Artificial intelligence (artificial intelligence) technology has made tremendous progress in recent years. It is now possible to assess their ability to perform specific tasks such as creating text, images and sound.
Now, what if we went further with more complex tests, like evaluating a job, for example, or more specifically, evaluating an AI system on its SEO ability?
Below, we’ll test Generative Adapter 3 (GPT-3) created by OpenAI.
Let’s keep in mind that an AI system will simulate the data it is trained on.
SEO was built along with search engine advancement, and everything is well documented in blogs, books, and interviews.
Therefore, the AI is likely to have learned from all available SEO content.
Therefore, the tests will measure the following indicators:
- Consistency in following the topic of SEO.
- The importance of SEO adjustments.
- The ability to answer SEO questions.
Consistency in following the topic of SEO
The following tests will focus on an AI system that understands the SEO situation and its ability to handle the topic clearly and consistently.
The first test will take examples of SEO horror stories—that is, short stories about SEO disasters—and see how AI performs.
I took two examples from Aleyda Solis’ website and here’s what was generated by the AI:
Two examples:
- When all the internal pages of the site have been translated to the main page. # horror stories
- When someone forgot to block the location of the test environment… that started indexing it. # horror stories
Results generated by OpenAI’s GPT-3 AI:
- When someone said, “I have a great idea! Let’s put our site in an irrelevant footer!” #horror stories
- When someone forgot to check the staging environment’s robots.txt file before going live. # horror stories
- When someone made changes to .htaccess without telling anyone #seohorrorstories
- When we switched domains and didn’t redirect old pages for more than 60 days. # horror stories
- When the customer generated thousands of pages of slave content with no value #seohorrorstores
- When the customer insisted on using relational database tables as URLs (? tbl = table & id = 2) #seohorrorstores
Analysis: The resulting texts are diverse and coherent.
Moreover, after a little research, we can conclude that they are unique.
Therefore, the AI understood that it had to simulate a short text that talks about SEO and add the hashtag #seohorrorstories each time.
This is a screenshot of the full example; We can create endless #horror stories.
In terms of creativity, we can go a little further with our tests and try to measure the consistency of creating a dialogue between two SEO experts.
For the next test, we will present the AI with a situation and instructions, and then ask it to have a discussion.
This is a situation where AI has to create a dialogue between two experts – one experienced, the other novice: when someone forgot to block the test environment of the site … that started indexing.
Results generated by OpenAI’s GPT-3 AI:
Aaliyah: Oh no! When did you realize this was happening?
Paul: I just checked Google Analytics this morning and saw we had some traffic coming from google.com.
Aliyah: This is not good. Did you check to see if the site test environment is still being indexed?
Paul: Yes, I did. and he.
Aaliyah: Well, we need to take care of this right away. First, let’s check to see if there are any pages that have been indexed.
Paul: There are a few.
Aaliyah: Awesome. Now we need to block the test environment until Google stops indexing it.
Paul: How do we do that?
High: We can add a robots.txt file to the test environment. This will tell Google not to index any of the pages on the site.
Paul: Well, that makes sense. Let’s do it.
Screenshot from OpenAI, April 2022Analysis: The dialogue is still very coherent
The AI system explains how to detect the problem and suggests a solution to the pre-production site indexing problem using robots.txt.
Next, let’s take one of the AI-invented #seohorror stories and create the dialogue associated with it.
Here is the new situation: when the customer generated thousands of pages of thin content without value.
Results generated by OpenAI’s GPT-3 AI:
Aaliyah: Well, that’s somewhat of a problem. I’ve created a lot of thin content pages that don’t really provide any value to users.
Paul: Yes, I know. I was just trying to put as many keywords as possible on the website.
Aaliyah: Unfortunately, it won’t help you with your SEO. In fact, it can actually hurt your ranking.
Paul: So what should I do?
High: You need to create high-quality content that provides value to your users. This is what will help you to rank higher in the search engines.
