Position: agentic AI orchestration should be Bayes-consistent

arXiv cs.AI / 5/4/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper argues that Bayesian decision principles are especially well-suited for the control/orchestration layer of agentic AI systems that choose tools, experts, and resource allocations under uncertainty.
It claims that agentic orchestration can maintain and update beliefs about latent task-relevant variables using observed interactions, enabling more coherent action selection.
The author contends that making the LLM itself an explicitly Bayesian belief-updating engine is generally computationally intensive and difficult as a universal modeling target.
The work proposes practical properties, design patterns, and examples showing how calibrated beliefs and utility-aware policies can improve agentic AI orchestration in collaboration with humans and other agents.

Abstract

LLMs excel at predictive tasks and complex reasoning tasks, but many high-value deployments rely on decisions under uncertainty, for example, which tool to call, which expert to consult, or how many resources to invest. While the usefulness and feasibility of Bayesian approaches remain unclear for LLM inference, this position paper argues that the control layer of an agentic AI system (that orchestrates LLMs and tools) is a clear case where Bayesian principles should shine. Bayesian decision theory provides a framework for agentic systems that can help to maintain beliefs over task-relevant latent quantities, to update these beliefs from observed agentic and human-AI interactions, and to choose actions. Making LLMs themselves explicitly Bayesian belief-updating engines remains computationally intensive and conceptually nontrivial as a general modeling target. In contrast, this paper argues that coherent decision-making requires Bayesian principles at the orchestration level of the agentic system, not necessarily the LLM agent parameters. This paper articulates practical properties for Bayesian control that fit modern agentic AI systems and human-AI collaboration, and provides concrete examples and design patterns to illustrate how calibrated beliefs and utility-aware policies can improve agentic AI orchestration.