Over 800 TAG members marched on Netflix & delivered a petition demanding the AMPTP keep #animation jobs #union, agree to AI protections & more. Show the AMPTP that you #standwithanimation. Sign your name next to the people that make the shows you love.

Give Animation Guild members a fair contract!

Members of The Animation Guild (TAG), IATSE Local 839, are facing unprecedented levels of unemployment despite animation outperforming on s

animationguild.org

chillyfeetsteak

Yesterday we marched to Netflix to deliver a solidarity petition in defense of our livelihoods! We are still in negotiations with AMPTP to get a better contract. If you love animation, have ever enjoyed a cartoon, or are just for workers' rights, please sign this petition alongside us!

aph-japan reblogged

hikari-m

oatmealspet

View on Twitter

Over 800 TAG members marched on Netflix & delivered a petition demanding the AMPTP keep #animation jobs #union, agree to AI protections & more. Show the AMPTP that you #standwithanimation. Sign your name next to the people that make the shows you love.

Give Animation Guild members a fair contract!

Members of The Animation Guild (TAG), IATSE Local 839, are facing unprecedented levels of unemployment despite animation outperforming on s

animationguild.org

aph-japan reblogged

koushirouizumi

jv

taikeero-lecoredier asked:

Hey,you make excellent points on the situation and all and i know you barely have any info as an ex staff but like. Do you think in case matt has enough of Tumblr hed just. Sell it instead of shutting it down if its considered ever?

jv answered:

I'm going to use this ask to answer half a dozen others that ask the same think but on anon (Sorry folks, these past few days I'm being kind of flooded with asks).

I'm sure he will. Actually, when last november someone leaked that Automattic was reducing tumblr staff to 1/3rd of what it used to be, one of my first thoughts was that Matt was taking weight out of tumblr to be make it easier to sell. Automattic won't want to lose a lot of people if tumblr goes away, but obviously you can't sell a company without sending some people who knows how to handle it going in the deal. So yeah, I'm pretty sure that Matt has been open to listen to offers since that point, at least.

Now, the problem is... who would want to buy tumblr? For starters, you need a lot of disposable cash. See, I'm sure Automattic would pretty much give it for free at this point. But still, even with reduced headcount, I don't think the monthly running costs of tumblr are less than 3 or 4 million US dollars. Any company willing to purchase Tumblr needs to be able to safely burn that amount every month and willing to do it.

Then, to do what? Tumblr has gone under 4 different corporate overlords during its history. And none of them, at any point, has made money out of it. Tumblr is a big black hole of money that somehow has been kept alive for 17 years. So any company willing to purchase tumblr would need to think they are smarter than the rest and have a different plan that could work.

When Automattic purchased it back in 2019, Verizon was about to close it for good. That's why Automattic managed to get it for a token amount. Automattic thought it was smarter than Verizon, than Yahoo, than David Karp, and that it could get it to profitability. It didn't.

I'm not saying there isn't any chance of a sale. I see how a company that has a steady income and is interested in a massive amount of traffic could be interested. But at this point I wouldn't bet on it, and even if it happens, the kind of company who fits that description won't be something we would like to have around (programatic ads companies, online gambling, etc).

So well, possible? yes. Likely? I don't think so. A good solution? I don't think so.

jv

Incidentally, that's why all those "didn't Automattic bought it for three mill? Let's crowdfund that amount and buy it from them!" posts make so little sense.

Sure. Maybe you can do it. You get people together and raise the money and buy Tumblr. And then next month you will have to do it again. And then next month. And then next month. And then next month. And then ...

jv

Also, it kind of explains my doomerism I guess. Matt fucking Mullenweg pays for this site to be kept open every month out of his pocket (not exactly, sure, but close enough). I'm not sure if right now he would be feeling like paying for much longer.

aph-japan reblogged

taichi-x-koushiro

zillychu

I've seen a couple people saying they're jumping ship for [other big socmed] after the AI announcement here, but like. Guys. Friends. You do realize all the other sites have been silently working with big AI companies for a while now. Bluesky has not implemented any acknowledgement or protection, and the CEO worked with crypto for years. They're just not broadcasting it or giving you an option to help remove your work from automatic scraping. Cohost has implemented similar levels of prevention compared to Tumblr.

