Fighting Slop With the Public Domain (and Why Slopfarms Perish Faster Than New Ones Appear)
We can combat the nonsense by producing more human-made works until the slop bubble implodes
In my humble assessment, slopfarms perish faster than people create them (or turn once-real sites into slopfarms). This trend may accelerate the moment all the major slop providers (so-called 'AI' companies) cease (dis)service on economic grounds.
It's not illegal to spread slop online (text, images, video, sound), but it is frowned upon and likely a copyright problem too. To us, Public Domain images (we spread more and more of them each day) can help 'drown out' synthetic nonsense or slop. Not all are equal and what counts most is visibility. A slopfarm nobody ever visits is as prominent as one that does not exist.
So try to produce and reproduce real works. It can help.
Thankfully we see many articles that speak of the long-term consequence or the overall cost of slop 'code' ('vibe' is a meaningful buzzwords) and it easily overwhelms some promotional nonsense from companies that try to sell slop tools to coders. There's growing pushback and scepticism. It's encouraging to see that. The propaganda cannot prevail if people speak out loud about the 'vibe' nightmare stories (technical debt and actual catastrophes).
We cannot emphasise strongly enough that one year ago and 2 years ago many sites we loved and cited had turned to LLM slop, with some slapping some slop images on top of garbage text (fodder). Some have "Walked Back" since then, others have "Shut Down". It was basically a failed experiment. Today in Google News, if we search for "linux", only one slopfarm can be seen: (WebProNews, formerly a real news site)
That's just SEO spam using slop (images and text).
We hope that by the end of the year all those remaining slopfarms will perish; my parents tell me that they're sick and tired of slop (in all its forms, they partly blame the Web at large). It's feasible; if enough people try to contain the spread or visibility of slopfarms and Serial Sloppers, they will eventually call it quits. Plagiarism by synthesis will always yield results vastly inferior to the original/s because the aggregation processes lack actual comprehension of what they ingest. Those aren't economically preferable either. They live in borrowed time and borrowed (from VC) money. Time is not on their side. █