Zeynep Tufekci is a self-described techno-sociologist.
While the job title is new to me, I admit after watching Zeynep’s TedTalk on machine intelligence, I think we’re going to need a lot more Zeyneps if we want to protect our social values.
Searching online, I discovered Zeynep has a small digital footprint. Perhaps it’s by choice but what matters is this former computer programmer, professor at the University of North Carolina and associate professor at Harvard University in Massachusetts is onto something important.
Her research warns of the perils of security and privacy vulnerabilities. The potential for abusing data and using it to influence politics and social media platforms is very real. In some instances, Zeynep is preaching to the converted in America who continue to worry about voter fraud and election interference by foreign countries like China and Russia.
Zeynep explains part of the problem might be in algorithms we don’t even understand. In today’s shared TedTalk, Zeynep warns of the potential for error in algorithms used by businesses for anything from employee recruiting to retention. In other situations where algorithms are used to determine fair prison sentencing, metrics that are supposed to cut through bias might actually do more harm than good because there are hidden biases within the algorithm.
Photo credit: Ted.com
Imagine the potential fallout in financial markets or undetected threats to national security if machine intelligence fails?
Okay, so maybe this is getting too heavy for Friday when most of us want to wind down from a stressful week. If you’re tired of big tech doom and gloom warnings and like your Hey Google machine, who can blame you for thinking it’s someone else’s problem?
Let’s consider this discussion from another angle. How do you feel about those viral videos that jump into your YouTube feed or trend on Instagram? It could be they aren’t telling you the truth. In one study by Verge that used data “scraped from 40,000 videos, the study found that creators, like Logan Paul, need to reach about 11 million views on a video before it hits the Trending section. Comparatively, segments from TV shows like The Tonight Show only need a couple hundred thousand views.” The Verge
Essentially, what you think is trending isn’t really because there is an unfair advantage. In fact, as much as 95% of content that is trending comes from traditional news outlets because there is an unfair advantage built into the algorithm that determines what’s hot. This means NBC only needs 400,000 hits to be trending versus another YouTuber who needs tens of millions of hits.
As we consider the convenience, wonder, and challenges of technology this month, give this techno-sociology some thought. What responsibilities do companies and our government have to protect us from unfair and manipulative persuasion?