Since making ChatGPT publicly available in November 2022, Google has struggled to adapt to the new world of generative AI, despite developing much of the technology behind it. The company’s recent response to AI content is already changing the way we browse the web, and could change the English language itself.
What is search engine optimization?
First, let me explain how the online content business works. Before the web, most people subscribed to newspapers. If you can publish your content in that newspaper,
who Newspapers had a predetermined audience, so they were always read. The trick is to persuade the newspaper editor to give you access, otherwise no one will read what you write.
Currently, at least in theory, there are no gatekeepers. Instead of dozens of newspapers, we have millions of websites. The problem now is not getting past the gatekeepers, but staying inconspicuous from the noise. Some destination sites still have built-in audiences, such as Blaze Media. But for the vast majority of independent writers, your best bet is search engine optimization.
SEO is essentially a large scale process that writers develop to help their content appear in Google search results, ideally on the first page of search results, and even more ideally at the top of the page. It’s a trick. SEO has a bit of a bad reputation in some circles, largely due to ugly “black hat” techniques to trick Google, such as keyword stuffing, where writers stuff their posts with target search terms. You’ve probably seen something like this in the past:
Looking for the best wireless headphones? Read our guide to the best wireless headphones. We have created a list of the best wireless headphones that will satisfy anyone who is shopping for the best wireless headphones. We carry wireless headphones at all price points to meet all your wireless headphone needs.
Google’s algorithm penalizes posts that are overstuffed with keywords, so we don’t see that much anymore. However, this will always be a somewhat viable technique, as Google must programmatically analyze billions of websites to determine their ranking in search results, with the goal of satisfying end users. And one of the ways computer programs can analyze these websites is by analyzing the keywords they hit.
Google wants you to think that the secret to getting your content seen and read in search results is simply writing good content. That would be great, but without some sort of SEO, everything you publish could be buried. If you hate SEO, you should blame Google for effectively owning the web and setting the rules, not the people trying to make a living from it.
How Google outsourced AI detection
When ChatGPT came out, SEOs quickly realized that it was:
Really I’m good at SEO. Visit ChatGPT, specify your target keywords, and we’ll automatically create a relevant blog post that will rank on Google in no time. This is not surprising. Because GPT learns languages the same way Google does. That is, it automatically skims the web, collects data from websites, and uses it to teach algorithms. GPT is designed to interpret the intent of a request and generate content that fulfills it, which is the same purpose as Google, so it’s as if it was created for his SEO.
This was clearly a threat to Google’s massive search business. People generally don’t want to read content generated by AI, no matter how accurate it is. But the problem for Google is that it can’t reliably detect and filter AI-generated content. I’ve used several AI detection apps, but their accuracy is at most 50%. Google’s Brain Trust could probably do a better job, but it’s still computationally expensive, and even the great Google can’t analyze every page on the web, so the company finds a workaround. There is a need.
This fall, Google
Helpful content updatesIn response, Google began to heavily emphasize sites based on user-generated content in search results, such as forums. The site with the most significant increase in search rankings was Reddit. On the other hand, many independent bloggers experienced traffic crashes, whether they used AI or not.
Google’s logic is very simple. Focus more on forum sites like Reddit and Quora. There, content is moderated by humans who can theoretically remove AI spam. Google essentially delegates the task of AI detection to forum moderators, who are often unpaid and it’s computationally cheap for Google to do so.
Interestingly, SEO is currently spamming these sites to gamify Google searches, and interestingly, Google recently started using user data to train Gemini, Google’s own AI project. We have signed an agreement with Reddit to acquire .
The great SEO heist and the last straw for Google
However, this update was not enough to deter AI-powered SEO. Google strongly dislikes SEOs who try to trick their systems, especially SEOs who make themselves look stupid. And I wonder if SEO mastering GPT is making them look stupid.Perhaps the most egregious example is
The big SEO heist in late 2023.
On November 24, 2023, SEO Jake Ward announced on X that he had pulled off an “SEO heist” that stole 3.6 million views from his competitors.The competition is
excel jeta well-established site with tips on Microsoft Excel.
In the X thread, Jake explained how he pulled it off. First, we extracted the Exceljet sitemap. A sitemap is a document that lists all the pages on your website. Most websites maintain a sitemap to make it easier for search engines like Google to crawl the site, which makes it easier for your content to appear in search results. You can view the Blaze News sitemap for yourself.
Jake then retrieved all the URLs on Exceljet’s website and extracted the “slug”, the descriptive part of the web address. So Jake created a list of articles such as:
And I entered them into a tool called
Byword.ai, mass-produces blog articles using AI. For a relatively small fee, Jake produced thousands of articles a day that reproduced the same topics and structure as his Exceljet. And it quickly surpassed Exceljet because its AI-generated content looks “perfect” to search engines.
But Jake made the big mistake of showing it off in public. Google immediately de-indexed his site. This means the site no longer appears in search results at all. He also deleted his X account.
The SEO heist was clearly the last straw for Google, but now they’ve announced even more powerful search updates to thwart the AI.
The latest on Google spam and how it affects the web
Google recently
New updates to spam policy Specifically, it targets “low-quality and unoriginal results” through a new “massive content abuse” policy. That means if you’re running a website built around publishing a ton of AI-generated articles, algorithm changes or “manual We expect traffic to drop to zero soon.
In fact, Google started issuing these manual actions almost immediately after the announcement.
Deindex hundreds of sites With almost no warning fresherslive.com and thejohnson.com. You can check this for yourself by going to Google and searching for the site.Domain name. If nothing appears for your domain, it has been removed from Google’s index.
For now, we don’t really know what the algorithmic penalties will be for sites created by AI, or whether they will be penalized at all. It will take several weeks for the update to roll out to all sites in Google’s index. For now, Google manually flags sites for removal from its index, and while there may be some innocent victims, most content is generated almost entirely by AI.
How AI paranoia will change the English language
AI paranoia has affected not only search engines but also social media. Indeed, both Twitter and LinkedIn are clearly full of GPT-generated responses. No matter how fast you engineer with GPT, it has a specific way of writing, and once you know how to recognize it, it’s easy to spot. The grammar is a little too perfect. GPT prefers long words and well-embedded phrases. GPT is free from misspellings and typos. And as GPT and other generative AIs become more sophisticated, that content will become increasingly difficult to find.
Already, I often see on social media that clearly human writers are accused of generating text with GPT because they use university-level words or quirky phrases. Ironically, as the output of generative AI becomes more human-like, good human writers may be punished by algorithms or human moderators who suspect their writing is slightly inappropriate.
Too good. I don’t think Google currently has the ability to scan billions of individual web pages for AI content, but it’s certainly a problem the company is working to solve. In fact, the future of Google search probably depends on it.
I happen to be a blacksmith by training, and one thing my teacher taught me is to never try to remove hammer marks or small imperfections from your work. This is because it shows that it was made by hand and not by a machine. Customers are paying a premium.
In the near future, writers will deliberately introduce spelling mistakes, typos, and grammatical errors into their writing just to leave a little “hammer mark” proving that the content was created by a real writer. Is there a possibility that it will become?
genuine human? Will these intentional misspellings and mistakes eventually become part of the English language as they become the norm?
What if it is our flaws and imperfections that ultimately make us special?
