13 | The Truth is Dead ! How AI is Putting At Risk The Trust That is The Fabric Of Our Society

0:00I've got a very scary announcement to make, and that is the truth as we know it is dead, or at least it's dying and it's in a very bad shape. And the reason I'm saying all of that is the capabilities of generative AI and how fast they're moving and the things that would enable anybody to do in the very immediate future. I've actually recorded this episode for a different podcast a few months ago. And since then, this has accelerated even further with companies like Metaphysic, which are using AI to enable amazing deepfake even in real time. At the end of the episode, I'm going to share some interesting news of the AI world that happened this week.

0:38And now to the episode, The Truth is Dead. In the next three years, AI technology will change our world dramatically. Whether you are a business executive trying to catapult your business forward, or just somebody who refuses to be left behind and want to advance your career, this is the show for you. I'm your host, Isar Maitis, a serial entrepreneur and an AI enthusiast. You'll hear invaluable practical tips from innovative business leaders, AI practitioners, and some of the brightest AI minds in our world today on how you can leverage AI in ethical ways to advance your career and grow your business.

1:24Hello and welcome to Leveraging AI. This is Isar Maitis, your host, and I've got a very scary announcement to make, and that is the truth as we know it is dead, or at least it's dying and it's in a very bad shape. And I know this may not be directly related to business, but it has huge implications on everything we know, including businesses and including our personal lives, obviously. And the reason I'm saying all of that is the capabilities of generative AI and how fast they're moving and the things that would enable anybody to do in the very immediate future. So let me give you a quick summary, and then we'll dive into the details.

2:05Generative AI enables you today to generate images that are near perfection and extremely realistic and is almost impossible to detect definitely with a naked eye, whether it's an actual photo or an AI-generated photo. And the same exact thing is going to happen to video. Anybody will create a realistic video of anything they want just by typing what needs to be in the video. That's it. Which means generating quote-unquote facts that will support whatever truth you want to tell will become at the fingertips of anybody who wants to spread any kind of news, stories, scientifical research, personal stuff, etc.

2:46etc. Every new thing will be able to be generated in seconds and be highly, highly realistic. Before we dive in, let's define what is truth. So if I look up dictionary definitions of what is truth, it says something like that which is true or in accordance with fact or reality, or a similar one, a fact or belief that is accepted as true. And you can see these two definitions are taking slightly different approaches. One says something is true if it's according to a fact or a reality. And the other says it's a fact that is believed to be accepted as true. Both are definitions of our truth and both are extremely problematic with those new capabilities that I'm talking about.

3:38Before we dive deeper, I want to say that when the concept of truth was born, the concept of lying was born with it. People have been using lies to achieve goals by deceiving other people since the beginning of time, all the way back to Cain and Abel. That being said, a lie from one person to the other on general information is easy to fake. But when it comes to, quote unquote, selling lies to the mass population is something that requires more and more resources. So the bigger the lie you want to tell, the more resources you need, because you need to be able to fabricate the relevant facts that will be supporting evidence to the lie that you're trying to sell.

4:22Hence why governments, when they try to deceive one another, have the resources to do these kinds of things. So the concept of manipulating the truth or telling lies or deceiving other people has existed for a very long time. In recent years, with advancement in technology and social media availability and the ability to generate content and to distribute that content through channels which are not monitored by anyone, has took a very serious swing at the concept of truth. And just a few examples, the recent war between Russia and the Ukraine, we've seen multiple videos that had nothing to do with actual war in Russia and Ukraine, but was published all over social media and in some cases by national media as videos from the war.

5:13And they're actually not. There are multiple instances of Iran launching rockets to show its capabilities when people very quickly caught up that the number of rockets launched are duplicated. So the same exact rocket is showing in more than one place just to show that they have more firepower than they actually have. We've seen numerous marketing campaigns using fabricated truths to promote different things from the very basic of people standing next to a Lamborghini that is next to an amazing mansion on the beach that they hired for 30 seconds paying a lot of money just to make you believe that they are really rich.

5:50And hence, you should follow whatever they want to sell you, all the way to much bigger deceiving campaigns that companies are pulling off in order to sell you different things. Another example is photos of models, right? We all know that every image that we see on a magazine or on social media and everywhere is being photoshopped. Like the model doesn't look in real life as good as it looks in the photos that are being shared. So all of these are examples of where people with a lot of resources are able to manipulate the truth and distribute it to a relatively wide audience in order to create the perception of something to promote a goal that they have.

