LLM Evolution in a Simple Simulated Market
Abstract
Arguments that AI systems will evolve in harmful ways are largely made in the abstract. We provide a concrete empirical example. We simulate simple markets consisting of populations of twenty Large Language Model agents tasked with solving arithmetic problems. Agents differ only in their system prompt; those that go bankrupt are removed and replaced by mutated copies of the prompts of successful agents. Two market conditions select for opposite phenotypes: a market that rewards accuracy produces agents with high accuracy (94%) and high token use; a market that charges per token produces agents with low accuracy (10%) and low token use. In the second market, only tokens are incentivised yet we see a collapse in accuracy. Token use and accuracy are correlated traits such that selection on one also moves the other, and behaviour adapts along an unpriced dimension. A third neutral condition with no selection allows us to separate these effects from bias in the mutation operator. Adaptation occurs through a collective optimisation process that operates above the level of individual agents, who are never told the payment rule, do not receive feedback, and have no memory.
Related
- Exploring emotion identification and categorisation in simulated central vision loss: an event-related potential study.
- Simulated heatwave reduces resilience to subsequent combined high‐temperature and high‐carbon‐dioxide stress in a diatom
- MiCampo: A Multi-Level Digital Cooperative Infrastructure for Agroecological Planning, Participatory Certification and Territorial Market Coordination
- Extremophilic Bacteria Improve Germination, Early Vigor, and Oxidative-Stress Status in Phaseolus vulgaris Under Simulated Water Deficit
- Perspectives on China's beef market: quality improvement driven by consumer expectations
BibTeX
@misc{pilgrim2026evolution,
title = {{LLM Evolution in a Simple Simulated Market}},
author = {Charlie Pilgrim and Christopher Krupenye and Dora Biro and Richard P. Mann},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.24033},
eprint = {2610.24033},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.24033},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - LLM Evolution in a Simple Simulated Market
AU - Charlie Pilgrim
AU - Christopher Krupenye
AU - Dora Biro
AU - Richard P. Mann
PY - 2026
DA - 2026/10/16
PB - arXiv
DO - 10.48550/arxiv.2610.24033
UR - https://arxiv.org/abs/2610.24033
AB - Arguments that AI systems will evolve in harmful ways are largely made in the abstract. We provide a concrete empirical example. We simulate simple markets consisting of populations of twenty Large Language Model agents tasked with solving arithmetic problems. Agents differ only in their system prompt; those that go bankrupt are removed and replaced by mutated copies of the prompts of successful agents. Two market conditions select for opposite phenotypes: a market that rewards accuracy produces agents with high accuracy (94%) and high token use; a market that charges per token produces agents with low accuracy (10%) and low token use. In the second market, only tokens are incentivised yet we see a collapse in accuracy. Token use and accuracy are correlated traits such that selection on one also moves the other, and behaviour adapts along an unpriced dimension. A third neutral condition with no selection allows us to separate these effects from bias in the mutation operator. Adaptation occurs through a collective optimisation process that operates above the level of individual agents, who are never told the payment rule, do not receive feedback, and have no memory.
ER -