Amazon's got an interesting new strategy they're pursuing: to become an infrastructure provider for the web-at-large. I'd mentioned this before - and its one of the places I had been trying to drag AOL for some time.
As this blurb in Tech Crunch mentions, its not without its substantial risks, but ultimately I think you want people rowing the boat with you, not against you. I think its why Yahoo still struggles, and Apple, despite flashes of brilliance, will always be (end up? :P) an underdog.
I think the biggest risk, in this area, is that your utilities get commoditized. Without some sustainable differentiated value, either in data services, user base, functionality, etc., its very tough over even the medium haul. Its great that Amazon is providing cheap storage for example, but more unique value propositions will be needed to create dependencies - otherwise their margins will get crushed.
Got to say this for Bezos - he rides the edge of the curve...