So, just how freaked out are consumers by governmental data scooping? Techies have known about many of these capabilities for a long time. But now, thanks to NSA whistleblowers up to and including Edward Snowden and his PRISM revelations, government data mining of personal information is the topic of cocktail party chit chat.
And it’s not just the government. Recent research shows that consumers distrust Facebook more than the NSA. Different perp, same problem.
So again, how worried is the public? Well, it’s not panic time, but concern is growing as evidenced by Pew Internet Research and other evidence, according to Kate Crawford, principal researcher with Microsoft Research.
What we still don’t know is the issue
It’s not panic time, but concern is growing by the day, said Kate Crawford, principal researcher for Microsoft Research, citing Pew Research polls. People are somewhat worried, but the biggest problem is we still don’t know what the real deal is, she said.
“The difficulty is most of what’s happening is extremely opaque,” Crawford told attendees at MIT’s EmTech 2013 conference on Wednesday morning. “First it was only about metadata and you wonder, should you care? But what’s interesting is that studies show that metadata is incredibly sensitive. We need to do an enormous catchup of where this data is going and how it’s being used.”
One big issue is that many programs and software applications rely on anonymized data, but it’s now clear that such data is not all that hard to crack. In her talk, Crawford noted that just four geospatial data points from anonymized phone records can identify 95 percent of people. So much for protecting your personal information.
A painful learning curve
Deb Roy, chief media scientist for Twitter and an associate professor at MIT, said the process of figuring out how to deal with these issues — and unintended consequences of big data usage — can be traumatic.
“The real challenge in the human-machine symbiosis is the human half has to figure out how to deal with these problems,” he said. “Long-term dependencies are hard to predict.” He cited retail as an example. Back when there were no big box stores or chains, shop owners had incredibly intimate knowledge of their customers’ preferences and patterns. When retail scaled out, that intimacy was lost and what’s happened since is that big retailers are using data to “claw back that customer intimacy.”
“The unfortunate part of that process is we learn by doing,” he said.
So what to do? Mitchell Higashi, chief economist of GE Healthcare, reinforced the need for informed consent. “Anonymized data has been used for years in healthcare. People have to decide — and the key word is decide — how much of their data should be used in healthcare.”
If they are educated about the potential upside of contributing their data for the greater good, and if they can be assured that the data is protected, all is fine. “If data is used to design a better society, people can be altruistic, but it has to be a personal decision,” Higashi said.
Ari Gesher, senior engineer with Palantir Technologies, said the upside of all this PRISM drama is it brings the topic to a broader audience. “What’s exciting is now we can have this policy discussion,” he noted.