Is this true though?

Sure, for a casual observer, these new methods for generating videos appear convincing, but is that the right bar to judge "ability to fake evidence"?

As far as I know, there have always been more sophisticated techniques and forensics to determine if an image is doctored, and likewise for video. I've not seen any research tackling fooling those methods yet, and I would bet that naive implementations of neural networks for generating videos would leave very obvious "neural network" artifacts. Of course, this is still new technology, so it will obviously get better at fooling our other tools over time too, but as of right now, I don't think the clamoring for "all evidence can now be faked" is all that justified.

I question becomes, does that matter? If everything needs to go through forensic experts, that works for courts. For media, who knows. Maybe it just becomes another way to challenge anything.

There's always been a certain value to letting people see things themselves, and use their non-expert judgement directly, en masse. Think of that iconic image of a Vietnamese girl escaping napalm or (more recently) the drowned refugee child on that Turkish beach. These had value past the factual information.

In any case, I think the Photoshop/stills case is a little heartening. We get an occasional false image, but overall it hasn't created some sort of massive truth-crisis. Other stuff did happen to truth, but Photoshop wasn't at the centre of it.

Does that matter?

On Internet the judgement is quick, long before the truth has revealed itself lives has been destroyed, jobs lost and reputation irreparably damaged.

But that was already true with words alone. Lives are already ruined every day by baseless claims with zero evidence, no video required. People who already research the source of words before forming an opinion will now research the source of videos, too. Those who are fooled by written and spoken lies will also be fooled by animated ones.

Words have almost no impact compared to visual imagery. Even if the words come from the mouth or pen of a respected individual, you can still doubt it, and usually people do.

However, whether we like it or not, everything we see, at least on a subsconscious level, we perceive as real.

My assumption would be that we have no evolutionary resistance to that. I don't know if there can be at all. People have been lying from our fist day of existence, but up until very recent times, there has never existed anything in nature that can create arbitrary images that look 100% real.

It doesn't matter if it's fake or not, if you see a convincing version of your president fucking a pig, you'll never forget about it, and you'll always attach that image to him. Similarly, people often can't differentiate between an actors character and the real person. Whether it's Macron on his campaign film or Gandalf on lotr, you associate that person with your simulated experience.

Animated lies are on a completely different tier of immersiveness and deception than speech or writing. You can't reduce this development to "just another lie". It's very close to the most convincing lies possible, and it's definitely very dangerous.

In addition to this, despite all our cutting edge forensics, we still fumble the ball in the courts, based on sham forensics, and send innocent people to jail.

As a defendant, you have to pay forensics experts -- highly -- to research and create such evidence of manipulation. If they are willing to take on your case; they have their own reputations and motives.

This all taking us further into an effectively tiered justice system. Can you afford to challenge the evidence, even if you think it's doctored?

If you are a cause celeb -- or a good vehicle for a non-profit -- maybe someone else will. Otherwise...

FYI, russian propaganda machine routinely uses videos from videogames and movies as "documents" in TV broadcasts. [1] Despite the fact that reports debunking these "proofs" appear in the next day or so, there's no backlash or retractions. Hell, even russian defense ministry used insanely crude photoshopped images in it's reports, and still - no backlash, at all. [2]

My point is, if you just want to brainwash the population, these new technologies are overkill. People aren't generally smart enough to require this sophistication from a fake.

[1]: https://gizmodo.com/russian-state-tv-airs-video-game-clip-as...

[2]: https://www.thedailybeast.com/kremlin-falls-for-its-own-fake...

On all sides, your side will always seem more innocent and correct despite evidence. We live in a very messy and dirty world.

Nikon used to provide "Image Authentication Software" which would authenticate images and show if they had been altered since taken by the camera. This was useful not only in news but also court cases.

This, obviously, could not be used on tweaked published images but would work if you had access to the original frame. Most photojournalists work to a code whereby they don't amend the images at all, even tidying up, but that's only viable if they and their editors are trustworthy.

The Nikon process was cracked a few years ago. This allowed edited images to pass the validation.

I'm actually really curious about this now. Couldn't we just hash the photo and compare the file to the signature to verify authenticity?

Yes, but at what point do you hash it? The Nikon software proved it was unaltered since the shutter press, anything after that would rely on it not being touched until the hash was generated.

No, this is an opportunity for new technology that adds additional data stream(s) to audio/video to ensure authenticity. Perhaps something like a one way hash of physical properties related to time/position and the rasterized data, in a way that cannot be faked.. or maybe a move away from rasterized data. Idk. Just brainstorming.

I’m confident there will be an innovation in this space since there is clearly a very urgent need. In fact, this could be a start-up opportunity. Maybe there is relevant academic research already on which a startup idea could be built.

A short-lived opportunity for sure. This is the very definition of an arms race, amd I can't see "real" winning out over fake. Anything that can be checked can be preempted, so "real" is always one step behind, as far as the present is concerned.

Historically, arms races have been pretty long-lived opportunities for weapons developers.

Or a DRM chip implanted in every human.

“Warning: the person in picture doesn’t have Widevine chip installed”.

Indeed, something like that would be the only way to implement this, and I hope that everyone will agree that we should not go there. To a certain extent, the ability to lie is foundational to human civilization.

