Co-Exist Newsroom
Content Module 3
Introduction to Machine Bias – How algorithms work to control and filter the information we see on our newsfeeds, and what we can do to take back control of our newsfeeds.
Total: 18 slides
Introduction to Machine Bias- Understanding and Mastering News Feed Algorithms
LEARNING OUTCOMES
After this lesson, you will be able to:
- Understand how algorithms are changing our news feeds
- How algorithms Works
- Adopt strategies to control them
Understanding Algorithms
Artificial intelligence and its algorithms are terms familiar to us all, yet their applications and functions may elude a clear understanding. We encounter them regularly, if not daily and far more frequently than we may realise. While algorithms indeed streamline numerous operations across various domains, it is crucial to acknowledge that these tools, though powerful, are not flawless and can, at times, inadvertently perpetuate discrimination.
Algorithms on social media
Your social feed algorithms are complex sets of computer rules and mathematical models designed to tier and prioritise posts, images, videos, and other types of content to display in a user’s News Feed.
Every click, every like, every share shapes our newsfeed – and by extension, our view of the world. Understanding this mechanism is the first step towards a more conscious and balanced use of social networks!
Under the Hood: How do algorithms work?
- Data collection: It all starts with collecting user data. Social media collects information about our online activities: what we like, what we click on, what we share, which accounts we interact with the most, and even how long we watch a certain type of content!
- Behavioural analysis: This data is then analysed to determine our preferences and interests. Algorithms look at patterns of behaviour to identify what captures our attention the most. They also consider the behaviour of our friends and connections…
- Modelling and prediction: From this analysis, social networks build user profiles that are used to predict what content will be most engaging for us. They use machine-learning models to constantly refine these predictions based on our ongoing behaviour.
- Filtering and Ranking: The algorithms then filter the available content, ranking it based on what they estimate to be our likelihood of interaction. This includes our friends’ posts, announcements, articles, videos, etc.
- News Feed Personalization: The result of this filtering process is a personalized News Feed, where each item is chosen to maximize the likelihood that we will engage it.
- Feedback and adaptation: Algorithms adapt in real-time based on our ongoing interactions. If we change our behaviours, for example by starting to interact with different types of content, the algorithm will adjust our news feed accordingly!
Social Impact: Filter Bubbles cases
This complex and dynamic set of algorithms influences not only the content we see but also the way we perceive the world.
A “filter bubble” is the personalised reality created by algorithms that the news feed offers us. Eli PARISER popularised this notion in his book “The Filter Bubble: How the New Personalised Web Is Changing What We Read and How We Think”.
According to PARISER, users often find themselves exposed to a restricted flow of information that reinforces their existing opinions and isolates them from conflicting opinions or different perspectives.
By filtering information and presenting us with a personalised reality, they have the power to reinforce existing beliefs and shape social opinions and behaviours!
As a result, users often find themselves exposed to a restricted flow of information that reinforces their existing opinions and isolates them from conflicting opinions or different perspectives. This can limit users’ understanding and knowledge on various topics, create echo chambers, and potentially contribute to the polarisation of opinions on important issues.
Case Study (1/3)
Real-life everyday examples showing how algorithms have influenced social media trends and discussions.
If you watch a video criticizing certain cultural or religious practices in a foreign country, the algorithm will recommend other videos that criticize the same country or other videos that portray similar cultural practices in a stereotypical way… You fall right into uniform thinking!
On Facebook, News Feed algorithms can amplify certain news. As a result, for example, particularly insignificant stories can go viral, while other, perhaps important stories become less visible. Example: In 2021, the video of David Allen dancing to the song “Stay” by Kid Laroi and Justin Bieber received 43.9 million likes and over 313 million views on TikTok.
Case Study (2/3)
Hashtags on Twitter. For example, the hashtag #MeToo has sparked a broad global discussion about sexual harassment and sexual assault, thanks in large part to Twitter’s algorithm that has favoured tweets using this hashtag because of their popularity and relevance.
On Instagram, algorithms tend to favour content from influencers with a large number of followers and high engagement, which can affect fashion trends and society as a whole and steer consumers towards certain products!
Case Study (3/3)
Algorithms can sometimes contribute to the rapid spread of disinformation, as content that elicits strong reactions is often more widely disseminated. For example, conspiracy theories and misinformation regarding COVID-19 have spread rapidly across multiple platforms,
There have been discussions about the misuse of algorithms in the dissemination of polarizing political content and in the potential influence on elections. For example, Russia’s Internet Research Agency (IRA) used targeted ads on social media to suppress black votes for Hillary Clinton to help Donald Trump in 2016.
Mastering Your News Feed
Strategies to regain control of your News Feed:
For example, you can choose to follow or mute certain accounts, or change how content is sorted.
with other types of content.
Don’t rely solely on your news feed.
your contact list can help narrow down.
Stepping away from social media can help reset your algorithms and give you a fresh perspective. Go on a digital diet.
Apps like Feedly or Flipboard allow you to create your own news feed by selecting specific sources and topics. This gives you more control over the content you see, regardless of social media algorithms.
Broaden Your Horizons & Take Back the Power Over the Machine
➜ By adopting these reflexes, you can have better control over the content that appears in your News Feed, making it more relevant and interesting to you.
➜ However, directly controlling social media algorithms is not possible for users, as these algorithms are owned and managed by the social media companies themselves. Never forget that behind a machine, there is always a human spirit (or even a very powerful economic empire) and that neutrality does not exist.
➜ While these tools and strategies do not allow you to directly control algorithms, they do offer to manage what you see on social media, to have a more personalized and controlled online experience.
RESOURCES TO GO FURTHER
Key figures/Article: the 30 most powerful brands on social media in 2023 https://www.fastcompany.com/90849097/most-innovative-companies-social-media-2023
Tools: here are browser extensions, such as F.B. Purity for Facebook, which allow users to personalize their social media experience. These extensions may offer options to filter certain types of content, organize news feeds differently, or block ads.
Books and studies
– Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy by Cathy O’Neil, Ed. Crown September 2016 (This book examines how algorithms, under the guise of objectivity, can reinforce inequality and bias).
– Algorithms of Oppression: How Search Engines Reinforce Racism by Safiya Umoja Noble, Ed. NYU Press 2018 (Noble explores how search algorithms, especially those used by large companies like Google, can perpetuate racist and sexist stereotypes).
– The Filter Bubble: How the New Personalized Web Is Changing What We Read and How We Think by Eli Pariser, Ed. New York, Penguin Press, 2011 (Pariser discusses how personalization algorithms on the internet create “filter bubbles,” limiting our exposure to diverse information and perspectives).
– Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor by Virginia Eubanks. Ed. St. Martin’s Press, New York, 2018 (This book focuses on how algorithms are used to manage social services, often to the detriment of the most vulnerable populations).
– Data and Goliath: The Hidden Battles to Collect Your Data and Control Your World by Bruce Schneier. Ed. W.W. Norton & Company, 2015 (Schneier explores the vast amounts of data collected about us online and how it is used, often in an opaque way).
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Education and Culture Executive Agency (EACEA). Neither the European Union nor EACEA can be held responsible for them. Project Number: 2022-2-IE01-KA220-YOU-