Amnesty International criticised for using AI-generated images
theguardian.com
theguardian.com
It’s already happening with the writers guild…
Porn already, for 95%+ of performers, pays terribly. Maybe one or two grand at best per "job" due to rampant piracy, ad-blocking, and the fact that nobody wants it on their credit card statement, and not everyone wants to figure out Bitcoin.
They hardly have the capital to start a union, or pay dues. But if AI porn takes off, those earnings will turn porn into purely a "because you want to" business. Probably not a bad thing considering the exploitative aspect in my opinion.
Even with Identity Verification, there's always the risk that someone will be trafficked, verified as a performer, and then later escape their trafficker, file a lawsuit, and win/achieve a settlement. Which has (allegedly) happened multiple times, at least with pending lawsuits.
What's safer than Identity Verification? AI-generated porn. Completely "ethically" sourced, with no humans involved to potentially get hurt. Infinite variety, instantly customizable to everyone's own preferences, and (theoretically) truly endless in style and form. Also, it kills one of the biggest secular arguments against porn: "If nobody could possibly be trafficked or underage, what's the problem again, outside of your religious arguments?"
The moment it really hits the fan, I expect MindGeek would immediately adopt it. They can't bear the idea of losing market share to some competing app with such features. More variety and lower risk? They never cared about their "performers" before, they are a liability that is currently just a cost of doing business. But if they aren't required for the business, they'll be dropped like a hot potato. Assuming porn companies would be concerned about morality is a mistake by any definition.
I would not be surprised, at all, if 10-20 years from now, the only porn on PornHub is a back catalog of AI-generated porn; with a subscription available for generating new porn to your preferences. After all, there is perhaps nothing that has more visual training data available for developing an AI against than porn.
> The moment it really hits the fan, I expect MindGeek would immediately adopt it. They can't bear the idea of losing market share to some competing app with such features.
MindGeek would just buyout said company. MindGeek has practically a monopoly on porn and porn sites.
https://medium.com/five-guys-facts/8-19-16-davis-ce677114158...
AI using AI for misrepresentations fits my very low opinion of them.
I mean, I'm all for effective altruism, for giving money to charities the charities that help the most people with it with the lowest overhead, and for doing your research before deciding who you donate to (I recommend GiveWell as a starting point); but "hiring PR agents to solicit donations" is hardly unique to Amnesty.
> 2021 was a record-breaking year with a total fundraising income of €357m.
And
> The largest proportion of the movement’s programme expenditure €26.4m (35%) was spent towards Goal One.
And
> 45% of total income spent on Human rights research, advocacy, campaigning, raising awareness and education
So ~20% (75m of 357m) was actually spent on real work, assuming everything reached end beneficiaries.
Interesting way of showing 45% of expenditure clubbing multiple heads, while showing 2% expenditure on "Maintaining our democratic system of governance".
Among other things, it sounds like their income was higher than their expenditures, eg that they had some money left at the end of the year; so your "75m of 375m" ratio isn't quite right.
Maybe I missed something.
And even 20% is probably very high, because I would guess people working on those would be paid some salary/ honorarium, and that also would be included in this expense.
The captions make them worse somehow. Like the images are supposed to underscore the point, but then they're fake? Such a weird choice all around by Amnesty International.
Also
> insulted women
The highest crime!
[EDIT] Archive of guardian link https://archive.is/4RSIb
Not sure why you're intentionally clipping the full context.
> I used to donate money every year to AI but will never do in future.
Took me a moment to parse AI as Amnesty International...
Surely a non-profit like this their number 1 priority should be accuracy and truth in the harm/crimes they are reporting. But nah its not about any of that is it.
Thay the money goes to employees rather than shareholders is at least as motivating.
I know a number of people in the non-profit space. Some acquire a belief that the end justifies the means, and therefore it’s okay to push the ethical limits as long as it’s for a good cause. They believe their own cause is good, naturally, so fundraising pushes can become a slippery slope of bending the rules.
https://eeditionnytimes.pressreader.com/article/281814288181...
https://en.wikipedia.org/wiki/Amnesty_International#Criticis...
https://en.wikipedia.org/wiki/Criticism_of_Amnesty_Internati...