6:30So the question you probably are asking yourself, so if this has been going on for so long from the Bible and definitely in the past few years, why are you jumping right now and saying that the truth is dead? And the reason I'm saying that is the latest advancement in generative AI technology have moved from government and large companies to literally anyone. So those of you who haven't seen generative AI tools, I urge you to forget about play with them, go and do a little bit of research and look at what they're producing today. So MeJourney just recently announced MeJourney 5. MeJourney is a company that allows you to write text and get an image out of it.

7:09You can write anything you want and the image gets created. You can ask for whatever style you want. So it can create a Renaissance painting. It can create a cartoon and it can also create a highly realistic photo. And their fifth generation will blow your mind because there is absolutely no way that the naked eye can tell the difference between an actual photo that was taken in a camera and an actual scene versus an AI generated image. Anything you can imagine, anything you can describe with words can become a highly realistic photo. But that's, like I said, it's just a part of the equation. The next generation of all of this is text to video.

7:52There are already two examples of this happening right now that you can see samples of. One, you could go to a company called Runway. They have been doing incredible things with manipulating video and still images and turning them into video. but now they've announced their second generation that will allow text to video. And do I think it will be highly realistic in Gen 2? I don't think so, at least according to the examples that I've seen, but Gen 3 will come very quickly after that and so on and so forth. The other version of this is just by going to Hugging Face and looking for model scope text to video synthesis.

8:31It's available and open to the public right now. You can go in there, type anything you want into the prompt, and it will produce a short video for you. Again, not fully realistic yet, relatively short, but the capability is there right now. So I'm going back to the definitions of truth. First definition was that which is true in accordance with a fact. But what is a fact? A fact, if I go back to the second definition, a fact or a belief that is accepted as true. So it's kind of like a circular thing, right? If you believe something is true, it becomes a fact, which then can be used in order to establish another new truth based on the fact.

9:16If I can fabricate facts by fabricating a video of an event that did not actually happen with anybody's face on it in any actual location on the planet, this is what is possible within the immediate future. So anybody will be able to create any facts they want in order to establish a truth that relays on that fact to establish it. And then it's in the eyes of the beholder to decide whether they believe in that or not. Very quickly, in the very near future, we will have an endless stream of fabricated news, videos, images, facts that people will use to promote whatever goals they want. Again, it's been done today.

10:01It's just going to be amplified by two orders of magnitude because of the ability to fabricate these facts on your home computer with zero additional resources, no cameraman, no editing skills, no photography capabilities, just by typing what your imagination wants in order to fabricate the facts you need in order to establish the truth or sell the story you want to sell. As I was getting to this realization in my journey through this process in my head, I wanted to go and check what the AI-generated content, so ChatGPT, will say about the topic. So I went to ChatGPT and asked it, act as an AI agent and futurist.

10:41Since generative AI is becoming so powerful, it will soon be impossible to differentiate between the real photos and videos and AI-generated ones. This means anyone in the world will be able to generate facts about events that never happened. Do you think this will alter the concept of truth? So that was my question. The answer that ChatGPT gave me is, as an AI expert and futurist, I believe that the increasing power of generating AI will have a profound impact on our concept of truth. With the ability to generate highly realistic photos and videos that are nearly indistinguishable from the real ones, it will become increasingly difficult to discern what is real and what is not.

11:25And it goes on and on about this and gives more examples like creating propaganda and spreading false information and so on, all the bad impacts that this happens and its impact to everything we know, such as what's the point in having a news media channel if anybody can generate news. So our trust in basically anything will erode over time. In the end, ChatGPT goes and adds the following. However, I also believe that we will develop tools and techniques to detect and mitigate the spread of fake photos and videos. For example, there will be new technologies that can identify and verify the authenticity of digital media, and there will be increased efforts to educate the public about the dangers of fake information.

12:12Okay, so there are already tools out there today who can allow you to detect text that is generated by an AI, whether it's ChatGPT or any other platform. That being said, there are already tools that allows you to take the text generated by ChatGPT and Copy.ai and so on, and manipulate it until those detection tools cannot detect it anymore. So I went ahead and asked ChatGPT about that as well. So I typed the following. There are already tools out there that can trick AI-generated text detection tools to think they are reviewing human-generated text. do you truly believe that future detection tools will be able to catch up to new developments of fake videos and images that will be trying to avoid detection?

12:58Not a long answer, but in summary, it says, in the future, I believe that there will be a continuous arm race between the creators of fake media and those developing detection tools to identify it. So what I think we're facing is think about the world of computer viruses, right? So new computer viruses are being created all the time. Then companies who are making money by detecting these viruses, aka antivirus software, are finding the new tricks, the new methodologies, the new concepts that these viruses are using, and trying to mitigate them with the new version of antivirus. And then the cycle goes on and on and on where in the end of the day, the people who are trying to avoid detection can avoid detection forever because every time they get caught, they find a new technique in order to avoid being detected.

