The fact that such things exist at all is an interesting evolutionary step in maturation of these system's architectures. Here you've got software-intensive systems each with a considerable user base and each with an incredible amount of very tangible information (e.g. Google's maps and Amazon's book listings) as well as exploitable metainformation (e.g. the online behavior of its users). Making such information available programmatically from a variety of sources creates an opportunity for combining that information in unexpected ways, ways that transcend each individual source but in turn elevate the value of each such site. While there is no question of the cool factor of what can be done, it remains to be seen how the business model for such mashups evolves: can you build an economically-interesting system on interfaces you can't control and that will likely change? will the providers of such interfaces seek compensation for the value of the information being used? are there unforseen problems of public privacy? Time will tell.
Quote of the day: