One Is Not Enough: How People Use Multiple AI Models in Everyday Life

TL;DR

This paper explores how people manage multiple multimodal AI models in daily life, revealing strategies for coordination, trust calibration, and workflow organization through user studies.

Contribution

It provides novel insights into multi-MLLM orchestration, an area previously underexplored, informing future tool design for better user coordination.

Findings

Users create hierarchies among models that change with context.

Personalized switching patterns help manage effort, latency, and credibility.

Users develop strategies to coordinate multi-MLLM workflows effectively.

Abstract

People increasingly use multiple Multimodal Large Language Models (MLLMs) concurrently, selecting each based on its perceived strengths. This cross-platform practice creates coordination challenges: adapting prompts to different interfaces, calibrating trust against inconsistent behaviors, and navigating separate conversation histories. Prior HCI research focused on single-agent interactions, leaving multi-MLLM orchestration underexplored. Through a diary study and semi-structured interviews (N=10), we examine how individuals organize work across competing AI systems. Our findings reveal that users construct primary and secondary hierarchies among models that shift over usage context. They also develop personalized switching patterns triggered by task aggregation to adjust effort and latency, and output credibility. These insights inform future tool design opportunities, supporting users to coordinate multi-MLLM workflows.