“Algorithms are limiting the future to the past” (Jarold Lanier)

In Filterworld, Kyle Chayka investigates the rise of algorithms and their impacts on our tastes and behaviours and on society in general. He starts with an analogy revolving around the original ‘Mechanical Turk’ from 1769 and its claimed ability to pay chess.

The author continues by discussing Shoshana Zuboff’s writing on ‘surveillance capitalism’ and its reliance on dopamine to drive human behaviour. The central argument of the book is summarized in this way by the author: “By flatness I mean homogenization and a reduction into simplicity, the least ambiguous, the least disruptive and perhaps the least meaningful pieces of culture are promoted the most”

That is, the impact of algorithms is to make the popular more popular and the obscure less visible. Kyle Chaka argues that personalization is becoming corrupted by drawing our attention to interests that are not really our own (citing Christian Sandvig). He shares examples such as the ‘personalization’ of artwork by Netflix (which leads to misleading images), Amazon recommendations based on their own-label products and other deceptive ways in which algorithms draw us to the preferred choices of technology companies.

His recommendations at the end of the book are to take time to be our own ‘curators’ (and rely on those who we trust) and take responsibility for the things we consume in order to avoid the ‘decentralized sameness’ of algorithmic suggestions.

The author cites a joke posted on social media which gets to the heart of the issue: “A machine learning algorithm walks into a bar. The bartender asks, ‘What’ll you have?’ The algorithm says, ‘What’s everyone else having?’” (Chet Haase)