RE: https://tldr.nettime.org/@tante/116605858023186072

? Offline

@joe @inthehands is there a coordinated effort that has a website? And/or server plugins that automate serving coordinated poison?

? Offline

@inthehands this is a fence-post defense against this, google Will Not Care

just start poisoning the data once you detect that google is the one fetching it, just absolutely fucking destroy their LLM output

? Offline

@inthehands also probably worth it to submit a pagemaster/webmaster request to them to directly tell them to deindex your site. Also DMCA takedowns to Google are usually effective. If you're in the jurisdiction of Australia you're potentially able to go after them iirc. (The Australian government went after them for embedding news articles in their output or something)

qurlyjoe@mstdn.social

@inthehands
What guarantee does one have that Google will abide by these restrictions?

? Offline

@mjd (thanks- this made me write down some of the thoughts I've been kicking around for a bit!)

"publish freely and for free" is also my site's situation, but it is a privileged one.

My mom's website is a funnel for her books, courses, services, etc. It is useful for her to publish helpful things for people to learn from; and to have those indexed in search; so that when people look for an answer to a question, they find out about her, and may wind up giving her money in the future.

Someone *can* scrape what she writes and re-post it, sure. And that's (in my opinion) a bad thing to do, because now the person who has done the work doesn't get any benefits from it: reputation, income, whatever. People *can* do that- SEO farms, image-sharing sites, whatnot- but as long as there is enough traffic going to the author/creator, there's still a viable business in "write for the web". (There is at least some legal consensus around this idea in the form of copyright law- that an author should be compensated for their work.)

What happens to that business model if "search" leads, not to the author's site, but to a hologram of it? Without providing the author any benefit, in citation (reputation), ad revenue, or sales funnel?

Some authors (like you, maybe? and like me?) will still write and publish. Wikipedia editors will still edit! The web doesn't fully disappear.

But...the incentives to provide that information *professionally* are degraded; the income has to come from somewhere else. And there's a lot of bad ideas that have better funding than some good ones.

If Anthropic scrapes the Crochet Answer Book (they did) and answers questions from it (it does), then who is going to write the next edition? Are Anthropic, Google, OpenAI going to pay for the one copy each they need to buy? Or is it necessarily the effort of someone who is not fairly paid for it?

If we get fully automated luxury space communism, fine, share everything for free; but, in the world we're in-- I'm worried that there's no longer a *financial* incentive to share reliable information openly. So some of those sources will dry up.

? Offline

@inthehands @joe gotta update my robots.txt real quick

? Offline

@inthehands IMO anything besides active denial measures (proof of work e.g. Anubis) is pointless. These companies don’t care about consent, and if they do that care is completely predicated on a few important nerds being in positions of power. That won’t last forever, especially in an environment where a larger and larger number of software engineers are people cashing a check rather than people who have a values based commitment to an open internet.

? Offline

@ShadSterling @mjd @cceckman yeah fair, I only commented because this is one place the distinction matters in that a social contract exists in aggregate as a set of expectations regardless of what an individual might expect or feel like they agreed to

? Offline

OK, a •lot• of replies need this reponse:

Yes, of •course• they will start ignoring robots.txt etc as soon as they think it hurts their business. Of course.

It is important to •force that fight•, rather than just capitulating in advance.

? Offline

Defeatism is a form of surrender. Cynicism is surrender. Despair is surrender. Nihilism is surrender.

Our job is to •care• and to •keep caring• and to •keep doing and keep building• and to •endure• longer than them.

? Offline

@ShadSterling @inthehands i don't know if there's a coordinated movement. there are prefab tools like https://lib.rs/crates/iocaine that are relatively easy to deploy, though i imagine they also lose some of their effectiveness as they become more popular and LLM providers start to counter them

? Offline

@inthehands @iwein HEAR HEAR!!

? Offline

@inthehands One of the things I've done recently is to bring enforcing robots.txt within my webserver engine. The /robots.txt itself still exists; the vast majority of it is a list of bots that are `Disallow: /` .

I still get a few of these bots attempting to hit the site, so it's definitely doing something.

? Offline

@inthehands you can block their bots at the network level

? Offline

@joe @ShadSterling
I share Joe’s concern that poison-in-box systems will become detectable, but they seem like a good place to start.

I’m even more a fan of bespoke one-off poison generators for those of us who have the means to write them. Both/and.

? Offline

@inthehands I guess “Search Engine Optimisation” has gained a new, contrary meaning.

? Offline

@inthehands @ShadSterling if you can live with exposing yourself to the demon core a bit, using a locally hosted LLM to generate the poison also seems like a good way to get a lot of hard-to-detect variability, and hopefully also slightly accelerate model collapse in the process

? Offline

@letterror

“search engine nopetimization”

? Offline

@inthehands Hey, I'm curious if you have an opinion on what that contract *should* say? I don't think "pay with traffic" has worked for a long time. What would?