I greatly prefer the transparency and the tiny bit of protection, no matter how flimsy. Tumblr is pretty awful, but it's still better than everything else so far. Which sucks but until we burn down the plutocracy, this isn't gonna be escapable.

(Also, no Tumblr did not quietly sneak this in. They literally announced it before implementing.)

And my thoughts on the CEO being... himself.

jv

Tumblr hasn't been transparent at all: they already sold your data, openAI already received them. The toggle only means that they won't send your data in the future, and they request openAI to stop using the data from you they already have (they are not obliged to comply). But the announcement had been worded in a way that makes it look not as bad.

Also, the protections Cohost and the rest have to avoid data scraping are in practice the same as Tumblr: the only thing Tumblr can do against data scraping is politely request to not do it, but if the scraping companies don't want to acknowledge the request (and they never do), they don't have any reason to do so, neither technical nor legal . Still, the staff announcement lets you think Tumblr is implementing an extra level of protection others don't have, when the only thing that actually does is letting you opt-out of them selling your private data in the future. Again, because you can't do anything to get it out of that first data sale.

So no, Tumblr is not being transparent here. It's being as sneaky as it could be.

aph-japan reblogged

not-terezi-pyrope

Often when I post an AI-neutral or AI-positive take on an anti-AI post I get blocked, so I wanted to make my own post to share my thoughts on "Nightshade", the new adversarial data poisoning attack that the Glaze people have come out with.

I've read the paper and here are my takeaways:

Firstly, this is not necessarily or primarily a tool for artists to "coat" their images like Glaze; in fact, Nightshade works best when applied to sort of carefully selected "archetypal" images, ideally ones that were already generated using generative AI using a prompt for the generic concept to be attacked (which is what the authors did in their paper). Also, the image has to be explicitly paired with a specific text caption optimized to have the most impact, which would make it pretty annoying for individual artists to deploy.
While the intent of Nightshade is to have maximum impact with minimal data poisoning, in order to attack a large model there would have to be many thousands of samples in the training data. Obviously if you have a webpage that you created specifically to host a massive gallery poisoned images, that can be fairly easily blacklisted, so you'd have to have a lot of patience and resources in order to hide these enough so they proliferate into the training datasets of major models.
The main use case for this as suggested by the authors is to protect specific copyrights. The example they use is that of Disney specifically releasing a lot of poisoned images of Mickey Mouse to prevent people generating art of him. As a large company like Disney would be more likely to have the resources to seed Nightshade images at scale, this sounds like the most plausible large scale use case for me, even if web artists could crowdsource some sort of similar generic campaign.
Either way, the optimal use case of "large organization repeatedly using generative AI models to create images, then running through another resource heavy AI model to corrupt them, then hiding them on the open web, to protect specific concepts and copyrights" doesn't sound like the big win for freedom of expression that people are going to pretend it is. This is the case for a lot of discussion around AI and I wish people would stop flagwaving for corporate copyright protections, but whatever.
The panic about AI resource use in terms of power/water is mostly bunk (AI training is done once per large model, and in terms of industrial production processes, using a single airliner flight's worth of carbon output for an industrial model that can then be used indefinitely to do useful work seems like a small fry in comparison to all the other nonsense that humanity wastes power on). However, given that deploying this at scale would be a huge compute sink, it's ironic to see anti-AI activists for that is a talking point hyping this up so much.
In terms of actual attack effectiveness; like Glaze, this once again relies on analysis of the feature space of current public models such as Stable Diffusion. This means that effectiveness is reduced on other models with differing architectures and training sets. However, also like Glaze, it looks like the overall "world feature space" that generative models fit to is generalisable enough that this attack will work across models.
That means that if this does get deployed at scale, it could definitely fuck with a lot of current systems. That said, once again, it'd likely have a bigger effect on indie and open source generation projects than the massive corporate monoliths who are probably working to secure proprietary data sets, like I believe Adobe Firefly did. I don't like how these attacks concentrate the power up.
The generalisation of the attack doesn't mean that this can't be defended against, but it does mean that you'd likely need to invest in bespoke measures; e.g. specifically training a detector on a large dataset of Nightshade poison in order to filter them out, spending more time and labour curating your input dataset, or designing radically different architectures that don't produce a comparably similar virtual feature space. I.e. the effect of this being used at scale wouldn't eliminate "AI art", but it could potentially cause a headache for people all around and limit accessibility for hobbyists (although presumably curated datasets would trickle down eventually).

