional requirements: safety, certification, service, warranty. That is where the Roadster will be decided, not on a ramp in Texas.
Shipping is the feature
I don't want this to read as cheap Tesla bashing. As someone who loves electric cars, I owe Tesla a lot. The original Roadster proved an electric car could be desirable, and the Model S forced an entire industry to take batteries seriously.
But on 15 October the lesson was clear. One company showed a car that flies. The other showed a car that will probably exist. As a CTO, I know which of the two teams I would want to work with.
I have written about the strength of things done with calm, and this is a perfect example. Porsche chose a pier, a piece of history and a rev counter with a needle. Tesla chose rockets, a livestream and a 2028 that, given the track record, is more a wish than a date.
In software as in cars, the most futuristic feature is still the same one: a release date that holds.