Meanderings, Siggy

Algorithms and Machine Learning: Are They Segregating Us?

Recently I watched President Obama on David Letterman’s new show My Next Guest Needs No Introduction on NETFLIX.  During the show, the question came up about Democracy. What fascinated me was President Obama’s innovative use of social media during his election campaigns; he essentially pioneered then mobilized the social media platform for political gain.  He later segued into algorithms, which is where my interest perked. 

Algorithm’s were essentially designed by organizations i.e. Google, Amazon, Facebook, Twitter, etc to capture information about the consumer which then creates a digital profile of the individual and the things they are interested in.  If I were an organization that sells merchandise, I would see incredible value in understanding my consumers profile then leveraging that information to reinforce their desire to purchase my product.  As organizations continue to create links to one another, the algorithms continue to refine their digital profile of our tastes in everything.

The account linking/merging and the EVERYTHING piece is quite scary.  For example, if I have a google account but not a Pinterest account.  CaptureI have the option to register a new account or link to another service which has my digital profile already.  If I choose to link the two, the algorithms are most likely working together and sharing more information about me therefore, multiple algorithms are refining their understanding of my interests.

This is where the design and structure of a non-threatening algorithm’s digital profile of a consumer gets weird. Because these algorithms are designed to understand our interests, the likelihood of items that are not interesting to the consumer becomes less likely to appear. What does this mean when this happens?  A consumer is less likely to be exposed to things outside of the algorithms digital profile of you, therefore, it reinforces your interests by exposing you to items that are likely to interest you first.

As we continue to use search engines, our digital profiles continue to be refined and the algorithms use that information to predict what you want to see.  These algorithms can essentially limit our exposure to sites or social media posts that don’t fall within the parameters of your digital profile.  An example would be a search of a news topic that leads an individual to a specific news organizations or videos with a political stance based on the algorithm’s understanding of you.

While it’s an educated guess, I believe that these algorithms have an unintended adverse consequence on our belief systems. Each time we like, share, comment, or hide posts and articles on social media, it reinforces the algorithms understanding of us, therefore, shows items of interest based on our digital profile. I believe this construct has also contributed to groupings of beliefs whether it is a political, religious, social, racial, etc. and it is not good for our society.  Rather than exposing ourselves to new thoughts and ideas, these algorithms contribute to self-reinforcement of our beliefs which then inadvertently causes us to slowly distance ourselves from thoughts and ideas which make us uncomfortable. 

We become comfortable in our bubbles and our exposure to other methods of thinking becomes nearly non-existent.  When people are exposed to the opposite view, meltdowns occur since our ability to consider other views has become irrelevant as a result of the algorithms digital profile.  I believe algorithms have trickled into our society so much so that it has contributed to the segregation of our society and as a result non-likeminded people can no longer value or consider that there are opinions and facts from both ends of the spectrum that can exist simultaneously.  

What do we do?