All in all a bit of a dick move that will make things harder for people in general, but I suppose that's the point, and what people who want to deploy this at scale are aiming for. I suppose with public data scraping that sort of thing is fair game I guess.

Additionally, since making my first reply I've had a look at their website:

Used responsibly, Nightshade can help deter model trainers who disregard copyrights, opt-out lists, and do-not-scrape/robots.txt directives. It does not rely on the kindness of model trainers, but instead associates a small incremental price on each piece of data scraped and trained without authorization. Nightshade's goal is not to break models, but to increase the cost of training on unlicensed data, such that licensing images from their creators becomes a viable alternative.

Once again we see that the intended impact of Nightshade is not to eliminate generative AI but to make it infeasible for models to be created and trained by without a corporate money-bag to pay licensing fees for guaranteed clean data. I generally feel that this focuses power upwards and is overall a bad move. If anything, this sort of model, where only large corporations can create and control AI tools, will do nothing to help counter the economic displacement without worker protection that is the real issue with AI systems deployment, but will exacerbate the problem of the benefits of those systems being more constrained to said large corporations.

Kinda sucks how that gets pushed through by lying to small artists about the importance of copyright law for their own small-scale works (ignoring the fact that processing derived metadata from web images is pretty damn clearly a fair use application).

aph-japan reblogged

hikari-m

inkyxward

It REALLY is fucking funny how there is genuinely a distinct third faction in AI discourse on this site that is not "Pro-AI" but "The anti-AI crowd is talking about divine sparks of creativity and shit and I'm not about that"

me

carlyraejepsans

#me #ai issues #ai discourse #xtian hegemony #cultural christianity #workers rights #labor rights #labor #us labor #uk labor #politics #strikes #labor strikes #identity theft #human rights

aph-japan reblogged

koushirouizumi

orlissa

Anonymous asked:

You don’t own fanfics. They’re inherently public domain because they aren’t your IP. Agree or disagree with AI, there are no grounds for “protection” from AI because it isn’t your IP to begin with. That’s what you chose when you chose this medium

orlissa answered:

Oh dear.

Okay, you get an answer, because at least you took the effort to write your ask out properly, even if you are hiding behind the grey, sunglassed circle.

Do I, or any fanfic author for that matter, have any legal claims to our work? No, not really, no. (Although if someone took a fic, filed off the serial number--deleted the fandom specific elements--, and then had it published for financial gain, yeah, that would be a case.)

BUT

Fandoms are built on a social contract that says we respect each others work, the effort people put into their art. We don't steal or disrespect the work of our peers. By feeding people's fanworks to AI you both steal and disprect it, and we need to make people realize that before it's too late--before fandom falls apart, because there will be no more real, actual fanworks.

Disrepectfully,

Orlissa

(i can't believe I have to say this)

aph-japan reblogged

pro-AI in the sense of "they taught a bread scanning computer to recognize cancer cells" etc etc

against AI in the sense of "we stole artwork from hundreds to thousands of artists, didn't credit them and didn't financially compensate them"

#ai issues #ai discourse #ai wank #ethical ai #medical technology #technology #medical advancements #cancer #cancer detection #art theft #artificial intelligence

aph-japan reblogged

dduane

EU approves landmark AI law, leapfrogging US to regulate worrying new technology | CNN Business

European Union lawmakers gave final approval Wednesday to a landmark law governing artificial intelligence, leapfrogging the United States o

CNN

Sorry, the original EuroNews link seems to have gone away. Here's a CNN one instead.

