I agree completely with the first part, and as for the second part they are working on a robot that can do just that at the University of Florida. It can also be controlled over the web. I'll toss you a link when I find it.

In 1993 I was telling my new boss that someday we would be able to create our own newscasts and watch them by selecting the topics we wanted to watch and in the order we wanted to watch them. That vision pretty much is the case today.

In the future, we'll see information more accessible and shareable across platforms and systems. Currently, there is still a hold on databases and the interfaces that connect them. For instance, it is hard for one system to share its information with another and back again to create a new information bank. Steps are underway to remove this obstacle, but for now, you have to hunt and fight to merge data and it slows down sharing of knowledge. Also, there is a lot of information being created everyday and not much is being done to organize it. Sure we have URL's and other taxonomy, but an idea you have might not reach the masses for quite some time until it reaches a certain level of importance. Eventually, all ideas will have equal weight and will be indexed for instance access.

From all of this, databases of information and individuals will be able to link together to reduce the amount of effort it takes to share this knowledge and build communities around that data. You will see communities devoted to information gathering and sharing. Crossovers between communities will further merge this data and allow it to be used and re-purposed for many useful applications. Applications will spring up immediately based upon the needs of the user. Sort of like a RNA double helix, people and information will be able to move up and down the chain depending on their needs and their preferences.

Other than that, I anticipate being able to direct the Coke machine down the hall to not only allow me to purchase a Coke, but send a robot to deliver it.