The Web Needs 'Pest Control' When It Comes to LLM Slopfarms
The goal is to discourage more sites becoming slopfarms
HAVING just manually checked recent articles from Bobby Borisov (I sort of 'audit' them once a day; something like this can be done without scanners), it seems like he realised it was a mistake to experiment with LLMs and stopped doing it. That's great news.
But there are still fake sites out there that produce fake 'stories' about Linux. Here's the Serial Slopper slopping away at authentic articles:
And Google News is still at it this evening (like this morning):
Fake images, fake text, just a third repeat in a third slopfarm - part of a network of slopfarms boosted by Google News.
The good news though is that slopfarms can be identified and treated accordingly; if we can prevent more sites becoming slopfarms, that's a positive step. Eventually slopfarms will perish and go offline as there's no financial incentive for them to exist and they rely on resources and costs of companies which are doomed to fail (no business model, just a stupid bubble). █