AI, Google, and the Death of the Web June 27, 2024 by Paul Byrne web AI How and when will the world wide web die? One of the first terms I learned when I became involved in ecommerce was SEO: Search Engine Optimization. This was Google’s first toxic gift to the world wide web. Google kept updating its ranking algorithms to the dismay of SEO ‘experts’ who kept having to update their strategies. “There are no solutions, only trade-offs.” – Thomas Sowell An entire class of SEO hackers continually outperformed good content with sneakiness while ironically, Google — through Matt Cutts — told people that the best way to make it to the top of Google’s search results was through good writing, the production of rich and engaging content (especially images and video), and good structure During the 2010’s, the war was won. Google’s strategy of ranking pages based on how often a search term appeared had significant unintended consequences. Google, whose stated goal was to improve the internet, was actually contributing to its decline. Blog content became unreadable as writers stuffed their pages with pointless repetitions of the same word. Massive lists of keywords were added and hidden with styling. Top-ranking articles were often the least readable. (Have you ever seen a top-ranking recipe page that didn’t annoy you with redundancies?) But, being the world’s foremost collection of overpaid math majors, Google would not be outdone. They started using the number of times other sites linked to your page as an important factor as a proxy for ‘authority.’ In response, SEO alchemists soon began creating link clouds. Entire hosting companies sprung up just to host fake websites to link to yours, making it harder to trace the source. Google then invented the ‘big hammer’. If you were considered to be cheating the system, your site would be de-indexed, potentially getting no traffic. Each change of the algorithm led to more banishings. I had multiple ecommerce clients who complained that if they didn’t play the game, the SEO ‘scammers’ would get all the traffic. Once they started getting some of the traffic, Google would put them in the penalty box and they would lose millions of dollars in revenue overnight. As Google relied more and more on authority, they were able to contain some of the spammy content issues. However, the same sites kept showing up over and over again, and they often had a specific ideological bias that aligned with Google’s Ivy League educated business people. The battle waged for an entire decade. I’m not a Google insider, so I don’t know when or how it happened, but I’m sure they threw in the towel and gave up the first page to sponsored content. Not that Google had become completely irrelevant, it just had moved from ‘organizing the world’s information’ to ignoring most of it. Google began spying on our digital lives, keeping millions of pages of data about us. Every time we searched for something, used our Android phone, visited a site with Google Analytics tracking on it (85% of sites you visit), used Google Maps, or just downloaded a Google app of some kind to our phone, Google’s huge vacuum cleaner siphoned up our data. Google protests that it doesn’t sell our data, but that’s really semantics. They ‘trade’ it with Facebook and other privacy pirates, they sell meta-data in bulk (which has been proven to reveal exactly who we are despite our name not being on it), they lease it to train AI models, and so forth. Google is notoriously cagey about the way they report revenue from the sale of your data, reporting it in the ‘advertising and other category in annual reports.’ (Compliments to ChatGPT for the illustration) Whatever, Google! Your business would completely collapse if you stopped violating our privacy. Call it what you want. From here on out So, let’s come to the reality (or irreality) of artificial intelligence. When we talk about AI today, we’re usually referring to LLMs: large language models. They are great at creating large volumes of fluffy, inaccurate content search engines love to float up. At the same time, they are really good at summarizing large volumes of information and prioritizing the topics for us. So, in Gmail-like fashion, we’ll soon consume the majority of our content via AI-curation. We’re already doing it, though. We have algorithms on Facebook and YouTube. In online meeting software, we now have AI bots taking notes and summarizing content. It won’t be long before they start notifying us of important content. But can we trust these AI summaries as much as we would have trusted an executive summary written by a human in attendance? If this is something you are fearful of, we have good news and bad news for you… but please decide for yourself after reading both of these other articles on the subject: The Rational and Irrational Fears of Artificial Intelligence and GPT4 is a f@&ing LIAR. Additionally, we are using AI to automate our responses to the flood of communication coming in. Let’s walk through the lifecycle of a piece of modern content created by AI: AI creates a piece of content AI distributes the content online via social media, email, and other messaging platforms Receiving AIs identify content we might be interested in and summarize it AIs generate response content –> back to step 1 When the human is no longer part of the information creation -> curation -> consumption -> response cycle, it all becomes completely pointless. Look at it from Google’s perspective. Your AI searches for information on Google, who serves up advertisements to your AI. Since your AI can’t make buying decisions, what is the point of that ad? The advertiser doesn’t want to reach your AI bot. It becomes computers trading data for the sake of trading data. AND THAT WILL BE THE DEATH OF THE WORLD WIDE WEB.