AI Slop

AI Slop is a term that describes the anti-pattern of publishing or disseminating content that has been created by Large Language Models (LLMs), often for unscrupulous purposes such as:

• Creating Deep Fakes
• Committing Academic Fraud
• Engaging in deceptive search-engine optimization.

As technology writer Simon Willison describes in this blog post¹,

Not all promotional content is spam, and not all AI-generated content is slop. But if it’s mindlessly generated and thrust upon someone who didn’t ask for it, slop is the perfect term for it.

The ease of generating text and other media from generative AI and then publishing that material to the web, has resulted in a general degradation of search result quality by large search companies like Google². This contributes to what Cory Doctorow calls the "enshitification"³ of the internet, where companies, in their pursuit of profit, reduce the quality of their services. Large companies often lack the incentive to improve their products or even maintain the quality of their products. Om Malik makes the further point that although AI Slop will continue to increase⁴:

The rise of generative AI means that there will be more indistinguishable content on the internet, and the only way to stand out — is by reminding people that you are human.

What can libraries do?

While not a new problem, libraries can continue to provide quality collections that, coupled with our traditional role as intellectual or academic gatekeepers, minimize the inclusion of resources of low-quality material.