...The EU AI Act outlaws social scoring systems powered by AI and any biometric-based tools used to guess a person’s race, political leanings or sexual orientation. It bans the use of AI to interpret the emotions of people in schools and workplaces, as well as some types of automated profiling intended to predict a person’s likelihood of committing future crimes. Meanwhile, the law outlines a separate category of “high-risk” uses of AI, particularly for education, hiring and access to government services, and imposes a separate set of transparency and other obligations on them. Companies such as OpenAI that produce powerful, complex and widely used AI models will also be subject to new disclosure requirements under the law. It also requires all AI-generated deepfakes to be clearly labeled, targeting concerns about manipulated media that could lead to disinformation and election meddling.

aph-japan reblogged

sarah-sandwich

seriously this is what we're doing now? If you get this report as spam. Don't take it to heart. It's spam. Report it so AO3 can take steps to block this shit.

aph-japan reblogged

izzyizumi

cilil

So today I got a rather unkind comment on AO3 (one could call it hate), but I believe it to be a bot for several reasons:

Guest account, but username attached
Said username exists but person is unlikely to be reading Tolkien fic (according to their Tumblr and AO3, they are in other fandoms)
Two grammatically correct sentences
Super generic text that could apply to any fic:

"I've seen better fanfiction written by a toddler. Get it together!"

I'm curious, did anyone else get comments like this? Let me know.

And to those who have gotten rude comments and are now worried/upset: Maybe it was just a bot too. Either way: You're awesome for putting your writing out there for others to enjoy and you don't deserve to get rude comments for it. If you want feel free to message me to compare cases and discuss details :)

cilil

For comparison, this is the one I received.

Here are some more examples fellow writers allowed me to share:

As you can see, these comments all the match the description above. Also they seem to be weirdly obsessed with AI and this entire operation may be an attempt to promote writing AI - which, if true, is disgusting on several levels.

(If any of the people with the AO3 usernames in question happen to see this: Don't worry, we're all sure it wasn't you, no hard feelings and sorry that a bot stole your name for hate comments. The names are only shown as proof that they are indeed stolen)

eccentricmya

Oh I was just wondering why I got a weird comment today!

This is mine. Looking at the pattern, it is definitely a bot. Stealing usernames and posting as guests under that.

I was upset for a hot minute, replied asking what did they mean... Turns out I shouldn't have bothered.

cilil

Thank you for the addition! This certainly paints a picture - they're claiming that the work of actual human authors is subpar (the toddler comment on mine and the human comment on yours) and/or bring up AI.

Sorry this happened and I hope you're feeling better now💕 personally, I do believe this work was written by a human being and a lovely too!🫂💗

moosemonstrous

I don't think they're so much trying to get people to use AI to write fics - there seems to be some sort of a scam where the bots encourage commenters to put a fic through their 'AI scanner' to check if it was written by a robot.

They're literally trying to get readers to do the work of scraping fics for them.

cacodaemonia

I'm sure other folks have added this already in different reblog chains, but AO3 has disabled guest comments for now.

Just in case, I would recommend not going to any of the AI writing sites these comments mention since it's possible they have malware on them. Or maybe the comments are just intended to get people searching for the terms and then that might make them pop up in google search's autofill?? I have no idea 😑

lttrsfrmlnrrgby

Locking your fics and comments to registered users eliminates this issue-- and hopefully limits bot scraping of the works themselves.

caparrucia

I hate this so much because some of the best recurrent readers and consistent commenters I've ever had have been anons.

