I think it's only just clicking for me that the death of the web is upon us. Instead of "googling" for stuff folks are searching on tiktok, instead of indexable forums there's closed discord communities. In fact apps and services are becoming locked within proprietary apps like discord. Paired with "ai" language models creating infinite web spam the whole thing is becoming less useful for the average person.
@mauve Yes. I believe that it's wider, a collapse of the current noosphere. Defensive exclusion by niche networks (physical and by protocol) will be the best we have until our disinfo immune systems ramp up.
With SEO and Google's ever-increasing enshittification of search results, increasingly, "well-behaved web pages" means "content optimised entirely for Google's walled garden."
Once it was an open standard, and optimising for that standard meant Google (and other browsers, and other everything) would handle your content well.
But we've passed that point. We build microformats for Google search into our sites. Google Enhanced Safe Browsing (et al) prevents customising or ad-blocking.
This is also why e.g. the horrible ecosystem of recipe sites is what it is: the most useful content has long ago stopped being what you find through Google. But Google needs a captive audience of regular ad-fee-payers, which in turn is who gets featured.
Much like Amazon only posting products from those vendors who pay, Google is simply selling your eyeballs directly.
That process of capture is basically complete, is what I'm saying, and has been for some time now.
@codefolio @mauve how do we get out of this?
There are folks who will tell you to stop using Google properties -- e.g. use DuckDuckGo, get off GMail. You'll notice that while they're ad-supported, competition *does* really help this.
But fundamentally it's a really hard problem because the advertisers (often the same as content producers) are a really important group here too. Google offers good tracking for those selling things.
But switching off Google is still good, if you can.
But, short version, if something is doing massive damage in the world, maybe try to use it less and fund it less.
LLMs are going to force our hand here, though. Old-style search is dead, it just doesn't know it yet. Instead, we'll get even-more-rapid centralisation and enshittification in a vicious new cycle.
If we're quick, old-Yahoo-style indices may be the answer, combined with vigorously rooting out LLM-generated content.
But it's hard, right? Everybody wants to be the person making a truckload of money on locking people into the new walled garden, very much including the incumbents.
Everybody wants to be the ones offering a small *taste* of "all human knowledge at your fingertips" and finding ways to not deliver, because it's far more profitable if you nickel-and-dime it, play it out a little at a time, at low quality.
The 90s Internet isn't coming back. Ad companies bought and killed it.
And when you try to look at maybe even *possibly* doing a non-ad-supported version, everybody complains that it wouldn't be free (which it wouldn't be.)
The current cost is just, y'know, removing the big benefit that we all wanted so much to have for free. Instead, a lot of it's gone and the rest is going.
But people *don't* want to pay for it, so... 🤷♂️
people dont have to pay for it.
peer-to-peer is the answer.
people create the value in the first place.
in peer-to-peer they can share it with each other.
Peer to peer doesn't fix the "people have to pay for it" problem. If you want always-up servers, people's random laptops (etc) aren't the answer.
You could imagine a setup where the costs were distributed differently and everything was reliable without a few big servers, but no such thing currently exists. Mastodon certainly isn't that.
I contribute to a Patreon so somebody else can run ruby.social, but it's very not-free. Especially in labour.
search is a different issue.
but i guess peers could crawl.
until there is p2p infrastructure with data it is hard to start experimenting. it is essentially a trust problem.
do you crawl and compute it all by yourself?
expensive.
do you let somebody or many do it and trust results?
it is a spectrum
true, but 1. many people have devixes which are almost alwqys online these days. 2. in many apps there is an option to have your friends or interested parties replicate your data to keep it available whipe you are not. 3. it makes it trivial to self host or use any number of small hosts, the more you use, the more resilient - having more hosts in parallel makes it better.
this is not possibpe at all with mastodon style federation or centralized
not *everybody* wants it, but it only needs a few who want it and manage to enclose quality content and pollute whats free.
it isnt a tragedy of the commons, it is a few players deliberately destroying thr commons for their own profit
i dont think this works.
it only needs a few dedicated people to pollute the entire web with bots and they become better and soon no captcha will save us.
@mauve yes. it is an attack by capitalists to enclose thebopen web and it is working.
It also leads to all new information being exclusively available to owners of the surveillance capitalists AI and also no useful information is left for anyone else to train models on the now totally polluted web.
it is really bad.
@mauve infinite web spam, thanks for giving a term to that feeling