Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

404media.co

Post reply on HN
Print view View original

181–190 of 320 posts

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#181

Post by int_19h »

RecycledEle wrote: post #80
everdrive wrote: post #6

This was one of the most obvious outcomes anyone could have and should have predicted. Even when people know an image is fake, it still seems to be able to sway them. The effect we've seen over the past few years is simply likely to become worse: more polarization, more extremism, more inter-group hostility, and more BS online.

> Even when people know an image is fake, it still seems to be able to sway them. Yes, this has been happening for a long time. This is why fiction literature exists. If images we know are fake could not move our emotions, nobody would have ever laughed at a cartoon.

I don't think OP meant "sway" in that sense.

One thing that I'm increasingly seeing is when you point a fake out to a person, and they just shrug and say, "okay, so this one was fake, but it's still a good illustration of how things are in general, so what's the problem?".

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#182

Post by dkersten »

reustle wrote: post #114
dissident_coder wrote: post #108

Earlier quoted context omitted.

This is why I refuse to turn off my adblocker on YouTube, and never will. I'm not rewarding YouTube with my money (for YT Premium) for being absolutely useless at vetting their advertisers. I'd rather just not use YouTube at all if it comes to it.

You eventually get to a scale where vetting is near impossible. Paying for premium brings them closer to a state where they don’t need advertising at all.

No, they can still vet, just doing so would make them less money. That is their problem, not yours or mine. If they can’t figure out a way to vet at scale, then they don’t deserve that scale.

Most of us could be more profitable if we decided to act unethically or illegally. Serving scam adverts (and my YouTube and Google adverts appear to be substantially more scam than not) is Alphabet saying they care more about profit than anything else and they should, quite frankly, be punished for it, both through things like fines from advertising standards bodies, and by more people taking matters into their own hands by using Adblock. I personally would also vote for any politician who put laws into place to make executives personally liable if their platforms are found to not do enough to get and combat fraud and scams on their platforms, because these companies actively profit from these scams through advertisement revenue.

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#183

Post by int_19h »

kube-system wrote: post #107
somenameforme wrote: post #92

I see a silver lining here. One of the worst parts about social media is image crafting. And it even seeps into real life on occasion where you see people at incredible locations spending all their time trying to craft some perfect pic to put up, instead of just soaking in the location or even really enjoying themselves. If anybody can [literally] image craft anything, then the entire point of online image crafting d…

...and then our social relationships are judged on a default presumption of distrust?

Aren't they already, given that the practice of "crafted" social media feed has been in the public awareness for over a decade now?

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#184

Post by carlosjobim »

philomath_mn wrote: post #170
carlosjobim wrote: post #151

Earlier quoted context omitted.

When the normal Internet becomes unusable because it is filled with AI spam, people will still find untainted information among the 20 million books and 100 million scientific papers in the shadow libraries. If we have a software that can make browsing books a little bit more like browsing the web, then usability will be better.

There are already many examples of AI generated books spamming Amazon -- how do we separate the wheat from the chaff?

People publish AI generated crap on Amazon because they make money from it and Amazon allows it because they also make money from it. People publish AI generated websites because they make money from it, and Google eagerly put blogspam on top because Google makes money from it.

The shadow libraries make no money by sharing books, instead the librarians risk prison for doing it. So there is no incentive for them to include AI spam books. There is also very little incentive for creators of AI spam books to try to have them included in the shadow libraries.

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#185

Post by int_19h »

keybored wrote: post #96
CrzyLngPwd wrote: post #14

Earlier quoted context omitted.

Maybe it's time to stop caring so much about things on the internet and focus on the truth that is in front of you, the real things.

This is the Postmodern Age. People have been writing about this phenomena decades before the widespread adoption of the Internet and before geniuses began penning opinion pieces about how it’s possible to believe untruths now.[1] Which means that (in turn) you can’t just “log off” and be rid of the problem since it predates... logging off. Neo thought he could just “log off” in The Matrix . But then then it turned ou…

Orwell wrote this in a letter in 1944:

"Everywhere the world movement seems to be in the direction of centralised economies which can be made to ‘work’ in an economic sense but which are not democratically organised and which tend to establish a caste system. With this go the horrors of emotional nationalism and a tendency to disbelieve in the existence of objective truth because all the facts have to fit in with the words and prophecies of some infallible fuhrer. Already history has in a sense ceased to exist, ie. there is no such thing as a history of our own times which could be universally accepted, and the exact sciences are endangered as soon as military necessity ceases to keep people up to the mark. Hitler can say that the Jews started the war, and if he survives that will become official history. He can’t say that two and two are five, because for the purposes of, say, ballistics they have to make four. But if the sort of world that I am afraid of arrives, a world of two or three great superstates which are unable to conquer one another, two and two could become five if the fuhrer wished it. That, so far as I can see, is the direction in which we are actually moving, though, of course, the process is reversible."

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#186

Post by philomath_mn »

cubefox wrote: post #27
philomath_mn wrote: post #15

I wonder if we will get to a point where the web is so saturated with this junk that the web itself starts to lose its luster. If AI can dream up a million impressive wood carvings, it seems like the real carving I saw in person is the only one that is actually interesting.

Perhaps the real wood carving is also no longer interesting, because you already saw so many (fake ones) before.

For me, the interesting part is the human who took the time to learn and apply the skill, along with the backstory of that human's choices (e.g. making a carving to honor a loved pet).

In the olden times, you could enjoy an account of a story like that on Facebook because it was supposed to be real (it often wasn't, but there was at least a kernel of truth). Now it seems like I can only experience "authentic" stories / content via first or very limited second hand experience (from a select group whom you trust)

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#187

Post by jvolkman »

yjftsjthsd-h wrote: post #176
jvolkman wrote: post #120

Earlier quoted context omitted.

YouTube offers an ad-free service. (yes, I know content providers can still embed their own promotions into their videos).

If the "ad-free" service features videos with inline ads, it's not really ad-free.

The inline ads are part of the content you're watching, not injected by YouTube. You may not like them, but there's a distinction.

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#188

Post by int_19h »

simiones wrote: post #63
CrzyLngPwd wrote: post #14

Earlier quoted context omitted.

Maybe it's time to stop caring so much about things on the internet and focus on the truth that is in front of you, the real things.

Well, if you care about your city or country, you have to care about many things that are not immediately in front of you. And the Internet is one of the main places where such things can be learned about.

Lately it feels more like things that aren't "immediately in front of you" are getting dragged into conversations where they don't really make sense.

This year, my town had a referendum on whether to raise taxes to fund a public pool to replace a 70-year-old one that is falling apart. The public conversation was only focused on actual pros and cons of that particular project rather briefly, quickly devolving first into a general fight over development vs lack thereof, and from there to national economic policy, crime etc. And I have a distinct feeling that most people who actually voted ended up being swayed (one way or the other) by those arguments - the partisan alignment of support & opposition at the end was very blatant.

Re: Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real

#190

Post by int_19h »

brvsft wrote: post #165
kleiba wrote: post #136

How is it possible to generate a stolen image?

Just more pearl-clutching about AI. The images are mimicking several original images of a person with some artwork they made. The 'stolen' images are of a very similar form, but with changes to the person and the artwork. Presumably, they make fake versions of the images to gain clout on FB or elsewhere without having to credit the original artist. While it is ethically dubious, I don't really care. The pearl-clutchi…

According to the article, they're using img2img to make those. That's very clearly a derived work regardless of your stance on AI art in general.
Post reply on HN