The difficult part is the "in a way that cannot be faked". If you can record it using a device you control, you can generate a fake made to appear as if it had been recorded with that device. If you do not fully control the recording device (say because it uses a secure enclave to generate signatures), you'll need some way to generate realistic inputs for fake data, but that might be as easy as photographing a high-resolution print.

If the effort to document the truthfulness of data is much higher than for unverified recordings, the majority of the evidence anyone can come up with will be of the variety that is trivially to fake for someone determined.

Tamper-proof HSMs in recording equipment might do the trick. The point is not that they can be broken. The level of sophistication has to be high enough that most bad actors will be incapable of doing that. This is why chip cards for security work.

> The era of verifiability-- when photographic and video evidence could plausibly be trusted-- is over.

If ease of producing convincing fakes of X means that the era of X as convincing proof is over, then why are signatures and written documents still taken as proof?

I think it just means that things will be scrutinized more carefully from now on, and that there will need to be a stronger document trail.

> We are back to a pre-industrial age of discourse, where rumor and hearsay and anecdote dominate again.

And I definitely don't think it's as grim as this.

Signatures _aren't_ hard proof, they're absolutely contestable. Images and videos once were hard proof, but now are becoming contestable.

I wonder how long it will be before we see journalism-focussed photography equipment automatically generating a secure cryptographic signature for each shot. A particularly fancy implementation could be signed by both the manufacturer and the photographer, offering two assertions the shot is valid and undoctored. Going back to a reputation system still isn't great, but it's a lot better than a "pre-industrial state of discourse"

Anecdotally,

Whenever I receive a signed-for delivery, I'm asked to provide my name and attempt to sign on the proffered touchscreen device - invariably scrawling a shape that in no way resembles my actual signature.

I recently accepted a new role, and as part of the on-boarding digitally signed my contract by typing my name into a text box marked "Digital Signature" - I'm kinda assuming it's just a regular input box with added legalese.

Signatures are absolutely no longer proof of anything, simply a token gesture with perhaps a dash of security-by-fud tacked on.

I think a signature is less about proof of identity, than proof of consent.

>why are signatures and written documents still taken as proof

I always thought this was tradition / theatre at this point.

On most important documents, they have to be notarized, which is a second level of authentication.

For any important document, the signing is almost always witnessed by someone else. My understanding is that the act of signing is primarily the legal act of accepting the document, not a security mechanism. The signature provides a quick reassurance that a document has been accepted, but if there were a real dispute, the witness could be asked to testify.

Yes, almost always signatures are only going to be relevant to civil cases, which are decided on the balance of probabilities. So all the evidence introduced is about pushing those probabilities around. A signature is neither necessary (except in some special cases) nor sufficient but it will usually be evidence for "acceptance", one of the main parts of a contract in law. A witness would shore up that piece of evidence but it rarely comes up. In most cases acceptance is not disputed because doing things you contracted to do is also pretty good evidence for acceptance and that's usual. e.g. if you're in court over a car lease, the fact you paid the first three months and drove the car is usually going to be more than enough "acceptance" for a court even if the signed document has been lost.

In current use, a physical signature doesn't really verify identity or authentication (i.e. that it was you who signed the document) but rather intent and consent - i.e., assuming that it's your freely given signature, this "ritual" marks that you intended to approve that document as opposed to words and gestures indicating that yes, maybe this is the right decision for some future time but might still want to think about it.

It's not as if people were fact checking CNN and Fox or running forensic analysis to scrupulously determine the plausibility of the stories in their newsfeed to begin with.

As far as society and its trust (or lack thereof) in digital media is concerned this won't change much. Most people believe what they want to believe, and doubt what they don't, trust the in group and doubt the out group.

Plausibly be trusted in what context? In a court of law? Or in the court of public opinion?

I think the general hypothesis among mass communication studies doctorates is that as soon as a technology is created and adopted by a mass audience, there are immediate examples of humans immediately using it to manipulate public opinion in some way. For example, as soon as telegrams became a thing used to manipulate public opinion (like the Zimmerman telegram), fake telegrams became a thing. As soon as online video became a thing, fake online videos became a thing (like Lonleygirl15).

Isn't what you are really saying is, "humans like to create fake things, and then a lot of people fall for those fake things." Right? Why should we buckle up for that? Isn't it already a foregone conclusion?

There are companies that make cameras that sign the images. I believe they are only used for legal proceedings now but I’m sure we could expand their use easily enough.

Any shopping would destroy the signature but that might be a good thing for dragging us back to reality.

On cursory inspection, deepfakes may pass muster. But on close observation it's still relatively easy to tell they are fake.

They'll get better.

And some people will believe even the most blatant photoshops because they want to.

Just look at the pizzgate incident. Some fool entered a pizza restaurant and fired is his gun because he beleived people on the internet that claimed it had a secret pedophile den in its basement.

Yeah, this is an interesting issue, but CCTVs are kind of invasive anyways privacy wise so maybe that's a good thing.

They're still useful for realtime monitoring though, which is good for security.

With powerful enough machines, could you generate fake video in near-realtime? Not today, I guess...next year?