Title: Just Ask the Group Chat: The Pitfalls of Algorithmic Personalization
Abstract: Algorithmic recommendation systems promise to curate our world, to reduce the problem of choice, and to show us what we really want. This promise is seductive: the overwhelming abundance of options available in modern life is exhausting, so why not let an algorithm do that work? Although offloading the work of deciding where to eat to a recommendation algorithm may seem trivial or unimportant, I argue that it is anything but. Recommendations play an important role in maintaining our interpersonal relationships. A friend’s recommendations holds you in an identity, a way of seeing you and relating to you that does justice to your complex values and preferences. Machine learning algorithms cannot do this, the personalization they offer is illusory. Worse still, when we consider the economic incentives of the recommendation systems, the moral risk is heightened. Some have argued that this threat is dire because it threatens our ability to be authentic, but I will argue that misplaces the problem. The real threat is assimilation. Instead of prompting LLMs, we should be texting our friends.
Audience
- Faculty/Staff
- Student
- Public
- Post Docs/Docs
- Graduate Students
Contact
Emily Berry
(847) 491-3656
Email
Interest
- Academic (general)