Cloud computing seems, to me, to be a level between personal computing and the thin client model.
A thin client (or dumb terminal) is a computer which has limited capabilities, instead leaving the heavy computing to remote servers. Thin clients use less power as a result of their lesser abilities compared to an average PC.
With cloud computing, you would use your PC’s web browser (or standalone web application) in order to use the capabilities provided by many (not necessarily powerful) computers working together for that purpose. I believe that this is functionally equivalent to the use of a thin client.
The similarity is apparent, but the difference is the more important aspect. With cloud computing, you don’t deny the capabilities of your own computer. There is no reason why your data must be stored remotely, on some unknown computer. Using a web application might require that the servers have a working copy of your data, but that does not imply that they need to store your data permanently. Supposing that one can rely on the organization providing the cloud service to keep the working copy secure and in tact, deleting it completely only when done with it, then the reliability issue is trivial. Of course there is still the need for a reliable connection.
Then again, the door is opened for true thin clients. What is a netbook? It is a limited capability notebook. To cut costs, the use of less capable OSes made sense for netbooks. After all, a netbook was meant to do but a few things: web browse, word process, email and very little else. Think about it; who needs a powerful PC when a cheap netbook can accomplish the same things by use of cloud services? Word processing can already by done online (e.g. Google Docs), and the only non-online aspect of email is choosing where you keep your messages stored (e.g. use of Outlook vs use of Hotmail).
Cloud computing really isn’t anything special if you already use email. The trust you place in a web service like Hotmail is considerably great, depending on the content of your email. It might not be that special, but it is an important part of our computing experience. To be able to access and modify documents where ever we are, whenever we want is a great convenience. A more professional entity (e.g. a private company) would most likely want its own servers handling its own special purpose computing, but the idea is the same, so long as some remote client is utilizing those special purpose servers.
In essence, cloud computing means that computing becomes more convenient. It also means that hardware requirements for the end-user aren’t a concern. Consider the idea of “cloud gaming” (OnLive). Imagine playing processor/graphics card intensive games without even owning a discrete video card. Of course, the main concerns relate to internet connection; how much bandwidth (speed), and how much bandwidth (caps) do you have? If the computing part of gaming is off-loaded, then that leaves the displaying part (relay of video/audio), and – depending on the quality – can be very bandwidth intensive.
I haven’t really said anything interesting, but it is satisfying to lay my thoughts out here. To sum this entry up (tl;dr), the concept of cloud computing isn’t anything special, but its implementation makes for interesting conversation.