Data is the new oil. It is highly valuable, and it is everywhere, even if you’re not aware of it. For example, it’s there when you use social media. Sharing pictures on Facebook lets its facial recognition software peg you and your friends. Thanks to that software, now anywhere you visit that has installed cameras, your face can be identified and your actions recorded.
It’s there when you log into Twitter, posting one of the 230 million tweets per day, which up until last month were all archived by the Library of Congress and will be made public for research. These social media data can be used to predict your political affiliations, ethnicity, race, age, how close you are with your family and friends, your mental health, even when you are most likely to be grumpy or go to the gym. These data can also predict when you are apt to get sick and track how diseases are spreading.
In fact, tracking isn’t limited to what you decide to share or public spaces anymore. Lab experiments show Comcast and other cable companies may soon be able to record and monitor movements in your house. They may also be able to read your lips and identify your visitors simply by assessing how Wi-Fi waves bounce off bodies and other objects in houses. In one study, MIT researchers used routers and sensors to monitor breathing and heart rates with 99% accuracy. Routers could soon be used for seemingly good things, like monitoring infant breathing and whether an older adult is about to take a big tumble. However, it may also enable unwanted and unparalleled levels of surveillance.
Big data is there every time you pick up your smartphone, which can track your daily steps, where you go via geolocation, what time you wake up and go to bed, your punctuality, and even your overall health depending on which features you have enabled. Are you close with your mom; are you a sedentary couch potato; did you commit a murder (iPhone data was recently used in a German murder trial)? Smartphone-generated data can be used to label you—and not just you, your future and past generations too.
Smartphones are not the only “things” gathering data on you. Anything with an on and off switch can be connected to the internet and generate data. The new rule seems to be, if it can be, it will be, connected. Washing machines, coffee makers, medical appliances, cars, and even your luggage (yes, someone created a self-driving suitcase) can and are often generating data. “Smart” refrigerators can monitor your food levels and automatically create shopping lists and order food for you—while recording your alcohol consumption and whether you tend to be a healthy or junk food eater.
Even medicines can monitor behaviors. The first digital pill was just approved by the FDA last November to track whether patients take their medicines. It has a sensor that sends signals to a patient’s smartphone, and others, when it encounters stomach acid. Some call it a snitch pill, medication with a tattletale, and big brother in your belly. Others see it as a major breakthrough to help patients remember to take their medications and to save payers millions of dollars.
Big data is there when you go shopping. Credit card and retail data can show whether you pay for a gym, if you are pregnant, have children, and your credit-worthiness. Uber and Lyft transactional data reveal what time you usually go to and leave work and who you regularly visit (Uber data has been used to catch cheating spouses).
Amazon now sells a bedroom camera to see your fashion choices and offer advice. It is marketing a more fashionable you, but it probably also wants the video feed showing your body measurements—they’re “a newly prized currency,” according to the Washington Post. They help retailers create more customized and better fitting clothes. Amazon also just partnered with Berkshire Hathaway and JPMorgan Chase, the largest bank in the United States by assets, to create an independent health-care company for their employees–raising privacy concerns as Amazon already owns so much data about us, from drones, devices, the AI of Alexa, and our viewing, eating, and other purchasing habits on Amazon Prime.
Big data is arguably a new phenomenon; almost all the world’s data (90%) were produced within the last 2 years or so. It is a result of the fusion of physical, digital, and biological technologies that together constitute the fourth industrial revolution, according to the World Economic Forum. Unlike the last three revolutions, involving the discoveries of steam power, electrical energy, and computers—this revolution is advancing much faster than the ones before and it carries both promises and perils for humanity.
Some people may want to opt out of all this tracking, reduce their digital footprint and stay “off the grid.” However, it is worth noting that data generation and storage can be used for great things — things that make the world better, safer and fairer. For example, sharing electronic health records and social media data can help scientists better track and understand diseases, develop new cures and therapies, and understand the safety and efficacy profiles of medicines and vaccines.
While full of promise, big data is not without its pitfalls. Data are often not interoperable or easily integrated. You can use your credit card practically anywhere in the world, but you cannot easily port your electronic health record to the doctor or hospital across the street, for example.
Data quality can also be poor. It is dependent on the person entering it. My electronic health record at one point said I was male, and I was pregnant at the time. No doctors or nurses seemed to notice. The problem is worse on a global level. For example, causes of death can be coded differently by country and village. Take HIV patients: they often develop secondary infections, like TB. Do you record the cause of death as TB or HIV? There isn’t global consistency, and political pressure from patient groups can exert itself on death records. Often, each group wants to say they have the most deaths so they can fundraise more money.
Data can be biased. More than 80 percent of genomic data comes from Caucasians. Only 14 percent is from Asians and 3.5 percent is from African and Hispanic populations. Thus, when scientists use genomic data to develop drugs or lab tests, they may create biased products that work for only some demographics. Take type 2 diabetes blood tests; some do not work well for African Americans. One study estimates that 650,000 African Americans may have undiagnosed diabetes, because a common blood test doesn’t work for them. Using biased data in medicine can be a matter of life and death. Moreover, if genomic medicine benefits only “a privileged few,” the practice raises concerns about unequal access.
We need to think carefully and be transparent about the values embedded in our data, data analytics (algorithms), and data applications. Numbers are never neutral. Algorithms are always embedded with subjective normative values–sometimes purposely, sometimes not. To address this problem, we need ethicists who can audit databanks and algorithms to identify embedded norms, values and biases and help ensure they are addressed or at least transparently disclosed. Additionally, we need to determine how to let people opt out of certain types of data collection and uses—and not just at the beginning of a system, but also at any point in their lifetimes. There is a right to be forgotten, which hasn’t been adequately operationalized in today’s data sphere.
What do you think happens to all of these data collected about us? The short answer is the public doesn’t really know. A lot of it looks like what is in a medical record—i.e. height, weight, pregnancy status, age, mental health, pulse, blood pressure, and illness symptoms— yet, it isn’t protected by HIPPA, like your medical record information.
And it is being consolidated into the hands of fewer and fewer big players. Large companies are selling data that originated from you and you are not sharing in the wealth.
A possible solution is to create an app, managed by a nonprofit or public benefit corporation, through which you could download and manage all the data collected about you. For example, you could download your credit card statements with all your purchasing habits, your Uber rides showing transit patterns, medical records, electric bills, every digital record you have and would like to download–into one application. You would then have the power to license pieces or the collection of your data to users for a small fee for one year at a time. Uses and users could be monitored and audited leveraging blockchain capabilities. After the year is up, you can withdraw access.
You could be your own data landlord. We could democratize big data and empower people to better control and manage the wealth of information collected about us. Why should only the big companies like Amazon and Apple profit off the new oil? Let’s create an app so we can all manage our data wealth and maybe even become data barons—an app created by the people for the people.