Siliconsciousness: Everything You See on the Internet is a Lie—Except for This

0:00To stay up to date on all the news that you need to know, there's no better place than right here on the DSR network. And there's no better way to enjoy the DSR network than by becoming a member. Members enjoy an ad-free listening experience, access to our Discord community, exclusive content, early episode access, and more. Use code DSR26 for a 25% off discount on sign up at the DSR Network.com. That's code DSR26 at the DSR Network.com. Slash by. Thank you and enjoy the show. Welcome to Siliconciousness, the DSR Network podcast focusing on the artificial intelligence revolution, politics, and policy.

1:02Hello and welcome to Siliconciousness, I'm David Rothkoff. You're host and this week as every week we're going to talk about something tied to AI this week also tied to breaking news. And that is disinformation, AI, current war in the Middle East. We're very fortunate to have with us Tal Haken who is an information warfare analyst and media literacy lecturer specializing in Ocent Investigations, Disinformation Analysis, AI Generated Media Verification, and Strategic Intelligence Research. How are you today, Tal? I'm doing great. Thank you for having me. It's a pleasure to have you. Well, let's let's let you give the overview as opposed to me. How do you think AI and AI driven disinformation is playing a role in the current conflict in the Middle East?

2:03In regard to its current role in the conflict at the moment in the Middle East, this isn't new or it's not something that came out of left field as a lot of different types of journalists or researchers are pondering at the moment in various different publications. I've been sounding the alarm for over two years as I've been doing this type of research and seeing the spread of AI generated content and how it's slowly becoming more and more a larger factor in the types of fake news that we're going to be encountering every single day. In this specific war, it has kind of gone into overdrive in terms of the amount of fake news, fake news related to AI generated content that is not just tricking the average viewer, which it's already been able to do for over a year, but now it's also able to trick seasoned journalists, researchers, people who self-proclaim themselves as experts, and so we're kind of reaching a new era when it comes to what people are able to identify. And this is like, this is due to two factors. One, people not understanding

3:03how to actually identify AI content. It's not really taught anywhere, no one's actually taught how to do this. You either have to invest a ton of time and effort, or you have to go willingly to different media literacy courses or lectures on AI identification, which people don't do. And the other factor is that AI is becoming not just cheaper to create, but also on higher quality. So if a few years ago you could create high quality imagery via AI, you would have to pay a ton of money. Only specific people would actually be conducting that type of work. But today, you can be a 12-year-old on TikTok and make a high high quality AI generated content as people are doing. And so it's a lot easier to access it, and it's a lot harder for you to actually identify what is to involve. So what we're currently seeing happening in the Middle East is kind of just a problem that's been going on for a while now, and just I think a lot more of the world is paying attention to it now, which is good. Yeah, it's certainly good, although I think as you suggest, the problem is only going to get more

4:04complicated as the technology improves, as people get more sophisticated in using it. But let's talk initially here. I think some of the first examples that we have seen of AI fakes in this particular conflict have been things like showing an attack on an aircraft carrier that didn't actually happen or showing different kinds of damage or origins of the damage and so forth. Do you want to walk through what you think of as good illustrative examples? Yeah, so different types of examples that we've seen that have been protruding more so during this war have been the destruction of Tel Aviv, one of the second largest cities in the state of Israel. And so that has kind of been a central theme to a lot of the different types of AI generated content, which is showcasing massive amounts of destruction either on Tel Aviv as a

5:06city or in Ben Gorion Airport, the only and largest airport in the central Israel. And they usually either showcase massive amounts of munitions falling down and destroying everything. Sometimes they show the aftermath of footage where everything is in rubble. And the type of narrative that's trying to be pushed with these types of AI videos is that while the United States and Israel are conducting large amounts of damage to Iran, Iran is not just only able to inflict a lot of damage against the Gulf States, but also against the state of Israel. Are you seeing it on both sides in this? I mean, or I mean, it seems to me as a casual observer, I'm seeing a little bit more AI-generated fakes coming from the Iranian side. But I am seeing a lot of weird use of AI coming from the US side. The president seems to like running AI quick cuts that involve him as a superhero and so

