Do deepfake videos undermine our epistemic trust?
journals.plos.org
journals.plos.org
Is it possible to insert deep fakes in there? Sure, but the technology is defonot good enough to not be able to tell it apart.
Obviously it got downvoted to "grey".
I'm always fascinated that hacker news, a forum for many developers and other, what I'd assume, "higher intelligent" people, fails to see this.
> 100% obvious everything media says is a lie
Having lived in an authoritarian state where everyone had pure "eat or get eaten" mentality I strongly disagree. Epistemic trust is not a requirement for a society to function.
It is much safer to assume everyone is there to get you (by scamming, misinformation, theft, or plans to hurt you physically just for the fun of it) than otherwise. Having tgat mindset has never stopped me from trusting people that have shown themselves to be trustworthy.
> This is why telegram and in a way twitter exploded as a source of news during this war.
By that logic, the Russians would have kicked Putin out of power by now.
Is this a bad thing? Is the photograph really a cornerstone of our civilisation?
It's not even just about deepfakes. Photographs taken by must smartphones are already heavily embellished by default. With just a few extra taps you can remove unwanted members of the public from your pictures. It's all fake, and the generations growing up with it will understand photographs just like we understand paintings. Nice, but not true.
This superficial information is becoming harder and harder to interpret. If you follow the trendline we will approach a world where will skeptical of every single thing you see on the internet with no way of verifying the truth.
For example, I told an AI to write this for me. And I also told the AI to insert some common spelling mistakes and grammatical mistakes to make it look real. Then I told the AI to lie to the best of it's ability and say something untrue and manipulative.
you are reading the output of what that AI wrote right now.
But when you dig deeper in the public narrative of many subjects you find out that the narrative is misleading and sometimes even totally wrong on almost any subject you look into. By this I mean that for example anything the newspapers write about on any subject that you know deeply about, they get wrong on some very important aspect. You just extrapolate this experience over all the possible subjects.
Then there comes a moment of realisation that you are spending lots of energy in maintaining all these public narratives in your head. And as the narratives are misleading or wrong, they are blinding you from seeing some of the facts. And as a consequence you do not see the situation as clearly as you could. And you may start clearing your mind off these narratives, releasing the energy you are spending to maintain them.
And then you might be able to develop a sense of "smell" or intuition when people are not being honest, but are trying to manipulate you or sell you some idea. And this helps in recognising the situations without the need to spend energy on being analytically skeptical.
"We found that much of the misinformation in our dataset came from labelling real media as deepfakes."
There is no real solution to this problem, but my thoughts on the matter are that in the future it will be common for cameras or recording devices to have a cryptographic signing key and the resulting footage would contain a signature. This would allow news outlets to prove to their audiences that the footage is untampered with. It could also be chained so that officially licensed syndicates could also offer provenance to their audiences, or that clips/re-encodes of the same footage could link back to an original recording.
This doesn't stop organisations or individuals from recording a deepfake, but it would put their reputation on the line (similar to how we have certificate authorities signing SSL certificates now).
It's up to you then if you want to trust the footage provided by AP vs. the footage provided by Infowars.
Certainly you could also just trust the photographer, but a chain of trust might be preferable to the general viewer (eg. I don't necessarily trust Photographer X, but I trust The Guardian and if The Guardian trusts this photo taken by Photographer X to be authentic, then I trust this image to be authentic).
Edit: I read more about the CAI and C2PA[1] and I think it would be sufficient for provenance and origin purposes, so what is missing is the trust network. It's likely going to a while for media outlets to establish trust networks and phase out the old, untrusted ways. There's definitely going to be room for both and I imagine outlets will start using terms like "we have received images from multiple sources, however their authenticity has not be verified by our editors".
[1] https://c2pa.org/specifications/specifications/1.3/specs/C2P...
Did you know, autistic people and advertisements don't go together: https://www.psychologytoday.com/gb/blog/the-fallible-mind/20...
In my life, I've always found advertising toxic fake and kind of offensive? Sometimes I would even personally boycott the product being advertised.
Like, most advertisements, especially super bowl ones are entirely CGI these days. I used to try to spot the CGI game but now its more of a 'spot the not cgi'
Then when it comes to deepfakes and I just see right through them. It's black/white, its like watching anime and thinking its real life or something. How can something like that damage trust? It doesnt bother me in the slightest.
