Artificial intelligence and Filter Bubbles
A squirrel dying in front of your house may be more relevant to your interests right now than people dying in Africa.
Mark Zukerberg
Cambridge dictionary defines a bubble as a situation in which you only experience things that you expect or find easy to deal with, for example opinions you agree with, or people who are similar to you. This definition is important as we have started to form a bubble of opinions around us, opinions that we agree with and AI has been instrumental in this.
Imagine if a newspaper was tailored to each and every reader, articles appearing only according to the bias of that individual making his bias stronger and less susceptible to other opinions, making that individual more and more radicalized with every article he/she reads, looks like something straight out of 1984 doesn’t it? Such a scenario would have been a far fetched hypothetical a few years ago but thanks to AI and prediction algorithms is rapidly becoming the reality we live in.
India has gone through an internet boom in the past few years thanks to ever increasing network penetration and dropping data prices, it is estimated that there will be 600 million internet users by 2021. With this ever increasing internet penetration more and more people have shifted their news sources from traditional print and news media to social media platforms on their smartphones, the problem with this is that news sources on social media do not have the same credibility as sources in print or news channels as they have authorities overseeing them. Combine this with the suggestion algorithms of these platforms, the way they suggest posts and you can see where I am heading with this. If you like a few pages that are subscribed to a certain ideology prediction algorithms ensure that you are bombarded with that kind of content and soon you may fall into the rabbit hole which is known as opinions (the irony of this being an opinion piece dawns on me, please bear with me here) and that will take you further from the reality of facts. What this does is that it leaves no room for critical thinking, your biases will become stronger and when you are confronted with something against your bias you may dismiss it as fake news or find it hard to believe.
This is less harmful when it is about things such as sport or entertainment, it is understandable that a person in India would want to know more about the Indian cricket team than the ice hockey team of Greenland. But filter bubbles can be specifically detrimental when it comes to the topics with much wider reach such as politics, economics and global issues such as climate change, religion, health or vaccination.
For a better understanding of this phenomenon i suggest that you go through the article on engineering filter bubbles with AI, they use the tensor flow embedding projector and have analysed over 48,000 facebook pages. The data is hypnotizing to look at and is extremely helpful in understanding social media bias.
I am not saying that prediction algorithms are useless and should be done away with, they are crucial to how we browse the internet and change it from a heaping mess of articles and data to something we can browse and enjoy, streaming apps such as netflix, hotstar, youtube would not be possible today without prediction algorithms and huge amount of user collected data is analysed every second to make these algorithms better and better, But when these algorithms start shaping our view of the society we have a problem. The term breaking the bubble is becoming ever so common and it stands for confronting your biases.
How to combat this issue? Really simple, First and foremost do not stick to one source. Every once in a while when you are browsing your favourite apps (facebook, tumbler, twitter , instagram etc) try to follow the pages you do not completely agree with ideologically they will give you a different perspective of things which is instrumental in broadening your mind. You can also visit websites like politico to get an analysis of the bias of your sources, If you are using a browser try to go incognito once in a while you will be surprised to know how many articles you miss.
