DSC Weekly Digest 01 March 2021

Some twenty years ago, as I was starting out my own career as both a technologist and a journalist, I had an opportunity to go to a conference on the intersection of the two in San Francisco.

The trip itself was eventful, at least in part because it happened, unbeknownst to planners of the conference itself, to coincide with a parade of nudists and nude bicyclists that happened to pass right in front of the hotel. Needless to say, the parade (if not the parade folk) ended up being well covered.

Besides the nudists, one of the topics discussed heavily at the conference dealt with the potential for automated filters to create different bubbles or world views. At the time, such filters were crude, but as AI has become considerably more sophisticated, so too have the sorting algorithms, to the extent that they ended up playing a big part in global politics in the last five years, creating growing polarization and cries of fake news on both sides.

In our newsletter for this week, contributing writer Stephanie Glen explores this phenomenon in depth in her article  Tweaking Algorithmic Filtering to Combat Fake News. In it, she makes the argument that the recommendation filters that determine what news a person sees (often invisibly) need to become more like how we determine our opinions in real life, via both actively seeking out novel content and taking in “background” content that interjects alternative potential viewpoints that at least challenge the dominant viewpoint.

However, this is an arms race, as both unscrupulous marketers and propagandists both have vested interests in making the “fake news” seem as real as possible, to convince the same AIs that they are in fact legitimate. It is this back-and-forth activity, comparable to the way that spam has forced the evolution of increasingly sophisticated spam filters, that is having the side effect of making AI more discriminating, more necessary, and potentially more dangerous. It is also forcing people working in this field, you, to acknowledge and even embrace the social and ethical aspects of what you’re working with.

This is why we run Data Science Central, and why we are expanding its focus to consider the width and breadth of digital transformation in our society. Data Science Central is your community. It is a chance to learn from other practitioners, and a chance to communicate what you know to the data science community overall. I encourage you to submit original articles and to make your name known to the people that are going to be hiring in the coming year. As always let us know what you think.

In media res,
Kurt Cagle
Community Editor,
Data Science Central