---
id: "quote-llm-entropy"
type: "quote"
source_title: "Don't Let AI Slop Muck Up Your Company's Processes"
source_url: "https://hbr.org/2026/06/dont-let-ai-slop-muck-up-your-companys-processes"
source_timestamps: ["¶17"]
tags: ["transformer-architecture", "information-degradation"]
related: ["concept-knowledge-entropy"]
speaker: "Matthias Holweg and Thomas H. Davenport"
speakers: ["Matthias Holweg", "Thomas H. Davenport"]
quote: "The greater the number of iterations of content through an LLM, the more it will depart from the original. Entropy can be managed, but not eradicated, as long as generative AI models use this underlying technology."
sources: ["execution"]
sourceVaultSlug: "hbr-seg-execution"
originDay: 8
articleStem: "hbr-sig-54-ai-slop-processes"
sourceUrl: "https://hbr.org/2026/06/dont-let-ai-slop-muck-up-your-companys-processes"
sourceTitle: "Don’t Let AI Slop Muck Up Your Company’s Processes"
---
# LLM Entropy

> “The greater the number of iterations of content through an LLM, the more it will depart from the original. Entropy can be managed, but not eradicated, as long as generative AI models use this underlying technology.”
> — [[entity-matthias-holweg]] and [[entity-thomas-h-davenport]]

The crisp statement of [[concept-knowledge-entropy]]: degradation is a fundamental function of the [[prereq-transformer-architecture|transformer architecture]], manageable but not eliminable without a step-change in model design. It underpins [[concept-generative-inbreeding]] and the 'game of telephone' in [[claim-sequential-ai-degrades-processes]].