unknown

AI personalisation systems from Netflix, Spotify, TikTok, and others have become extraordinarily sophisticated, but they raise serious concerns around consent, filter bubbles, and privacy. The consent frameworks embedded in GDPR and the EU AI Act are undermined by dark patterns and the practical impossibility of opting out without losing core service value. Filter bubbles demonstrably increase polarisation by ~15% and reduce content diversity, though research shows nuanced results. Transparency through explainable AI helps but faces fundamental limits. Privacy-preserving techniques like federated learning and differential privacy offer technical paths forward, but economic incentives in attention-based business models work against adoption. The piece argues that solving these problems requires regulation, competitive pressure, and a cultural shift toward optimising personalisation for user-defined goals rather than platform engagement metrics.

Personalisation Kills Choice

Exploring the intersections of artificial intelligence, decentralised cognition, posthuman ethics, society and culture.