Deep Fakes Apr 14 2023

Joe Rogan, a popular video podcaster, recently commented on a deep fake of his podcast. Deep fakes are created using artificial intelligence to copy someone's voice and image and make them say things they never said. This technology has become more accessible, allowing anyone to create deep fakes. Deep fakes are made using deep neural networks and face-swapping techniques. They have been used for illicit purposes, such as non-consensual pornography and creating misleading political videos. It is important to be skeptical and verify information to avoid spreading false information. The truth is crucial for a successful life and society. Hello, my name is Gordon Wilson, and here is my commentary. What prompted this commentary was a Twitter post from Joe Rogan, who commented on a deep fake of his podcast, so you may have a couple of questions at the moment. Who is Joe Rogan, and what are deep fakes? Those of you who are fans of the MMA, that's Mixed Martial Arts for the Uninitiated, know who Joe Rogan is. He's a frequent commentator at MMA Fights. Years ago, Joe Rogan also started a video podcast that just interviewing people. He had no political agenda, he simply had a curious mind and enjoyed talking with people who had distinguished themselves in any discipline that he found interesting. His video podcasts have grown so popular that they are watched by the millions, making him the most popular video podcaster on the internet. What are deep fakes? A deep fake is taking the rapidly developing technology of artificial intelligence to copy both the voice and the image of a person, and then giving them words they have never said. The advanced state of computer technology has reached the point where the processing power of computers is now able to process millions and millions of instructions in a microsecond. Now, this is an oversimplified explanation, but in essence, it means a computer program can integrate new information taken from any source, and in essence, learn new information. The technology to copy what an individual looks like and to reproduce their voice has existed for quite a while, but the technology also required massive input from a very skilled person to produce the results. Now with artificial intelligence, this ability is made available to anyone who wants to employ it. In fact, this morning's news broadcast and everything you have heard up to now has been produced by AI. There's a website that you can go to and enter samples of your voice, and the website will clone your voice. I did that about three days ago, and what you hear now is the result. Now I'm going to switch to my own voice for the remainder of this commentary, and this is what I sound like in real time with my real voice. Now I'm sure you can tell the difference immediately that the cloned voice wasn't quite the same as my real voice. For one thing, it talks faster than I normally do, but nevertheless, it was a very good replication of what I sound like, and I produced that in a matter of about 15 to 20 seconds. So clearly, I'm a complete novice at using this technology, and also the programs I used were completely free. If you want to have available more sophisticated features, you have to pay for it. There are several methods for creating deep fakes, but the most common relies on the use of deep neural networks that employ a face-swapping technique. You first need a target video to use as the basis of the deep fake, and then a collection of video clips of the person you want to insert in the target. The video can be completely unrelated. The target might be a clip from a Hollywood movie, for example, and the video of the person you want to insert in the film might be random clips downloaded from YouTube. The process guesses at what the person looks like from multiple angles and conditions, then maps that person onto the other person in the target video by finding common features. Another type of machine learning is added to the mix, known as Generative Adversarial Networks or GANs, which detects and improves any flaws in the deep fake with multiple rounds, making it harder for deep fake detectors to decode them. Though the process is complex, the software is rather accessible. Several apps make generating deep fakes easy even for beginners, such as the Chinese app Zayo, DeepFaceLabs, FakeApps, and FaceSwap, and a large amount of deep fake softwares can be found on GitHub, an open source development community. Deep fake technology has historically been used for illicit purposes, including to generate non-consensual pornography. Deep fake video has also been used in politics. In 2018, for example, a Belgian political party released a video of Donald Trump giving a speech calling on Belgium to withdraw from the Paris Climate Agreement. Trump never gave that speech, it was a deep fake. That was not the first use of deep fake to create misleading videos, and tech-savvy political experts are bracing for a future wave of fake news that features convincingly realistic deep fakes. I'm producing this commentary now so that you can be aware of how easy it is to put words into somebody else's mouth. There's an old adage that says believe nothing what you hear and only half of what you see. That pithy little piece of advice is now more important than ever to follow. How do you spot a deep fake? Well, the first thing is don't be gullible. Train your detective mind, learn to see those things that don't seem consistent with what you know about the person or the events in question. With video deep fakes, there is obviously more things that you can look for than a simple voice clone like I employed as an example. Check and double check the information you are being given. Don't be guilty of spreading false information even if it's information you would like to be true. Jesus, speaking to his disciples, made the statement, the truth shall set you free. On the other hand, the New Testament tells us that Satan is the father of lies and his purpose is to destroy life and not save it. Be careful you say and speak the truth. A successful life and a thriving culture cannot succeed when the underpinning foundations are based on a lie. Well, that's my opinion for this Friday afternoon, April 14th. My name is Gordon Wilson, and this is my real voice.