Solid piece—but I'd push back on software as the durable moat: AI is eating software. What took Tesla 4 generations to build is exactly what's now being automated and commoditized. Software is a staging post, not the destination.
Yes, safety-critical automotive software moves slower than general software—so the moat lasts a while longer. But "longer" isn't "forever," and a 4-generation lead on a commoditizing asset is a shrinking lead. Betting the industry on it is betting on the thing that's deflating.
The real moat is the one AI can't eat: the physical world. You can't generate a battery pack, a chassis, or a factory. AI needs a body. So the scarce asset becomes AI integrated with physical production at scale—which is precisely what Germany is good at.
That flips your conclusion: mechanical engineering isn't the drained moat, it's what carries you through wave 3. The risk isn't that Germany is behind on software. It's that Germany is behind on software and doesn't realize its physical base is the real asset.
The problem is that the customer does not have a choice in many aspects : the bureaucrats in Bruxelles chose for him. That camera that films the driver permanently in every new car and sends the footage to the carmaker. Or the automated braking systems. The automatically forced respect of any speed limit and of other driving rules is coming.
On the buttons instead of screens : that may be a European trend like the persistence of the manual gearbox.
Tesla is an American company manufacturing American cars. The Mustang never was a rival to Ferrari or Lamborghini in spite of the hype in films.
Consolidating 100-plus supplier-sourced ECUs into a handful of domains is less an engineering rewrite than a renegotiation of who owns each subsystem's release cycle, which helps explain why the top 8 are software-native rather than just electric. Legacy makers didn't just build cars around hardware, they built supplier contracts around hardware, and that's harder to unwind than a codebase.
Organization-first shows up in pure software companies with no supplier ecosystem to unwind. Amazon rebuilt a core cloud service with six engineers in 76 days against a budget of 30 to 40 engineers over a year, by writing the specification before the AI wrote any code and keeping a human gate on every change. The Gartner leaders sound like they did the organizational equivalent before touching the electronics.
Solid piece—but I'd push back on software as the durable moat: AI is eating software. What took Tesla 4 generations to build is exactly what's now being automated and commoditized. Software is a staging post, not the destination.
Yes, safety-critical automotive software moves slower than general software—so the moat lasts a while longer. But "longer" isn't "forever," and a 4-generation lead on a commoditizing asset is a shrinking lead. Betting the industry on it is betting on the thing that's deflating.
The real moat is the one AI can't eat: the physical world. You can't generate a battery pack, a chassis, or a factory. AI needs a body. So the scarce asset becomes AI integrated with physical production at scale—which is precisely what Germany is good at.
That flips your conclusion: mechanical engineering isn't the drained moat, it's what carries you through wave 3. The risk isn't that Germany is behind on software. It's that Germany is behind on software and doesn't realize its physical base is the real asset.
All this “technology” only equals one thing: control.
And not for the one who purchased the vehicle.
No one wants this in a car no matter how they hype it.
They are already backtracking on the screen tech by putting in manual buttons again due to customer complaints about the silly screen controls.
And Tesla’s are crap btw.
Terribly made and will not last.
The problem is that the customer does not have a choice in many aspects : the bureaucrats in Bruxelles chose for him. That camera that films the driver permanently in every new car and sends the footage to the carmaker. Or the automated braking systems. The automatically forced respect of any speed limit and of other driving rules is coming.
On the buttons instead of screens : that may be a European trend like the persistence of the manual gearbox.
Tesla is an American company manufacturing American cars. The Mustang never was a rival to Ferrari or Lamborghini in spite of the hype in films.
Consolidating 100-plus supplier-sourced ECUs into a handful of domains is less an engineering rewrite than a renegotiation of who owns each subsystem's release cycle, which helps explain why the top 8 are software-native rather than just electric. Legacy makers didn't just build cars around hardware, they built supplier contracts around hardware, and that's harder to unwind than a codebase.
Organization-first shows up in pure software companies with no supplier ecosystem to unwind. Amazon rebuilt a core cloud service with six engineers in 76 days against a budget of 30 to 40 engineers over a year, by writing the specification before the AI wrote any code and keeping a human gate on every change. The Gartner leaders sound like they did the organizational equivalent before touching the electronics.