#AI-Generated Deepfakes

[scamandfraud]

Deceptive Chameleons: Unveiling the Multifaceted Nature of AI-Generated Deepfakes

AI-generated deepfakes are not simply manipulated videos or audios. They are, at their core, synthetic chameleons: digital creations crafted by artificial intelligence that seamlessly blend the real and the fabricated. But unlike regular chameleons blending into their surroundings, deepfakes aim to deceive, masquerading as authentic representations of reality to manipulate our perception. The…

[reasonsforhope]

"Major technology companies signed a pact on Friday to voluntarily adopt "reasonable precautions" to prevent artificial intelligence (AI) tools from being used to disrupt democratic elections around the world.

Executives from Adobe, Amazon, Google, IBM, Meta, Microsoft, OpenAI, and TikTok gathered at the Munich Security Conference to announce a new framework for how they respond to AI-generated deepfakes that deliberately trick voters. 

Twelve other companies - including Elon Musk's X - are also signing on to the accord...

The accord is largely symbolic, but targets increasingly realistic AI-generated images, audio, and video "that deceptively fake or alter the appearance, voice, or actions of political candidates, election officials, and other key stakeholders in a democratic election, or that provide false information to voters about when, where, and how they can lawfully vote".

The companies aren't committing to ban or remove deepfakes. Instead, the accord outlines methods they will use to try to detect and label deceptive AI content when it is created or distributed on their platforms. 

It notes the companies will share best practices and provide "swift and proportionate responses" when that content starts to spread.

Lack of binding requirements

The vagueness of the commitments and lack of any binding requirements likely helped win over a diverse swath of companies, but disappointed advocates were looking for stronger assurances.

"The language isn't quite as strong as one might have expected," said Rachel Orey, senior associate director of the Elections Project at the Bipartisan Policy Center. 

"I think we should give credit where credit is due, and acknowledge that the companies do have a vested interest in their tools not being used to undermine free and fair elections. That said, it is voluntary, and we'll be keeping an eye on whether they follow through." ...

Several political leaders from Europe and the US also joined Friday’s announcement. European Commission Vice President Vera Jourova said while such an agreement can’t be comprehensive, "it contains very impactful and positive elements".  ...

[The Accord and Where We're At]

The accord calls on platforms to "pay attention to context and in particular to safeguarding educational, documentary, artistic, satirical, and political expression".

It said the companies will focus on transparency to users about their policies and work to educate the public about how they can avoid falling for AI fakes.

Most companies have previously said they’re putting safeguards on their own generative AI tools that can manipulate images and sound, while also working to identify and label AI-generated content so that social media users know if what they’re seeing is real. But most of those proposed solutions haven't yet rolled out and the companies have faced pressure to do more.

That pressure is heightened in the US, where Congress has yet to pass laws regulating AI in politics, leaving companies to largely govern themselves.

The Federal Communications Commission recently confirmed AI-generated audio clips in robocalls are against the law [in the US], but that doesn't cover audio deepfakes when they circulate on social media or in campaign advertisements.

Many social media companies already have policies in place to deter deceptive posts about electoral processes - AI-generated or not... 

[Signatories Include]

In addition to the companies that helped broker Friday's agreement, other signatories include chatbot developers Anthropic and Inflection AI; voice-clone startup ElevenLabs; chip designer Arm Holdings; security companies McAfee and TrendMicro; and Stability AI, known for making the image-generator Stable Diffusion.

Notably absent is another popular AI image-generator, Midjourney. The San Francisco-based startup didn't immediately respond to a request for comment on Friday.

The inclusion of X - not mentioned in an earlier announcement about the pending accord - was one of the surprises of Friday's agreement."

-via EuroNews, February 17, 2024

--

Note: No idea whether this will actually do much of anything (would love to hear from people with experience in this area on significant this is), but I'll definitely take it. Some of these companies may even mean it! (X/Twitter almost definitely doesn't, though).

Still, like I said, I'll take it. Any significant move toward tech companies self-regulating AI is a good sign, as far as I'm concerned, especially a large-scale and international effort. Even if it's a "mostly symbolic" accord, the scale and prominence of this accord is encouraging, and it sets a precedent for further regulation to build on.

[dailyfigures]

Wait, just out of curiosity, but do you actually support AI art or is that one person just being weird

oh no sorry if that was unclear, i DO NOT support AI art!!!

i just had a problem with someone accusing real artists of being AI based on nothing, and then continuing to be rude when i told them they were real people. it really was just about someone being mean for no reason.

