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

High-tech deception of ‘deepfake’ videos

apnews.com

Post reply on HN
Print view View original

51–60 of 70 posts

Re: High-tech deception of ‘deepfake’ videos

#51

Post by abruzzi »

hurpaDurpa wrote: post #45
mjw_byrne wrote: post #37

Deepfakes are surely no more a "problem" than photoshopped still images? Everyone knows that a still image can be altered to (convincingly) depict something that never happened. So, a still image is not considered good evidence of anything unless it has a verifiable provenance, chain of custody etc. Presumably that wasn't always the case, but now that photo editing tech is universally accessible, we all know we can't…

I'd have to agree. In fact, I'm actually glad to see some pushback against the seemingly indelible veracity of photographic imagery. Photos, and by corollary video, haven't been incontrovertible for at least two decades, and Hollywood movies are a clear demonstration of that fact. We won't really be able to rely on pictures as proof, at some point in the immediate future. Unfortunately, in an ever-more terrifying far…

I think the pictures-as-proof is a straw man. There are a lot of people that believe what they see if it reinforces their existing beliefs. If you make videos of Trump saying something bad, people on the left will believe it, even if the provenance is weak, and people on the right will be inclined to disbelief. A similar fake video with Obama or Clinton will have an opposite effect. Much of that effect can be had without any provenance at all due to the general ability of videos to circulate, usually though social media, with little to no real “authoritative source” to validate them.

Re: High-tech deception of ‘deepfake’ videos

#52

Post by mjw_byrne »

roywiggins wrote: post #42
mjw_byrne wrote: post #37

Deepfakes are surely no more a "problem" than photoshopped still images? Everyone knows that a still image can be altered to (convincingly) depict something that never happened. So, a still image is not considered good evidence of anything unless it has a verifiable provenance, chain of custody etc. Presumably that wasn't always the case, but now that photo editing tech is universally accessible, we all know we can't…

The volume alone might be a problem. It takes some Photoshop skills to pull off an acceptable fake. If there's a turnkey website that can spit them out en masse, that's something different than having to actually hire a human to do it by hand. Previously it took at least one CGI artist to convincingly fake video, probably several. This inherently limits how much of a problem it can be, because it doesn't scale. AI sc…

I agree, but I think my original point covers this - if video proves to be very easy to fake at scale (e.g. using your example of a turnkey website) then that should result in equally large-scale scepticism about video as evidence. The internet may drown in annoying deepfake-based memes but as long as everyone knows its all fake, there seems little potential for the kind of serious harm people are worried about (reputational damage, election meddling etc.)

Re: High-tech deception of ‘deepfake’ videos

#53

Post by gnicholas »

mizzack wrote: post #36
gnicholas wrote: post #31

Earlier quoted context omitted.

I would think that there'd be enough deepfakes on both sides that everyone would recognize that iff there are tools that accurately detect deepfakes, it would be useful to have them enabled on social networks. They don't need to censor the videos — just flag them as having deepfake characteristics.

Like when Facebook (briefly) tagged articles from certain sources as being disreputable and it led to an _increase_ in clicks to those articles? People will believe what they want to believe. We're also seeing that they don't like mega corps acting as arbiters of truth. So, I'm not sure what, if any, effect this moderation/curation effort would have.

I think what we really need is an embarrassing term for "someone who was fooled by a deepfake" (like "catfish"), so that there will be a social stigma that incentivizes people to not be tricked by them.

I don't currently have any better ideas that "getting deepfaked". Suggestions?

Re: High-tech deception of ‘deepfake’ videos

#54

Post by save_ferris »

mjw_byrne wrote: post #52
roywiggins wrote: post #42

Earlier quoted context omitted.

