If you web traffic is drying up because of generative-AI content scrappers, here's how to get even
Hasten the coming of model collapse
Let me go straight to the punch: the above video sums up what is happening to the Internet. In this video, an unhealthy, flabby man is getting his nutrition by connecting a tube from his stomach to his mouth. It is obvious that this is not healthy, and his body will eventually consume itself.
This image is analogous to what is happening to the Internet today.
Websites are dying
Take a look at this article from 404 Media,
… new research from the internet infrastructure company CloudFlare suggests that OpenAI is crawling 1,500 individual webpages for every one visitor that it is sending to a website. Google traffic has begun to dry up as both Google’s own AI snippets and AI-powered SEO spam have obliterated the business models of many media websites.
This general dynamic—plummeting traffic because of AI snippets, ChatGPT, AI slop, Twitter no workie so good no more—has been called the “traffic apocalypse” and has all but killed some smaller websites and has been blamed by executives for hundreds of layoffs at larger ones.
To give you an example, of this, I just did a Google search on “AI model collapse" (a topic particularly relevant to this article). This is what I saw:
To the left of the search results are Google’s AI Overview. To the right are the websites that Google’s AI Overview got its information from.
In the past, Google’s search results show the right side. Then the user would have to click on the links on the search result to go to the website that provided the information.
Today, with AI Summaries, there is no need for the user to visit the websites that provided the information. They can simply just read from the AI Overview and move on.
The implications of Google’s AI Overview is dire for websites. It means traffic is drying up. As this NYMag article reported,
It all amounts to a kind of traffic apocalypse in which it seems all spigots for traffic are being turned off, affecting news organizations big and small, new and old. It hurts outlets heavily reliant on digital advertising but also those that draw revenue from product recommendations and subscriptions. The whole premise of internet publishing — that you could reach audiences far and wide — is starting to crumble, forcing publishers to reevaluate what kind of stories they produce and what kind of readers they want — and, ultimately, to think smaller and more bespoke.
And note the bold in my above quote:
"think smaller”
This means that a lot of Big Media are going to become small media. Many businesses that publishes content on their websites are going to disappear. Those who survive will become much smaller. For example, in May, Business Inside laid off 21 percent of its workforce.
Dying businesses doubling down on AI
Big Media and Big Websites are finding that they are trapped in deep holes. You would think that the most sensible option is to stop digging further. But instead, that is exactly what they are doing. As this 404 Media article reported,
Despite the fact that generative AI has been a destructive force against their businesses, their industry, and the truth more broadly, media executives still see AI as a business opportunity and a shiny object that they can tell investors and their staffs that they are very bullish on. They have to say this, I guess, because everything else they have tried hasn’t worked, and pretending that they are forward thinking or have any clue what they are doing will perhaps allow a specific type of media executive to squeeze out a few more months of salary.
Big Media are doubling down on AI by adopting the AI-first ‘strategy’. Basically, one of the objectives (among many other objectives) of this strategy requires their employees to use AI to further crank out even more content. Their rationale is that using AI will increase productivity, and therefore, improve profitability, which in turn is the key to saving their businesses from oblivion.
But using AI to churn out more content is precisely the wrong strategy to pursue!
AI model collapse
When more and more of the Internet are resorting to using generative-AI to mass produce content, than those AI tools will train their output on synthetic AI-generated content. This forms a positive-feedback loop- people use generative-AI to create content, which in turn is fed into training AI, which in turn is used to create even more AI-generated content.
There are credible research that as this positive-feedback loop continues will degrate in performance over successive generations. Eventaully, you will get what is called “model collapse”, where AI is producing unreliable, untrustworthy and poor-quality content.
You can do a Google search on “AI model collapse” and its AI Overview will give you a nice summary of what it is (I understand the irony here).
Hasten the coming AI model collapse
Like it or not, this is the direction that the world is going. It is going at neck-break speed for the AI-first ‘strategy’. I doubt any of us individually can do anything to stop it.
But what we can individually do is to hasten the arrival of AI model collapse. How do you do that?
By all means, use AI to generate content to be fed back into Big Tech’s AI training machines. Go all out to do that. This will give you a temporary edge in ‘productivity’. If you are skilled at using AI to generate content (text, images, videos, music, etc), sell your AI skills to the highest bidder while you still can.
One day, when AI model collapse happen, we are going to have yet another AI winter (there already are quite a few in the past, for those in the academic circles). This is when the world will learn the lesson of its folly. And unfortunately, the world has to learn it the hard way.
You can hasten the day when the world will learn it the hard way by going all out to spew out AI-generated content to be fed back into Big Tech’s AI machine learning algorithms.
Bring it on.
Note: the video above is generated by AI. Unfortunately, this article is still hand-written. I am not doing enough to hasten the coming AI model collapse.