Spam, the real scourge of the Internet, has found a very annoying replacement: slop. A practice that emerged with the rise of generative AI and which has invaded social networks. But what exactly is it?
Like many people, you are probably inundated with unsolicited messages every day, whether by email or SMS. This is called spam – also called junk, a contraction of junk and email. Some are purely advertising in nature: they essentially seek to attract you to a commercial offer and, incidentally, to track you by establishing your consumption profile, which they can subsequently resell. Nothing very serious, except the clutter of the messaging. But others are really dangerous: these are phishing attempts, the sole aim of which is to extract money from you with a more or less well-crafted scam. But after having fought against spam for years, Internet users are now having to face a new practice that is just as problematic: slop.
Slop: pay attention to the productions of generative AI
As explained by New York Times, this funny term originally refers to the mud which collects at the bottom of the tanks of oil ships. But, little by little, it took on a new meaning, first on forums popular with Internet enthusiasts like 4Chan and Hacker News, before landing on YouTube in the comments area. Today, slop refers to low-quality, intrusive and unwanted content generated by artificial intelligence. These may be fake images, fake videos made in AI, absurd responses given by chatbots or even mediocre online articles written in mass production by ChatGPT, without relevance or veracity. Its goal is to deceive the Internet user into believing that the content was created by a human, to generate advertising revenue and to direct the attention of search engines towards other sites.
For example, there are many AI-generated images on Facebook representing crazy photos, highlighted by the social network’s algorithms. Pretending to be real photos, they seek to attract an uninformed audience to external sites, generally to fake media filled with advertisements, dropshipping sales sites or corrupted pages intended to steal their personal data (see our article ). Likewise, it could be cheap books on Amazon written using ChatGPT or even Google which suggests you add non-toxic glue to your pizza so that the cheese sticks…
Slop: a clear term to combat this danger
British programmer Simon Willison was one of the first to promote the use of the term “slop” because, according to him, it is important to put a name to the phenomenon, in order to give the public the means to precisely define the problem , and thus warn Internet users of this danger. The word gained popularity last May, when Google incorporated its Gemini chatbot into the results generated by its search engine – it didn’t take long before the AI made a series of blunders, for example declaring that astronauts found cats on the Moon.
“Before the term ‘spam’ became widespread, it wasn’t necessarily clear to everyone that unwanted marketing messages were a bad way to behave. I hope ‘slop’ will have the same impact: making understand to people that generating and publishing unreviewed AI-generated content is bad behavior”concludes Simon Willison in the columns from the Guardian.