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

What Happens When AI-Generated Lies Are More Compelling Than the Truth?

behavioralscientist.org

Post reply on HN
Print view View original

81–90 of 111 posts

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#81

Post by A_D_E_P_T »

brookst wrote: post #41
A_D_E_P_T wrote: post #29

Earlier quoted context omitted.

The watermark would still be there. You can take a photograph of the image with an old Polaroid and, if the resolution is high enough, the watermark would still be there.

If it can be detected, it can be removed. Asking that false things carry claims of falseness is a dead end.

Let's say you want to release a deepfake image. Your AI image, as generated, is not watermarked or hashed. Difficulty: Story-mode. You can release your image in two minutes flat.

Let's now let's say your AI deepfake image is watermarked/hashed upon generation. Difficulty: Intermediate to hard. You might be able to get it done, but not without real effort, and removing all traces of the watermark without leaving artefacts might be so difficult as to be nearly impossible.

...So it doesn't eliminate the possibility of fakes, but it hugely raises the cost and effort bar. Only the most motivated would bother.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#82

Post by neepi »

bayindirh wrote: post #78
neepi wrote: post #75

Earlier quoted context omitted.

Used to work on archival. You want to store your images in TIFF and use FEC if you really care. That withstands significant bitrot. Signing will just tell you that the image is broken but not allow you to recover it. You can do fine with SHA256 and 3-2-1 backup.

Thanks for the heads-up, didn't know TIFF/FEC. However, I can't because the images are .ARW (Sony's camera raw format). I keep multiple copies with hashes of them currently. I check the disks periodically. Mid-term plan is to build an UNRAID array with patrolling. Will probably do backups with Borg on top of it, and keep another "working copy", so I have multiple layers of redundancy with some auto-remediation. UNRAI…

You have to convert to an archival format from whatever your raw format is. The issue with raw files is that they are compressed so any bitrot in those leads to more than just losing a pixel or two here or there but sometimes the entire file. TIFF is not compressed at all. FEC allows single pixel corrections to be done accurately. We had a metric shit ton of LTO tapes, changers and online disk cache that handled all that.

I don't do that at home. I have 600GB of Nikon raws. I keep one copy on my Mac. I have a time machine backup in the house (in another room) on an Samsung T7 shield and I have an off site backup which is a quarterly rsync job to another Samsung T7 shield.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#83

Post by psychoslave »

>Once we take the counterfeits for granted, we may begin doubting the veracity of all the information presented to us through media.

Hmm, that's not a totally new stuff. I mean, anyone taking take time to document themselves about how mass media work should already be acquainted by the fact that anything you get in them is either bare propaganda or some catch eye trigger void of any information to attract audience.

There is no way an independent professional can make a living while staying in integrity with the will to provide relevant feedback without falling in this or that truism. Audience is already captured by other giant dumping schemes.

Think "fabric of the consent".

So the big change that might occurs here, is the distribution of how many people do believe what they get thrown at there face.

Also, the only thing that you might previously be taken as only reliable information in a publication was that the utterer of the sentence knew the words emitted, or at least had the ability to utter its form. Now you still don't know if the utterer had any sense of the sentence spoken, but you don't even know if the person could actually even utter it, let alone have ever been aware of the associated concepts and notions.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#84

Post by rightbyte »

neepi wrote: post #27
HPsquared wrote: post #2

There was a brief window where photography and videos became widespread so events could be documented and evidenced. Generative AI is drawing that period to an end, and we're returning to "who said / posted this" and having to trust the source rather than the image/message itself. It's a big regression.

I await cryptographically signed photos coming out of cameras. Actually I’ve said that should be the case for the last 20 years. You should be able to follow a chain of evidence towards an unprocessed raw image potentially.

Wouldn't compression make the signature invalid? Feels kinda easy to fake anyways.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#85

Post by psychoslave »

j-bos wrote: post #53
RcouF1uZ4gsC wrote: post #34

Earlier quoted context omitted.

Remember the fairy photo hoax that fooled Lewis Carrol? There was never a time when authenticated photo and video could be trusted without knowing the source and circumstances.

It fooled Sir Arthur Conan Doyle, a motivated believer, not Lewis Carrol. People will believe what they want. Trust is the fundamental issue.

Lewis Carroll also was fooled by it, when Churchill showed to him. Abraham Lincoln who was there at the moment it happened confirmed that to me, I can show you the original email he sent me about it (bar the elements I'll have to hide due to top secret information being included in the rest of the message).

