Is Optimizing for Engagement Changing Us? | by Luke Thorburn | Understanding Recommenders | Medium

— Luke Thorburn, Jonathan Stray, Priyanjana Bengani

12 min read Oct 10, 2022 -- — Luke Thorburn, Jonathan Stray, Priyanjana Bengani Press enter or click to view image in full size Of the many potential harms of recommendation algorithms, perhaps the most unsettling is the idea that they might be changing us — our opinions, preferences, beliefs, and behaviors — in ways we don’t understand. Our concern here is not marketing or propaganda, but the strange ways that automated systems can affect us in their algorithmic pursuit of maximization. For example, many are familiar with the idea of a social media “rabbit hole”, in which the algorithm…

related reading

• Effect of recommending users and opinions on the network connectivity and idea generation processarxiv.org
• The Myth of The Algorithm: A system-level view of algorithmic amplification | Knight First Amendment Instituteknightcolumbia.org
• Discovering Language Model Behaviors with Model-Written Evaluations — LessWronglesswrong.com
• 2301.00993v1.pdfarxiv.org
• [2605.07912] Sycophantic AI makes human interaction feel more effortful and less satisfying over timearxiv.org
• How to Prepare for the Deluge of Generative AI on Social Media | Knight First Amendment Instituteknightcolumbia.org
• [2607.18966] Measuring Reward-Seeking via Contrastive Belief Updatesarxiv.org
• Models Don't "Get Reward" — LessWronglesswrong.com
• What is algorithmic amplification and why should we care? normaltech.ai
• Bridging Systems: Open problems for countering destructive divisiveness across ranking, recommenders, and governance | Knight First Amendment Instituteknightcolumbia.org
• [1904.06813] Personalized Re-ranking for Recommendationarxiv.org
• Aligning Recommender Systems as Cause Area — EA Forumforum.effectivealtruism.org