6:11forth. Is this a two-way street in your view? I think I'm being able to actually identify what percentage of AI on different sides of the aisle, either you could say the pro-Iranian axis or the pro-American Western perception really requires a in-depth analysis that different types of posted are online. The reason I can't tell you is because my algorithm and what I see is based on what I'm researching based on what appears in my feed. And so what I see day-to-day might be something that's different for other people. But definitely, it's AI is affecting every single aspect of the war, both from a propaganda standpoint as you put with President Donald Trump, with those types of AI clips are just meant to you be used essentially as a new wave of propaganda, used to have posters that were paintings of propaganda. Now you have AI generated videos. The other end of it is though is creating a false reality where the individual thinks that what they're seeing on the screen is real. And so that we the majority of the content that I personally have seen

7:15has had to do with trying to showcase Israel destroyed. But I have seen some AI generated content when it comes to Iran, one specific type of AI that I have been seeing slowly garner more and more popularity and being shared more often is AI generated strikes on different Iranian soldiers. And so you have photos, real photos, or real videos of Iranian soldiers sitting around at the checkpoint. And then somebody will come and they will essentially copy onto the video and AI generated part of it, kind of like extending the clip and making it as if they have been blown up, which which isn't true. And so these are the types of things that we are viewing me and other researchers. So what one of the problems I think that contributes to this situation is the evolution of what I once called 20 years ago in an article, the info sphere, but this sort of information environment. And by that, I mean, we have some of the trusted sources of

8:26visual data that we would get like news organizations cutting back on their coverage. So we have less coverage. We have militaries in the US military and the Israeli military have both been guilty of this, being very, very draconian and keeping journalists away from pointing cameras at what's going on. So there's less of the sort of reliable sources, people who are supposed to be representing the truth here. And at the same time, there's this explosion of other sources, whether it's unverified cell phone video content, or it is actively generated AI content or edited content. And so there's less and less of the stuff you can rely on and more and more of the stuff that's hard to trust. And that seems to be worsening daily. Is that your perception? Am I overstating

9:28the problem? So I see things a little bit differently. I think that, as you said, with the info sphere, what has happened is that in the past like 20 to 30 years, there's been an absolute massive increase in the amount of information that is accessible to the average individual. And it used to be that you would have to just look at one of several different new stations and whatever those new stations showed you or whatever photo was presented in the newspaper the next day, that's all you got of what actually happened in that war and that conflict and that terror attack and that's stabbing whatever. Now, you have multiple different angles. You have entire timelines that are built by researchers online. You can get every little aspect of data. And so we've also become accustomed and privileged to having all this data at our fingertips. And so when we don't get access to all this information, we don't like that. And so for example, it's pretty accustomed that for example, you're not going to film where interceptors aren't launching in order to destroy incoming ballistic missiles because then individuals online can geolocate essentially,

10:31they can triangulate where that video was taken and then share the coordinates of that defense system. 20, 30 years ago, that wasn't even a thing that people thought about that you, that's something that you could do. Now, that's something that happens with every single photo of a video that goes online. So the types of protocols that are enacted against journalists or against pretty unoppressed is different because of the new dangers that arise when it comes to the freedom of this type of information that's out there. On the other hand, because people want this constant access to information immediately, it also means that it's easier for them to fall for false information because while a lot of legacy news organizations, which they have their problems, but a lot of them still will wait for confirmed information. They will wait to gather as much information as possible before they presented to the viewers because there is accountability, even if it's not from external forces that's from within, that the journalists will be fired. So they're a lot more careful. People online from anonymous profiles, alternative news channels that don't care. They will just post whatever they see online first in order to gain clicks and to gain

11:32the viewership. And me as a viewer, if I want to understand what's happening in the war in the Middle East, I don't want to wait for the New York Times or the BBC or CNN to show me a clip 24 hours later. That's confirmed. I want to see now, and they said that there was an explosion five minutes ago. I already want to see that explosion in front of me. And if the main and if verified news outlets can't provide you with that, I promise you there'll be a heap of other individuals that will provide you with AI images, old images in order to get your clicks. This podcast is underwritten in part by the US Embassy of the United Arab Emirates. Its editorial content is completely independent, and the views expressed are exclusively those of participating experts. It is presented live without editing. For further information about the UAE's efforts in the areas of artificial intelligence and technology, go to the website of the Embassy at www.uaedashembusy.org and search for UAE-US Tech Cooperation. We thank them for their

