I’m writing this article from pure frustration over some of the content our competitors, clients, or businesses in general are attaching their names to. It makes me wonder… do marketing managers ever read what is being posted on their blogs? Can you truly stand behind some of the news articles that your organization is spreading across all of its social media?
Not all AI content is bad and we as a marketing agency use AI in ways to support our business, add efficiency, and improve our overall performance. But no, we do not use it to create blogs and news article content for our clients. Why? Let’s start at the very beginning.
Why using AI does not help your organization when writing original content.
The definition of what would be considered “original content” is any copy, text, images, or artwork that has not previously been published and it’s original to the creator. Now let’s look at how AI content is being generated. Per ChatGPT “AI content generation tools use machine learning algorithms, often based on natural language processing, to analyze large datasets of text and learn how to generate new content that is similar in style and tone.” This clearly states that none of the content created by AI is original and consists of scanned copy that is rewritten to avoid plagiarism to deliver “newly” created content.
1. Does Google Like AI-Generated Blogs?
Do you run each article you write using AI through Undetectable, GTPZero or Copyleaks? Why do you do it? Because you’ve re-written a portion of the copy and you’re hoping that it will read as a 60% or more human text. Is that the best way to deliver great content for your site or your clients? Think about it. If all of the above companies can create these detection tools, do you think Google utilizes a better one? Yes, most likely. Although Google very much supports AI technology, the premise behind the company is to deliver the most accurate content in their search results. In order to stand behind its mission, the company will hopefully continue to prioritize the best quality content available. If Google understands that AI-generated articles can possibly come from other AI-generated content that has been made public, what information remains the most reliable and accurate? Reading hundreds of AI articles I can safely say… it’s not top-notch content and worst of all – it’s not always accurate. So let’s move to #2.
2. AI Content is Not Always Accurate
We learned this while reading a few articles where historical data seemed not to align with general knowledge. AI can mix historical events and refer to content that should be attached to a different person or time in history. We know that most of the topics avoid these references as AI rarely includes specific content that mentions businesses or individuals. We did a simple search where we asked in Google search who was an architect for a certain building. The Generative AI that’s very much experimental as Google states, brought an answer that was 100% inaccurate. It attached the name of an electrical contractor as an architect on the project on the basis of a snippet that mentioned on that electrician’s website the actual architect’s name. Without going to the actual web page, you were delivered completely misleading information. This happens often as AI does not comprehend the copy the way humans do. And although the tools are continuously improving, there will always be a lot of errors that come from AI-delivered content as you’re working with a machine that scans and translates data. If you are using AI, please at minimum fact-check the content.
3. All articles use the same structure and provide poor reading experience
One of our clients asked us to rewrite about 10 AI articles generated by another company for that client. It was not an easy task. The topics for these articles were most likely also AI generated which means they had absolutely no value to that business and covered the most trivial topics. Since they were already shared on social media, the client decided to keep the URLs and this content alive. Each article was unnecessarily long (4,000 to 5,000 words) mostly due to an endless amount of filler words as the topic did not support the need for it. They all started with an “Introduction”, many H2 section tags, a “Summary” that repeated word-for-word the content above. And let’s not forget the dreaded statement “in this article we will explain” and ending with “FAQs” list, which seems to be a dead giveaway for any AI generated blog. This company is an expert in its domain and the quality of that content was very much on an elementary level. It made me wonder how a business that positions themselves as an authority in the area would ever deliver an article like this. The buzz of improving SEO and creating fresh blogs just to have a new post date on your news page seem to cloud the original premise of it. I believe the art of writing is not dead and I truly hope that sooner than later penalty from Google will be severe enough where all of the SEO generated AI content writing will simply stop.
4. Lack of creativity, added value, new knowledge
This is probably the biggest downside of the AI generated content. The internet space is ballooning to unimaginable size at a very fast pace while new, valuable, helpful and groundbreaking copy gets buried under layers of 5,000 keyword-stuffed articles. What I love about reading factual papers and columns is that you always learn something new, expand your horizons, get new outlook on the problems and issues you are facing. This “rewritten”, reworded and reworked content by thousands of the same AI generators is drowning our society in mass misinformation. Many of us poses very unique knowledge that they acquired while excelling in their field. I hate to think that there will come a day that this information no longer leads Google search results. That our destination will be a hybrid of articles that collectively don’t make much sense but cover similar topics. And if that’s the case, will we stop writing? What will be the point if our knowledge will no longer be shared and rewarded?
So what’s next in AI content writing
Like I said initially, not all AI content is bad. Short prompts, ideas for bullet points to include in your well-research article and to enhance its message are absolutely brilliant. Using freelance writers that mass-produce content even if they don’t understand the language they are writing in – that’s not the future of the mighty digital universe. So how can we tame this and prevent it from happening? Start with yourself. Your own business. Training your own marketing team. Always focus on bringing to the table what’s most valuable to your reader. If we start there, I think we will be ok.