[alwaysbewoke]

[chimaera0803-blog]

Umm do the criminal minds season 17 writers know about rule 34? I'm enjoying the show but this BAUGate bit is really uncomfy with the developments and misuse of generative AI. I feel like they are really tempting fate and putting the actors at risk, especially AJ Cook

[cardboardheartss]

What in the Pluto in Aquarius is going awnnnnn!!!!

[siduribythesea]

Nature's version of Ai-generated deepfakes 😳

These coastal shrubs appear to blooming profusely with pink flower clusters in various shades and stages of development...from a distance.

Just like a closer look at an Ai-generated picture reveals subtle discrepancies (wrong number of fingers, uncanny valley, oddly placed fixtures, or physically incorrect optics), a second glance at these real-life deep fakes shows similarly incongruous details.

Meet Exobasidium, a fungus who hijacks the flowering function of a particular species to produce fake fruit/buds (galls) and obscenely engorged flowers (fruiting bodies) to trick pollinators into spore dispersal.

This mycological organism, without the benefit of "eyes," has taken mimicry to the next level by acquiring enough biological data to come up with a good, but not quite convincing, mime of the real thing. It's a bizarrely unsettling phenomenon and yet another example of the prowess of fungal minds totally outpacing the bounds of human tech.

"Deepfakes" are apparently as old as the first mycological surveillance on angiosperms...

[crispsandkerosene]

Probably kind of a hot take, but I wish tumblr people would maybe make the nuance between "data scraping and companies doing whatever morally bankrupt corner cutting to maximize their profits are bad" and "the very concept of AI is evil, I hope Disney sues and makes artstyles copywritable, stuff you tell to chat GPT will be known to all AIs ever!!"

[thealogie]

is there any group of lawyers building an effort to protect jobs in the arts from ai layoffs?

That’s not really something that can or should be handled by a group of lawyers. Well, at least not in the sense people are thinking. Many law makers (ie members of Congress and senators) are lawyers and they should be doing something. It’s also something that unions have and will continue to fight to put in contracts (and they’re aided by lawyers but the primary activists are actually not lawyers).

Lawyers are however very much working to deal with the effects of AI on existing legal issues like copyright, privacy/sexual harassment, right of publicity etc.

[youwontbelievethethingsiveread]

Anyone else ever think about how in Abaddon's Gate, Clarissa Mao spent a minor fortune on some software that could make a fake video of Holden that only the best pros could maybe tell might not be real based on some very subtle clues - ten years later in real time any idiot can go online and make a deepfake AI video for free of just about anything, and suddenly Clarissa spending money of this is just weird 😆 just an example of where the technological development overtook the writing I guess

[kotori-mochi]

Good news but bitter sweet.

Good news is the US government finally realized that deepfake po/n from AI generators is indeed an issue.

The annoying thing is it took them a year and this is the reason why they are taking action.

Yup it's because someone did it to a popular celebrity with a giant fan base. Not the evidence of deepfakes of child p/rn or the women being harassed by this but Americas sweet heart.

I will add at least now they are noting this but it's still just annoying.

Also to anyone who doesn't know deepfake po/n has been a big issue for a year and there are AI bros who see no issue with it.

Also how it works is someone can use your image or your children's or siblings, put it in an AI generator and then they can turn your photo into por/n.

Above is a real Ad on X/twitter I have actually seen some pop up in my comments or feed. It's very disturbing but hopefully with the government noticing this something will be done to stop this.

Fingers crossed.

[scamandfraud]

Deceptive Chameleons: Unveiling the Multifaceted Nature of AI-Generated Deepfakes

AI-generated deepfakes are not simply manipulated videos or audios. They are, at their core, synthetic chameleons: digital creations crafted by artificial intelligence that seamlessly blend the real and the fabricated. But unlike regular chameleons blending into their surroundings, deepfakes aim to deceive, masquerading as authentic representations of reality to manipulate our perception. The…

[69constellationsinatrenchcoat]

https://www.rnz.co.nz/news/world/507694/taylor-swift-ai-deepfakes-spark-calls-for-new-legislation

Look, I'm not a swiftie, but what the Fuck.

WHY WOULD YOU D O T H a T ???

As someone who's had images shared online that I did not consent to, this is Foul. Absolutely foul. The repercussions on someone's reputation is absolutely disgusting.

[jamieandgeorgedashigz]

This AI apocalypse has genuinely fallen out of hand. I guess I'm a bit worried for Nicole's fate, lol.