12:39support. We thank everybody who's supporting this podcast for their support and we look forward to it developing and growing over time because the issue is so important. Okay, well that was putting it a lot better than I put it. But I still think it describes some of the same phenomena. It's interesting to me. It's a bit ironic. Today here in the United States, I don't know if you pick up on news like this, but the new ownership of CBS announced that they're shutting down CBS News Radio, which was a very big news gathering operation. It's been around for 100 years. And I remember when I was a little kid, and there would be a war in the Middle East, and there was no CNN. What we would do is we'd listen to news radio reports around the clock on a radio that we had in a corner in our kitchen. And then of course CNN became the norm. And now

13:46the norm for most people for consuming this information is a social media feed. And that changes everything because the context is so different. And you would think that going along with that, there would be a big caveat, empty or component, right? There would be a big component of, well, we're getting a lot of garbage in through that. And so we're going to be better consumers. Are we better consumers or are we more gullible consumers because of the way we are getting most of this info? I think the problem is that we consume news now as we as our parents and grandparents did. And that's the issue because it used to be that of course there's always going to be certain caveats when it comes to issues within the programs or biases or agendas. But for the most part, for people who don't know if you've ever been inside of any type of major news network, they're massive. There's like hundreds of people. There's experts of every single field you can

14:48think of. And so journalists can go and reach out to all these different experts. They can reach out to their verification teams, their digital analysts, everything before it ever gets to actual broadcasts. And so what you as a viewer saw was the end of that filter. So there'd be an attack. There's tons of different videos or claims or witness testimonies coming out. And by the time that got all the way to you on the television screen, it went through a million different specialty filters, also legal filters to ensure that there was no defamation or other different cases. And then it finally got to your screen. And then your job is literally to just consume. Today we do the same thing. We just consume. But the problem is that the information we're consuming isn't going through any filters. It's just straight to your feed. The biased witness testimonies, the unreliable witness testimonies, the video that was filmed from a random camera that was passing by. All this information is put on to you. And now you're the entire newsroom that has to decide for all that information and come to what is actually happening. And people don't know how to do that. That no one was taught how to do that. Media literacy isn't something

15:48that people really actually know. They think they know how to analyze online content. They don't. I can tell you from experience doing this for almost 11 years in the field of media literacy, that it doesn't matter if it's a kid who is in seventh grade that I've spoken with or professors who are in their 50s. Everyone has the exact same issues. I've had my professors from university come up to me and show me a video that's very obviously fake to me, but they thought it was real. Not because they're lesser than me or they're dumber than me. No, it's because they weren't taught how to actually analyze the media that they watch. And that's the problem. Go on, admit it. They're lesser. But there are built-in problems here. For some of them, I've had a government official and done a bunch of other things in my life. But for a period, I was the editor of Foreign Policy magazine. And so we would get these complicated stories. And you would say, well, is this a trusted source? Is the article that we're getting sourced? Is the editor knowledgeable on the article?

16:50There would be all of these filters in order. And we wouldn't run something unless it had multiple sources and so forth. Exactly. Those are the questions that aren't asked by the average viewer. No one asks that. Those questions when they see a piece of content. No, well, exactly right. But what I was getting at is it's actually worse because if it's a social media feed, then a lot of the, quote, filters who are directing content towards it are your friends. And so if you happen to have a lot of gullible friends, you're going to get more junk. And by the way, I see this even on, I banned and X when Elon took it over because I just thought it was going to become the kind of propaganda hub that it was. And I moved over to Blue Sky. And Blue Sky's got all, it's got its own biases. But there are loads of journalists and people I trust. And all day long, every single day,

