I have been following up, as probably everyone in the planet, the US elections and one thing that was mentioned many times by many is how could all the predictions fail and poll results come out empty. Do not worry, this is not a political article, we all had our share of those in the past months. I am using this as an example of how what we think and what we read about comes back to impact how we see or, in this case, not see the reality beyond. This is of course the filter bubble of content we consume online and particularly on social media.
The Filter Bubble
If you are not aware of the filter bubble, I reference this article that I have written a while back on the topic. Back then it was all about personalization of the content and the social feed on channels like Facebook and even search results from Google to show relevant content based on patterns users have shown. What is happening now, I think, is more of an information overload. Where many people are reading the same content and sharing more than back then when the filter bubbles were set in place.
With information overload, many might be just picking through the piles of content and, as a result of their own prejudice, filtering out any opposing articles. Then comes social media, where people tend to express themselves in controversial topics like politics and religion wanting to prove their point and to show that many articles confirm what they have already concluded. This creates the social bubble especially that we, as humans, generally connect and friend like minded individuals and when we are all sharing that same content we are naturally being blocked out of what the opposition is doing.
Social Bubble and AI
I have recently covered #AI and how it would make sense for marketers to pursue it now rather than #VR or #AR. However, with the social bubble and how #BigData collected on people’s opinions and behaviors can support the machine build patterns and subsequently make well informed decisions, artificial intelligence is skewed. What if those sharing on one topic outweigh the others just because one party is more passionate or even more tech savvy and thus they use the Internet and social media more to express themselves and share. On the other hand, the opposing party consumes more and is content with reading what has been written vs sharing across to overshadow what might seem as the prevailing argument?
How are we to build more intelligent machines that can balance out the lack of data in this particular case? And if the machine is to do that, how are we going to get proper informed decisions from data manifested by the machine and not necessarily derived from the situation on the ground?
Can we burst out?
My analysis of the situation, as a human being, is that we want to remain in our bubble of thoughts and we like to be surrounded with affirmatives of what we know and believe in. So, we will continue, if we are socially enabled naturally, to consume and push out more content that further emphasizes our own wishes. We might even take it a step further and dig-out content from the other side of the fence just to see how many are actually listening, but we will stop there and naturally wouldn’t share across to remain true to what we think and believe in.
For the machine, well, we need to validate. So if the data says 9 billion people are talking about this topic. The machine should know:
- There are only 7,432,663,275 people in the world in 2016 accordingly to WorldoMeters.info
- Only 40% of the population have Internet connections according to Internetlivestats.com
With that, calculations to verify the sources and filter out repeated content might at least show that one side is more active and thus what we all see is basically efforts but not actual results as the majority has not yet spoken and what might be considered as a slam dunk win is just the insider view from our own bubble !