While I don't know if Anil (and Co.) is right about web trends, since predicting the future is always...well, one should always be wary of those that say they know the future.
Upon saying that, I'd like to bring up what I think there is a future for. I think there is a potential market of electronically customizable cars. If there was a car company that made a car that had a programming platform that provided access (in forms of drivers or APIs) to non-critical components of the car, I think developers would be all over it. While it would be a niche market at first, it would be something that developers and hackers would find interest in playing around with. And in playing around with it, they will build tools for it. If they build something that the average consumer will like, then there will be more of a drive to buy the cars that run the software.
This is mostly the usual platform vs. developers dilemma, much like the chicken and egg. Why write a platform if there are no applications for it? And you can't write applications where there's no platform.
The parts that are required for this to take off are almost there. Wireless access to the internet is becoming more prevelant with Wifi and also 3G, and mapping software is becoming more readily available with google earth. When that happens, and a platform is built, perhaps the price of positioning systems will go down, and other applications be developed for them.