I can think of many examples. A recent fun one was Ngozi Fulani's (of false allegations of racism) charity Sistah Space. Exposed using only publicly available information as a total grifting sham. Remember Camila Batmanghelidjh? There are god knows how many more like those two.
Hope Not Hate - a labour party activist controlled disinformation & propaganda generation entity run as a Ltd funded entirely by a special purpose charity entity bankrolled by people like Sigrid Rowsing. You will note their "work" receives almost not interaction on social media - which is the majority of their raison d'être
There are quite literally hundreds of "local" charities in places like Tower Hamlets that don't come close to meeting an normal persons definition of charity. They simply funnel money, often from local councils, to themselves.
I've also attempted to report a charity before here in the UK. The CC refused to accept my anonymous info and if my information didn't lead to a prosecution (and had to sign a document to this effect) that i would be prosecuted myself. That was some years ago, so the process may be different.
A great many charities raise money in a charity entity, then funnel all that money to a small Ltd controlled by a small number of people. The oversight, and legislation are appalling.
There are a number of "charities" in the UK alone claiming to be tackling "long covid".
Who knows how many evangelical church scams there are enriching some classic conman preacher.
I remember once being on a video call with a quasi-VC but for social enterprises and I asked what I thought to be an entirely banal question: "how do you measure the effectiveness of the organisations you fund and ensure that they're not misappropriating funds?" I may as well have spat in their tea, the video call ended soon afterwards!
Some very very modest achievements i might have had came in a room of people where I politely, professionally, pointed out amidst a silence that the offer they were about to buy was based on quicksand.
If i had done so rudely they would have spent the money and fired me
That said, I don't think this changes anything. Someone was a bit lazy or arguably over-dramatic. I don't see it as some egregious violation of public trust. Even with real pictures, they would have been cherry picked for effect.
If they wanted to use AI-images, they could have used any number of filters/effects to reduce the phot0-realism and make it clear the images were illustrations (vs "photos").
[1]: The chain of causation is that Charlottesville wants to unconstitutionally limit the Unite the Right protest, ACLU threatens litigation, protest goes ahead, protest supporter kills protest opponent. Obviously the ACLU isn't the main one responsible, but I don't think it was crazy for this to prompt reevaluation.
It's a faux agency used for corrupt western influence in the guise of "human rights". India banned it for good reasons.
The best possible interpretation is that it was hopelessly naive ("why don't they just camp in a empty field so Russia can kill them more precisely?") but more likely they tried to both-sides a war which simply does not have two valid sides.
Just so we're all clear — what has been happening in Ukraine for a long time is russian soldiers have been using force to try to annex Ukrainian territory.
The kremlin denied this for the better part of a decade, but have recently admitted that those little green men were actually russian soldiers.
Amnesty International don't get even half the criticism they deserve.
"I wish had I known that the people fighting the power were actually just fighting for the power."
I'm pretty sure Amnesty International has been accused of being a partisan political movement for as long as it's existed, by those who don't like what it's critisizing.
If you ever thought they weren't political it's because you weren't part of the groups they were fighting.
Or at least that's their goal, certainly sometimes they have failed, because of biases. Is this one of those times?
The article you posted says that Amnesty says international laws and standards of war say you should not put military bases in residential areas. Have they been consistent at calling this out for all countries? I think so? Are these in fact standards under international human rights laws? I think so?
So this is not a departure for how Amnesty usually operates.
People often think that such laws and standards ought not to apply to the side "in the right"; or that they ought not to apply when the side in the right needs to violate them in order to win. Or that a given standard is wrong and should not be considered part of international law and human rights standard. Or you can think that the framework of international law is unrealistic and pointless and isn't even the right framework. In general, or in this case.
Some of these may be reasonable positions. But Amnesty has always operated from that framework, and has consistently critisized any side that puts military equipment in residential areas.
