Do Algorithms Decide What You Believe?

Vocabulary

Algorithm – A set of rules or instructions that a computer follows to solve a problem or make decisions, such as choosing what content to show you online.

Example – Social media algorithms decide which posts appear at the top of your feed based on what you have clicked on before.

Polarization – The process of dividing people into two opposing groups with very different opinions, often with little middle ground.

Example – Political polarization has increased in many countries, making it harder for people with different views to have calm conversations.

Echo chamber – A situation where someone only hears opinions and information that agree with what they already believe.

Example – If you only follow people who share your political views on social media, you may be living in an echo chamber.

Engagement – In social media, the interactions people have with content, such as likes, shares, comments, and time spent viewing.

Example – Posts that make people angry tend to get higher engagement than posts that are calm and balanced.

Curate – To carefully select and organize content or information, often based on specific criteria or preferences.

Example – Streaming services curate your homepage based on what you have watched before, so two people rarely see the same recommendations.

Do Algorithms Decide What You Believe?

Every time you open a social media app, something invisible is making choices for you. An algorithm is deciding which posts, videos, and news stories appear on your screen. It is choosing what you see first, what gets buried, and what never reaches you at all. But does this digital gatekeeper actually shape what you think and believe?

The concern is not new. For years, researchers have worried about “filter bubbles,” the idea that algorithms trap people inside information bubbles where they only encounter views that match their own. If a person clicks on one political opinion, the algorithm feeds them more of the same, creating an echo chamber that gets narrower and narrower over time. According to this theory, algorithms are quietly pushing society apart.

There is real evidence behind these concerns. Research from the Wharton Business School found that content evoking anger gets shared significantly more than other types of content. A Harvard Business Review study confirmed that “high arousal emotions” like anger, fear, and joy drive the biggest social media responses. Since platforms make money from engagement, their algorithms naturally push emotional and divisive content to the top. A calm, balanced article about tax policy will almost never compete with an outraged headline about the same topic.

The effects are measurable. Social Media Today documented that after Facebook introduced algorithmic ranking of feeds in 2013, terms like “fake news,” “woke,” and various conspiracy theories gained significant traction online. The platforms did not create these ideas, but their algorithms gave them a megaphone. When divisive content gets more clicks, the algorithm treats it as “good” content and shows it to more people.

However, the story is more complicated than it first appears. A major literature review by the Reuters Institute at the University of Oxford found something surprising. Their analysis of multiple studies across several countries concluded that “algorithmic selection generally leads to slightly more diverse news use, the opposite of what the filter bubble hypothesis posits.” In other words, algorithms might actually expose people to more viewpoints, not fewer.

How is that possible? The researchers identified two key mechanisms. First, algorithms create what they call “automated serendipity.” When you search for something, the algorithm may return unexpected sources alongside expected ones. Second, people encounter news by accident while using platforms for other purposes, such as watching funny videos or checking on friends. This “incidental exposure” introduces them to topics and viewpoints they would never have sought out on their own.

The Oxford research also found that true echo chambers, where people are exposed only to one political perspective, are much rarer than assumed. In the United Kingdom, only about 2% of people occupy left-leaning echo chambers and about 5% occupy right-leaning ones. The vast majority of people maintain diverse media diets. The United States was the only country studied where more than 10% of people relied exclusively on partisan sources.

So if algorithms are not trapping everyone in bubbles, what is driving polarization? The researchers suggest the answer is more human than technological. People who end up in echo chambers typically choose to be there. They actively seek out sources that confirm their existing beliefs. Political leaders and media figures play a larger role in polarization than any algorithm does.

This does not mean algorithms are harmless. Even if they don’t create echo chambers for most people, they still amplify the loudest and most extreme voices. They still reward outrage over nuance. And they still make it harder for thoughtful, moderate content to reach large audiences. Some platforms are now testing features that let users manually control their feeds, but adoption has been low. Most people prefer the convenience of letting the algorithm choose for them.

The debate continues, and it matters. As more of our information comes through algorithmic filters, understanding how these systems work becomes an essential skill. The algorithm is not going to tell you what it is hiding from you. That awareness has to come from you.

Discussion

How much time do you spend on social media each day, and have you ever noticed that the content you see seems designed to keep you scrolling?

The article mentions that anger-driven content gets shared more than calm content. Why do you think people are more likely to share something that makes them upset?

Do you think you are in an echo chamber on any of your social media accounts? How would you even know if you were?

Some people argue that social media algorithms should be regulated by governments, while others believe people should be responsible for managing their own information diet. Which side do you lean toward, and why?

The Oxford researchers found that most people actually have diverse media diets. Does this surprise you? Do you think the “filter bubble” problem is exaggerated?

If social media platforms offered a button to turn off the algorithm and see content in simple chronological order, would you use it? Why or why not?

How do you personally decide whether a piece of information or news story you see online is trustworthy?

Sources

Reuters Institute (University of Oxford): Echo Chambers, Filter Bubbles, and Polarisation: A Literature Review

Social Media Today: 2026 Planning: Algorithmic Polarization

Harvard Business Review: The New Science of Viral Content

Wharton Business School: What Makes Online Content Viral?