Google Warns Over "Any Method" Used to Auto Generate Lots of Content For Rankings

[engine]

Google's Search Liaison, Danny Sullivan, has warned over the use of "any method" to auto generate content primarily for search rankings may fall foul of its systems. This message appears to be prompted over the most recent experimentation of AI-generated content by SEOs.

Clearly, if you're not careful with your content creation you could end up with problems.

For anyone who uses *any method* to generate a lot content primarily for search rankings, our core systems look at many signals to reward content clearly demonstrating E-E-A-T (experience, expertise, authoritativeness, and trustworthiness).

[twitter.com...]

[Brett_Tabke]

Got to give them credit for trying to scare people.

Unfortunately for Google, it is just not possible for them to stick their finger in the hole in the dam.

AI generated content will is going to dwarf the human generated content within six months to a year.

Already there are big sites putting up AI generated content.

[BigKat]

As Google continues to display more ads, creators of original content displayed in organic will realize diminished returns and some will undoubtedly turn to AI to fill the void.

[mhansen]

Haha - This is a joke, right? I stumbled upon a serp earlier this week where I rank very well, for a high-ish volume, commercial intent query.

I believe Google has turned it's authority screws a bit too hard on this latest update, as the #1 organic result, while definitely related to the query, has no less than 100 auto-generated FAQ's with AI written answers at the bottom of an otherwise 400 word page. I plugged 5-6 of the answers into an AI detector and they all come back as 100% AI written. Being curious, I did the same for each of the 6-7 paragraphs in the content section and the entire page is 100% obvious AI according to the tools used.

[engine]

>stick their finger in the hole in the dam.

Exactly!

G seems to have been caught spending too much time bean-counting and not enough time leading.

[Brett_Tabke]

It's worse than that - it is the act of a monopolist. It is anti-competitive.

[EditorialGuy]

It's worse than that - it is the act of a monopolist. It is anti-competitive/

In the U.S., Google's search results are protected by the First Amendment (as courts have ruled repeatedly over the past several decades).

[Sgt_Kickaxe]

Wow... and Google sent that threat out to their liason team, too. Google created the market for ranking spam, paid links and other bad experiences... so hold the threats. Ironically the threat illustrates the problem of Google being both keyholder and gateway to information.

I think it's time Google stop trying to police what is online because they only control what is on Google. They are free to ban who they want, you know the risks, but as long as it's actually legal according to the law, publish what you want. Google is not the law.

I suspected Google would start crying foul after they called a code-red, even after they created an entire industry dedicated to helping creators jump through their hoops.

Food for thought - Google, in terms of stopping spam, does the same job as an ad blocker like uBlock Origin, expect Google punishes sites you DO want to see, too. They'd best remember that, quickly, as new ways emerge to find information. Regardless of how dominant and wealthy they are they, too, can be the next Netscape or AOL.

.

[Sgt_Kickaxe]

One other thought - if that warning was for real, was it on Twitter only? It's got errors, too, like "to generate a lot content...".

Coming from Engine, responded to by Brett, and being on the official Twitter channel, I assumed it was legit. Now that I investigate to learn more I can't find the warning on Mastadon etc.

Was it for Twitter only? Trolling? If not, why is it not on the blog or other sources? Nothing about this thread is Good for Google.

I only need to do some simple Google searches to find all the "this content by AI" type disclaimers, and there are MANY, all from sites Google trusts so Google is sending mixed messages.

Really, if this is code-red related, it's not good. I assume the warning was about the BankRate question a Googler asked about Google? All this is just more evidence that Google may have peaked.

Edit: The tweet has been edited as of a moment ago. AI content is fine, unless you use it to rank, so I can't wait to hear how Google reads minds. Don't be a problem for Google and they'll treat you right, apparently. That's the bar, guilt and intent will be decided for you, by Google, for Google.

Thank god the other engines aren't lashing out prematurely.

[tangor]

There's irony in all this: g has been pushing AI for some time. Now AI is bad... Go figure.

Waiting for act three to see how this plays out.

[superclown2]

A major American magazine is now ranking for multiple search terms in my UK vertical using boilerplate content that is clearly either AI generated, or created by a spinner. It seems that the top three criteria for ranking now are authority, authority and authority. Experience and expertise? Laughable if it wasn't so tragic.