-Jamie

[imagominus]

Every existing argument that speaks for generative AI can be countered with a very logical counter-argument. Let me show you:

"It is an art-tool like any other" - No, it's not, because a tool doesn't do your work for you like gen-AI does, tools enable you to do things more easily and to do more things at once, but gen-AI does the entire thing, and not even that good.

"AI is faster than any artist" - Maybe, but that comes at the cost of creativity and quality. You may be able to fool those who only take a short glance at it, but the moment they comprehend the creation, it gets exposed!

"AI is the future" - It is very clearly not! Gen-AI is very resource-heavy, depends on real artists' works, and doesn't have the quality to be of use once the userbase learns to be more skeptical of online creations. Not to mention the AI feeding of its own creations, resulting into WORSE generations than before.

"AI can generate anything you want" - No, it can't! Everything it generates it can do so only because it has been fed with according existing creations, so once you tell it to generate something it has never been trained to generate, it fails. Try making it generate anything outlandish, like irregular numbers of appendages, or chimeras. More often than not, it will disappoint you.

"AI helps me with inspiration" - Sure, it could be used for inspiration. But do you know what else does? LITERALLY EVERYTHING! You can look to your left, and take whatever you see there as inspiration. You can look at a specific object's texture and use bits of that as inspiration! You could even take a piece of paper, crumble it, make a photo of it, trace the visible folds and edges and use even THAT as inspiration! Inspiration is everywhere, though whether it can be used in the given context is another story, but the point still stands. You do NOT need AI for inspiration. Oh, and remember how gen-AI is trained on other people's creations? That means, that whatever the AI puts out ALREADY EXISTS! You just have to search for it, so just take whatever prompts you just had in mind, and formulate a (google?)-search based on it. Just like that, you may have just found some images to inspire you, if not an entire ARTIST TO SUPPORT!

"AI is still very fun to play around with thanks to its fast generation process!" - Oh, sure, it can be fun to see what the AI spits out, but those super-fast results are their own problem! Not only could you play around with anything else and have even more fun than during the waiting period for the gen-AI to finish generating, but you could have even more fun with stuff that takes longer to do. Oh, you don't? Well, that just means you have fallen for what i call the "modern curse", that is instant gratification. I'd know, i want to commit to digital art for the sheer possibilities of it all, but i can't enjoy the art process since i find more joy in any other game or social media. The need for instantaneous results from anything is incinuating us to not wait for anything. It hurts our patience and sense of time and robs us of any non-instant fun to be had. I, too, have fallen to this modern curse, and am trying to find my way out of it, so i can actually commit to something worthwhile, which gen-AI is NOT! And it's not just our enjoyment it is hurting, but our community as well! With feedback coming more from the instant responses of gen-AI and systems designed to keep us hooked to their stuff, we ultimately communicate with each other less and rob ourselves of social connectivity and personal growth, as we stray away from the necessary criticism to learn from mistakes and grow as people. If you enjoy gen-AI as much as you say you do, i assure you, you NEED to get off the web a bit, talk to people, play longer lasting games, and, for lack of better word, touch some grass.

"AI is a great investment!" - And with that statement, you just exposed yourself as someone willing to scam people out of well earned money with non-human creations. Oh, you're not? Then why are you calling gen-AI worthwhile? The resources and energy requirements to keep the gen-AI running as a service are too much to maintain over longer timespans, and you hardly make any money back running gen-AI. Not only that, but if you generate images and other creations and advertise them as your own creations, you are abusing people's trust that their works are legitimately man-made, with human creativity. By that point, you would already be a fraud, and even worse if you decide to SELL the creations as your own art. You would not only betray people's trust, you'd also scam them out of their money, and the moment they learn that it's AI generated, they wont think that "oh, how far AI has come", they'll think that "that mf just scammed me with their AI-generation bullssss, i am never buying anything from them again!" If you're in it for the money, you may get some money, but at the cost of people's trust in you, and forget about getting famous for it. Even the big tech companies aren't spared from the harsh criticisms of the public, so why would you be any better?

I could go on and on and on, but the point still stands. Generative AI is only hurting our community, especially once scams and AI-gen-spam comes into play (don't even get me started on AI-spam and content-farms). It isn't worth any finances, passion or mental energy. It has only hurt the userbase, betraying people's trust, betraying built-up hype, betraying the artist community, and as time goes on, big tech will learn about their mistake of throwing away artists' trust for some machines.

Hmmmm, how do i end this one? You know what, i have a hot take for you! Here it comes: Generative AI should be forbidden BY LAW! Straight up! It may seem harsh, but we are too ill prepared for gen-AI to make it as big as it did, whether as large projects or as public tools. The sooner, the better, get rid of gen-AI, and i mean ALL OF IT!