#fandom #fandom issues #fandom harassment #bad media criticism #ai issues #spambots #ao3 spam #archive of our own #ao3 issues #fic writers #fic writers issues #fanfiction #(IF YOU ARE ON AO3 {Even IF not actively using it as much anymore} re the usernames BE AWARE)#(Me ' Theyre stealing peoples USERNAMES now?!??!? ')#for the record the comment harassing spam i got back when kizuna was much newer was NOT this type of 'Guest' comment #they had whole blank accounts some of which were made a year+ before and everything #(but passing the news along to anyone on AO3)#i guess now theyre just ~~casually~~ insulting highly specific Jewish cultural holidays most Goyim donteven know like Shabbat too huh all??#(because I posted a serious+reflective *shortfic* in *Daisukes character tag* {bc DAISUKES POV} after not using the tag in a while h u h??)#basically my experience since circa 2k20 #(This Is What Actual Spambot Spam Might Look Like)#(The fact that anon comments seems to actually be Disabled because of it now sucks so hard though wow...)#remember to always report spambots for SPAM #(And harrassing comments for Harassment but like im glad they temporarily turned off comments its just... Shocking they actually had to)#(FYI but these Issues have also been why my stuff is mainly locked to archive even if I temp allowed Anon comments in *past* yEA)

aph-japan reblogged

neil-gaiman

ashlovesfiction

I saw a very blunt Instagram comment today that told a writer, "AI is going to steal your job soon. You may want to choose something else." It was so nonchalant and casual, like what was just said wasn't heartbreaking to hear.

Can we writers just make a pact just to... not quit? Can we not give in so easily? Can we actually fight to keep our professions and continue to share our own original work? I will never expect writing to be my main source of income, but that does not mean I'm so willing to give it up for the sake of some robot.

charliejaneanders

I seriously would not worry... First of all, in order for AI to steal anything from anyone, we'd have to develop AI first, and we're nowhere near doing that. (What we have now is basically what Ted Chiang calls "applied statistics," or what others call "enhanced autocomplete.") And writing, as a practice, contains hundreds of tasks that require human creativity. Seriously, don't believe the hype.

#neil gaiman #ai issues #ai discourse #art theft #art #creativity #humanity #careers #jobs #ai misinfo #misinformation campaigns #fearmongering #fear mongering #applied statistics #seriously #i dont even have a job to steal

aph-japan reblogged

mclennonyaoi

it is insane how people do just . not know what objective is. like if you are trying to argue there is some objective difference between “good art” and “bad art” and your first argument is that “well objectively good art makes people feel good and bad art makes them feel bad” you . have no clue what art is. and also have no clue what objectivity is.

#ai issues #ai discourse #fandom wank #arts #creative arts #art #writing #performance arts #crafting #objectivity #yeah #seriously #not up for debate #good commentary

aph-japan reblogged

prokopetz

Like, I'm not gonna argue that AI art in its present form doesn't have numerous ethical issues, but it strikes me that a big chunk of the debate about it seems to be drifting further and further toward an argument against procedurally generated art in general, which probably isn't a productive approach, if only because it's vulnerable to having its legs kicked out from under it any time anybody thinks to point out how broad that brush is. If the criteria you're setting forth for the ethical use of procedurally generated art would, when applied with an even hand, establish that the existence of Dwarf Fortress is unethical, you probably need to rethink your premises!

aph-japan reblogged

dduane

A WordPress ‘Firehose’ Allows AI Companies to Buy Access to a Million Posts a Day

There is a complex chain of companies buying access to WordPress and Tumblr posts through a company called SocialGist.

404 Media

I stopped using Jetpack when they started charging for it. WordPress friends: here's a reason (if you're still using it) to adjust your settings. (tl:dr; turn off "enhanced distribution".)

What this means (or at least suggests) for Wordpress users, whether their blogs are self-hosted (like all the ones in our household) or hosted at Wordpress.org:

Go through your working plugins, from whatever source. Start investigating whether, and/or to whom, they are selling your data for training. This is going to be more small-print-reading than I remember signing up for... but (sighing at the echo of the voice of Abe Lincoln), "The dogmas of the quiet past are inadequate to the stormy present. The occasion is piled high with difficulty, and we must rise with the occasion."

And if you have the time, email your plugin providers and ask for a clear and straightforward answer to the question "Are you using my content to train AI apps?" If they won't give you an answer... dump them.