17:52somebody's promoting something that's just garbage. It's just not true. And so I don't see progress being made on this stuff. And I'm just wondering when you're talking about it, when you're dealing with consumers of it, whether it's one of your professors or a younger person, what are the warning signs that you're looking for? That were the processes for consuming this that you're advocating? It's asking really simple questions. That's really it. You know, at the end of the day, of course, there are certain types of research and investigations that I'm able to conduct because of an expertise that I've grown over the years. But basic media literacy and 90% of the fake information that we come across on a daily basis is really easy to debunk by asking really, really simple questions. What am I actually looking at, for example, a video that I saw online a few days ago, the guy was saying that it was actually an AI video. Like he was saying that it showed Ben Gorion Airport on fire.

18:56All you see is like a big like nothing in the background. It's like a pitch black and you see some flames in the background. That's it. That's all you're seeing. You're not seeing anything else. And so the simple question that I ask is, what do you actually see in the footage? How do you know it's Ben Gorion? Just what? Because the caption told you it was Ben Gorion? That doesn't mean it's from Ben Gorion. And those simple questions can almost destroy the majority of the false information that comes across my feed. And it's also how, despite the tons of information that I consume on a daily basis, I'm able to still filter through a lot of it because I just ask a really, really simple question. Am I seeing actually what the caption tells me on seeing? For example, based on a report. Okay. By who? Who reported it? Like what's the source? Oh, it's anonymous. Okay, so maybe I shouldn't like trust this because anonymous doesn't really mean anything. Sometimes it's also, you know, oh, also over time you start to learn what to look out for. So I know for myself, for example, can I see how different news agencies quote me, for example, when they quote my work?

20:00And I noticed that it changes depending on the different news organization. They emphasize different aspects of my work. And so what I learned from that is, for example, when they're saying a US official said a US official could literally be anybody. When you think of US official, you're thinking of some guy who's like the commander in chief who's right next to the president or an advisor to President Donald Trump, it can literally just be a secretary who works at some random department on the other side of the country. That's who the official US source could be. As in, you have to ask yourself when they're these different types of anonymous sources are or any type of source. What it actually is the individual that we're talking to. And one big thing that I do, and I a bit of a ramble here, is that there's a huge difference between facts and narratives. Very, very two very, very different things. There is factual information that we build our world of views on. We build our narratives on. But we need to know how to decipher. I can watch CNN, BBC, ABC, Fox News. Doesn't matter to me because when I'm watching them, I'm taking the

21:06narratives that they're pushing and I'm throwing them aside and I'm listening for the factual information. So if there was a, they're saying that there was a shooting, okay, at a school. Okay, I know that there was a shooting at the school. Great. That's the factual information that I'm gathering. Then they say that his agenda was he was a right wing or he was left wing. My only question is, okay, based on what? If they just said that he was right wing or left wing that I don't care what they have to say. But if they say that he was right wing because he hears the manifesto where he wrote that he was pro this and this and this. Okay, so that sounds like a factual information. It's based on that. But if they just randomly throw out that the individual is left wing or right wing, but there's nothing following that to factually base that, I know that's a narrative piece and I just ignored and move on. But you know, those are certain kinds of judgments that are appropriate for all journalism. But we're specifically dealing with the challenges associated with getting content that's AI generated here. Because it looks like when it's done well that you're seeing something

22:13as it is actually happening or as it actually happened and that therefore it is a fact because you believe your eyes, right? And I think, as you noted, the AI techniques are getting more and more sophisticated. And it raises a question for me. Who helps us sort this out? You're giving some guidelines for each of us. Look at the source, look at the context, look at the timestamp. If there's a timestamp, look at, is it an old image or a new image? You know, what is it that supports what they assert it to be and so forth? That's one set of things. But say I'm looking at an image on a platform, on TikTok, on Instagram, on Blue Sky, on X, on Facebook, on, you know, pick, pick, pick, pick your platform. Do you think it would make sense

