Practically every day, for a few months now, a type of video generated by various AI tools has been circulating, with the perfect representation of a human being in their appearance, mannerisms, knowledge, voice, and movements. These videos with avatars that are close to the perfection of human representation share attention with others, entirely created by AI, even placing beings (can we call them that?) that do not exist in reality on the job market.
If you haven’t seen any of this type and want to find out what we’re talking about before reading on, here are some examples:
- Reid Hoffmann, co-founder of LinkedIn, recently released an interview with REID AI, his avatar nicknamed “digital twin”, created with some AI tools (interview here).
- How about the opposite, transforming digital characters into humans? AI allows this too. A trailer from the cartoon The Simpsons™ starring humans in the 1950s can be seen here.
- Another example has been seen in profiles of digital influencers created by AI, some of which can earn, according to news, USD 11 thousand per month (news here).
These few examples illustrate how everything seems possible to be created by AI today, from papers to books, films, artworks, and publications by content creators. For a group of people, some of these videos are scary. For others, fun. For others, still, visionary and useful in a creators’ economy, allowing a type of time optimization that was previously inaccessible. All of these positions have their positive and negative points, which help us reflect on the ethical and legal limits necessary for the use of AI.
If we think that the objective is entertainment and art, AI constitutes another form of expression, but, even in this context, challenges arise in the area of intellectual property and also in the role of work: what will become of cinema, with its entire group of actors, producers, costume designers, professionals of all types who have practiced this art with so much emotion for more than a century, but which also normally has high costs, when AI programming allows the production of virtually anything?
When, on the other hand, the purpose of using AI is efficiency and time optimization, it is seen as highly useful. Imagine being able to create a script, environment, and gestures with AI for a video in your area of activity without having to go through an entire process of editing, cuts, re-recordings, insertion of subtitles, new takes for every small error? This use allows us to optimize time, physical effort, and editing work, but, on the other hand, it can generate negative effects such as the elimination of jobs, misuse of images, voices, propagation of fake news, in addition to the possibility of giving space to evil actions against children and other vulnerable audiences.
Without some rule that impose identification of content generated by AI, for example, we may be subject to deception by artificially generated content. There are already several examples published in the media, such as the case of a company employee who transferred 25 million dollars at the request of a deepfake created by AI, vividly posing as the company’s CFO (reported here). Imagine the effect that malicious videos with false information (but appearing to be real) could also have, for example, on an election campaign or on a person’s image!
Like everything in life, the use of AI also needs balance and limits so as not to extrapolate its space indiscriminately. Without a minimum of control, rules, values to be observed, we could start to live in a world in which the emotion of an actress on the screen will be replaced by an avatar; where the work, technique, and effort of a painter will be passed over by untouchable digital arts, although extremely valuable, as occurs with some NFTs; where knowledge acquired with effort throughout a lifetime will be replaced by images generated by tools with a superhuman ability to synthesize information.
In order to achieve such a balance, however, some challenges need to be faced, including the balance between technological advancement and the preservation of rights, guarantees, and minimum ethical, moral, and cultural standards. How far can one of those go without impeding modernization or without offending rights such as privacy, intimacy, copyright, and principles such as ethics and transparency?
We are already living in a time where AI suggests music, films, books, replacing, for some, that recommendation from the book seller or work colleagues; where robots serve us – or try, unsuccessfully, to serve us – to resolve issues with product and service suppliers. It seems that we have reached a time in humanity and technological development in which we will have to get used to watching avatars, thinking that they are people, seeing representations of people who did not actually participate in the filming or who are not even alive, hearing voices that seem familiar, but whose authorization for use is questionable.
We need to remember that AI is a human invention and must be subject to control and supervision by human action. Without such balance, it is not difficult to imagine that we will reach the extreme scenario of receiving orders from robots and impalpable creatures, in a typical science fiction film. Could it be possible that, one day, even the wine we drink will be made by AI? Well, that’s a subject for a future publication.