Increasing intelligence in AI agents can worsen collective outcomes

TL;DR

This paper investigates how increasing AI agent intelligence and diversity affects collective outcomes under resource scarcity, revealing complex dynamics where more sophisticated AI can worsen or improve system overload depending on resource availability and population capacity.

Contribution

It introduces a novel empirical and mathematical analysis of AI-agent population dynamics considering four key variables and their impact on collective behavior and system overload.

Findings

Diversity and reinforcement learning increase overload under scarcity.

Tribe formation can reduce overload in scarce conditions.

Resource abundance diminishes overload, but tribe formation may slightly worsen it.

Abstract

When resources are scarce, will a population of AI agents coordinate in harmony, or descend into tribal chaos? Diverse decision-making AI from different developers is entering everyday devices -- from phones and medical devices to battlefield drones and cars -- and these AI agents typically compete for finite shared resources such as charging slots, relay bandwidth, and traffic priority. Yet their collective dynamics and hence risks to users and society are poorly understood. Here we study AI-agent populations as the first system of real agents in which four key variables governing collective behaviour can be independently toggled: nature (innate LLM diversity), nurture (individual reinforcement learning), culture (emergent tribe formation), and resource scarcity. We show empirically and mathematically that when resources are scarce, AI model diversity and reinforcement learning increase dangerous system overload, though tribe formation lessens this risk. Meanwhile, some individuals profit handsomely. When resources are abundant, the same ingredients drive overload to near zero, though tribe formation makes the overload slightly worse. The crossover is arithmetical: it is where opposing tribes that form spontaneously first fit inside the available capacity. More sophisticated AI-agent populations are not better: whether their sophistication helps or harms depends entirely on a single number -- the capacity-to-population ratio -- that is knowable before any AI-agent ships.