23:20for those platforms to require AI content to be identified, to use AI content, to use AI to help identify questionable content? So I have always been a very, very big advocate for legislation and policies to be enacted. And regulations that are then enforced by various platforms, I don't believe that the various social media platforms have AI incentive to do so, because at the end of the day, I add generic content or fake news, brings in a lot of views, because people like chaos and usually fake news is derived from trying to make something more radical, extreme, that's not like reality, and so people click on it. And so they get views that platform gets more money, so they have no incentive to change this in my opinion. And there's no reason to. If I was in their shoes, I don't know if I would want to change it. And so what really needs to happen is you need governments to change it. You need different types of, you need different government legislation to put policies that if you want your platforms to be in our countries and to work in our countries,

24:22you need to enact policies that ensure that AI generated content is labeled as such. You have to, there's also different types of policies that I personally support. For example, that there should be a identifying marker for anybody who creates AI generated content. If you create AI generated content, I want that connected to your name. You don't get to just create whatever you want and then just throw it into the void. And then there's no consequences. If you made something nice, you just made some nice little scenic route showing a cat running around in the field. Who cares? It's fine. Enjoy it. But if you're making content that's made, for example, to show a defamation against somebody, taking somebody that you know in real life and making them appear naked, then yes, that should be connected to you so that there can be consequences against you for what you did. What I do online and when it comes to media literacy is a bandaid. I'm very much aware that what I'm doing online isn't going to change it for the majority of people. It's very much abandoned. It's a crutch for the time being. AI generated content or AI in general advanced so quickly for all of us and no safety guidelines were put into place to stop what we're currently facing.

25:26We're currently in the Doomsday era of AI when it comes to the types of consequences that we're going to face as a result of unregulated AI. I believe hopefully that by the time I have grandkids, we won't live in that reality anymore. You won't be able to just create whatever you want. You won't be able to just create images of AI that will be able to just be put all over the platform with no consequences. I think eventually it'll change. Like anything, it'll eventually be regulation. But right now we're suffering the consequences. Do you think AI could help solve its own problem? In other words, So I was going to get to the next point with that is that I actually currently at the moment, I do not trust any AI identifying engines. I don't trust that. I've used all of them. I don't trust them. They're unreliable to me because they become a crutch. They're a tool. The way in which most AI generated content works, both at the lens, large language models like chat to BT or perplexity or Gemini and all that. And these different AI identifiers is they're not that amazing actually at the end of the day. They're incredible as a tool. And I think that's the issue that people

26:27have when it comes to AI, like artificial intelligence, is at the moment what we have as artificial intelligence isn't a superhuman. It's not a thinking being that's able to understand complex topics like the human being. It's a probability. It's essentially just probabilities statistical analogies in a box. That's it. And so if you know if you use, for example, if I take an AI AI tool and I look at a image and I see that it has a 76% chance of being AI, for me as a researcher that doesn't tell me much. That's just an indicator that it might be AI generate content and I need to do more research to figure it out. But the average viewer will not do that. The average viewer will see the 73% and say, oh, this is AI. That's it. And that's the problem that with how AI a lot of the time is used is it's used as an all-purpose solution and not as an additional tool in our tool set. So are anything you trust in this area? At the moment, no. I don't think that anything is advanced enough. And I also think that a lot of the current regulations that we do

27:27have in place are just not adequate enough. A great example of this is the current watermark systems with different AI platforms. A great example is Sora AI. Now you probably have seen it. It's the AI that has the little Sora going from the top right, the middle left, and then bottom right. The biggest issue is it doesn't say AI. It's a Sora. Also the Gemini. The Gemini logo at the bottom right that little like diamond shape. That doesn't say AI. It's just a little shape. So if you don't know what Sora is or you don't know what that little symbol is with the bottom right side of the page, you don't know the video or photo is AI. People need to be spoon-fed in remation. You cannot rely on them understanding these complex or nuanced topics or understand all the different types of AI systems out there or what their names are. You need to spoon-fed them. So if there's AI generated content, you have to have a massive banner in front of them that says AI generated. And I can show you millions of the examples to back up this claim that I'm making. Go into TikTok. There's a channel that I follow in TikTok. I forgot it's name that literally

