AI summaries cause ‘devastating’ drop in audiences, online news media told

[rcjordan]

https://www.theguardian.com/technology/2025/jul/24/ai-summaries-causing-devastating-drop-in-online-news-audiences-study-finds

[ergophobe]

I've mentioned that Google Photos basically destroyed the business of some photographers I know. These are not people who sell ads or depend on pageviews per se. They sell fine-art prints. When Google switched from showing tiny thumbnails to showing a large enough picture that you could see roughly what was there, traffic collapsed.

This is the word equivalent of that except worse. Most people would rather have the high-res image but are willing to make do with the low-res version. But I think most people would rather have the summary and are willing to wade through the "high-res" version of the article if they have to.

[rcjordan]

I've been meaning to post about this for a month now, but can't quite get it into words. Let me try to poorly sum up my thinking.

We know G's history of crowding the top-o-the-serps with their assets --maps, local businesses, etc.  This is the ultimate crowd-out.  You provide the content to create the serp (and their ads) but no need to click through ...G AI just did everything in the background and shows you their summary (and their ads). Most users will see no reason to go beyond that.

[ergophobe]

>> ultimate crowd-out

I think so in the sense that it has the potential to create an answer engine that never has to bleed traffic.

Let's assume Google's ultimate goal is to never have someone leave Google except by clicking a revenue-generating link.

They can't achieve this, because in the end, all content sites will block Googlebot if Google simply steals the content and never sends traffic.

Since they have to allow some unpaid leakage from their site, that in the end limits their ability to provide summaries that perfectly answer questions, but assuming their AI continues to improve, that's their main limitation.

This process of enshittification can't continue indefinitely, but experience has shown us that it can continue a lot longer than we think.

I know I've said this several times, but I thought 30 years ago that universities as we knew them would be radically changed by 2025 because they couldn't keep enshittifying (sadly, I did not have that word) their product and charge top dollar for another 30 years and yet, in the US, everything that was bad in 1995 is much worse today and costs have outpaced inflation.

So I keep thinking that at a certain point major, important sites will simply block Google.

But thus far, there is no obvious alternative to traditional universities or Google.

And that is also why I do not make big bets based on my ability to foretell the future. Experience has shown that isn't very good either.

[Brad]

My websites are neither large nor important but I could block Googlebot and not really miss the traffic because a lot of Google traffic is of poor quality (they hit one page and leave.)  Some few explore deeper.  Now I have no advertising nor am I selling anything so I don't care except that Googlebot hangs around a lot and the traffic does not seem worth it for the bandwidth they use and linking info they gain.

I get much more traffic from Bing based engines (DDG, Yahoo, etc. and some from Bing) and I have no idea how that came about.  And I probably get more traffic from Mojeek, Marginalia and Wibey combined than Google.  *shrugs*

So I'm thinking a lot of Indieweb content sites would block Google and not look back.  Some already have but they are hardcore anti-Big Tech.

[ergophobe]

The thing is, the Indieweb sites could block Google and Google wouldn't care.

The other thing is, so many Indieweb sites are barely maintained or maintained by someone who doesn't know what a robots.txt or a meta noindex is. Or, for that matter, a Wordpress upgrade.

I think for it to hurt Google, it needs to be big sites that a lot of people want to visit.

[Brad]

Of course it's not going to hurt Google.  Google is doing a darn fine job of killing itself without any help.

[ergophobe]

Google is doing a darn fine job of killing itself without any help.

We may not *like* how Google has evolved, but that is different than "killing itself" - a company with only two quarters of negative revenue growth in 15 years looks pretty far from needing to call the priest for Last Rites

The problem for me is not so much that Google is killing itself, but just how rewarding enshittification is for Google, Amazon, etc.

[ergophobe]

>> killing itself
>> far from needing to call the priest

And just today Cory Doctorow has a riff on exactly that tension. Nothing new, really, but it summarizes why the company is going to go as close to killing the golden goose and then, maybe, if they have good leadership, pull back, but just enough to keep from dying. Not enough to ever make it the product you once loved.

https://pluralistic.net/2025/07/28/twiddlehazard/