#diane duane #wordpress #matt mullenweg #tumblr ceo #(late adds tags)#(yeah the SAME CEO is involved)#ai issues #art theft #fanworks theft #fandom #fandom issues #copyright

aph-japan reblogged

koushirouizumi

renthony

At this point I'm just assuming everything I ever create and post to the internet is going to be stolen. People have been stealing, reposting, and adding their own pay links to my art for years now, without the help of AI.

I've made D&D themed stickers that are now all over "free clipart" sites, despite me filing requests to have them removed. I've seen my graphics ripped off and included in someone else's art without credit. I've had people tell me that an ACAB image I made showed up as a sticker getting put up around Seattle. Facebook meme pages crop my username out of my posts all the goddamn time. Voice actors on YouTube use my posts for "dramatic reading" videos constantly, and only one has ever asked me permission or given me any cut of the profits from their video.

I see my art out in the wild with no source back to me, and I'm a tiny creator compared to a lot of others. People repost shit constantly, whether it's here, Pinterest, Facebook, Twitter, TikTok, YouTube, whatever. I remember the old tumblr days of "We Heart It is not a goddamn source" PSAs.

I think people are right to be concerned about AI, but at this point I'm much more concerned about it from the perspective of "companies want to use it to cut labor costs," and less "it's theft."

People didn't need AI to steal my art before now. I'm more concerned about trying to freelance in a market full of "oh, we can just get ChatGPT to write and illustrate our articles."

#me #ai issues #fanworks theft #art theft #writing theft #writers issues #writing #art #artists #its basically this #mood #us labor rights #labor rights #this is pretty much where im at rn #(and I dont even get that level of reposting {without credit} happening to me that OP describes but)#good commentary #people here have still stolen my gifs outside of the contexts they were made for multiple times +for series negative posts though #with no apologies like literally 0 apologies from people i didnt contact myself about it like because they had asks closed etc #which i hate with a burning passion {way more than when used on positive posts}&i hope none of the ppl involved ever come near me again #unless they actually mean to apologize genuinely #(because I *have* gotten *genuine* apologies before and I'll acknowledge those but theres still instances where it *hasnt happened*)#and they know it #(Also part abt Youtubers here is especially f*cked up & I seriously hope those people own up and apologize someday VAs community BE BETTER)

aph-japan reblogged

koushirouizumi

zillychu

I've seen a couple people saying they're jumping ship for [other big socmed] after the AI announcement here, but like. Guys. Friends. You do realize all the other sites have been silently working with big AI companies for a while now. Bluesky has not implemented any acknowledgement or protection, and the CEO worked with crypto for years. They're just not broadcasting it or giving you an option to help remove your work from automatic scraping. Cohost has implemented similar levels of prevention compared to Tumblr.

I greatly prefer the transparency and the tiny bit of protection, no matter how flimsy. Tumblr is pretty awful, but it's still better than everything else so far. Which sucks but until we burn down the plutocracy, this isn't gonna be escapable.

(Also, no Tumblr did not quietly sneak this in. They literally announced it before implementing.)

And my thoughts on the CEO being... himself.

jv

Tumblr hasn't been transparent at all: they already sold your data, openAI already received them. The toggle only means that they won't send your data in the future, and they request openAI to stop using the data from you they already have (they are not obliged to comply). But the announcement had been worded in a way that makes it look not as bad.

Also, the protections Cohost and the rest have to avoid data scraping are in practice the same as Tumblr: the only thing Tumblr can do against data scraping is politely request to not do it, but if the scraping companies don't want to acknowledge the request (and they never do), they don't have any reason to do so, neither technical nor legal . Still, the staff announcement lets you think Tumblr is implementing an extra level of protection others don't have, when the only thing that actually does is letting you opt-out of them selling your private data in the future. Again, because you can't do anything to get it out of that first data sale.

So no, Tumblr is not being transparent here. It's being as sneaky as it could be.

#social media #social media issues #tumblr #tumblr issues #ai issues #current events #bluesky #cohost