last day (15 days later) » 

06:56
0
Q: How might deepfake video be weaponized politically?

BurnsBAA few weeks ago the New Yorker ran a story on image and video synthesis. The article discusses some of the technical aspects involved in verifying legitimate images, creating new video from old footage, and some of the social aspects surrounding these issues. There's a proof of concept video at t...

"...verbal claims require other evidence to be taken seriously. " Maybe, but such things could be used, therefore to ruin the reputation of the claimant. This has the potential to become a broad discussion, about faking evidence and the essential chaos and ultimatley anarchy (or totalitarianism) that that could bring on (in an essentially opinion-based way). How can we narrow this down to stop it getting closed?
Looks like a very broad & opinion based question.Such can certainly be used to ruin an opponent's reputation; can be used to suggest things about an opponent; can be used to preserve one's reputation in advance of a likely deepfake video attack; can be used to saturate the social media market with red herrings, straw men and windmills of every description. VTC until narrowed down a bit.
"Suppose there is an earth-like world some 10 or 20 or 50 years in the future such that this technology has matured (audio synthesis is implicitly assumed). I will call this deepfake:" You're seriously behind the times.
If a quality "pee tape" can be fabricated, it would certainly will be.
@Duckisaduckisaduck I tried to address that in the question, that this technology will be so mundane it will be useless. Perhaps you have a suggestion of how I can clarify?
@elemtilas I tried to say that none of those things will happen, perhaps you have a suggestion for how I can clarify that?
06:56
We will probably be finding out the answer to this very question in under 5 years, as machine learning moves forward very rapidly.
@BurnsBA Genuinely not sure, damned good question though, I'll be wracking my brains tonight as I sleep, could be a big headache (the subject I mean).
"Basically, if fake videos are common, then no one will expect any negative video on it's own to hold much value", why would it be limited to negative videos? If I'm a dictator and create a deepfake video to show how great I've made the country that could be just as false as a deepfake video to show how horrible the country was before
The term "weaponized" bugs me here because it has so many possible meanings. I think you mean 'to discredit political opponents'...but maybe you mean 'to induce violence against a specific target' or 'disinformation' or something else entirely. It's vague and confusing.
Since politics is already full of lies, including plausible visual lies, I'm not sure I understand the question. There seems to be an assumption that images on video must be true, which has never been valid (see anything with stunts by Buster Keaton).
I dont think this is a world building question. Technology like Deepfaking is already available and anyone who is computer savy should be able to set it up themselves. It will follow similar lines to Hacking and Piracy.
@Shadowzee I don't think that's true at all, perhaps there is some misunderstanding, I've edited to clarify
@user535733 Yes, that's might point. I edited to clarify though.
06:56
@BurnsBA I'll stop talking about the Deepfaking as a technology and address the issues with your question. I've voted to close because this question is both broad/opinion based and its about an existing technology (Used in the Movie Industry) and its impact (if readily available) in our existing world. How can you weaponize the technology is far too broad a question. I hope you can understand that point. There would be an innumerable number of ways to sabotage someones life and career. Fake someone being racists, doing illegal stuff, etc --cont
@BurnsBA I'm not sure about if this is worldbuilding because this is a current and existing technology. Just look at the movie industry. It is becoming more common now, since computers have better processing power (which was the biggest limitation). People are already capable of face swapping which is the hardest step and this is already starting to be an issue.
@shadowzee my question is saying there not innumerable ways to sabotage someone's career. In fact, my question makes an assumption the valid number is zero. That's what I'm asking, is there something overlooked? And try as i might, I cant find any examples of this technology as I outline in my question. Any suggestions on how I can further edit my question to make this more clear?
@BurnsBA I think you underestimate the scope of what is possible. If someone were to release a video of Trump and Putin having a discussion about Dismantling NATO and post that as news, its going to cause a huge stir, especially if you can't tell if its real or fake. Similar things would happen if you had Duterte smoking weed while his offers shoot drug dealers. Its true extreme things may be overlooked because they are obviously fake, but there are many ways to attack someone that are neither extreme or outrageous (especially if they turn out to be true in the first place).
@Shadowzee: No, if a random someone releases the video, and it is known that such videos are easy to fake, nobody will believe it. If CNN releases this video, some people will believe it, but others will point out that Putin and Trump's schedules (which are public) have only a few opportunities to meet, and all of them were in known rooms which do not match the room in the video., or the clothes they were wearing at the time. Which would make CNN ratings drop, so CNN would never post it in the first place.
@BaldBear Assuming the deepfake technology is working fine, recreating the entire environment is a trivial problems which could be solved. The public schedule makes it even easier to have a potential leek because its secret footage of a private meeting. You also missed the point that Putin could of been the one to release the footage and claim its true. Damage is done instantly and takes a while to resolve, but the damage was already done.
If you're saying there are so many hundreds of fake videos crowding any topic that no one cares it feels like you've answered your own question and there's nothing left to say
 
10 hours later…
16:33
@Shadowzee This question is about a hypothetical future where such technology is common. It's not about what the world looks like today but a magical (i.e., non-existent) place where people have grown up with accessible video creation software. Imagine the holodeck from star trek. How does that affect political ads when anyone can make a holodeck program about anything? The question isn't about the world today, but this future world. I've further edited my question to highlight this.
@BaldBear How much does it even matter if public figures have known schedules, how many people know where every public figure is every day? Sure a news outlet may know that and avoid showing a deepfake video but social media surely won't and when your friend reshares that fake video are you going to dig back to find where it originally came from?
@BKlassen I'm not sure that's true. I think there are a few different approaches. I've edited my question to sketch a rough outline of some of these.

last day (15 days later) »