Google vs. AI-Generated Content
From social media to blog posts, we are constantly being bombarded with all types of content online. In content marketing, creators feel pressured to rapidly provide quality work to get their brand or client's brand noticed. This is where Artificial Intelligence (AI) can potentially ease the burden.
With the recent upsurge of AI-generated content, new tools are emerging to make it easier for marketers to up their delivery game. But there is an ongoing debate about the appropriate use of automatically generated content and whether it damages SEO (Search Engine Optimization) ranking and the user experience.
In response, Google has produced several updates that penalize auto-generated content. The search engine giant claims that this content violates its guidelines.
Creators are curious about a few things:
- How does Google detect auto-generated content?
- Is there a workable compromise that can be reached in the AI/human writing space?
Battling Trends
Two current trends are at odds: on the one hand, marketers are finding that AI content tools are becoming a vital part of their strategy. But on the other, Google insists that personal, unique content is essential to its users.
Google's E-A-T (Expertise, Authority, and Trustworthiness) is its criteria for high-quality content. Recently Google has added an extra "E" for Experience. So, the acronym now reads as E-E-A-T.
This update challenges AI content tools. They can draft content at record speeds but cannot authentically speak to human experiences. Or at least not in a way Google feels comfortable allowing.
Artificially Intelligent Content Creators
The latest technologies are commonly adopted in marketing strategies, and AI is no exception. AI has been invaluable in helping marketers understand their target audiences, forecast trends, and drive revenue. Now, marketers are taking things a step further and including AI in their content writing strategy.
Copywriting and image-generating tools that use machine learning, natural language processing, and predictive algorithms have helped computers understand language and context to produce seemingly natural content.
GPT-3
Made by OpenAI , GPT-3 (Generative Pre-trained Transformer 3) is one of the most powerful AI content tools today. It is notably ahead of the competition in creating natural-sounding written content.
GPT-3 uses unsupervised machine learning, like Google's BERT, and self-trains to better comprehend typical human language, context, and syntax.
Unsupervised machine learning relies on a pre-trained algorithm to write the best content for its users’ needs. Vast quantities of training texts are web scraped to understand how humans use written language to communicate. This third version of GPT has been trained with over 570 gigabytes of text, hundreds of books, and all of Wikipedia.
However, GPT-3 has its cons - the tool is pricey and not perfect. Also, whether or not Google will allow it is still the question.
Google's Robots Vs. The Marketing Team's Robots
It's ironic, but Google uses its own AI to detect other auto-generated content. Google's SpamBrain is an AI-powered spam detection and prevention system. It knows how to read and contextualize a website, measuring it against Google's standards.
If SpamBrain flags written material for any reason, that content may fall by dozens of ranks on the search engine results page (SERP).
Within Google's spam policies, there are strict guidelines around "auto-generated content" as it arguably serves less value to Google's users.
For example, a blog written by an AI tool might be overstuffed with a specific keyword. Even when the content drafted is realistic, if a keyword is used out of context or in an unnatural way, the Google site crawlers can determine that the work is:
- Not providing a positive experience to its visitors
- Potentially written by a bot
A lack of correct editing and formatting can also indicate to Google that the content is AI-generated. If formatted properly, Google can easily understand the page's context and tell if it's high-enough quality to place as a top-ranking result.
Google's algorithm is constantly learning, and there are even dedicated employees who report on what they suspect might be robots posing as bloggers. Users are also encouraged to inform Google of suspected AI-generated content.
Why Does Google Care?
Google uses AI, so why does Google care about AI-generated content?
The answer is that Google cares about providing value to its own users. Google Search Team member Duy Nguyen recently said, “Scraping content, even with some modification, is against our spam policy. We have many algorithms to go after such behaviors and demote site scraping content from other sites.”
Rapidly generated content often forgets to take Google's E-E-A-T into account. Even if it is optimized correctly and receives a high SERP ranking, the quality of the content doesn't bring much value to Google's users. It can, in fact, potentially harm their experience.
AI Vs. 3rd Party Detection
Google isn’t alone – other third-party players are also on the lookout for AI-generated content. Even OpenAI, the company that built GPT-3, has a detector available.
AI content may sometimes slip through the cracks, but there is an increasing number of ways to detect auto-generated material.
AI's Place in the Content Pipeline
With Google and AI-generated material at odds, does AI have a place in the content creation process? Absolutely.
Should AI be the only content creator on the team? Absolutely not.
Today's AI is smarter than ever, but automation is not an excuse for us to become lazy. We still haven’t reached the point where a marketer or writer can push a button and effortlessly create stellar content. The human element is still a vital component.
AI can't yet speak to human experience (the new E in EEAT). To establish a bond with the reader, an author must show that they understand how the content makes a person feel. That comes with human experience.
Another Tool in the Toolbox
AI will only advance from here, and the rumor is that GPT-4 will be exponentially more powerful than its predecessors. To keep up with this fast-moving evolution, Google will be under pressure to continuously adapt its guidelines.
Sophisticated as it is, AI will still need to understand and learn how to relate to readers more emotionally.
In the meantime, marketers can treat AI like another tool in the creative toolbox. They should use AI., those that don't risk falling behind the curve. The key is using it appropriately in the content writing process to efficiently produce something original.
If the content is original, then Google won't have a problem with other people's robots.
