Both search engines and social media networks make up the primary gateways for most people on the web and mobile to get information.
Knowing that the internet is experiencing a massive increase in information due to the fact that more and more people are contributing to the flow of data, no online companies can deal with such huge amount if they rely on humans only.
They need lots of computing power to deal with the information, and algorithms play a huge role in those computation.
Using algorithms allow search engines and social media networks to put the heaviest burden of sorting information to computers.
This is an advantage, given that computers can deal with massive amount of information at any given time, 24/7, and way faster than even the most capable of humans working in a team.
The disadvantage however, algorithms are problematic because they are primarily designed to serve people what they want the most.
On the surface, this is nothing to worry about.
But going deeper, in the world where fake news is abundance, the more the people are drawn to sensational posts that are mostly associated with fake news.
As a result, the more likely the algorithms will recommend those people with more fake news.
The main reason this can happen, is because search engines and social media networks are generally platforms that are free to use.
The only thing the companies want in return for providing their services, is user data.
It's through obtaining the massive amount of user data, that include and not limited to: real name, email address, location, demography, age, activity, and lots more, that the companies can profile users to understand their interest.
And it's by knowing users' interest that search engines and social media networks can make money by selling ads, and tracking users to also sell their data through real-time bidding.
If the web is full of goodness and love, the algorithms can be the best thing the companies have. Sadly, the web is also full of hatred, bullies, racists, trolls, extremists, alt-right people and anything in between, including those who create fake news.
When people on the internet are visiting sources for misinformation because of their desire for sensational information or because of their entertainment value, algorithms which have their primary objective to learn from users' habits, can consider that the people in question do have interest in sensational content.
Both search engines and social media networks don't have real contents on their own. Most of their contents come elsewhere and from their users.
This is why search engines and social media platforms that are ad driven, will reward contents that have been clicked, or are enticing users to click.
For example, when users of search engines or social media click on a result or a post inside a feed, the algorithms will learn that the link they've clicked is relevant to their interest. This is called relevance feedback, and this helps the algorithms to give higher weight to that link for that query/user in the future.
And if enough people click on that link in a given time, the companies will see this as a strong relevance feedback, and the algorithms will make that website/post to rank way higher in their search results/feed.
And because users are more likely to click on links that are shown higher up, simply because they're more visible, this creates a positive feedback loop.
Another way of saying it: the higher a website is showing up, the more clicks it will get, and that in turn will make the algorithms keep the website high, or move it even higher.
Search engines and social media companies love doing this, because it drives more engagements and make them earn more money.
In other words, those companies value popular contents because they can boost their business' metrics.
The problem happens when those contents that are highly-valued by search engines and social media platforms are fake news.
This dance between algorithms and the human nature can further foster the spread of misinformation.
Starting this point, the bubble that is created will worsen things the bigger it gets.
The Confusion Between True And Relevant Information, And Content With Entertaining Value
The reason why this happens, the more a web page or post is engaging, it will be in the search engines' and the social media networks' best interest to give their users what they want to read, watch, or simply click.
Both search engines and social media networks collect vast amounts of data from websites and posts that crawl or have indexed, and judge their value using an array of metrics.
Because algorithms are also designed to evaluate how users react to headlines, titles, and snippets, algorithms that act as the recommendation system, create a list of items to show to users, while calculating the likelihood that they will click on the items.
Traditionally, the goal is to bring out the information that is only the most relevant to users that come from the most-relevant sources.
But in the modern days of the internet where any sort of information can be presented and shared, and in a place where spamdexing thrives, the notion of relevancy has gotten fuzzy.
This happens because people are using search engines and social media networks to search and find entertaining search results and posts, and not only to find truly relevant information.
Things can be harmless at first.
But if people get distracted from time to time and click on results that aren’t relevant to their query, they will get drawn to exciting images and sensational headlines.
Because the algorithms think that certain people love being entertained and not being showed truly relevant information, the algorithms will close the loop by burying truly relevant information, and just show those people the contents that entertain them.
The algorithms are designed that way.
It's not a design flaw. It's just that the algorithms couldn't correlate and can have difficulties in understanding the differences between a fact that people need at a given time, and a content with entertaining value they need to see at other times.
This "confusion" can make things like conspiracy theories and sensationalized news, to show up more often.
Before search engines and social media networks can stop relying on engagements as one of their priorities, their algorithms will inadvertently drive users down a complex rabbit hole they cannot easily escape.
Further reading: How Search Engines Process Your Queries Determines Your Satisfaction