top of page
Writer's pictureRay Alner

AI Will Cause an Informational Dark Age

Ads Run the Internet

Ads. The bane of our modern capitalistic monstrosity we’ve created for ourselves. It’s not only annoying but also frustrating when you’re trying to read an article and assaulted by so many ads you can read maybe a sentence at a time at best before having your attention grabbed by some video about why you need a $30 deodorant.

But we can’t ignore them. They are the necessary evil of the free information we want when we are scouring the internet for some obscure fact we only care about for a fleeting moment. Or when we are looking for the “best wallet” amongst the other myriad of knockoffs we can purchase.

AI Shakes Up the Internet

AI. The wonderful new product that is supposed to make everyone more productive and allow for information transfer that surpasses any known previous innovation, while skirting the ads we all hate. It is allowing for companies to lay off huge swaths of people and become more “productive” with AI, even though it may just mean belching generated content to grab the algorithm.

We can’t ignore the impact of AI, generative AI, or the myriad of new tools that is supposed to make us productive powerhouses. Just like the way “the book” generation had to search for information in books at libraries, “the Google” generation had to search for information in a curated list of links, and finally, “the AI” generation will have information served to them in a procedurally generated method that could eventually fit the way that user best digests that information.

Here’s the problem. Unless there’s a revolution on how information is monetized, “the AI” generation will cause an information dark age that will be based on an echo chamber of other AI generated responses.

Oops… AI Shakes it Up Too Much

We’ve started to see this happen already. An article published by The BBC discusses this issue. One company, housefresh.com, was obliterated by the changes Google did to their search algorithm, favoring content from bigger lifestyle magazines. Because Google’s algorithms are proprietary, there is nothing easy these smaller content creators can do to improve their rankings, causing them to make hard changes to stay afloat. In this case, housefresh.com released much of their staff that spent time testing this equipment, because a company that could figure the keywords out better, NOT test the equipment better got ranked higher.

As AI continues to permeate every aspect of our lives, there's a risk that the rich, unbiased information currently used to feed these AI systems will become scarce. Instead, we may end up with AI responses that are largely based on the information from biased blogs or articles that are sponsored or owned by product companies, promoting their product as the "#1 hair ties" rather than tested and verified products reviews.

Now, there are some efforts to improve the data flow from opinion pieces, like the Reddit & OpenAI partnership, but it’s highly likely the AI will get a sense of entitlement and opinion, rather than facts or impartial testing that support a product, or idea.

I’m afraid that as AI becomes more ubiquitous, the thing that pisses people off the most, but the necessary evil that keeps the lights on for independent unbiased reviews on products shutter because generative AI that scours the internet will not provide the revenue from ads like a real visitor to the site does.

How we can Fix It

Now, there are ways we can solve this, heres a few ideas that may kickstart a conversation to monetize sites:

  1. Add advertising to generated responses that is returned to the companies or websites that the information was scraped. I know this a “bleh” option. I myself like using ChatGPT because of its lack of advertising, but hey, gotta make money somehow.

  2. Do a music payment model with independent sites receiving a cut of either ad revenue or subscription revenue generated by the responses, or direct click links through the generated responses if someone finds the information provided helpful.

  3. Allow companies to create the same agreements that ChatGPT has with Reddit to provide information via API from well researched, well documented experts, providing better responses from the AI rather than opinion pieces likely scraped AI generated systems.

  4. Provide a quality score on the information that can be adjusted based on the quality requirement you need for your response - like Daniel Miessler’s Fabric tool. An addin like this could cause junk responses to be filtered out, requiring higher quality information coming from AI, maybe tipping the scale to require quality information.

Just like the music industry evolved as new technology and ideas allowed for them to stay in business, I’m hoping that the generative AI industry will allow for the rest of the industry to evolve before it gets choked in its own vomit of generated echo chambers.

9 views0 comments

Comments


bottom of page