Every single one of us is striving to create such a communication space where we would feel the most comfortable. We build family relationships and connections at work; choose where to live and who to be friends with; fabricate our own environment for interacting with others. Social networks are fantastic for this purpose, because they conveniently let us build borders to delineate this space of communication. You can arrange your virtual world in such a way, for example, that you can speak and be heard, but not be spoken to. Or vice versa: you can stick to listening and never speak your own mind. All sorts of fine-tuning is possible, and it works instantly. You can ban a specific person, become invisible to them and, for good measure, make sure that they don’t see you either. Wouldn’t it be perfect if we could do all these things in real life, too?
In this project I create synthetic images of people to study the possibilities of increasing communication comfort. If we can choose a communication partner, why not choose this -–real or synthetic – partner’s voice, the color of their eyes or hair, their looks. I want to understand how a person feels when they have a choice, and not only between the real and the synthetic. This project is a step from “I can improve the quality of communication” to “I can design my communication partner”.
The project was inspired by the fact that I had spent the last two years staring at a computer screen. The virtual space has become the place where I work, teach, meet people and relax. Those people I met are not synthetic, of course, although I often don’t see their faces. On the other hand, the faces of real people that I do see, can well turn out to be synthesized. Some of us alter our faces to seem more handsome or trustworthy. And although I am perfectly happy with my face, I, too, do this from time to time. I also replace other people’s faces on the photographs I take. At times because I don’t want to recall this or that person, at other times – because they didn’t give me a permission to capture their image, at other times still – to tell a story. So, at one point I thought that I could replace the faces of people I was talking to, and decided to experiment with the way they’d look like.
In the real world, I really miss this opportunity to change the identity of those who dwell in my communication space. Wouldn’t it be nice to be able to choose a TV anchor or an online lecturer, to change the voice of a person reading your audiobook or even the voice of a person you are talking to? It will be possible sooner or later, and once the technology becomes available, I'll be among its first adopters.
Some of these things can already be done: I used one neural network to generate pleasant faces for non-existent people (having to choose one among hundreds or even thousands!); I used another one to adjust the faces I had chosen so that they would look even more attractive for me. Yet another neural network helped me animate these images. I ended up synthesizing about 60 images and named them all ‘Rachael’ after the character from Blade Runner.
The idea of the title for this project comes from a scene in Blade Runner where Deckard first meets Rachael, one of the synthetic humans known as replicants. According to the plot, Deckard visits the company that creates the replicants to check if he can apply a test that distinguishes between a human and a replicant. The company’s CEO asks first to check the test on his assistant, Rachael. After a test that has been running much longer than usual, Deckard finds that Rachael is a replicant who believes she is human. The CEO explains that she is an experimental model who has been given false memories.