What Will Be The Shape Of The Cloud?
The term cloud computing was already everywhere for a while, but with Google announcing a cloud based printing service it seems to becoming more of a reality than ever. One idea behind cloud computing is that many applications can be offered through a web interface rather than on local computers. Clearly, the most famous example is webmail, but there are also applications that are less web-reliant, like photo editing, that are ‘moving into the cloud’. In this post I will highlight some of the changes in user experience that may emerge because of cloud services and some of the economic changes a shift to cloud based services may bring about. Finally I will discuss a variant of cloud computing that I call radical cloud computing, which might be one of my worst ideas so far.
What is ‘cloud computing’ anyway? Although there were ancestors, the term cloud computing was coined in August 2006 by Eric Schmidt of Google. He used the term to describe the architecture Google is using for its search engine. Google’s data storage is distributed across many computers which can be seen as ‘a cloud of computers’. Nowadays the term is widely adopted and has gotten a broader meaning. Often the term is used as a fancy synonym for ‘the internet’ or for the technical infrastructure behind it. However, for this post the idea of Software-as-a-Service is the most important connotation. In short, rather than running software on my own computer (software-as-a-product) and accessing the internet occasionally, I can access the functionality of applications like my word processor trough the internet without installing word processor software on my local computer (software- as- a- service). The actual word processor software runs on a group of servers (the cloud) somewhere in an Amazon or Google datacenter. My computer becomes an interface to that remote word processor and stops being a host for my local word processor. The bulk of the ‘computing’ is no longer done by my ‘computer’ it is done by ‘the cloud’. In this Ted talk Kevin Kelly explains the idea very well. Those who are still skeptical of the possible performance of browser based software should watch the Google Wave demo.
Cloud computing has a couple of advantages. For example, for data storage the internet turns out to be a relatively safe place. While companies that offer data storage in some way (such as social bookmarking tools such as delicious, webmail services like Gmail and generic online storage services like Dropbox) can go bankrupt and although online data does get lost for other reasons, almost everyone is subject to a computer crash or theft at some point. I would argue that on the average online data is ‘safer’.
A second advantage of online data storage is that online data can be accessed from multiple devices in a way that is most convenient for the context of use. At one point I may want to look at my twitter stream in a side bar, allowing me to survey it but not to interact with it intensively, while at another point, maybe while I am on the move, I do want to interact with it. Similarly, I should be able to find my bookmarks made in the mobile device when working at my home computer again and so on. Although there is room for improvement on the integration of online data in the user experience of the different devices (and their operating systems) this seems to be a valuable development and creating this more extensive holistic web experience will be one of the most prominent UX challenges for upcoming years. Googles cloud print has many advantages for mobile web users. My mobile device may be a remote control allowing me to print my cloud based documents on the nearest cloud enabled printer taking care of the authenticitation and possibly providing an interface to set some preferences for the printing itself.
Cloud computing may also change the market for computer hardware. When more and more computing is done at the client side the client’s hardware can be less powerful and cheaper. Net books (small laptop computers) already tap into a market for people that mainly use their computer for accessing the internet. If more tasks can be done remotely and if internet connection is fast enough, more users may be able to escape the software/hardware loop. Nowadays many people buy new hardware because new software is making it slow. But if the computing and most of the storage is done remotely, this buying incentive is gone. Nowadays, consumer computer market is still very performance driven (focused on computing power and memory as most important buying considerations), but this may shift drastically. In a cloud driven market other considerations for computing hardware (such as design) will become more important and the consumer computer market may diversify as a result of this. More than in today’s performance based world classic ergonomics may become important, and users may own more devices for more different contexts of use.
A big question for cloud computing is how users will pay for the software services. Currently Google is the nr1 supplier of cloud software and Google offers this software free of charge, but I do not consider it likely that this model will scale to other suppliers. The reason is that free software is paid by advertising and there is a limit to that model. Take Google docs for example. It can be paid because trough my use of their software, Google is able to get to know me better, and is able to help their advertisers to target me specifically which in turns pays for my software. This model seems to be a lot less viable for photo-editing. How is my photo editing helping Google to get to know me, compared to my document writing? Moreover, there seems to be a general limit to the matchmaking ability of data-mining algorithms.
Another business model that seems to be suitable for software-as-a-service solutions is the freemium business model. In a freemium model a subset of the software functionality is offered for free to anyone, while the company makes money with a more extensive, paid function. This model is applied successfully for I-phone apps but there is a big challenge for companies to seduce their free users to pay for the extras. A different change, in how we pay for cloud services is that the software industry may shift from a product payment model to a subscription payment model. Maintaining a cloud of computers doing the computing costs money. Currently most software payment is ‘product’ oriented: once you have paid for software you are allowed to use it indefinitely. Because of the task of maintaining the infrastructure software companies might want to shift to a subscription model and ask you for a regular payment during use. When many competing software companies are trying to seduce users to pay more (for more features) and more often, I have little hope things will stay transparent to users. Cloud computing may give room to the emergence of an industry of intermediaries helping users with software subscription packages or sorting out their software expenses.
One option that I have given some thought, but for which I did not find to many sources, is the idea to use p2p software for cloud computing. I will call this variant radical cloud computing. An example is the SETI@Home project where non utilized computing power of end users (in this case of a screen saver) connected to the web is used to help NASA find extra terrestrial intelligence. Also Trojan horses, and p2p clients like music sharing program KaZaa, to use the computers of end users as a base for the cloud. While such an approach carries difficulties it could carry some of the UX advantages of cloud computing while end users are still paying for the computing trough the hardware. Possibly I could not find sources for it because it is a ridiculusly bad idea, possibly for other reasons. Fill me in if you can.
Filed under: (re)thinking media, discussion | Leave a Comment
Tags: Cloud Computing, Dropbox, Freemium, Gmail, Google Wave, Holistic Web, P2P Computing, Seti@Home, User Experience