Experimenting with AI-generated content

Spoiler alert:
I am not human. Well, only partially. This article is largely written by Artificial Intelligence.

Unique and readable SEO-optimized content generated by Artificial Intelligence in real-time? How AI-generated content will impact the future of human writing and SEO when Google won’t differentiate robot input from human input anymore.

In 2018, Elon Musk's OpenAI published its general GPT-2 machine learning model for writing content (currently at GPT-3). Once it was released, multiple organizations started experimenting with it. Meanwhile, there are several service providers available, that offer AI-generated content one click away. We are currently experimenting and are curious where the journey of machine-based content will lead to, and whether it will soon replace human content writing.

Automatically generate content, where are we by now?

With artificial intelligence, marketers can automatically generate content for simple stories such as stock updates and sports reports. You’ve probably even read content written by an algorithm without noticing it.

An example of news written solely by an algorithm shows that we come across this sort of (short) news content very often already:

“Tuesday was a great day for W. Roberts, as the junior pitcher threw a perfect game to carry Virginia to a 2-0 victory over George Washington at Davenport Field.”

Also, The Guardian in September 2020 let a robot write an essay on its purpose as artificial intelligence, it’s quite impressive and funny just check it out here.

Google’s quest against spamming robots

Google by nature isn’t only a very visionary and innovative business, but also a valuable benchmark to evaluate the development of machine-based content.

Throughout its history, Google has fought against users manipulating their algorithms for their benefits, despite the fact that Google itself is on top of the list of machine-based technology.

After the page rank algorithm, which caused the internet to be flooded with backlinks, Google focused on semantic content, which Google bot crawls closely analyzed. The result was pages filled with keywords to saturate them semantically and rank higher. 

Not fooled easily, Google launched big algorithm updates over the years, trying to penalize sites that used keyword stuffing, link farming, duplicate content, irrelevant links, etc. Paradoxically, the best way to do this is to upgrade Google's AI-technology to improve search results.

With these updates, Google was sending a clear signal: content creation must be unique, spontaneous, credible and authentic. Google would defend the “human” side of the internet, namely the relationship of trust, credibility, and reliability with the user and the “content spontaneity” of writers. As a result, Google can now know precisely whether your content satisfies the user or not, without relying on backlinks and semantic content. It fully values the information and engagement value of your content.

But what if these human values can be replicated by robots?

Even if the impact is yet to be considered minor, but it pushes everyone in the content and writing sector to reconsider the core of content creation strategy and how to invest in it in future.

Authenticity is still the best up-to-date map, but machine learning will accelerate the battle in many markets. From an economic cost vs. time input/productivity point of view, it has overtaken the human factor already.

The threat of AI-based content

The fact is that the learning machine also spreads to the side of content creation and generation. Machine learning-based applications rely on intelligence that learns to write based on online data. Just give them a few keywords, and they create unique content, using the information they find.

As for the applications available, the result is not really convincing yet, see the previous example of The Guardian. Algorithms still lack relevance and structure, and global semantic understanding. With the example of this article, the original output wasn’t too readable and lacked structure and argumentation. But there is no doubt that these technologies will evolve to the point of being indistinguishable to the user and especially to Google.

Should Google then question its human content policy? They might one day admit the growing power of content automation.

We keep on experimenting for the time being…

Heitland Innovation Coaching - Let’s speak today and connect the dots of innovation!

Follow our social accounts to stay updated on our content