28:27in the name it says AI. The channel description says AI generated. The description of the video says AI generated. There's a little tag on the bottom by TikTok saying AI generated. And all the comments are like, this can't be real. Is this real? This is amazing. Like no one, no one understanding it. This is a completely fake. I was in there all questioning whether it might be real or false. Despite the there being a million different indicators that it's fake, unless they actually have a massive banner in front of them saying this is AI, people fall for it. And so that's the issue that we face with this type of content. Yeah. At one last point, we're running out of time here. But one thing that does strike me is we've spoken this whole time about AI and video. And of course, AI is very useful in faking other things as well. And in fact, it's better at faking other things. And so, you know, documents that you may see transcripts that you may see letters that you may see. It's it's it's it's it's much much easier now to create a complete believable fake that isn't video. Also using AI, right?

29:30No, of course. And the one of the best ways in order to decipher whether information is true or false when it comes to even if it looks 100% real is other indicators. And you said it earlier when you're saying when you were an editor that you have to look for multiple sources, the same thing. Okay, this one individual is sharing this one document that is showcasing a resignation of a leader. Okay, the leader of this country right on any of his platforms that he was resigning did any major network and talk about the fact that this person was resigning or is the only proof that this individual is resigning is his one document. The photo of this one document. If that's the only evidence that we have, then there's a high likelihood that it's false. And that's how you have to look at these things as a probability. What's the likelihood of it being true based on how many indicators do we have? So if you see one little in one indicator, one photo of an individual, for example, in a very embarrassing position, for example, a politician that they put in a very embarrassing put. So if you're singling one image going around and nobody talking about it, it doesn't really make sense because of the

30:33background or what he's wearing and doesn't make sense to who the person is in real life. And there's likely not real. Yeah, I think that's a really important point. And I think it moves us to a different mindset because I think a lot of people, you know, who look at things particularly from media platforms, start with the assumption of trust. And I think as you go into these things, you need to start with the assumption or the attitude of skepticism where you say, who is the source? And this is what I do when I see something like that is, first of all, I ask who's posting it. And then I look at the comments. And I look for credible comments from people who are confirming it, or I look for multiple sources. And I just won't, you know, repeat something because it tickles something in my brain until there are multiple confirming reliable sources just because this stuff is so prevalent. Yeah. And it's never going to be 100%. Maybe everybody confirms and it turns out to be true or false. And we're never going to be able to get this 100%.

31:33But by the way, kind of expanding on the comments or the replies, everybody else, you know, everyone in the world, we all have our own expertise, we all have our own experiences and life experience and anecdotes. And a lot of the times with different fakes, different individuals can help us. So if I see something, a video that they're showcasing this bridge and they're saying, this is from Australia, this is from Sydney, Australia. And suddenly I see someone in the comments who's from Australia saying, this isn't Sydney, Australia. And then another person saying, oh, that bridge is from my hometown. I see that he lives in Croatia. Then that's an indicator that something's wrong here. Now, it doesn't mean that it's true. And I can give an example of this actually from today. There was a video that went around that people were trying to claim was Ben Gurion Airport. And it was actually from Denver, from the Denver International Airport in the US. Now, somebody replied to me saying, I'm from Denver. I've been to the airport multiple times. This is not the Denver International Airport. Okay. So he's coming from a position of authority. So what did I do? I looked up images of the Denver International

32:37Airport, found multiple angles that matched the video. And I said, well, you're wrong. And so that's another thing. And just because the comment says, as in you said, we're being skeptical and not just trusting, just because the commenter said that he was in Denver, doesn't mean he's right. Just because we're from a certain place or just because we were in the military, just because we were journalists or just because we are an expert, it doesn't mean we could not be wrong. And it's also something important to remember. Super helpful. And I really appreciate these perspectives. And I hope that everybody else does as well. I'm sure they will. Perhaps we can have you back at some point in the future to talk more about this. For now, thank you, tall. Thank everybody for listening. Be skeptical. Be it informed consumer of information going forward, especially in emotionally charged volatile times, where there's a lot of incentive for people to share misleading information. We'll be back with more on this and other subjects here on Siliconsciousness. Join us for that. Until then, bye-bye.

33:41This was Siliconsciousness, a production of the DSR Network.