Crowdfunding vs Bitcoins: donations are better investment than digital tulip mania
When Identifying "Low Performers" and "PIPs" Aren't About Improving Performance But Reinforcing a Clique in Your Company/Organisation

Slop and Flop (IBM), Slopfarms and Hybrids (Linuxiac)

posted by Roy Schestowitz on Feb 08, 2026

Slopfarms are sites/blogs - or sometimes whole domains - that are 100% LLM output. We coined this term about 2 years ago and many terms that we coined in the past caught on and spread, including "openwashing", "social control media", and "GAFAM".

Here's a great example of slopfarms:

Slop or fake: What Is AppArmor? A Practical Look for Linux Admins

Slop or fake: Linux Security Hardening Guide 2026 SSH Backup Strategies

They use slop images, slop text, and in more recent years (2025 onwards) even some author names were pseudonyms.

Here's a slopfarm - a former news site (yes, it used to be a real, legitimate site) - that has infested Google News for about a year already. It was in the indices before it became a slopfarm; then it changed and became just a firehose of slop:

Slop or fake: GOG Bets Big on Linux: Inside the DRM-Free Retailer’s Ambitious Push to Win Over Open-Source Gamers

Slop or fake: Breaking Linux on Purpose: The Art of System Recovery Without Reaching for the Reinstall Button

Those are just some of the latest ones. They're not good; they get many basic facts wrong.

As we noted last year and again this year, Linuxiac also experiments with becoming a full-time slopfarm. Some articles are still real, some are 100% fake/plagiarism, and this very latest one in slopfarm is seemingly a mix:

A recent discussion on the Linux kernel mailing list considers whether machine learning could assist kernel subsystems. Viacheslav Dubeyko, an IBM engineer, proposed creating a generic infrastructure to enable kernel subsystems to interact with machine learning models operating entirely in user space.

And since ML immediately brings to mind AI, I want to make the following point clear. The idea does not involve embedding AI or ML code inside the kernel. Instead, kernel subsystems would expose data to the user space, where machine-learning models normally run, and receive optional recommendations in return.

What compels a formerly OK site to dabble in slop?

Did Bobby Borisov assume he would never get caught?

We'll continue to monitor this.

Meanwhile, in relation to IBM, someone has just posted a good comment in relation to a likely paid puff piece about slop:

The proof is in the pudding. The claim is that they have an order book chock full of AI consulting business. That may be true but there are already a lot of consulting choices available and more are emerging all the time. There is nothing about massive success stories related to watsonx or WCA. Those successes would possibly foretell actual future growth.
The author(s) should be required to use IBM internal AI AskHR, AskIT for a week (or even a day) before declaring that IBM is an AI darling. Actual IBM AI today is similar to when IBM was flogging Notes and Freelance when the rest of the world decided that they were garbage.

IBM is not an AI darling; it's an AI poseur.

IBM tried - and failed - with slop. Companies don't profit from slop; watch what happened to Google and others.

They can take the "masses" (horses) to water by drowning GMail or Windows or whatever with slop, but they cannot make them drink. █

Slop and Flop (IBM), Slopfarms and Hybrids (Linuxiac)

Other Recent Techrights' Posts