The volume alone might be a problem. It takes some Photoshop skills to pull off an acceptable fake. If there's a turnkey website that can spit them out en masse, that's something different than having to actually hire a human to do it by hand. Previously it took at least one CGI artist to convincingly fake video, probably several. This inherently limits how much of a problem it can be, because it doesn't scale. AI sc…

I agree, but I think my original point covers this - if video proves to be very easy to fake at scale (e.g. using your example of a turnkey website) then that should result in equally large-scale scepticism about video as evidence. The internet may drown in annoying deepfake-based memes but as long as everyone knows its all fake, there seems little potential for the kind of serious harm people are worried about (repu…

> ...as long as everyone knows its all fake,

There's a shockingly large number of people who aren't even convinced that election meddling took place in 2016, or that propaganda was disseminated to many people on social media, despite unanimity from US intelligence.

If we as a society can't all agree that these things even happened, how can we hope to properly educate people on the existence/prevalence of fake video? It seems like common sense to many people, but we can't assume that everyone will accept these truths based on the political rhetoric we've seen in the last few years.

Re: High-tech deception of ‘deepfake’ videos

#55

Post by function_seven »

mjw_byrne wrote: post #37

Deepfakes are surely no more a "problem" than photoshopped still images? Everyone knows that a still image can be altered to (convincingly) depict something that never happened. So, a still image is not considered good evidence of anything unless it has a verifiable provenance, chain of custody etc. Presumably that wasn't always the case, but now that photo editing tech is universally accessible, we all know we can't…

The problem isn't so much that fake videos will be used to lie. It's that valid video evidence can be dismissed as Fake News with some credibility. This could be bad for the overall definition of "truth".

Video of the Prime Minister/President taking bribes? Eh, you can't trust these videos these days!

Re: High-tech deception of ‘deepfake’ videos

#56

Post by mjw_byrne »

abruzzi wrote: post #51
hurpaDurpa wrote: post #45

Earlier quoted context omitted.

I'd have to agree. In fact, I'm actually glad to see some pushback against the seemingly indelible veracity of photographic imagery. Photos, and by corollary video, haven't been incontrovertible for at least two decades, and Hollywood movies are a clear demonstration of that fact. We won't really be able to rely on pictures as proof, at some point in the immediate future. Unfortunately, in an ever-more terrifying far…

I think the pictures-as-proof is a straw man. There are a lot of people that believe what they see if it reinforces their existing beliefs. If you make videos of Trump saying something bad, people on the left will believe it, even if the provenance is weak, and people on the right will be inclined to disbelief. A similar fake video with Obama or Clinton will have an opposite effect. Much of that effect can be had wit…

I think the general public is learning how to be sceptical, albeit slowly. It used to be the case that merely being published counted as authority (I remember reading about UFO abductions and thinking "this is a real book! There must be some truth to it!" In my defence I was eight years old) but the internet is forcing everyone to be less naive.

That being said, movements like the anti-vaxxers and the flat-Earthers are testament to people's ability to believe whatever the hell they want.

Re: High-tech deception of ‘deepfake’ videos

#57

Post by mjw_byrne »

save_ferris wrote: post #54
mjw_byrne wrote: post #52

Earlier quoted context omitted.

I agree, but I think my original point covers this - if video proves to be very easy to fake at scale (e.g. using your example of a turnkey website) then that should result in equally large-scale scepticism about video as evidence. The internet may drown in annoying deepfake-based memes but as long as everyone knows its all fake, there seems little potential for the kind of serious harm people are worried about (repu…

> ...as long as everyone knows its all fake, There's a shockingly large number of people who aren't even convinced that election meddling took place in 2016, or that propaganda was disseminated to many people on social media, despite unanimity from US intelligence. If we as a society can't all agree that these things even happened, how can we hope to properly educate people on the existence/prevalence of fake video?…

US intelligence doesn't enjoy a reputation for trustworthiness, to put it mildly. But if people can just go online, upload a video and get back a "hilarious" deepfake, that's an irrefutable demonstration that video-meddling is easy and widely-available, isn't it? They can literally see for themselves how accessible the faking tech is.

