February 20, 2024

Computer Science Pioneer Jaron Lanier Discusses Big Data, Privacy at NYPL


Jaron Lanier at New York Public Library, NYPL Live October 10, 2013The original developers of network technology wanted to democratize access to information, but while networks have succeeded in improving access, the ways in which governments and corporations are now gathering and using personal data has been an unfortunate consequence, argued author and computer science pioneer Jaron Lanier during a LIVE from the New York Public Library (NYPL) event on October 10. In a wide-ranging interview with NYPL’s Director of Public Programs Paul Holdengräber, Lanier also discussed the role that libraries play as people find it increasingly difficult to keep information about themselves private.

“Throughout history, whenever there’s some creep who wants to consolidate power, that [has meant] controlling the flow of information—controlling who can say what, know what,” Lanier said. “So we thought we’d make a distributed digital network, where information can flow and can’t be blocked, then that would inoculate humanity against this dysfunction.”

The Internet has indeed made it much more difficult for governments to restrict information. However, the free flow of data has also given a new sort of power to institutions that are best equipped to pool that information and analyze it—the corporations, banks, and government agencies with access to the biggest computers and sharpest algorithms. The trend is manifest in everything from targeted online advertising to recent revelations regarding the extent of the National Security Administration’s (NSA) domestic surveillance programs, Lanier explained.

It is problematic enough that corporations are now profiting from data that individuals surrender for free on social networking sites such as Facebook, for example, but the trend also has more insidious implications. Lanier related a conversation that he once had with an executive at a health insurance company who said, “In the old days, if I wanted to grow our business I would insure more people,” Lanier recounted. “But now that I have computers with data from all of these people, I can start to analyze them and predict [potential health outcomes] and only insure the people who won’t need it. So my incentive structure is now to insure as few people as possible.”

In this way, the current shutdown of the federal government over the Affordable Care Act is actually related to big data policy, Lanier said.

Algorithms can often find trends in big data that may be totally random, he added, yet these algorithms have a tangible impact on people’s lives, in some cases affecting their job prospects, how much money a bank is willing to lend them, or how much they pay for different types of insurance, to cite just a few examples.

“Where does this data come from that tells you who to insure and who not to insure? You might think it’s only [data regarding] medical conditions, but these correlative algorithms are kind of blind,” Lanier said. “If you look at how they work, there may be a weird thing where people who have dogs get a certain type of disease more often or people who buy a certain color of car. The algorithms don’t care if [a correlation] makes sense, and a lot of times these random relations end up being consequential…. The thing is, while you’re playing in this trivial way with ‘what am I going to ‘like’ [on Facebook] who am I not going to like, maybe I’ll mashup this and that, I’ll send somebody this link,’ all of this is being fed into algorithms that are actually of consequence to you. It’s truly strange.”

Libraries and Privacy

Lanier described libraries as the last remaining bastion of intellectual privacy in this new era of data collection.

“There’s a remarkable thing about the public library,” he said. “If you go to the public library to learn about something, and you do it with paper books, it’s the only instance in which you can learn in our society today…[where] you aren’t under observation.”

However, during a Q&A session in which an audience member asked whether libraries could potentially perform intermediary functions that could help patrons produce anonymous content or preserve more of their anonymity online, Lanier was doubtful.

“I think, really, it comes down to the economics and the politics…. Could librarians resist the pressure of the various spying classes and spying operations? To our horror, it turns out that our tech companies couldn’t, because it’s the law. That might be the case for librarians, too. I think it’s better to try to think about shifting the overall system than to have a particular exception case, because it just becomes a target.”

Speaking on the future of libraries more generally, Lanier cautioned libraries against knee jerk responses to broad trends, and said that to plan their future, individual libraries mostly need a clear understanding of the role they play in the present.

“The question of what should happen to libraries is a hard one…. Not all libraries are the same,” he said. “Some libraries have a particular culture of scholarship around them and that’s what they should be about—that culture. Some libraries have an urban culture around them or a community around them, and they should be about that…. I think the thing to do is to not think of the library as an abstract category, but to look at what [a specific library] is actually achieving and how it matters to people, and try to understand that. What in that should be preserved or should be a seed for what comes next?”

Matt Enis About Matt Enis

Matt Enis (menis@mediasourceinc.com; @matthewenis on Twitter) is Associate Editor, Technology for Library Journal.