So, it's possible they made a mistake here in some way, but it's in a way that is pretty consistent with how they have behaved for decades and how they do things. So I don't see any reason to say they have "morphed", and that they are now "partisan" in a way they weren't before. Their theory is that they *avoid* being partisan by critiquing both sides of a military conflict to the same standard, you can disagree with this stance, but I don't see any reason to think they acted from "partisan" motivation rather than from trying to stick to their usual way of doing things. The complaint seems to be that some people wanted them to be partisan on the part of Ukraine (which may be reasonable), while they tried to apply same standards to Ukraine and Russia. There is no "morphing", this is what they do. You can think what they do is the totally wrong approach, for any number of reasons -- but it's not new.
Has a definite “only write what I can google” vibe to it and is not acceptable for such a well funded organization.
Another good rule of thumb is that the evidence which is proven to be fabricated is not the only evidence that was fabricated, and not the only situation in which it was.
We rely a great deal on “appeal to authority”, this organization being one of them. The epistemology of a lot of common belief breaks down very quickly once you start to see this. But if it goes too far, once the broken belief system affects the identity of the beholder, it reverts to blind faith and anger. We have a lot of Tooth Faries forced into our collective identity by school teachers.
I have a lot of sympathy for the local folks in dangerous situations who took real risks to contribute to Amnesty reports. Amnesty's actions unfairly devalue their work.
Another good rule of thumb is to assume stupidity not malice.
Ultimately rules of thumbs are just that, and shouldn't be over relied on for things that matter.
> Gareth Sella was blinded in his left eye when a police officer in Bogotá shot him with a rubber bullet at the protests. He argued that hiding the identity of protesters makes sense to protect them from ending up in jail on inflated charges. “As the UN has documented, the state has continued pursuing protesters and more than 100 are in jail, many with disproportionate sentences, such as terrorism, to make an example of them. Hiding our identities seems sensible to me given that two years on we continue living in the fear that we could be jailed at any moment or even that they come after us on the streets,” Sella said.
People in first world countries often underestimate just how dangerous political protests are in other parts of the world. It's not just about getting arrested or injured at the event itself, but about being identified and targeted many weeks afterwards. The protesters are certainly willing to risk a lot, but they are taking every step possible to not martyr themselves.
But yeah, it's a hole bunch of nothing here. Nobody was lying or trying to mislead.
> To avoid misleading the public, the images included text stating that they were produced by AI.
So doesn’t this really only impact the people… careless enough not to RTFA?
And then we get this hackneyed parting shot, parroting common myth/misconception about AI generated images:
> AI-generated images stitch together photographs previously taken by humans to create new ones…
Which, no, that makes it sounds like a collage,
> …raising questions of plagiarism and in photojournalism and the industry’s future. Torres said Amnesty’s use of AI images was an insult to the photojournalists who cover protests from the frontline.
“Won’t someone please think of the photojournalists!”
I am not impressed with this article.
It seems to me that if this were an illustration by an artist no one would care. So it seems to me they should just use "in the style of".
Which... sure, there might be human rights violations, I don't necessarily doubt there are innocent people caught in the crossfire. However, what they don't mention was that El Salvador had a murder rate as high as 103 per 100K in 2015, the highest murder rate in the world by country by almost double the next highest. Is that not a human rights crisis worth mentioning? It was so bad that BBC reporters were seeing dead bodies in the streets.
After the crackdowns (and, if you watch El Salvador's video, easily 90%+ of prisoners in their TCC have visible gang tattoos on them), they are now sitting at a murder rate of 8 per 100K. Not surprisingly, the president has a 90%+ approval rate.
