How do you even prevent your personal website being scraped for ai training?
 It's All Fucked Shirt $22.14 |
 |
It's All Fucked Shirt $22.14 |
How do you even prevent your personal website being scraped for ai training?
How do you even prevent your personal website being scraped for ai training?
It's All Fucked Shirt $22.14 |
|
It's All Fucked Shirt $22.14 |
https://glaze.cs.uchicago.edu/
At least poisons the AI Image algorithms
>AI Image algorithms
how do i prevent robots from scraping my personal blog?
https://sizeof.cat/post/block-chatgpt-scraping/
No solid proof of this working beyond cherry picked graphs so it's likely snake oil.
Make accounts mandatory? Bots have a much harder time scraping if there's a strict API it has to bypass
don't put anything online. print flyers of your profile or whatever you want to show off to the world and throw em off of high buildings
> they take a picture
>it's on X (formerly twitter) now
>be BOT
>spend entire life harping against intellectual property and copyrights
>complain that AI is stealing your shit
You can't make this shit up. I mean, read the What is Glaze page:
https://glaze.cs.uchicago.edu/what-is-glaze.html
>Style mimicry produces a number of harmful outcomes that may not be obvious at first glance. For artists whose styles are intentionally copied, not only do they see loss in commissions and basic income, but low quality synthetic copies scattered online dilute their brand and reputation.
THESE ARE THE SAME SHITTY ARGUMENTS THE RIAA USED DECADES AGO AGAINST "PIRACY"!!!
The creators of this shit should have their hands cut off then be given blood thinners and hung upside down.
>drawgays are BOT
i cant help but notice a certain inconsistency in your approach
>Seeing the artistic style they worked years to develop taken to create content without their consent or compensation is akin to identity theft.
so stupid
its kinda true.
but i still going to agree with you bc frick drawgays.
theyre all fricking pedos drawing hentai anyways
>its kinda true.
How so? Just being inspired by a style isn't really the same as outright plagarism
the style was learned from the creator of it, without his consent.
if i'd be a drawgay, or if cp wasnt their no1 product i would be angry about that.
but since thats not the case-
kek. frick em. tell em to learn to weld
>the style was learned from the creator of it, without his consent.
oi you got a loisence for that style?
exactly. thats a good fricking question.
i think i am feeling english today. they might be onto something.
All artists learn from other artists to at least some extent.
if that were true, there would be no one leading and in making new techniques, which does happen
The point I was making is that everything is at least somewhat derivative
That fact that AI is easier is irrelevant in this regard
>That fact that AI is easier is irrelevant in this regard
i disagree.
the market does too.
when you look at handmade shit it is much more expensive than its mass produced counterpart.
the problem with ai imagery is that soon there will be no difference between ai generated art and the original.
i mean, i would stand for construction workers.
whose jobs will soon be taken over too.
so i would stand for drawgays. if they werent piece of shit subhumans, aka "artists"
context:
>the problem with ai imagery is that soon there will be no difference between ai generated art and the original.
Its literally copying bits, look at the "ghosts" in the images. Ann Grahm Lotz
The point I was trying to make is that both AI and humans copy styles to some extent and that artists should not claim exclusive ownership over a style. It is worth mentioning that styles have never been copyrightable.
actually... plagiarism is something thats well-defined in music.
i think drawgays got the short end of the stick in this matter bc lawmakes didnt find a way to quantify style in drawing, as they did for melody in music.
(cosmic justice i guess. roundhouse kick drawgays into the concrete)
*lawmakers
Copying music note-for-note is equivalent to copying part of a picture exactly. Musicians are allowed to be inspired by other musicians (i.e. genres)
ackshually...
its a matter of 5 notes and rythm iirc. in music its a matter of air, not even a whole song or style.
>style
thats 100% subjective too. when does style end and where does an air begin?
quick example: apocalyptica.
is it doom? is it metal? or is it classical with doom/metal airs?
music is created of discrete component parts. so an air could be defined, and a reasonable threshold could be drawn.
its not the case with drawing.
doesnt mean that the creative process is any less involved in that (when its not hentai)(which is 95% of art production nowadays)
>when does style end and where does an air begin?
I think judges and juries would be able to "know it when they see it"
kek
>I think judges and juries would be able to "know it when they see it"
>judge + jury all Black folk
>youre white
>not even blue hair
yeah but its a human thats the beneficiary and then the raw skill comes into play.
also thats something that earned, it comes with training and dedication.
with ai, the beneficiaries are corporate bc the software engineers are paid pennies on the dollar
and it "just happens automagically".
also disloyal competition with humans. if drawgays could classify as such.
Incredibly based post Anon. Artists that defend copyright and IP need to rope
>THESE ARE THE SAME SHITTY ARGUMENTS THE RIAA USED DECADES AGO AGAINST "PIRACY"!!!
This is the same board defending Musk brain implants because based and redpilled. They will never learn.
BOT isn't anti-copyright, its pro-do-what-i-want
ill pirate because frick israelites, and ill poison AI datasets because frick israelites
problem?
Every social media site is spammed by bots, and BOT is no different.
AI giants are pushing for AI regulations. These will keep small competitors out while Microsoft, Google, etc., will share the enormous profits that AI will surely bring.
I don't think there's a single real human bean in this board that is against AI because it "steals le images!!".
The problem isn't regulations, which we desperately need. The problem is regulatory capture.
I have the same problem, anon.
I have thought of poisoning the html with random chunks of non-rendered random text. For images, I guess downloading them obfuscated and drawing them into a canvas element with javascript will prevent bots from scraping them. it's definitely an arms race. good luck, anon.
They don't fricking care about *your* personal website, so you have nothing to worry about. I don't understand why people think data sets consist of randomly scraping the internet for anything and everything rather than having quality control and picking specific sources
nta
but its bc you gotta have something to label, right?
do you realize how much data goes into a generative model?
>I don't understand why people think data sets consist of randomly scraping the internet for anything and everything rather than having quality control and picking specific sources
that's what they do though. they use classifiers to throw out spam text so they can have minimal noise
>ken-watanabe.jpg
LET THEM SCRAPE
I've never seen a ai be better than a human at any creative thing.
Just at things that are solved, like chess. With imagery, its not as good either. There's nothing behind it.
#no-scrapesies
Actual answer,
Block their user agent.
Block IPs that download only HTML and won't download js/CSS/images (using fail2ban)
If someone requests 10 pages on my site without getting the other resources, they get a knuckle sandwich
>Block their user agent.
i LOVE spoofing my user agent
>Block IPs that download only HTML and won't download js/CSS/images (using fail2ban)
i LOVE following all i can find with wget -r
kys Black person
>How do you even prevent your personal website being scraped for ai training?
Why care? What's the issue?
Ask politely on robots.txt?
robots.txt
redirect all refers to meatspin or tubgirl
by hosting 'p and a wordlist of slurs.
ai scrapers are trained to avoid that.
>robots.txt
It's basically an honor system though, so results may vary.
>implying anything you post here isn't already being scraped for ai training
I don't post anything important here
So no I'm not implying that