AI's dead end: AI SEO
The alternative title for this post could also be - how to become average as quickly as possible.
We are told that AI - meaning the Large Language Models e.g. ChatGPT - are now at the PhD level and coming up with novel ideas. Or if not now, very soon. We are talking about a Self Driving Tesla soon! In order to achieve this feat the models have to get just a little bit bigger and we still need more content for training.
In past issues of Recurrent Patterns, I mentioned that this thought is flawed. While feeding the data into the LLM during training, there is no context provided which would identify under which circumstance the information is accurate or relevant. Further, the volume of the information is so vast there is no way to verify the accuracy of the information nor do we know how to replace the wrong (or old) information with the correct, most up to date information. There was also a widely cited academic paper suggesting that we will run out of content for training sometime around the year 2030. Unfortunately, people didn't finish reading the paper to the end where the authors said: 'or maybe not'.
These are fundamental issues which will always result in the (in)famous fine print - 'ChatGPT can make mistakes. Check important info.'
The insatiable appetite for more content is also creating several other problems. At the beginning of the Internet or more accurately the World Wide Web, everyone was racing to put out there as much content as possible to be discovered by various search engines. A few years later after the Big Bang when stars shined and collapsed, the Death Star appeared and captured most of the content for free and in return provided traffic to the actual source. At the same time, it provided the content creators with the ability to monetize the traffic through advertising - conveniently provided by the same company.
It is debatable if that was a 'fair deal', because it decimated the media business. Many newspapers and magazines disappeared, discontinued the print edition and are in survivor mode. The stronger brands erected paywalls. Still, they kept using the only marketing channel - Google - which kept them alive. Everyone focused on optimizing their content for that marketing channel.
Appearing at the top of the Google search results, without paying for expensive ads there, became a lucrative industry and an untold amount of money went to people who claimed that they know how the Google ranking algorithm works and how to adjust the website to constantly rank at the top. For every 10 top winners - e.g. first page of Google search with the top ten links - there are thousands of losers. Who would ever check anything beyond page 2 or 3 ...
Time to time, Google would change the algorithm and new winners would be declared and a new opportunity to sell Search Engine Optimization (SEO) services was created. The world was in balance.
Then, we got AI, a technology which will change humanity. A technology which will understand us better than we understand ourselves. The makers unleashed a new army of bots to siphon any and all content and trained their models so they know everything. One of many side effects is that there is no longer need for a search engine, since these models became something better - the answer engines. With a voice of authority, they answer any question with confidence and for good measure some even include a link to the original source for a reference. But once you read the answer, why would you ever click on the link?
That broke the promise. AI companies are getting - or better say, taking - something and giving back nothing. Actually, on top of the free content, the AI companies are getting back something else - they are getting sued by media companies.
A simple search on Google - 'traffic to websites is plummeting because of ai' returns the following answer
'Yes, some websites are experiencing plummeting traffic due to AI, primarily Google's AI Overviews, which provide direct answers, reducing the need for users to visit external sites. Publishers, especially news sites, have reported significant traffic declines, with some losing over 70%. This shift is causing a widespread disruption for content creators who rely on organic search traffic for revenue, as AI-generated summaries now dominate search results, pushing original content lower on the page.'
with the following links next to the answer:
Google AI Search Shift Leaves Website Makers Feeling ‘Betrayed’ - Bloomberg
News Sites Are Getting Crushed by Google’s New AI Tools - WSJ
The answer is all you need to know, especially when both articles are behind a paywall.
It is an existential problem for all the content producers and it also affects the SEO cottage industry. They too are fighting for survival since it is harder and harder to justify any money spent on something which is no longer relevant. These companies are now pivoting to AI SEO. Optimizing the content for the AI.
What is AI SEO? I am glad you asked. Why would I even bother to come up with the answer, when Google summarized it for me:
'AI SEO uses artificial intelligence to improve a website's visibility and ranking in search engines by enhancing keyword research, content creation, technical analysis, and performance monitoring. It leverages technologies like machine learning and natural language processing to understand user intent and search engine algorithms, enabling more effective, predictive, and adaptive strategies compared to traditional reactive SEO methods. Key aspects include generating AI-optimized content, discovering high-volume keywords, and tracking performance in AI-powered search results.'
Yes, it is a nonsense. Sounds very sophisticated but it is just a pile of garbage glued together.
There is more but one statement really underscores the lunacy of it:
'Content Optimization: AI tools can help generate high-quality, relevant, and engaging content that aligns with search intent and ranking goals, improving organic visibility.'
Here we have a technology, which will produce content, which will then use that content for its training, which will then create content .... Now you tell me how, from this loop quickly spiraling down to averageness, we get PhD level ideas.
From the same query, we also learn that AI SEO tools will provide 'Content Optimization: AI tools can help generate high-quality, relevant, and engaging content that aligns with search intent and ranking goals, improving organic visibility.'
And that in turn will lead to 'Increased Efficiency: Automates tedious tasks like keyword research and content generation, freeing up SEO professionals to focus on strategy.'
Who can argue with a sentence which includes the word 'strategy'?
Just to top it up, to make sure that these AI SEO companies get to the top to grab your attention, they use ads to beat any 'organic' results with these promises 'Create SEO-optimized content fast with AI trained on ranking strategies and brand voice.'
#chokingonirony
The recurrent pattern? AI, as any other technology, can provide a temporary competitive advantage. Because the technology is available to everyone it is no longer anything special. It can increase efficiency, but doesn't fix any strategy flaws. It will get you faster to being average. To be better, you have to be smarter. It is you, all you.