13:50And I think the same exact thing will happen over here. Yes, there'll be a multitude of tools that will allow us to detect fake videos, fake images, and try to avoid them. And at the same time, there will be more capabilities to generate videos that will avoid detection. and hence, we will end up totally, totally eroding trust and the concept of truth. What does this mean? What does this mean to us as individuals? What does it mean on our personal lives? What does it mean to our business lives? What does it mean to society? The truth is, I don't have a clue. And the bigger and worse truth is, nobody has a clue.

14:27And yet, the strain of powerful AI capabilities just keeps on accelerating faster and faster. and it may push us all off a cliff. We don't know what cliff it will be and when it's exactly coming, but it's definitely questioning the fabric on which societies are built, right? So trust and truth are some of the basic things on which societies and human relationships are built. And this will be put to question with these new tools. So while I don't know what it means on the bigger picture, like I said, I don't think anybody knows. I can say a few things for certain. One is the importance of being aware of what I just said and looking for tools that will enable you to detect what is true and what is not and to do additional fact checking with additional sources beyond the one or two sources that come to you through social media or through your news feed or even through national media.

15:20Go and check several different sources in the hope that that video or source that you're watching has not been populated yet across all the other channels. So So that's number one. The other thing that I see as inevitable is the growing importance of human connection and human relationships in the future. And this is the only thing that I see that can mitigate what I just said, meaning if you are in person speaking to people, shaking hands, having dinner, going to a conference together, meeting in an actual business meeting with actual people, you can know that what's happening there is actually happening versus anybody sending you a video of something that presumably happened will be and should be questioned as far as whether it actually happened or not.

16:09Even today, as human and social creatures, we have a higher level of emotional impact where we're actually seeing and meeting people, whether it's on social interactions, personal interaction, romantic interactions, or business interactions. All of this will be amplified dramatically when people will be able to generate videos and images of stuff that presumably happened. I know this episode might be a little depressing to people who has not been in this field and has not been following what's going on. I'm sharing this information not to depress anybody, but to prepare you for what's actually coming and to have you aware of what's actually happening and is available and doable right now, so not sometime in the future, and to get you to think about these topics and share it with other people so as a society, we are better prepared for what's coming.

17:00I hope you found this interesting, and I would love to hear your thoughts and feedback on this. Please connect with me on LinkedIn, Isar Maitis, I-S-A-R-M-E-I-T-I-S, and let me know what you think on this topic. I would really love to hear as many opinions as possible. And I really want to try to spread a way for us to mitigate the negative impacts of this AI wave that is coming. I hope you found this episode valuable. It's different than other episodes that are released that are a lot more practical. This is way more theoretical, but I think it's critical that we have these conversations in order to, as a society, be more prepared for what's coming.

17:35And now to this week's AI news. The first big news is related to plugins in ChatGPT. In the previous show, I told you that OpenAI announced that they're going to open plugins for everyone, but the rollout took quite some time. But at least everybody that I know now have access to the plugins. So if you are a paying member of ChatGPT, you should have access to plugins, which is obviously a big news because it's a huge expansion of what ChatGPT can do. The other big news related to ChatGPT is the fact that they now have an iOS app. So if you have an iPhone, you can now download a ChatGPT app and run it on your phone.

18:13It has some cool features that make it more phone friendly, such as the ability to use a voice to text and actually speak to ChatGPT and get the responses. Another cool thing is that it buzzes the phone when the answer is done. So you know the answer is there. So you don't have to look at it the whole time. So there's a few phone related cool features. The downside, still no plugins. So you can get the plugins now on the web browser, but you don't have them yet. And I'm sure that's probably the next version, but you don't have them yet on the iOS. Another interesting piece of news that's related to Apple and ChatGPT is that Apple now banned its employees from using ChatGPT, which comes at an interesting timing when they just launched the iOS app.

18:58But in doing so, Apple joins a growing list of companies who have banned chat GPT usage, such as Amazon and JP Morgan and Samsung and a bunch of other companies that fear that sensitive business information will leak into the training models and through that to the access of anybody in the world. So that's obviously a big concern that you need to be aware of. If you're using these tools, there's an opportunity that your data, whatever you provide it, will be used for training of the model, which means it's now available for people to explore or exploit when they're asking the model specific questions.

19:36So if you are not aware of that, please be aware of that. There's a lot of other small pieces of news, but these are the biggest news that happened in this past week. And until next time, go explore AI, play with it, try things out. If you find something cool, share with me on LinkedIn. And until next time, have an amazing week.

19:58you