Contrary to their claims, Google doesn't like AI content at all and won't even rank them

Waifu_tatyana

Elite Member
Joined
Apr 20, 2021
Messages
2,479
Reaction score
1,772
No matter how heavily you edit your AI articles google refuses to rank them. All AI articles are currently at "discovered: not indexed" for over a month.

I have tried manual link submission, and forced indexing.

Before anyone says "Maybe your prompts suck, " they don't suck at all. I have created a custom workflow that generates quality long-form articles with <5% hallucinations and even passes most so-called AI detectors with grace. The content is extremely informative too.

None of that matters in the end. GSC data shows google crawls my site every 2 days and updates my indexed pages list but the indexation rate stays the same.

I plan to rewrite the articles from scratch manually and resubmit them.
Is anyone else facing this issue?
 
Last edited:
I've only just started doing this and still setting up my first site, but posts are definitely getting indexed, and they are the first ones that are admittedly crap.

What is your domain like, in terms of what links it has? If it's a new domain then that might be your issue.
 
I've only just started doing this and still setting up my first site, but posts are definitely getting indexed, and they are the first ones that are admittedly crap.

What is your domain like, in terms of what links it has? If it's a new domain then that might be your issue.
Yeah, my domains are like 2-3 months old each. But indexation shouldn't be an issue irrespective of domain age.
The link profile of the domains is fine with no spam links.
In fact, one of the domains is in the niche where spam links are very normal and textbook shitty domains with garbage thin content and shitty link profiles are ranking fine.
 
Try to add a bit of humor to your article ... Oops! this is my secret for making AI content more engaging ... Well this secret is open for everyone now
Not really. If you know a little bit about NLP customizing a unique AI language model from an existing model is quite easy.

I have created very sophisticated workflows for my AI articles that spit out human-like content (at least it reads very human-like to the naked eye).

Here's a raw snippet of an article introduction for the keyword "Beginner's guide to Omegle" produced by my tweaked model

Code:
Omegle is an excellent way for meeting new people as well as speak with strangers. By using the Omegle video app, you can have a text chat or a web chat similar to the Omegle website.
Omegle users have the option to remain anonymous and are not required to divulge their identities. The conversations work by randomly selecting a user and starting a conversation. A chat can also be stopped at any time and a new one started.
By adding his interests, one can make Omegle's talk more fascinating. The discussions can then be started based on shared interests.
 
Last edited:
Content has to be accompanied by videos and text also. Most content isn't indexed by default. Also you must do keyword research to find topics that aren't overly competitive. If you have a million competitors then you won't rank for a highly competitive keyword. I rank #1 for a lot of seriously hard and competitive keywords all day. I use AI...
 
Content has to be accompanied by videos and text also. Most content isn't indexed by default. Also you must do keyword research to find topics that aren't overly competitive. If you have a million competitors then you won't rank for a highly competitive keyword. I rank #1 for a lot of seriously hard and competitive keywords all day. I use AI...
I use videos, infographics, outbound reference links, and everything. Nothing helps.
 
focus on EEAT also build topical authority
 
Heard the so called "randomness algo"? When you can't figure it out why a domain is weird on indexing and ranking, maybe change another domain, or build more sites, I guess Google has sth to let some domains perform well while others shit no matter what you do/think. No idea how to call this phenomenon.
 
Not really. If you know a little bit about NLP customizing a unique AI language model from an existing model is quite easy.

I have created very sophisticated workflows for my AI articles that spit out human-like content (at least it reads very human-like to the naked eye).

Here's a raw snippet of an article introduction for the keyword "Beginner's guide to Omegle" produced by my tweaked model

Code:
Omegle is an excellent way for meeting new people as well as speak with strangers. By using the Omegle video app, you can have a text chat or a web chat similar to the Omegle website.
Omegle users have the option to remain anonymous and are not required to divulge their identities. The conversations work by randomly selecting a user and starting a conversation. A chat can also be stopped at any time and a new one started.
By adding his interests, one can make Omegle's talk more fascinating. The discussions can then be started based on shared interests.
Now check out the below revised - humor added article ... seems more interesting to me!

Well, well, well, Omegle sounds like the perfect place for introverts to socialize without having to put on pants! And who doesn't love the thrill of randomly being matched with a stranger who could be your next best friend or your worst nightmare? Just remember to use your inside voice and avoid revealing too much personal information, unless you want to be the next star of a Dateline NBC special!
 
Heard the so called "randomness algo"? When you can't figure it out why a domain is weird on indexing and ranking, maybe change another domain, or build more sites, I guess Google has sth to let some domains perform well while others shit no matter what you do/think. No idea how to call this phenomenon.
On all 3 sites? The sites have different hosting, registrar and GSC accounts too.
 
Now check out the below revised - humor added article ... seems more interesting to me!

Well, well, well, Omegle sounds like the perfect place for introverts to socialize without having to put on pants! And who doesn't love the thrill of randomly being matched with a stranger who could be your next best friend or your worst nightmare? Just remember to use your inside voice and avoid revealing too much personal information, unless you want to be the next star of a Dateline NBC special!
That sounds like BuzzFeed :D
Will probably try this out. Thanks

Here's the content detection stats of your content

Screenshot_2023_0510_160427.png
 
Never faced that issue with either indexing or ranking the AI articles.
I use expired domains on most of the websites.
I also use fresh new domains on some sites.
And all my keywords are low competition.

All my articles are auto optimized for SEO as they are generated by AI based on the top 5 serp results.
SEO optimization and topical relevance is the key to ranking.
 
Not really. If you know a little bit about NLP customizing a unique AI language model from an existing model is quite easy.

I have created very sophisticated workflows for my AI articles that spit out human-like content (at least it reads very human-like to the naked eye).

Here's a raw snippet of an article introduction for the keyword "Beginner's guide to Omegle" produced by my tweaked model

Code:
Omegle is an excellent way for meeting new people as well as speak with strangers. By using the Omegle video app, you can have a text chat or a web chat similar to the Omegle website.
Omegle users have the option to remain anonymous and are not required to divulge their identities. The conversations work by randomly selecting a user and starting a conversation. A chat can also be stopped at any time and a new one started.
By adding his interests, one can make Omegle's talk more fascinating. The discussions can then be started based on shared interests.

From what I read from that snipped, your tweaked model sounds like default GPT-3.5 lol.
 
Back
Top