That's what I don't get Amnesty International. From 103 to 8 per 100K. Thousands and thousands of lives will be saved every year. Is that worth the risk of a some innocent people with obvious gang tattoos being swept up in it? I'd say... yes, because we don't live in a perfect world, and El Salvador was quite literally in a state of war. But Amnesty International is more focused on their story than the people who obtained freedom from having to live in one of the most dangerous countries on earth. Think only about potential innocent victims in prison, but the 6,500 people who were murdered in a single year shouldn't be considered victims in the equation. I think it's pretty reasonable to say that, if there is a crackdown that arrests 60,000 (alleged) gang members, and has (let's say, hypothetically) 5,000 people innocent arrested (which, from the video, I think is high, but let's just say), but results in over 6,500+ lives saved every year (let alone countless rapes including of children and other crimes), there's a just cause.
I would have had much more respect for them, and I think the people of El Salvador would have had more respect for them, if they had accepted the crackdowns as a necessary evil under desperate circumstances; and then worked to identify and free anyone incorrectly swept up in the arrests. They could have done that - they could have started working hard to free the actually innocent, on a case-by-case basis. Instead they just denounced the whole thing... which, surprise, hasn't done anything. It definitely makes them feel good, even if it does literally nothing.
We all laughed when we learnt this, but didn't think it was a big deal. Their privacy policy makes a ton of sense - if I was a foster child, I wouldn't want that to be advertised publicly. And the organization would have hamstrung itself if none of their posters featured any children. People don't donate or volunteer based on facts and logic. They do it based on empathy. Not having children's pictures at all, or only blurred photos, would be disastrous for fundraising and getting the foster children their much needed resources.
To be clear, I do not endorse lying at all. Lying makes things much worse in the long run. But the above non-profit wasn't lying in any sense. They simply showed photos of average looking kids in neutral settings, displaying neutral expressions. There was no misinformation in any of their photos. I haven't met a single donor to date who felt lied to or betrayed by their posters.
I can see the parallels between what Amnesty Intl did and the above. They have a very credible reason for not using real photos. The people protesting themselves have requested hiding their identities to avoid life-changing persecution. And Amnesty states that they explicitly labelled their images as having been produced by AI. They have clearly been acting in good faith.
In an ideal world, non-profits wouldn't need to rely on AI generated images... and people wouldn't require pictures to mobilize and get involved. But let's be realistic - there's minimal rationality in people's engagement with non-profits. I skim through the news everyday, and didn't even know there were protests happening in Colombia. The only possible way for organizations like Amnesty International to get their message out there is by using attention-grabbing images. And as long as they make a good faith effort to portray them realistically and label their images as AI-generated, I don't have a problem with it.
> To avoid misleading the public, the images included text stating that they were produced by AI.
> Gareth Sella was blinded in his left eye when a police officer in Bogotá shot him with a rubber bullet at the protests. He argued that hiding the identity of protesters makes sense to protect them from ending up in jail on inflated charges. “As the UN has documented, the state has continued pursuing protesters and more than 100 are in jail, many with disproportionate sentences, such as terrorism, to make an example of them. Hiding our identities seems sensible to me given that two years on we continue living in the fear that we could be jailed at any moment or even that they come after us on the streets,” Sella said.
One difference is that there is no controversy over whether foster children exist. And, as you say, they used neutral pictures of kids with neutral expressions -- what if they had instead used (staged) photos of kids in distress, to get more sympathy? (And still, nobody denies that foster kids in distress exist, but this already seems questionable, right?)
The things Amnesty highlights, some people want to challenge the fact that they happened, that they even exist, or happened like Amnesty said, or were as bad as Amnesty said. Credibility as being honest and accurate in reporting on what happened is pretty crucial to Amnesty's ability to succeed at it's goals.
But perhaps the Amnesty staffers were thinking about it like you, and that it was the same. I don't think it is, at all.
Certainly they have no obligation to put people at risk by including pictures in which people can be identified. There are many options that are not using manufactured pictures, including pictures without visible faces or obscuring faces. Using real-looking but manufactured/staged pictures is a really bad choice. Even if it would get more attention or dollars or support than other choices, the consequences are disastrous for Amnesty's credibility at exposing things that some forces want to say didn't happen or weren't as bad as Amnesty says.
This is high-school skill, don't be fooled.