Zizek in the Cloud

A few days ago on Inside Higher Ed, Žižek published a piece on cloud computing, a topic that he has begun drawing attention to in his lectures recently. I think there are a number of problems here, which I will try to draw out.

Žižek describes the “Cloud era” where computers use network connections to access vast processing power that’s geographically remote, an architecture that’s also called client-server: the servers process code on behalf of the client and then return just the results back to it. This is in contrast to “pre-Cloud” computing, where the code resides on the client computer and the computations are made locally. There are two issues with his account that should be clarified: first, the features that he goes on to point to as problematic are not really new to the latest technology advances that distinguish cloud computing from earlier data centers. These advances make it cheaper and easier for companies to build client-server applications compared to data centers, but these differences don’t really impact the end user. The difference is whether the servers are owned by the company that runs the service, or they purchase computing time by the hour on servers owned by a different company. This is similar to how factories in the early industrial era owned their own power generators (windmills, water mills, and so on) but today they simply purchase it by the kilowatt-hour from specialized power companies.

Cloud computing can be viewed as an intensification of the trend Žižek describes: in the beginning, users ran code on machines they owned (“locally”); then, the internet came along and we started accessing websites and the code ran on the servers owned by internet companies; finally, those internet companies ran code on servers owned by a different company. The process of offloading computation to another entity was repeated, first from the user to the internet service, and then from the internet service to the cloud service.

But this nice linear progression is disrupted when we realize that it did not all begin with locally-running code. Bill Gates became wealthy for realizing his vision of a personal computer on every desk and every home. Prior to this, a common arrangement was very close to the “new” client-server computing that’s used on the internet today: users with relatively underpowered terminals entering data and submitting requests over a network to large, powerful computers that ran the code. This “totalitarian” model of computing is what Apple alluded to in their famous 1984 commercial - the personal computer was marketed as a liberation from central authority and control. Žižek rightly points out how cloud computing is marketed as a liberation from the need for control, we can trust the benevolent experts to take care of all of that for us, but still, he seems to repeat the earlier criticisms of central control made by Bill Gates and Steve Jobs to market their products.

But this marketing of cloud services is not addressed at ordinary end users, but at the owners of internet services. The difference for the end user is about the same as whether a business generates its own power or buys it from a power company. What’s interesting is that this marketing exists to persuade internet services to give up control of their data centers, but no similar marketing was directed at consumers to convince them to use internet services rather than running applications on their local machine.

The advertising for Intuit’s Mint service does not mention the benefits that come from not running local applications like instant bug fixes and feature enhancements. Žižek claims that the movement from local applications to cloud computing (or in data centers) makes things more abstract for the user, but this is not really true, at least not at the naive phenomenological level. There is not too much difference in my experience of running calculations in Excel on my computer compared with running them in Google Spreadsheets, the real process of computation is equally hidden from me in both cases. To the extent that this increased abstraction does enter my awareness, it makes my experience worse, not better. For example, I have to log in to Google and deal with password and security issues that I don’t have with Excel, and the dependency on a network connection may cause my changes to be lost if the network goes down at the wrong time.

So I think it’s not true that computers become easier and non-alienated for users as the system itself becomes more alienated, but this logic does work if we think about non-alienation as sociability. The ability for us to form and maintain connections, share emotions and feelings on the internet does depend on an increasingly alienated, abstract and pervasive background of networked computers. The problem of cloud computing is not that I, the end user, am further separated from the code that does the real work - this was never accessible to me to begin with. The problem is that the data that I produce is now physically under corporate control, and sometimes legally as well. The terms & conditions contract that you agree to when signing up for many internet services often includes signing over intellectual property rights to what you upload or create using the service.