Published 
Updated
‍🕑 6 min read‍
AI writers are growing at a huge rate. You see them everywhere. And it is not just about ChatGPT, Gemini or other tools such as TypeTone.ai, but AI writing in general.
A wealth of AI content is being released. Even so, creating AI content rarely gets you the results you expect straight away.
This time we experimented with two well-known AI tools, because a new AI tool comes out almost every day and one quickly overtakes the other. Think of an AI prompt and how you can get the most out of it.
Does AI content get you more search traffic, or does it lower your search traffic? Does this content produce more traffic for the time you put into it, or less? Is AI content really better than content written by people?
In this 6-part series, I discuss the most relevant points with you.
How many people use AI? AI is used for all kinds of things, from legal to financial and from business to retail.
For example, Capilex uses a fully automated process to provide business loans. One of the first with full automation. It saves a lot of hours on compliance.
AI is not a trend, AI is here to stay.
Back to the question:
- almost 30% say they use AI daily at work
- 41% expect to be using AI daily by 2030 Â
Source: research by BSI among 1,000 Dutch people (2023).
There is also something to say about AI use within businesses. In 2025, 29.8% of Dutch SMEs already used AI, as did 66.2% of large companies (CBS, Statistics Netherlands, March 2026). Below are the figures for 2023: Â
According to figures from Statistics Netherlands (CBS).
When you look at AI through a marketing lens, most marketers use AI tools for content creation, writing persuasive copy or finding inspiration for ideas, and some use it to analyse data or create images.
‍‍
The biggest use case we see is content creation, according to a Salesforce survey.
‍
Bear in mind that when you analyse marketing data, some inaccurate results can come out. It can help you with the rough work, but once you move on to complex statistical analyses, the outcome often falls short.
This also depends on the quality of the input and the task you give it.
The first important thing to know is how you create content with AI tools such as Gemini and ChatGPT. These large language models (LLMs) do not make new content. All they do is reorder and combine content that already exists.
So nothing that comes out of ChatGPT or Gemini is new.Â
How exactly does a language model work? In 4 simple steps:
1: collecting data
- Collecting relevant text data from multiple sources
‍
In plain terms: imagine we go on a treasure hunt and everywhere we look, we collect treasures. Those treasures are bits of text from different places, such as books or the internet, that help us learn something new.
2: Creating tokens (identification)
- Breaking the text down into smaller “tokens” (words, phrases or even individual characters)
‍‍In plain terms: it is as if we have a big puzzle and cut it into small pieces, such as words or even letters, so we can look at them better and understand how they fit together.
3: Paraphrasing and feature extraction
- Capturing semantic relationships
‍
In plain terms: imagine words are friends who work together to tell us a story. Some words are like best friends and have a special connection, which helps us understand what they mean together. This helps us learn more than if we listened to each word on its own.
- Extracting relevant features such as word frequencies or other numerical representations‍
In plain terms: imagine we are making a magic potion. We count how often we use certain ingredients, such as words, and this helps us understand what matters in our potion. That way we can see which flavours are the strongest.
4: Text generation
- Creating new text based on the input from the prompt or existing content
In plain terms: this is just like painting a picture after someone gives you an idea or when you look at an image. You use what you see or hear to make something completely new with your brush.
No. Most people cannot tell the difference between AI content and content written by people.
Astonishing.
Even specialists find it a challenge. These linguistic specialists failed to spot the difference 61.1% of the times they were shown AI content.
A remarkable new development since the launch of such AI tools.
Another remarkable thing is that younger people have more “trouble” telling AI content apart from human content. The 65-plus age group seems to be the best at this.‍One explanation is that the younger generation accepts what they see more quickly, while the older generation is often more cautious with content.
People often cannot see straight away whether content comes from AI or was written by a person. Can search engines? Or a better question, what do search engines such as Google think about it themselves? Does AI content rank higher than human content? Or lower?
The answer is simple: Google does not mind.
Whether the content is written by AI or by a person, Google applies the E-E-A-T framework. Maybe you have heard of it before. It stands for Experience, Expertise, Authoritativeness and Trustworthiness.
In plain terms: imagine you have a big cake. That cake is made from four special ingredients. Experience and Expertise are like knowing exactly how to bake the tastiest cake, Authoritativeness is like everyone knowing you as the best baker, and Trustworthiness means everyone trusts that your cake is safe and tasty to eat. Google looks for cakes (or stories) that contain these four ingredients, whether a robot or a person made them.
‍‍AI content can rank well too, provided it is of high quality and relevance. Have a specialist check it for coherence and accuracy.
I always advise a check before you publish content written purely by AI. From my own experience, it helps to be clear about what you are using that content for. Is it for a series of landing pages? Then use AI content as a base and give every page something of its own, because Google treats pages made at scale with little added value as spam. Is it for your homepage? Then let AI assist you with suggestions.
There are risks with AI content.
I will highlight the main ones. They are lack of transparency, cybersecurity, bias and discrimination, intellectual property and privacy.
So make sure you have screened for these points to limit any risks you face. Do this especially before you submit these pages for indexing in Google Search Console.
‍
‍
More in this series on creating content
The next step
Discover your biggest growth opportunities with a free opportunity analysis (worth €416) — a tailor-made report, free of charge and with no obligation.
About the author
Dominic Pietersz
Your dedicated point of contact
Dominic helps SMEs grow through Google and other channels. With his international experience, he turns challenges into concrete growth.
Follow Dominic on LinkedInInspiration · On a rainy Sunday, out of curiosity, I went to test two AI writers. The result became this piece on AI content and how Google handles it.
