As we grapple with spam in our emails and junk mail in our postboxes, a new digital nuisance has emerged on the internet, known as "slop." This term is gaining traction to describe the vast amounts of artificial intelligence (AI)-generated content that clutter the web, mimicking human-created material but often serving little to no real user need.
"Slop" refers to non-interactive AI-generated texts and images that primarily aim to populate the internet with faux content to garner advertising revenue and manipulate search engine rankings. Unlike a chatbot, which interacts and responds to user inquiries, slop is static, often misleading, and serves as digital filler material.
Simon Willison, a tech developer and early proponent of the term "slop," stresses the importance of identifying and naming this issue. "When 'spam' became a household term, it helped people understand and combat unsolicited email. By defining 'slop,' we can highlight the problematic nature of unchecked AI-generated content on the web," Willison notes.
One notable example of slop was an AI-generated travel article mistakenly listing a food bank as a top tourist destination in Ottawa. Errors like these can range from harmless but absurd to dangerously misleading. The latter is exemplified by incidents involving AI-authored books on mushroom foraging sold on Amazon, which contained incorrect and potentially hazardous information.
The issue extends to visual content on platforms like Facebook, where AI-generated images often go viral for their absurdity. These include bizarre depictions such as a figure of Jesus Christ with prawns for limbs or elderly women claiming unrealistically high ages. These images, while sometimes humorous, contribute to the spread of misleading information and clutter the digital environment.
The term "zombie internet," coined by Jason Koebler of 404 Media, aptly describes this phenomenon where bots, inactive user accounts, and real users mingle, leading to a disjointed and often nonsensical online experience. This environment is problematic not only for users but also for advertisers who fund much of the internet's free content. Misidentifications of legitimate ads as AI-generated slop by users are increasing, notes Farhad Divecha, managing director of digital marketing agency AccuraCast. This trend could undermine the trust and effectiveness of digital advertising.
The challenge of addressing the spread of slop is daunting. Google's recent rollout of "AI Overviews" in search results, which provides AI-generated summaries, illustrates the complexity of the issue. While intended to enhance user experience by summarizing content, it also risks increasing the presence of slop by blurring the lines between authentic and AI-generated content.
Combating slop will require a concerted effort similar to the initiatives that curtailed email spam. However, this task is complicated by the economic incentives of major tech companies that benefit from the proliferation of AI-generated content.
As digital citizens, it is crucial to stay informed about the content we consume. Recognizing slop when we see it, demanding transparency from content providers, and supporting initiatives aimed at maintaining the authenticity of online information are steps we can all take. By doing so, we ensure that the internet remains a space for genuine human interaction and valuable information exchange, free from the clutter of misleading AI-generated content.