✔️ Google: Machine Or AI Generated Content Still Not High Quality

Google has no way of determining whether or not an AI wrote the article, but at the time they were thinking about how to filter out AI generated content the GPT-2 was highly inaccurate. So they created algorithms for determining the accuracy of statements knowing that eventually there would be a flood of AI generated content on the web.

They already had a trust threshold that your site had to pass before they would start to evaluate your content (by sending traffic to it). It was a pattern they looked for in terms of what legitimate business entities would do when starting a new site versus a spammer with 100 sites.

It's only after your site passes that initial trust phase that each page gets it starting value. You can write as many articles as you want, but without passing that trust phase those pages don't pass any real juice to the rest of the site so it can rank or become an authority. That was their initial way of combating doorway pages.

Now, they just try to funnel as much SERP traffic to the bigger authority sites as they can by consolidating SERP results. You try searching for a tiny longtail and in many niches you'll get back a vague article from an authority site in the niche instead of a smaller site that might have been more accurate.
 
It's only after your site passes that initial trust phase that each page gets it starting value. You can write as many articles as you want, but without passing that trust phase those pages don't pass any real juice to the rest of the site so it can rank or become an authority. That was their initial way of combating doorway pages.
So what's the way out?
 
Google itself claims that they basically nailed it now: https://ai.googleblog.com/2022/10/ul2-20b-open-source-unified-language.html
 
Google itself claims that they basically nailed it now: https://ai.googleblog.com/2022/10/ul2-20b-open-source-unified-language.html
hyprocrites.
 
Back
Top