The Algorithm That Knew You Better Than Your Wife
カートのアイテムが多すぎます
カートに追加できませんでした。
ウィッシュリストに追加できませんでした。
ほしい物リストの削除に失敗しました。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
The Psychological Power of Data
Episode Overview
This episode explores the transformative power of data in shaping our digital interactions and the implications of psychographic micro-targeting.
Key Points
The "Ocean" Model: The podcast introduces the Big Five personality traits: openness, conscientiousness, extroversion, agreeableness, and neuroticism, which were used to build a model of human personality traits.
The Power of Likes: Research demonstrates that an algorithm can predict personality traits with greater accuracy than a person, starting with just 70 Facebook likes.
Psychographic Micro-Targeting: The firm Cambridge Analytica harvested data from 87 million Facebook users to create "psychological warfare tools" for political campaigns.
Shift from Demographics to Psychographics: The podcast highlights a crucial shift from targetting people based on demographics (like age or location) to psychographics (like specific psychological vulnerabilities).
The Need for Awareness: The episode emphasizes that self-awareness is the only real defense against psychological engineering, urging listeners to be as analytical about the content they consume as the algorithms are about them.
Key Takeaways
Data Footprints: Our online interactions create deep and revealing data footprints that can be used to model our personalities, preferences, and vulnerabilities.
Privacy Concerns: The episode underscores the profound consequences of data-driven behavioral engineering on privacy and democracy, challenging listeners to think critically about how their data is used.
The Challenge: The core challenge is deciding how much of our inner lives we are willing to let remain un-scraped in an increasingly quantified world.
This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit donwoods.substack.com/subscribe