[EditorialGuy]

Let's step back for a moment and be coolly logical. Google has never been hostile to automatically created content per se: The devil has been (and will remain) in the details.

Things like currency values, the status of an airline flight, weather and climate data, are nearly always generated automatically, yet they rank just fine in Google. It sees pretty likely (to me, anyway) that the application of artificial intelligence to such content wouldn't necessarily be a bad thing in Google's eyes. If FlightAware can use AI to predict the likelihood of a flight delay for Delta Flight 1234 (as opposed to an historical percentage that may or may not be relevant under today's air-traffic and weather conditions), what's not to like?

On the other hand, it doesn't seem unreasonable that Google might be skeptical of an AI-generated foreign-policy paper, a comparative analysis of coronary stents vs. coronary bypass surgery, or even a guide to the London Underground that purports to be (or that has the look and feel of) human-created content--at least at the current state of the art. It all boils down to context, IMO.

[Lurker 2point0]

This whole thread is misleading. The OP, who is an admin? did not tell the full story of the conversation ongoing on Twitter. He failed to post this quote by Danny Sullivan.
We haven't said AI content is bad. We've said, pretty clearly, content written primarily for search engines rather than humans is the issue. That's what we're focused on. If someone fires up 100 humans to write content just to rank, or fires up a spinner, or a AI, same issue..."

[twitter.com...]

Nor did he bother to read this article on Search Engine Roundtable. [seroundtable.com...]

The whole twitter thread, was in response to BankRate dot com blatantly admitting they are using AI content.

So I hope that was a wet blanket that has caused you guys to stop letting your hair be on fire. ;)

[Brett_Tabke]

@Lurker2point0 The original post is on point because this is bs:

> content written primarily for search engines

There is zero content on the web today that isn't written for the search engines as well as humans. If it wasn't for the engines, there would be a robots.txt block on it. Google is trying to have it both ways here: Scare the seo's with misleading tweets by multiple googlers, and let everyone know it is ok if it 'isn't for se rankings purposes'.

8 more major news agencies have been found (including CNET) using many articles they post from ai's. Bankrate is irrelevant.

[futurism.com...]

also, WSJ is playing with it: [narrativa.com...]

[Lurker 2point0]

You do not understand the context of the word "primarily." If your first intent is to get search engines to rank you, rather than your first intent is to inform your readers, then yes, they have a problem, but like Danny said, it doesn't matter if you hire 100 human writers, spin out 100 versions of the same article, or use AI. If the content is bad, then it is a problem.

[tangor]

The real problem is these recent AI chat programs aren't "bad"! For the most part they actually achieve real world goals at the press of a button. This falls into the "close enough" category to encourage a lot of lazy folks looking for web income to keep pressing that button, making g work that much harder, and value, over all for everyone, becomes homogenized and indistinguishable.

Perhaps there is a future "battle of the bots" on the horizon?

[Lurker 2point0]

"Perhaps there is a future "battle of the bots" on the horizon?"

Yes, prepare for major disruption of the entire ecosystem. How people get their information, how people produce content and everything in between.

[tangor]

@Lurker 2point0... forgot to say Welcome to Webmasterworld .... however, your name suggests you have been here before (lurking!). :)

[Brett_Tabke]

>You do not understand the context of the word "primarily."

99% of the content on the open web today was written with the intention of getting traffic out of search engines. <full stop>

The real point here is that Google is attacking ChatGPT as a competitor, because most ChatGPT output I have seen is better than most of the human output on blogs I see.

Now, what if we had a Wordpress plugin that would take a list of keywords and automatically generate ChatGPT posts based on those keywords? What if it was actually better than the same thing that a human site does? What if it ran the text through an article 'spinner/tweaker/grammerly' and it was undetectable as ai? (The plugin exists and was in part written by ChatGPT)

[Sgt_Kickaxe]

This reminds me of the Demand Media fiasco. Google is looking the other way for some, still.

- "Hey G, since CNET didn't write that content, only vetted it, can everyone ask AI to paraphrase it all as a basis to then improve on it?"

- "Oh, I saw your message. Vetting AI content is enough for some, but not for all, got it."

