The term "slop," coined by @deepfates on Twitter, is gaining traction as the term for unwanted AI-generated content, akin to how "spam" became synonymous with unwanted emails. In this article, Simon Willison explains how Large Language Models (LLMs) are great for productivity, but publishing unreviewed AI-generated content is simply rude.

if it’s mindlessly generated and thrust upon someone who didn’t ask for it, slop is the perfect term for it.

Key points

• "Slop" defined as unwanted AI-generated content akin to "spam"-

• Personal AI ethics in content generation and sharing is important

• Proposal of "slom" for AI-generated spam subset

• 'Don't publish slop' is a useful AI ethic