In the final analysis, one can only say 'one knows' about those things one has personally verified to be true. The rest is hypothetical.
And it is going to be more interesting when we will have the ability to manipulate videos (near-)realtime.
It was near impossible to create convincing fake videos of people speaking ( https://youtu.be/X17yrEV5sl4 ) and so even those who would not take an altered screenshot at face value might be more inclined to believe one such.
> “Early in the Reticulum-thousands of years ago-it became almost useless because it was cluttered with faulty, obsolete, or downright misleading information,” Sammann said.
> “Crap, you once called it,” I reminded him.
> “Yes-a technical term. So crap filtering became important. Businesses were built around it. Some of those businesses came up with a clever plan to make more money: they poisoned the well. They began to put crap on the Reticulum deliberately, forcing people to use their products to filter that crap back out. They created syndevs whose sole purpose was to spew crap into the Reticulum. But it had to be good crap.”
> “What is good crap?” Arsibalt asked in a politely incredulous tone.
> “Well, bad crap would be an unformatted document consisting of random letters. Good crap would be a beautifully typeset, well-written document that contained a hundred correct, verifiable sentences and one that was subtly false. It’s a lot harder to generate good crap. At first they had to hire humans to churn it out. They mostly did it by taking legitimate documents and inserting errors-swapping one name for another, say. But it didn’t really take off until the military got interested.”
> “As a tactic for planting misinformation in the enemy’s reticules, you mean,” Osa said. “This I know about. You are referring to the Artificial Inanity programs of the mid-First Millennium A.R.”
> “Exactly!” Sammann said. “Artificial Inanity systems of enormous sophistication and power were built for exactly the purpose Fraa Osa has mentioned. In no time at all, the praxis leaked to the commercial sector and spread to the Rampant Orphan Botnet Ecologies. Never mind. The point is that there was a sort of Dark Age on the Reticulum that lasted until my Ita forerunners were able to bring matters in hand.”
Haven't read Anathem and couldn't find in any summary who the Ita are. Stephenson's story being about a post-societal collapse I fantasise that that ITA are the Information Truth Authority.
Well, if Stephenson didn't mean that, I'll just make it up, right? :) Because anyway, that's the thing we need to be mindful of and avoid in the coming years.
In some ways I see LLM-AI and DeepFakes as being the undoing of the printing press and the camera respectively.
Both inventions "liberated truth". Gutenberg's press eroded the authority of the church and kick-started the real scientific revolution. The camera changed popular conceptions of geography and history. Together they created "News" as a concept in modernity, even before electronics brought instantaneous television.
If AI undoes this, a good way to understand the worst future is to rewind to a pre-Gutenberg and pre-photographic era. A key dystopian feature would be a singular centralised truth and information authority, essentially replacing the church circa 14th century.
IT Administrators :-)
Baseless really? I was shown the photo of a hospital bombed in Gaza. I was told there was 500 dead. I was told it was an IDF bomb. The NYT relayed Hamas propaganda without a second thought. How is it baseless?
Trust in journalism is at an all time low for good reasons. And I agree with you that deep fakes can't do much more damage. It is just the 2020's photoshop.
The root cause of that fubar in the heat of reporting is well documented - western media sources echoed each other when one first used an english mistranslation of an arabic tweet - an estimate of 500 injured was rapidly "reported" in english media as 500 confirmed dead.
That's typical of the shitty social media driven infotainment bend that's come to be called "journalism" but really isn't.
Sadly the first wave of "in first before ..." is what everybody remembers.
The answer should have been yes after about Terminator 2 was released, and has nothing to do with AI.
"We found that much of the misinformation in our dataset came from labelling real media as deepfakes."
That's basically the liar's dividend at work.
Also some of it probably has to do with generally feeling bad, anxious or uncertain about the future. Some might take solace in inflicting those same feelings upon others who seem to have it better.
But it is also the case that 'the other side' often rallies behind obvious nonsense dictated by the identity politics of the day (like IQ isn't real, math is racist, etc.). There are probably similar feelings at play there, a want for belonging of some sort.
Because he sure does in a deepfake I saw.