- "Wait a minute, their AI took content they didn't write from everywhere to create a pre-vetted version..."

What a mess, I'm glad it's google's mess.

[edited by: Sgt_Kickaxe at 1:47 am (utc) on Jan 14, 2023]

[Lurker 2point0]

Tangor,...Yep, been here for a long time. Since 2011

Brett, I have seen a couple of the plugins. At least the ones on the repository. I have also played around with ChatGPT. It does good for a 500-800 word blog post with one command. It doesn't put in headings. With multiple commands, you can get a catchy title, detailed outline, and a 2000-word article on evergreen topics that do not require "fresh content."

I'm afraid I have to disagree that Google is attacking ChatGPT. That was the whole point of the clarifications put out by Danny Sullivan. They were being confronted because of Bankrate and its use of AI. They said it is "Spammy" content they don't like. In fact, they quietly changed their section on auto-generated content and included the descriptor "Spammy." If you cannot tell the difference between spammy content and content that is useful and helpful to readers, then I don't know what to say.

You are right. I am a professional in my niche. I do what I write about for a living. As a writer, I am sad to say that AI content (GPT-3 and ChatGpt) writes better than I do, and with the tests, I have run, my readers like the content better. It gets a lower bounce rate, more page views, and a longer average time on site. I was a stickler about facts and the message given out with the content, but still, it performs better than I can do.

[tangor]

Will this lead to a return of the "best viewed in "x" browser" kind of thing?

THIS SITE IS chatGPT free, 100% human.

[SEOGuyDubai]

This age is gonna be AI Based age soon. They may have some parameters for tracking it but they won't be able to track it fully. Also, if AI-generated content is been modified by a dedicated resource, what would they do? So, the thing is they are changing and every thing but they won't be able to control all the things.
In fact, they are making everything difficult for small & tiny businesses to keep up online.

[Atomic]

The biggest issue for me is that these sites are using chatGPT to produce content that makes no sense for them to put on their sites. They're just abusing their authority to manipulate rankings. Isn't that against guidelines?

[tangor]

Isn't that against guidelines?

Begs the question: Whose guidelines?

[Atomic]

Google's guidelines which I am sure you've read: https://developers.google.com/search/docs/essentials/spam-policies

[tangor]

Heh!

Rest my case.

[Brett_Tabke]

> If you cannot tell the difference between spammy content and
> content that is useful and helpful to readers, then I don't know what to say.

Then you aren't reading the same stuff. Most of what ChatGPT puts out is far better than what you find in the average post on Sullivans own site (searchengineland).

q: What grade level in the US do you write at?
a: I am a general-purpose language model and can write at various grade levels depending on the task and prompt given to me. I can write at a level suitable for elementary school students, high school students, or college students and beyond. My default is to write at a level that is easy to understand for most people, but I can adapt to a specific grade level if provided with more information.

Ask it a medical treatment question and request ADA format and it gives an answer that would be home on any university site.

btw, did you see that ChatGPT passed Google AdWords certification?



Also ChatGPT has Passed:

• The SAT [econjobrumors.com...]
  
• The bar exam [twitter.com...]
  
• Google ad certification [twitter.com...]
  
• Amazon Web Services Certified Cloud Practitioner [twitter.com...]
  
• The Turing test [mpost.io...]
  
• Some health care tests [abc.net.au...]
  

(if you know of more certifications or tests it has passed, please let me know)

[Lurker 2point0]

Brett - No argument about Search Engine Land. I agree. But I wouldn't call it spammy, just like when used right, I wouldn't call GPt-3 or ChatGPT spammy. Which was my whole point. The title of this thread was "Google warns over "any method" when what they did was clarify issues concerning using AI content. It should have read "Google clarifies that AI isn't the problem, spam is."

[Atomic]

My team and I have been examining content produced with ChatGPT and find it lacking in many ways. There's zero voice and the word choice leaves a lot to be desired. I expect it to improve. But right now this content is sub-par in every respect. Key details are missing from every article I've seen so far. How helpful is that? I also find it interesting that many of the articles I've seen do not rank yet are indexed.

So, who thinks Google wants the total number of indexed web pages to double? Those datacenters aren't cheap.

I am shocked anyone is impressed by the quality of the content we're talking about. What does your content look like?