By Christo Lute, Director of Advanced Analytics,
There can be no doubt that new data technologies have the potential to wildly improve human lives. From medical decision making, to real-time translation of foreign language speakers, to music playlists curated to your tastes.
But there is a darker side to data science, uses for new technologies that do not improve our lives and violate some of our values and feelings of fairness. Each of these systems is possible with today’s current technologies and some have even begun to be implemented:
The Arab Spring was a revolutionary wave of protesting, riots, and civil wars in the Arab world that began in late December 2010 and continues to influence the region today. One of the more noteworthy activities of these revolutionary actions was the significance of social media on these revolutions. Social media enabled speedy coordination among groups and allowed governments to communicate directly with citizens.
These conversations are still available today, fossilized on Twitter and Facebook, available for anyone to read. With all that data available, some interesting questions arise, among them, “Can we use that data to predict revolution?” For a team of data scientists, this question is not as difficult as it may at first seem. Design a web scraper to capture all of the social media data from the months leading up to revolutions and throughout it, correlate that data to events, and analyze the text to see if keywords or phrases or any other variable is predictive of an upcoming revolution.
If successful, suddenly anything that someone posts on social media could be predictive of revolution. For rulers of totalitarian regimes, this is useful information. For those seeking to destabilize those regimes, this is useful information. For those who wish to profit off of conflict, this is useful information.
With the rise of data technology, we have entered a new era of performance management and employee engagement. We can now regularly and easily measure employee engagement and track engagement over time. These engagement metrics can be used to predict when an employee becomes disengaged at work, ultimately leading to churn or poor work quality. Ideally, these metrics would be used by employers to improve the quality of the work environment, build up team cohesion, and disrupt the spiral toward disengagement.
But better work environments are expensive. What if the environment was just good enough to keep an employee from churning, but not good enough to make them feel engaged? How costly is that? Instead, it might be cheaper to use employee engagement metrics, coupled with employee retention metrics to determine where the average minimum employee engagement mark rests. Instead of producing a better outcome, companies could use data technology to produce a work environment that is “just good enough” to stay and work, but not good enough to hit satisfaction marks. The worries over employee metrics at companies like Amazon are just beginning of the race to the bottom.
Send a large drone up over a city with a powerful camera pointed downward that can take one picture per second for the length of its flight. What can you do with that data? It turns out you can solve all sorts of problems, ranging from discovering who planted an IED in Iraq to tracking down kidnappers and killers in Ohio.
You can also track anyone’s location and whereabouts during that time. Anyone who steps foot outside will be seen by this “eye in the sky.” Need to know if your husband is having an affair? Hire a private investigator and use the drone data to track his whereabouts. If you want to get a leg up on the competition in the delivery business, you can discover their main routes and activities. Governments, corporations, and individuals can all utilize this information to know where people are, where they went, and at what time. Combine this with a tagging protocol for cars and individuals, store this information into a database, and you’ve captured the movements of an entire city. You can run predictive analytics on this data and sell the results to the highest bidder.
The intention of discussing these uses of data technology is not to instill fear over privacy violations or to encourage you to stay off the internet, but instead to show that like all new technologies, data technologies cut both ways. They are neutral tools to be picked up and manipulated to the ends of the person using them. Data technology isn’t inherently good, and it’s not inherently evil, but it is inherently powerful. Which is why it’s so important to invest in analytics professionals (yourself, your coworkers, your team) by integrating leadership training into their larger skill set. Alongside thinking flexibly and creatively, we must think ethically about how to shape these new technologies for both the good of our businesses and the public good.