The Lock-in Hypothesis: Stagnation by Algorithm

TL;DR

This paper hypothesizes that the feedback loop between large language models and human users creates an echo chamber effect, leading to entrenched beliefs and reduced diversity, which is supported by empirical simulations and real-world data.

Contribution

It formalizes the lock-in hypothesis and empirically demonstrates the impact of the human-AI feedback loop on diversity using agent-based simulations and GPT usage data.

Findings

Sudden drops in diversity after new GPT releases

Empirical evidence of belief reinforcement in LLM interactions

Support for the echo chamber hypothesis in AI-human feedback loops

Abstract

The training and deployment of large language models (LLMs) create a feedback loop with human users: models learn human beliefs from data, reinforce these beliefs with generated content, reabsorb the reinforced beliefs, and feed them back to users again and again. This dynamic resembles an echo chamber. We hypothesize that this feedback loop entrenches the existing values and beliefs of users, leading to a loss of diversity and potentially the lock-in of false beliefs. We formalize this hypothesis and test it empirically with agent-based LLM simulations and real-world GPT usage data. Analysis reveals sudden but sustained drops in diversity after the release of new GPT iterations, consistent with the hypothesized human-AI feedback loop. Code and data available at https://thelockinhypothesis.com