LLMs as Cognitive Viruses: Population Dynamics
emergentmind.com · 2,278 words · saved by 1 readers
Analyzes how LLM use can alter collective cognitive organization with a dynamic, population-level model.
The paper develops a mathematical model to explore how interaction with Large Language Models (LLMs) can alter collective cognitive organization, highlighting the potential for uses of LLMs to become captive cognitive bundles and lead to abandonment of unaided cognition in favor of persistent reliance on LLM-mediated processes. Large language model-mediated practices exhibit bistory-dependent bistability and hysteresis in transitioning forms of cognition On the parameter space of transmission, recovery, and reinforcement coefficients, varying the transition pressure ($\lambda$) that takes…
saved by
related reading
- Large Language Model: world models or surface statistics?thegradient.pub
- Towards a Typology of Strange LLM Chains-of-Thought1a3orn.com
- [2603.24676] When Is Collective Intelligence a Lottery? Multi-Agent Scaling Laws for Memetic Drift in LLMsarxiv.org
- Large Language Diffusion Modelsarxiv.org
- Generalization Dynamics of LM Pre-training — Jiaxin Wenjiaxin-wen.github.io
- Solipsistic Superintelligence is Unlikely to be Cooperativearxiv.org
- [2605.27288] It's Not Always Sycophancy: Measuring LLM Conformity as a Function of Epistemic Uncertaintyarxiv.org
- Cognitive Biases in Large Language Models — LessWronglesswrong.com
- The social AI hypothesissubstack.com
- LLMs Get Lost in Evolving User Intentarxiv.org
- 2301.06627arxiv.org
- BCIs and the ecosystem of modular mindsberen.io