Human-AI Interactions and Societal Pitfalls

TL;DR

This paper introduces a Bayesian model to analyze how user-AI interactions influence societal outcomes, highlighting risks of homogenization and bias propagation, and proposing solutions for personalized, efficient AI collaboration.

Contribution

It presents a novel Bayesian framework for understanding user-AI information sharing trade-offs and societal impacts, emphasizing the importance of flexible interaction to prevent homogenization and bias.

Findings

AI training on AI-generated content can cause homogenization.

Biases in AI outputs may propagate societal biases.

Reducing interaction friction enables personalized, productive AI use.

Abstract

When working with generative artificial intelligence (AI), users may see productivity gains, but the AI-generated content may not match their preferences exactly. To study this effect, we introduce a Bayesian framework in which heterogeneous users choose how much information to share with the AI, facing a trade-off between output fidelity and communication cost. We show that the interplay between these individual-level decisions and AI training may lead to societal challenges. Outputs may become more homogenized, especially when the AI is trained on AI-generated content, potentially triggering a homogenization death spiral. And any AI bias may propagate to become societal bias. A solution to the homogenization and bias issues is to reduce human-AI interaction frictions and enable users to flexibly share information, leading to personalized outputs without sacrificing productivity.