Professor Dan Schiller was invited to deliver several lectures at the Global Fellowship Progam at Peking University October/November, 2016. Below is his interview on “big data” with Wang Jianfeng from Chinese Social Sciences Today (CSST) after his lectures. This interview was originally published in Chinese Social Sciences Today (CSST) on January 19, 2017.
Wang Jianfeng: How do you evaluate the trend toward what is called Big Data?
Dan Schiller: Big Data refers not just to the scale or volume and diversity of data that are now being created but also to the need to make sense of these data through data science, through network analysis, and through other specialized disciplines that are trying to grapple with this challenge. One problem is that this often accords a new priority to an old emphasis, which is empiricist. Anything can be data, so let’s just look for patterns. We don’t care if they amount to anything meaningful. Let’s just see what seems to be related to what. There is, however, a deeper issue. An instrumental purpose is typically encoded in the accumulation and subsequent analysis of Big Data.
Thus, we have a problem because we need to know whose instrumental purpose it is and what goals it serves. If the goal of Big Data is to preserve the fishing grounds of the people who have been fishing in some part of the ocean, maybe that is ok because maybe it can be used beyond that to preserve the fish as well as the fishermen. If, however, the goal of collecting and analyzing Big Data is to extract profit from any area of human interaction, direct or mediated by machines, then I am not so sure. Actually, I am sure: It is wrong. Because then Big Data is organized around the instrumental purpose of profit maximization, which is not only exploitative but also often carries what economists call externalities. It may have all kinds of other effects beyond the immediate goal of profit‐ making, but nobody pays for these—except the rest of us. Disease, environmental despoliation and inequality are primary examples.
So how do we build the system of organizing Big Data if we need Big Data? And I am not sure we do need Big Data, because much of the data collection that is happening should not be occurring. We need a process of what in Europe and Canada they call data protection. I am not sure that is the right term, but I am sure that we need a policy or a structure of decision‐making for data collection, as well as for data analysis.
And this poses wholly new problems of political organization. Who should be making the policy, and on what grounds? Big Data thus poses profound questions. Because on the one hand, it gives new power to the units of big capital that are learning to exploit it for profit‐ making, while, on the other hand, it takes away power from everybody else, often without anyone knowing what, specifically, is happening. So we have a really big problem of balance—a power disparity—and, looking ahead, of a need for political creativity.
The issue is partly about education. People know now that when they go online, they are giving up their data. They know that, but they don’t realize that when they turn on their washing machine, or when they open their refrigerator, or when they take a shower, or when they go to bed, they are giving data. We need a forum for the discussion and decision‐making about which data ought to be generated and collected, by whom and for what reasons. Until we have that, we don’t have an answer to the problem of Big Data.
Wang Jianfeng: Now that information is regarded as a commodity, could overcapacity in the informtion industry take place? Do we have too much information?
Dan Schiller: We have too much of the wrong information and not enough of the right information. So there continues to be a desperate need for more information on the environment, more information on workers’ safety and occupational disease, more information on epidemiology and public health, more information on the social conditions of working people and the inequalities that prevail across society. We don’t have enough information on any of these, and where we do already possess the information, it isn’t widely circulated. There is indeed a huge information deficit.
However, in some contexts, there is also too much information. There is too much information being extracted from the everyday interactions that people have as they use technology that is embedded not just in smartphones and tablets and computers, but in the Internet of Things. So I think the question needs to be reframed in terms of what information we need and what information we get.