The Moral Concerns Related To Deep Fake Videos
The world today is of digital transformation. Technology has been improving day by day within the glimpse of an eye. The advancements of Artificial Intelligence and deep learning has been improving a way faster than it was originally thought. 10 years ago the word “fake” would normally be associated with plastic surgery or cheap DVDs. As the world is slowly moving more and more online where anyone anywhere can create and publish a story, we have a question that arises in us “Is this fake?”. Deep fake videos has created a stir with its AI generated clips which were conceived by the Reddit user, deepfake, who used open source libraries (Tensor Flow), Google image searches, social media websites like Instagram, stock photos, and Youtube videos to create a machine learning algorithm which allowed him to insert people’s faces onto pre-existing videos frame by frame. Today free apps like the fake app make Computer Graphical Interface (CGI) face mapping very easy. It’s a simple tool that anybody can operate for free for face swapping. This tool can save hundreds of dollars and a face swapping video can be made as fast as in 30 minutes. This cuts the cost of making videos and effort by many orders of magnitude and also puts a powerful tool in the hands of an everyday person. There is an old saying that “the camera never lies”. When we someone in a video then we have the perfect backup for any idea or action that we want to attribute to that person and it makes our understanding better but these all changes with the rise of Deep Fakes. So with this new emergence of technology, a question arises in us “ Is “deep fake” morally correct? What impact will it have on us and the people around us?”. Now let’s take a look at 2 popular positions of morality.
One which is called Consequentialism and the other is Deontology. Consequentialism is about achieving the greatest good for the greatest number. It was the philosophy of people like John Stuart Mill, classical liberals, whom today we would probably call libertarians. Consequentialism pertains to costs or benefits. We can also think about it as dealing with results, outcomes and as the name suggests, consequences. Consequentialism believes in the theory of whether an act is morally right depends only on the consequences. On the other hand Deontology, with its roots in Kant, determines rightness by features of acts other than their outcomes. Deontology refers to a certain set of principles or rules and moral values. It has its certain set of limits and bindings which cannot be crossed under any sort of circumstances. Deep fake video technology has can save hundreds of thousands of dollars but it also has some real-world consequences. Let’s discuss each of them both from a consequentialist and deontological point of view. Let us look at the television industry for an example. Deeper fakes are altered videos usually of famous people produced by neural networks like superimposing a face onto someone else’s body so that it exactly looks like the same person and then they alter the content making the facial movements and match it with the audio.
Film studios have been able to swap faces for years. Oliver Reed did it for some scenes in Gladiator and even after the accidental demise of Paul Walker during the shooting of Fast and Furious V he was impersonated till the end of the movie with the help of Computer Graphical Interface (CGI) face mapping. In so far as the level of satisfaction of a consequentialist mentality goals with the help of deep fake video technology the industry was able to carry out the task and this was done solely for the entertainment of the people but for a person who is deontological he would also enjoy the performance but also the content might be misinterpreted. He would be misguided and might have an opinion that whatever the actor didn’t do was featured in the film. There have been action scenes where the hero jumps from the 16th floor to save the heroine. Have we ever thought that the film industry will take such a big risk and will make the actors do these stunts by themselves? These dangerous scenes are performed by a stunt performer who is a trained professional who performs stunts. These stuntmen perform the stunts and later they the edited using the deep fake video technology portraying that they were performed by the original actors featured in the film. For a consequentialist, this has been done solely from an entertainment point of view and he would happily enjoy the performance but to a deontologist, he will also be overwhelmed with the performance but the sequence of action will also put them in bewilderment to distinguish from the real deal. One of the major negative consequences of DeepFake videos is in the Adult Industry. Existing porn scenes have celebrity faces imposed on them. All we need to do is to find an actor with the same build and most of the work is done by the AI algorithms. Not only a celebrity it can be any person photo which is available over the social media. To a deontologist, this is something which is terribly wrong as it against all the boundaries as such things have a terrible impact on the persons mental health and career. There have been cases where due to this the person has even committed suicides and have been in trauma for years.
The mass accessibility of deep fake software has many wearing effects that are very hard to ignore. How with the help of a regular Joe one can make realistic videos of anyone saying anything they want them to say and it is very difficult to find the truth out of lies. From a politician saying words that were not spoken against a celebrity, doing things that were not done by him. The treat of these deep fake videos is something that can no longer be denied. If we are not careful enough the next fabricated scandal that can threaten our national security or sway public opinion is just waiting around the corner. To a deontologist, this would have an adverse effect on the personality and dignity of that politician and the celebrity concerned. Also, it depends on the content of the video in which the politician has been portrayed in. If it is a funny video in which then the consequentialist may feel good. However, in my opinion, those who are having great personality cult shouldn’t be mocked at. Doing this a consequentialist may enjoy themselves for the time being in this field of recreation but to a deontologist, it will have a serious impact. Such videos can lose the reputation of a political leader, loss of trust by the people and even the loss of elections. In thinking that this sort of entertainment will put the national or global security at stake. Again thinking from a different perspective, if you can make a politician say anything using the deep fake technology and if these videos become commonplace perhaps you give some ground to the politician to deny something that they actually said then it becomes a question of how even to trust an evidence given.
This would lead to a confusion in the minds of the people as to what to believe and what to trust in. There are some fields of deepfake video technology where the conclusion is drawn in such a platform that there is no tug of war between the consequentialist and a deontologist. Let us look at some examples. Dimensions in Testimony (DiT) is a collection of interactive biographies from USC Shoah Foundation that enable people to have interactive conversations with the hologram of a Holocaust survivor and some other genocides. The answers are already pre-recorded in the studio but the effects are astounding. The special thing about the human interaction is that it is magical and it gives a more profound impact than what books and history could teach us. So a model was developed using the deep fake tech that talks like the survivors and the model look exactly like the photos and videos available on the internet. We can imagine a scenario like Albert Einstein himself teaching the theory of relativity or our grandparents talking to us, wouldn’t that be so magical! And the best part about this is that it is all about us. It’s all about the questions which we want to ask and to know what are the topics we are interested in. And for both a consequentialist and deontologist this would lead to utmost happiness and this would lead to a single platform in the field of the dimension of knowledge. This is what is called a good aspect of deep fake technology. The deepfake technology can also be used for facial scarring and masking deformities online. This makes the person look good on the television screen and this also helps in boosting the confidence of the actors as they feel good and is also a visual treat to the viewers. The deepfake video technology can also be used in blackmail attempts, they could be used in extortion, implicate innocent people in crimes and in civil code proceedings. These videos can be used to carry out all kinds of fraudulent claims and defenses. As we enter this video misinformation, perhaps our biggest weapon is awareness.
So before jumping to any conclusions, we should understand the full potential of deep fake video technology, the working of the model and the consequences of it. If we start recognizing the fact that this powerful video manipulation is now widely accessible and can be easily used by anyone we should start being more critical and mindful of all the video content that we encounter every day. We should try our best to educate people so that they can understand the potential of face-swapping, deep fake video technology including its consequences on others.