Re: High-tech deception of ‘deepfake’ videos

#58

Post by mjw_byrne »

function_seven wrote: post #55
mjw_byrne wrote: post #37

Deepfakes are surely no more a "problem" than photoshopped still images? Everyone knows that a still image can be altered to (convincingly) depict something that never happened. So, a still image is not considered good evidence of anything unless it has a verifiable provenance, chain of custody etc. Presumably that wasn't always the case, but now that photo editing tech is universally accessible, we all know we can't…

The problem isn't so much that fake videos will be used to lie. It's that valid video evidence can be dismissed as Fake News with some credibility. This could be bad for the overall definition of "truth". Video of the Prime Minister/President taking bribes? Eh, you can't trust these videos these days!

In the court of public opinion, yes, this is true - video is no longer proof that something happened. I guess that's a price we pay for technological progress. Presumably at some point it will be possible (and easy and accessible) to generate body fluids with someone else's DNA in them, at which point DNA evidence will suffer a similar hit.

In a court of law, presumably video will now be subject to the same scrutiny as photos. IANAL but I imagine photos do not "stand alone" - the side that uses them has to make a convincing argument that they are trustworthy.

Re: High-tech deception of ‘deepfake’ videos

#59

Post by prepend »

creaghpatr wrote: post #44
prepend wrote: post #27

Earlier quoted context omitted.

The bigger risk would be PR types where there is no court and prosecuted. For example, the Clippers were forcibly sold after the owner-Donald Sterling[0] was outed as a racist. A few fake videos like that would have high impact and wouldn’t need to go through court. [0] https://en.wikipedia.org/wiki/Donald_Sterling

I agree, but would-be fakers should be deterred by the airtight defamation case that would cost them millions. That is, of course, if the plaintiffs can prove the video is in fact, fake.

I think it’s actually hard to launch such a case. I mean take Sterling. The tape was illegally leaked and was there a defamation suit against the leaker? California requires dual consent for recording [0], and no criminal charges have been brought or civil charges. Not the same, but shows there’s no repurcussion in this type of situation.

The issue is that it will be really hard to prove as fake so even if an adversary didn’t just anonymously leak the material. In a PR situation you don’t need authenticity, just confusion.

I expect that you could use this pretty effectively for stock scams (eg, video posted on Friday that Bezos is launching electric car tanks Tesla, etc). Or even more sinister fake videos of things that may be true (video of Musk assaulting some woman).

What I’m actually scanning for are small markets with small $10-100k upside potential like a fake video of Tom Cruise that impact Mission Impossible opening that can be shorted through mediapredict or other markets that aren’t SEC regulated.

Not to do it of course, but to see trends in how these videos are developed.

[0] http://www.dmlp.org/legal-guide/california-recording-law

Re: High-tech deception of ‘deepfake’ videos

#60

Post by pdkl95 »

majos wrote: post #43
mjw_byrne wrote: post #37

Deepfakes are surely no more a "problem" than photoshopped still images? Everyone knows that a still image can be altered to (convincingly) depict something that never happened. So, a still image is not considered good evidence of anything unless it has a verifiable provenance, chain of custody etc. Presumably that wasn't always the case, but now that photo editing tech is universally accessible, we all know we can't…

This is a good point, I hadn't really thought about it but the credibility of photos has cratered in the past decade. Deepfakes' destruction of the credibility of videos is still worse because it totally removes the ability of a random citizen witness to capture evidence. This has been useful in the past, e.g. for documenting police brutality or just-developing events like shootings. In a way this is a very undemocra…

> it totally removes the ability of a random citizen witness to capture evidence

The police/etc have the same problem.

> putting the power of deepfakes in the hands of ordinary people is going to make everybody a less credible witness, no?

"Ordinary people" have a major advantage: parallelism and redundancy. Making one fake and disseminating it to many people is efficient and cheap. Faking many videos in a way that avoids easily discovered contradictions is very hard. Evidence always needs to be corroborated.

Post reply on HN