Hacker News Reader

Comment by zelphirkalt | Hacker News Reader

The idea could also be to first post something obviously fake, in order to make everyone think, that their deep fake video tech is not addvanced enough to create really good fakes. Later they post something, where they put in maximum effort to make it seem real, having people think back to the previous fake and think: "Oh but this looks real and they are not able to produce better deep fakes than that other video."

View on HN

That seems self-defeating via Streisand Effect or at least Cry Wolf effect.

Most of the people who’d believe a video saying the opposite of what they’d expect is real have never been aware of deep fakes.

And now thanks to this poor deepfake they are aware of deepfakes and are primed to question future videos, even if they seem real.

Good point. Hm, it might be my bubble making me think, that of course people will have heard of usage of AI to fake videos, even if they don't necessarily call them "deep fake" or stuff like that. I wonder how likely it really is, that people have not heard about the idea of videos being fake, involving making people say things they never said and making them look realistic.

My parent's generation all forward "articles" to each other's inboxes with full text in the message body. None of them, as far as I'm aware, have yet questioned who wrote the article, where it was published, what other sources have run the story, or whether the article content in the email message body has been manipulated.

Videos being fake? That's asking an order of magnitude more scepticism from them.

> And now thanks to this poor deepfake they are aware of deepfakes and are primed to question future videos, even if they seem real.

They may also be primed to question real videos. Aka “flood the zone with shit”.

"flood the zone" makes it difficult to assess new information, so people fall back on their priors. That would be counterproductive for Russia rn, because the narrative is against them.

Fair point. I'm not sure what would be productive for them at this point though.

Years ago I read somewhere that the way Russia does cyberwarfare is less organized than most people suspect. I'm not sure if it's true, but the argument is that it's more of a lousy goosy free agent type deal. Some of these agents are very clever, but others not so much, so you have a range of competencies just trying stuff.

That could be wrong, or they may have evolved a lot since, but amateurish deep fakes could be a fit. Some random kid just trying stuff and hoping for a pay day / bounty? Pure speculation though.

The general class of 'They're pretending to be incompetent... as a trick!' theories about Russia all seem less likely by the day.

A trick? Doesn't make sense.. For what purpose? I think Russians have a really hard to to pretend to be weak, so for me this scenario was dead on arrival.

I had the feeling Russia hasn't used their best men and equipment yet, which also seem less likely bu the day

There may be something to that with specific regards to jet aircraft, but even then I'm not convinced. More generally with regards to manpower, I think you're probably wrong. VDV have been ground up like cannon fodder repeatedly, and there are reports of redoubled conscription of Russian students. I don't think they're holding their best men back, I think they've already gotten their best men killed with little to show for it.

That sounds like a bad strategy. All this does is make people think about how there are deepfakes they need to watch out for. So now people will be heavily scrutinizing videos of Zelenskyy.

We (Ukrainians) have been warned about upcoming deep-fakes, disinformation, ... since pretty much day 1.

But the Russians haven't.

Reply on news.ycombinator.com