www.Tobold.com
Tobold@GMail.com
A blog with my thoughts regarding games I am playing and other stuff in life. Please read my Terms of Service
RSS Feed
Thursday, July 17, 2025
Death of the link
I like to keep myself informed about what is going on in the world, not just locally, but globally. I use sites like Google News a lot. However, my user experience over the years has deteriorated. These days, more often than not, if I click on a link to a news story on Google News, I hit a paywall. Now if that was a Google News paywall, and I could pay some amount of monthly subscription to read all the news I wanted, I might actually do that. But instead it is a thousand different paywalls from a thousand different news organizations. First of all, it would be prohibitively expensive to pay for all of them, or even just a decent selection. But second, every news organization has its own bias, and by subscribing to one and not to another, I'd be locked into that specific bias. I prefer to read different views of the same story, which means different news organizations, which means different paywalls. So the links from Google News are becoming decreasingly useful to me.
At the same time, a lot of my other use of Google is asking Google a specific question. And the main Google search site has changed in the last years too: Google now very often gives me an overview powered by AI, and very often that overview already answers my question. Yes, this isn't 100% reliable and sometimes the AI answers are just plain wrong. But especially in the domain of questions about video games, following the link to the article is a lot worse: SEO-optimized web pages that bury the answer to your question somewhere near the bottom of the page, so you scroll through all the advertising. The AI overview extracts the actual short answer I need, and is a lot more useful. It is only a question of time that Google News will display an AI overview of all news stories, and I won't have to follow links there anymore either.
Of course this follows the course of events that Cory Doctorow described as enshittification: First a platform like Google News offers a great service for free and lots of people sign up; second the service becomes less good for the users, in order to enable business partners to make money from those links; and finally Google decides to screw the business partners too, and just steals their content for their AI overviews without sending customers their way. AI has the potential to break the fundamental business proposition of search engines and the link.
I don't think that AI bears the sole responsibility for this. It isn't AI's fault that if you search today for "best product" in any given category, you'll only see ads and sponsored content, and honest reviews are increasingly hard to find. Google doesn't give an AI overview if you ask a question like "What is the best 42 inch OLED TV?", just advertising and links to articles that are clearly sponsored content. Just for giggles I asked the question to ChatGPT, but as ChatGPT uses the internet as a source, it also comes back with an answer that recommends the brands that have spent the most amount on ads and sponsored content.
But through AI we are clearly in the middle of a major development of how the internet fundamentally works. And while right now AI is giving free answers that are sometimes more useful than links, I don't think this is going to last forever. News organizations and other websites are either going to find ways to block AI scrapers, or are going to come up with new technology to create AI paywalls, so that AI companies have to pay something to use the content. Right now the AI companies have the upper hand, and content creators are somewhat flailing with ineffective lawsuits, but anti-AI technology for content protection is sure to arrive.
- posted by Tobold @ 9:31 AM Permanent Link
Comments:
<< Home
Newer› ‹Older
I think that enshittification is bullshit bingo as well.
The fundamental reality is that value needs to be compensated and both value and compensation depend on the user in question.
A platform that is "good" and attracts users is because the value provided is manifold higher than the compensation.
If you create a news aggregator for yourself that crawls the net and collects news, that is great value as you don't need to do it yourself.
It is also great value for a lot of other people and at some point, your costs for running the server will reach a point where you want to capitalise on the value.
Do you charge your users? That's tricky online when the users are used to a free service and can just jump to the next platform that is still free.
You can run ads, but those will pay less and less.
You can bill the news sites for favoured ranking. That will work for as long as the sites are generating a surplus from the traffic you drive.
At some point the sites will turn to paywalls because ads no longer pay and other pages just scrape their content.
Now driving traffic no longer works because people don't want to click paywalled links.
So to stay profitable, you have to find a new way of providing the content to your users.
That isn't enshittification but companies realising that there is more value to their products when an aggregator is able to turn a profit and subsequently shifting their approach.
You can see the same with Netflix. Big studios were selling their movies and made Netflix big because they didn't think streaming would work that well.
Then it worked well and studios wanted either more compensation or turned to their own streaming service.
Yes, the result is that the service for the user deteriorates - but that isn't because the provider wants to ramp profits but because digital overhead is barely existing. There isn't enough margin in providing an aggregator when you can spin up your own version.
If you compare it with regular work were you produce a thing and sell it, you need a place to sell it, you need to transport it to the place, you need someone selling it, you need ..., So there is a reasonable calculation to give up on some value of your product in favour of not doing all these things, selling wholesale to someone who will squeeze the margin.
But you can't squeeze bits and bytes. Copying a link to your page or streaming a movie doesn't come with enough overhead that justifies you turning a large profit. So companies will adjust their way and your service suffers.
As for "best product": that is the result of human optimisation.
Nobody wants a random product if they can get the best.
So they search for the best.
Thus sites advertise as the best to drive eyeballs to their ads or affiliate links.
Thus every site showcases the best.
Thus companies that want to sell, will sponsor content that shows them as the best.
Thus every product is highlighted as the best.
Thus "best" loses its value as a signifier.
AI isn't going to change the Internet. As you said, AI is just another aggregator in a world where there aren't enough hurdles and thus no cost to aggregation.
The fundamental reality is that value needs to be compensated and both value and compensation depend on the user in question.
A platform that is "good" and attracts users is because the value provided is manifold higher than the compensation.
If you create a news aggregator for yourself that crawls the net and collects news, that is great value as you don't need to do it yourself.
It is also great value for a lot of other people and at some point, your costs for running the server will reach a point where you want to capitalise on the value.
Do you charge your users? That's tricky online when the users are used to a free service and can just jump to the next platform that is still free.
You can run ads, but those will pay less and less.
You can bill the news sites for favoured ranking. That will work for as long as the sites are generating a surplus from the traffic you drive.
At some point the sites will turn to paywalls because ads no longer pay and other pages just scrape their content.
Now driving traffic no longer works because people don't want to click paywalled links.
So to stay profitable, you have to find a new way of providing the content to your users.
That isn't enshittification but companies realising that there is more value to their products when an aggregator is able to turn a profit and subsequently shifting their approach.
You can see the same with Netflix. Big studios were selling their movies and made Netflix big because they didn't think streaming would work that well.
Then it worked well and studios wanted either more compensation or turned to their own streaming service.
Yes, the result is that the service for the user deteriorates - but that isn't because the provider wants to ramp profits but because digital overhead is barely existing. There isn't enough margin in providing an aggregator when you can spin up your own version.
If you compare it with regular work were you produce a thing and sell it, you need a place to sell it, you need to transport it to the place, you need someone selling it, you need ..., So there is a reasonable calculation to give up on some value of your product in favour of not doing all these things, selling wholesale to someone who will squeeze the margin.
But you can't squeeze bits and bytes. Copying a link to your page or streaming a movie doesn't come with enough overhead that justifies you turning a large profit. So companies will adjust their way and your service suffers.
As for "best product": that is the result of human optimisation.
Nobody wants a random product if they can get the best.
So they search for the best.
Thus sites advertise as the best to drive eyeballs to their ads or affiliate links.
Thus every site showcases the best.
Thus companies that want to sell, will sponsor content that shows them as the best.
Thus every product is highlighted as the best.
Thus "best" loses its value as a signifier.
AI isn't going to change the Internet. As you said, AI is just another aggregator in a world where there aren't enough hurdles and thus no cost to aggregation.
Good points, it's easy to blame greed when it's more about free services being unprofitable in the long run.
Post a Comment
<< Home