[starblaster]

"There was never any support back then, when these leaks came out. There was never any support, you know, it was completely blacklisted. And there's always been, like you say, a convenient reason to disparage Assange and WikiLeaks, either because of shifting goalposts [...] about the meaning of journalism and, you know, whether or not he's publishing or a "journalist" [...] or his own political views or gossip about his personal life [...] And all of it, all of that shit, all it does is do the work of state power to de-legitimate WikiLeaks and turn Assange into a pariah, which has happened. You know, none of it has anything to do with the fact, and I keep repeating, that every single thing he has published is true! The same cannot be said about anybody at MSNBC, The New York Times, Fox News, CNN, or The Washington Post. So, apparently, there is no solidarity for political prisoners of the state.

But this is what fucking gets me! None of these cowards are going to be persecuted, none of them. All of these people that are shaking in their boots — like Chris Hayes [...] — you're fine! Who's going to persecute you, the fucking government that you parrot the lines of? The government that you continue to let have advance notice of the story that you're saying, in case there's anything they wouldn't like you to publish? The 'anonymous' intelligence sources you continue to give cover to? The bullshit partisan hackery you spit out night after night after night? What state secrets are you revealing? What actual journalistic work are you doing? This is why we call journalists scum [on TrueAnon]. Motherfucker, you are the mouthpiece of the state! The government needs you to sell their lies. You are the one who obfuscates power, you aren't expendable, you are the one who is necessary. You aren't going anywhere, you fucking tool.

And these assholes, the ones who guarantee their prime time TV slots by molding their views and their voice[s] into what the networks and their corporate overlords and the bankers and the U.S. government [...] want you to hear, these are not the people who will be persecuted. No, they get promotions. They get awards, they get celebrated, they get retweeted, and liked, and byline and byline and byline; they climb the ranks and they shape the shit that is now permanently lodged in your fucking brain that they pathetically sell to you as 'the news' and you've pathetically convinced yourself is the truth.

No, the people who get persecuted are anyone who dares question the United States war machine, who dares reveal: any of the coups, the assassinations, the inner workings of the security apparatus, the well-funded networks of fucking so-called democratic political power, the abuse that millions suffer at the hands of the U.S. military and all the paid-off NGOs, the corporations, the warlords, the cartels; the sexual abuse, the rape, the murder, the towns that we bomb to oblivion, the countries that we sell off to the highest bidder, and all the governments that we demolish. Like, what we do every day, every day, every day. Those are the people that will be persecuted.

People like Chris Hayes have fucking secured their spot. They're not going to be attacked, he is not going to be arrested. And now they sit back and fucking feign concern for Assange; "Oh, what this means, for the free press! Oh, no! What this means!"

This has nothing to do with Assange or WikiLeaks or anything WikiLeaks has revealed! This is about the media's precious industry which is dying, ironically, because of the very corporations that they continue to cover for, that are eating the news rooms alive. And you know what? The public is right to fucking despise you [journalists]. They are right to despise you for the obvious bootlicking and all your brazen career-ism, for years of covering up all the lies and all the obfuscation of how real power operates in this country; selling the wars and the PR campaigns of the State Department and running fucking interference for the bankers and all the politicians bought off by the bankers who stole all the jobs and all the homes and all the fucking futures out from under the world.

How many bloggers, who are now calling themselves aspiring journalists — or, I guess, what you'd now call "professional opinion-havers" — how many out there look up to Chris Hayes or the Chris Hayeses of the world and their career arc? How many stare longingly at the fucking New York Times op-ed page, at the shiny bylines and the prestige publishers, at the fucking social power that comes with the fawning adulation of their peers jealous of their little positions in the clout hierarchy of the media machine? How many of these content producers, editors, self-appointed political taste-makers, take-havers, tweeters — and, guess what, yes, fucking podcasters — how many of them just want these precious fucking spots? Like, what would they do? What would they do? What do you think the would do? What do you think they would sell you to get one of those spots? Do they even fucking realize when they do it?

But how many of them, now, actually look to [Julian Assange]? How many of them will even fucking publicly support the guy who is literally, currently, being tortured — who will die of starvation, or maybe madness, or maybe he'll just simply be kicked in the fucking skull in the damp basement of a supermax prison in the middle of fucking nowhere — because he dared to temporarily embarrass our highest-ranking officials and mildly inconvenience the functionaries of the American empire's unstoppable machine?

Who are you gonna fucking look to?"

Liz Franczak TrueAnon, "Episode 106: #FreeAssange" October 8, 2020