Yup. I’m in Automotive, I work for a company that makes software for basically any car brand you can think of. I just recently left an internally developed project that aimed to create a personal assistant in the car. It was terribly ran and will go nowhere, but other departments in other companies will probably have more success, especially since the rise of chatgpt.

To add to your point though - the main idea on how to sell this assistant to car makers was the features, but the driving force behind developing the project was customer data. Collect a huge amount of info from customers, info that is shared with the car brand, but also accessible to us. To give some credit, discussions were never about using it for evil purposes - imagine a secretary knowing their boss’ schedule, our software would make suggestions like “you can’t make your 1 PM luch appointment with the client, would you like to reschedule it” and “I see you’re headed to Chicago and will arrive in 2 hours, should I make a reservation at that restaurant you like?” or some shit like that. But we all know that it’s not the engineers who decide what the company does with the data once access to that data exists. And knowing where a user eats, having access to their calendar, having access to their phone… This shit can get out of hand so easily when a budget-oriented executive type decides it’s time for this project to be even more profitable by selling the data to advertisers.