Quantum leap CarPlay/Siri could be a big win but, even as an Apple fan in general, have no particular interest in an Apple Car absent things like self-driving that blow everyone else out of the water--which seems a pretty big ask.

Also, what would the margin be?

They could probably do full development from scratch for under $10 bil if they were frugal and patient, or more if they want to go fast, and farm first product out to a manufacturing house like magna. This is their MO already (they don't want to own a plant).

In the current era, it's probably cheaper to develop a car then to build out sufficient AI datacenters - which is also a negative ROI segment today for AI companies.