4 ms·
I agree. This allows you to switch on autonomous driving when driving to unfamiliar places, long distance, and for cases where you would make more than $8.50/ho
by prennert 5y ago
I agree. This allows you to switch on autonomous driving when driving to unfamiliar places, long distance, and for cases where you would make more than $8.50/hour by working while the car drives. Or just because you had a drink and want to be driven home in your own car.
I am sure many people would opt to not use it most of the time, but having the option is great. As long as not using the option does not make the car a worse car than an non-automomuous one.
I feel most negative comments are coming from the dark-pattern aspect of it. How transparent is the charge going to be, are people going to be trapped in bills (yearly subscription outlasting their actual car-ownership) they wont be able to afford?
I also think that if few people use the feature regularly, but all cars need to have the hardware to support the autonomous driving, everybody is going to essentially overpay for features they are not using. Because one way or another VW needs to break even on the hardware at least.
Therefore, I think it could be an amazing business model if the promises of SaaS are fulfilled, a product that constantly improves and that you only pay for when you use it. In this case the incentives of customers and VW are aligned.
I think what people are afraid of, are dark patterns, subscription bundles and general in-transparency make the car a lot more expensive than people bought into.
- TheCondor 5y agoI agree that it seems like an interesting approach to bringing it to market. Imagine living in rural somewhere where they don’t have the same quality of data that they have for Silicon Valley, you don’t have to pay for the service that doesn’t work as well. Also, I expect there will be a service component whereby mapping a models are updated regularly so some fee for that makes sense. Then what if insurance was covered in that fee?