Submitted 10 hours ago by theangriestbird@beehaw.org to technology@beehaw.org
https://www.404media.co/the-un-made-ai-generated-refugees/
In early tests at a workshop attended by humanitarian organizations, refugee aid groups, and nonprofits, Albrecht and Fournier-Tombs said the reactions were strong and that many were negative. “Why would we want to present refugees as AI creations when there are millions of refugees who can tell their stories as real human beings?” one person said
I love how the article then proceeds to not answer this question. What a dumb idea. What a waste of UN funds.
I feel like the article answers the question, or rather it gives the researchers a chance to answer the question:
When I spoke with them, both Albrecht and Fournier-Tombs were clear that the goal of the workshop was to spark conversation and deal with the technology now, as it is.
“We’re not proposing these as solutions for the UN, much less UNHCR (United Nations High Commissioner for Refugees). We’re just playing around with the concept,” Albrecht said. “You have to go on a date with someone to know you don’t like ‘em.”
Fournier-Tombs said that it’s important for the UN to get a handle on AI and start working through the ethical problems with it. “There’s a lot of pressure everywhere, not just at the UN, to adopt AI systems to become more efficient and do more with less,” she said. “The promise of AI is always that it can save money and help us accomplish the mission…there’s a lot of tricky ethical concerns with that.”
She also said that the UN can’t afford to be reactive when it comes to new technology. “Someone’s going to deploy AI agents in a humanitarian context, and it’s going to be with a company, and there won’t be any real principles or thought, consideration, of what should be done,” she said. “That’s the context we presented the conversation in.”
The goal of the experiment, Albrecht said, was always to provoke an emotional reaction and start a conversation about these ethical concerns.
“You create a kind of straw man to see how people attack it and understand its vulnerabilities.”
So if you read the headline and have the obvious visceral reaction, if you are asking yourself that question from the article, it kind of sounds like that is the point. They’re doing it now so that if people see it and say “that’s stupid”, hopefully that stops xAI or someone else from trying this to profit on the suffering of poor people. Alternatively, if people see it and say “wow this actually helped me understand”, that is also useful for the world at large. It doesn’t sound like the latter is the case, but that’s why you test a hypothesis.
Those are kind of non-answers… “Why the fuck are you doing that?” and the answers are all “Well, somebody’s probably doing it at some point, so why don’t we do it now?” or “you gotta try stuff” as if that explains anything. Like, no, there are some things that don’t need to be tested. This is arguing on the level of “Caaaaarl, that kills people!” You don’t need to punch people in the face to know that’s a dumb thing to do. You don’t need to spill milk to know it’s a dumb thing to do. And you sure as fuck don’t need to date somebody you dislike to know that fucking them is a dumb thing to do or create ai refugees as the UN to know it’s a dumb thing to do! Like, what argument is that? We’re not talking to three-year-olds that have never touched a candle! The UN should be able to anticipate the consequences of their actions! ESPECIALLY IF THEY HAD WORKSHOPS WHERE PEOPLE TOLD THEM IT’S A FUCKING DUMB THING TO DO!!
sleepundertheleaves@infosec.pub 47 minutes ago
Oh my fucking God.
Privileged people in the West already struggle to believe that refugees are actually people.
So the UN creates another barrier to understanding, and lets privileged Westerners tell themselves “this is only a simulation, actual people don’t have it that bad”?
Silencing the real voices of actual refugees seeking help, and replacing them with a fucking autocorrect algorithm?
I feel like I can handle most AI bullshit calmly and objectively at this point. But AI generated fake refugees? I didn’t think I could get this angry about AI anymore.
And I can’t even wish horrible fates on the assholes who implemented these AIs because I feel like they had the best of intentions, to raise awareness about actual refugees and so on and so forth. They were just so fucking stupid they thought this was a good way to do it.
Jesus fucking Christ.