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#86

Post by alpaca128 »

Elaris wrote: post #66

This got me thinking. Sometimes it feels like a story doesn’t have to be true as long as it feels right, people believe it. And if it spreads fast and sounds good, it becomes “truth” for many. That’s kind of scary. Now that anyone can easily make something look real and convincing, it’s harder to tell what’s real anymore. Maybe the best thing we can do is slow down a bit, ask more questions, and not trust something j…

True, people tend to believe things more if they are stated more confidently. That's basically how many things from scams to conspiracy theories and even cults work. Now with LLMs you have machines that can produce thousands of words per hour in flawless grammar and sounding as sophisticated and confident as you want.

The famous quote "A lie can travel halfway around the world while the truth is putting on its shoes" is older than the internet, so this asymmetry was already bad enough back then and whoever coined the quote couldn't have imagined how much farther it would shift.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#87

Post by rightbyte »

SupremumLimit wrote: post #16
rightbyte wrote: post #4

Earlier quoted context omitted.

Videos and photos have been faked for a long time. Nothing has changed in that regard but decreasing the effort required somewhat.

Your comment follows two persistent HN tropes: (1) ignoring the article, which deals precisely with why the cost of production matters, and (2) steadfastly refusing to recognise that quantity has a quality of its own - in this case a monumental reduction in production cost clearly leads to a tectonic reshaping of the information landscape.

I did read the article and it does deal with fake photos being an old thing and name drops e.g. Stalin's former companions being removed one by one from photos.

"There was a brief window where photography and videos became widespread so events could be documented and evidenced."

Photos are video have in themself not been evidence. You need trust the photographer or publisher too, since the camera was invented.

Moving the cost from big $$$ level to my neighbour level makes scams more personalized, sure.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#88

Post by drweevil »

What is a lie, and what is the truth? These are age-old questions, not some recent phenomenon. The Spanish-American War was at least in part precipitated by the infamous "yellow journalism" of the time. Propaganda and disinformation played a large role in the events leading to and including WWII. And what is The Truth? Using photography as an example, lies can easily be told by omission, even without any dark-room chicanery. What is the photographer's subject? What is off-frame? Which photographs did the editor select for publication? What story is not being told?

If anything, the idea that one can take information as "true" based on trust alone (what does the photograph show, what did the New York Times publish) seems to be a recent aberration. AI will be doing us a favor if it destroys this notion, and encourages people to be more skeptical, and to sharpen their critical thinking skills. Forget about what is "true" or "false." Information may be believed on a provisional basis. But it must "make sense" (a whole subject in itself), and it must be corroborated. If not, it is not actionable. There simply is no silver bullet, AI or no AI. Iain M Bank's Culture series provides an interesting treatment of this subject, if anyone is interested.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#89

Post by rightbyte »

jdiff wrote: post #7
rightbyte wrote: post #4

Earlier quoted context omitted.

Videos and photos have been faked for a long time. Nothing has changed in that regard but decreasing the effort required somewhat.

"Somewhat" is doing an awful lot of heavy lifting. Fakery has gone from handcrafted in small batches (or at worst, produced through much effort in sweatshops) to fully automated and mass produced.

Ye well "alot" might have been a better wording.

However, the real breaking point in my view was when shills and semi-automated bots become so prevalent that they could fool people into believing some consensus had changed. Faking photos doesn't add much to that in my view.

Re: What Happens When AI-Generated Lies Are More Compelling Than the Truth?

#90

Post by salawat »

grues-dinner wrote: post #72
lnrd wrote: post #68

Earlier quoted context omitted.

What prevents anyone to take a signed picture by photographing a generated/altered picture? You just need to frame it perfectly and make sure there are no reflections that could tell it's a picture of a picture and not a picture of the real world, very doable with a professional camera. All details that could give it out would disappear just lowering the resolution, which can be done in any camera.

With a bit (OK quite a lot) of fiddling, you could probably remove the CCD and feed the analog data into the controller, unless that's also got a crypto system in it. Presumably if you were discovered you would then "burn" the device as its local key would be known then to be used by bad actors, but now you need to be checking all photos against a blacklist. Which also means if you buy a second hand device, you might…

Any problem that requires cryptographic attestation or technical control of all endpoints is not a solution we should be pursuing. Think of it as a tainted primitive. Not to be implemented.

The problem of Trust is a human problem, and throwing technology at it just makes it worse.

Post reply on HN