After "spam," artificial intelligence presents us with its own version of digital garbage: "slop." Soulless content, mass-produced for monetization, that threatens to flood the internet.
What is a slop?
The term slopSlop, which translates to "garbage" or "junk content," refers to texts, images, videos, and other materials automatically generated by generative artificial intelligence tools, without human intervention or quality control. Unlike AI-generated or intentional content, slop does not aim to inform or entertain. Its sole purpose is to appear as human-created content in order to attract traffic, inflate statistics, or generate advertising revenue.
The slop is not interactive and does not respond to a real need. It is pure automation, designed to exploit the weaknesses of algorithms and passively monetize with the minimum possible investment.
Why is this a problem?
Just as spam once became a constant nuisance in our email inboxes, slop is now starting to pollute our searches, social media, and content platforms. Its proliferation is driven by how cheap and quick it is to generate thanks to artificial intelligence.
The problem is that, although nobody wants to consume this type of content, the internet business model incentivizes it. Generative AI has facilitated the creation of texts, images, and videos on an industrial scale, regardless of their quality or usefulness.
Examples of slops in the real world:
-
A travel article recommended visiting a food bank as if it were a tourist attraction in Ottawa.
-
Books about fungi published on Amazon, written by AI, with advice that could be lethal.
Some of these examples may seem absurd, but many others are simply frustrating or a waste of time. Worse still, they force us to filter through increasingly junk content before reaching truly useful information. In the long run, this erodes user trust in all types of content, even legitimate content.
The (insufficient) response from Big Tech
The technology industry has reacted slowly to the threat of the slop. Some actions have begun to be implemented:
-
Meta (Facebook, Instagram) requires labeling AI-generated content.
-
TikTok automates the labeling of this content.
-
Google It is testing automated summaries in its search results, which, ironically, could worsen the problem by further blurring the line between human and automated.
Simon Willison, developer and one of the first to use the term slopHe argues that defining and labeling this phenomenon is crucial to raising awareness of its dangers. Just as the term “spam” helped combat junk mail, openly discussing slop could be the first step toward understanding and regulating it. However, he cautions that eradicating slop will be much more difficult than eliminating spam.
Zombie internet?
The proliferation of the slop adds to an even more disturbing phenomenon: what some call the Zombie Internet. Jason Koebler, a journalist at 404 Media, defines it as a “dysfunctional and meaningless” digital environment, “where there is no longer a true social connection.”
This new ecosystem is composed of bots, inactive accounts, real humans, and AI-generated content, all blending together in a chaotic landscape where it's no longer easy to distinguish the real from the fake. Empty internetA theory that has been circulating for years makes more sense than ever: we are navigating a sea of appearances, where authenticity is becoming increasingly difficult to find.
Whats Next?
The slop is here to stay, at least for now. Artificial intelligence isn't the problem in itself, but rather its indiscriminate and automated use for purely profit-driven purposes. The challenge lies in how we respond as a digital society: from users to developers, including platforms and regulators.
Naming the problem, as happened with spam, is the first step in tackling it. And today, that name is... slop.
Original article by Javier Lacort | June 16, 2024 published in Engadget.