Screenshot from OpenAI, April 2022Analysis: Once again, the AI managed to generate real dialogue by explaining why creating thin content with SEO didn’t work and even recommending solutions.
We can conclude that the AI system can strictly and accurately talk about SEO, but can it also make SEO-related adjustments to the text?
As it happens, GPT-3 provides a text editing mode that has been available since March 2022.
Let’s try.
The importance of SEO adjustments
The first test will ask the AI to replace the link anchor with something more marketable to better highlight the product.
Results generated by OpenAI’s GPT-3 AI:
Screenshot from OpenAI, April 2022Analysis: The test is relatively successful.
Here, the AI replaces “mahogany-paneled steak” with “best New York steak,” keeping in mind the context.
The solution works regardless of the number of links.
Then, the second test consists of completing two simultaneous tasks: asking the AI to change the link anchor while adding an H1 title tag to the beginning of the document.
Results generated by OpenAI’s GPT-3 AI:
Screenshot from OpenAI, April 2022Analysis: The test worked again, and you can apply the results to thousands of documents. A title that summarizes the document is added in the appropriate place.
Finally, we will ask the AI to add a list of three relevant links using tags
- .
Screenshot from OpenAI, April 2022Analysis: Again, the format is well respected and the anchors are consistent.
The links are fake, but GPT-3 can easily be selected from the link database to use real links.
In short, GPT-3 is good at contextually editing text when the correct instructions are given. This can be a huge time saver for large-scale edits.
The ability to answer SEO questions
We are now going to test the SEO knowledge of the AI system.
you choose to use BrightEdge’s SEO Testbut any test could have been used:
As with each of the previous tests, we used GPT-3 OpenAI technology to answer the questions in our SEO MCQ. The number of correct answers measures the quality of the test.
Screenshot from OpenAI, April 2022Below, we see how each question is presented to the AI; The green shaded text represents the answer they selected.
BrightEdge provides a table of questions with the correct answers in one column and the answer chosen by the AI in the adjacent column:
Screenshot by the author, April 2022GPT scores -3 are encouraging: the AI scores 16/20, and even then, it lacks the context in which it goes wrong.
Now, let’s ask AI for recent updates from Google regarding the release of Google MUM on May 18, 2021.
Screenshot from OpenAI, April 2022What is the first purpose of Google MUM?
Again, as you can see, the AI answers perfectly.
In short, we have a true SEO response engine if we ask very specific questions.
restrictions
Now, we can take a look at the limitations of AI based SEO.
The following is a non-exhaustive list:
The ability of AI to fix unknown or poorly documented problems.
Oftentimes, in the SEO field, understanding what actually happened when a problem occurred is time consuming and tedious, and sometimes there will be no explanation. An AI cannot find solutions with data that is not part of its analysis.
The ability to take into account everything that exists.
There are often many limitations to what is actually done or tested on a site or set of sites. Therefore, it is currently impossible to find all the source code and project documentation for the AI system to find an accurate answer that takes into account all the information available.
The ability to gauge the true quality of content or an idea by its originality or representation of the current situational context.
AI systems don’t work in real time yet, and they lack originality and creativity. As a result, the true measure of content quality is currently scant. Humans are often the ones who rate high-quality content. AI systems generally detect spam and poor quality content.
To summarize, an AI system can outperform a novice SEO or someone with little SEO knowledge, but they still have a long way to go to replace an experienced SEO expert.
In general, the duo of AI and humans will always be the best combination, with humans focusing on tasks with the highest added value.
In conclusion, all seasoned SEO experts are way above AI SEOs.
More resources:
- Google says AI-generated content is against guidelines
- Google explains how AI keeps searches safe via MUM & BERT
- Advanced Technical SEO: A Complete Guide
Featured image: Shutterstock / metamorworks
![7 Ad Copy Tests To Boost PPC Performance [With Examples]](https://altwhed.com/wp-content/uploads/2023/01/1672844873_7-Ad-Copy-Tests-To-Boost-PPC-Performance-With-Examples-390x220.png)
