I started learning AI with Decohere and then moved to Midjourney, Runway and others. I have been experimenting for a while and also AI photography brings a lot of satisfaction with high quality images, I was bothered by the fact that AI images are not real. At the same time, I could see the tremendous effort behind the quality of images and I loved how simple I could render a professional image without much efforts. I could see also how animating AI images could help in my projects, such as documentaries and especially illustration of open source investigations. Images carry a powerful message but I needed to find my own style and identity.
I started to take online classes for Midjourney and the software is quite satisfying that it lets you progress pretty fast. I wanted to create a model for a character movie documentary and ended up with a fascinating series of images that would emotionally react with their environment. It did not appear at first glance, but exercising to create a muse from scratch and applying styles to her, her face would react with the style and the environment. I just needed to exploit this property and master how it works. The idea of writing books came through my experiments designing Martha. I had to organize my styles, I wanted to record how I created them and be able to replicate how I did it. Making books became a way to organize a timeline of events that led to Martha be emotionally reactive. In the process, I made this video from Midjourney presets.
If you are an AI designer, you will find the subtilty behind those images, the lines of the face, the wrinkles, how light attaches to those lines and how colors are organized. I learnt so much from Midjourney, the way the robot interacts is really satisfying, even if sometimes I get frustrated. I will talk more about my frustrations later.
I designed Martha as an elderly woman because I wanted to create the character of a Director of Photography in her 50s, someone I could identify with and work behind the scene to carry my message. To design a woman in her 50s with Midjourney was not easy. Either the character looked too old, or too homey, or too grandma, or too feminine. I mean, woman are either represented with an apron or lips stick, but there is never in between. I wanted a woman in her 50s who is dynamic, travelling the world, investigative journalist, engaged, curious, someone who has empathy, who can feel the events and can show that she feels. I wanted someone emotionally reactive to tell without words how a situation shows when I am making a documentary.
My documentaries are about poverty, homelessness, price gouging, forestry, deforestation, biomass abuses, mafias, political scandals, overall the large aspects of the society that significates sufferance. Some things I have to say cannot be live recorded because it would jeopardize my investigations. My sources need confidentiality and it is my belief as a photograph that I don’t need to treat people like wild animals to talk about the struggles of the society. People need to preserve their own dignity and I don’t believe that showing everything tells everything because the audience is not prepared for receiving and understanding what they are shown. My role as a photographer is to translate a complex situation into simplicity and rather than show raw images, I prefer to process my images with filters to only show what serves the purpose of my work.
Literature has done documentaries for ages by building fictions but making those fictions so real that people could relate into their own life and vicinity. That’s what I want to do with my documentaries, to relate, narrate and explain real things that people can know are real, but not showing the absolute reality and instead, filter reality with AI.
Developing my own style with AI has become the purpose of saying “this is a fiction made with AI” rather than showing AI photography and letting people to believe that this is true. By showing an illustration of reality with filters and exaggerated sceneries, I am able to trigger the switch in people’s mind to lead them try to understand the message rather than hiding behind what they believe to be reality. An example, when I talk about homelessness to wealthy people, they always react to pinpoint the failure of individuals and if they think homelessness as a group, it is still the failure of a group. In the contrary, I am questioning the failure of the society, how the society created the group.
As any scientific approach, there is a postulate and I have a postulate as an investigative journalist, a researcher, or even an artist. I can tend to be neutral and explain things with neutrality, but it usually involves showing things on different sides and from different sides. Real images with a camera show the subject but never the cameraman, never the interviewer, never the biased aspect of the question while in my case, I am testifying that I am biased, I am agreeing that I am biased, I am showing how I am biased and how it affects my perception but I am giving the tools for the viewer to correct me. AI helps me to build a narrative, an approach and showing how this approach can never be reality.
Martha became older as I worked on the character of Sofia from the following video, but Martha took this sweet face of older people when they know so much things that we don’t. I wanted to use this character as a filter. If I blend Martha with someone else, they will be this old soul. You will find on the video above many representations of those old souls, what means that pictures aren’t just pictures. Photography can capture way more than people’s images and that’s usually what I seek with photography. I don’t look for the “Midjourney’s perfection” but the Midjourney’s possibilities and I find them with imperfection, when the robot got emotional too. You hear me, some features of rendering show the incapacity of the robot to answer on everything and that’s usually when the image becomes interesting, because the robot wants to narrow information within the sample of an image, injecting inside the image the DNA of emotions. When you zoom out on those emotions, a brand new world opens.
I am still at the beginning and learning, but this is why making books became so urgent. Learning, I am still in this phase of organizing and knowing what’s important when you start with a model. When you got too experienced, you lose this ability to prioritize essential features and you also loose accidental results while with art, accidental results are usually the mark that something happens and as an artist, you try to replicate and master the technique. My books are not meant for newbies, I mean not only. My books are meant for advanced artists who want to discover the world of inspiration with AI. I make them for myself, to not forget about simple mistakes I made that turn out to be fantastic and to not forget to sometimes improvise wildly.
I am the person wild behind the camera. I can ferrously track a subject if I want to say something, find the right angle, reach the angle, get to this point of view. Poverty is almost invisible in a city and even if I sight it, I might not be able to stay decent and respect the privacy of people. AI helps me to contour this struggle and styles bring the wilderness, how from photograph, I become investigative. How from OSINT, I can build a non fictional movie with fictional pictures. How also my portfolio helps in getting images because I can impregnate my ethics into the AI camera. As a woman, I have a notion of intimacy, respect, “pudeur”, danger, security, concern, that filters my images. Even if I can reach one image, being a woman probably means more than just release and it is not about the robot telling what I can show. It is not about gender either. It is about, just like Martha, having this ADN of women in history, in biology and in experience, being able to restrain from showing the crude raw aspect of a situation.
If you like the content of this website, please consider to make a donation. I have setup a Gofundme page, a Givebutter page, an individual Stripe page, a business Stripe page (for donations from businesses with tax information), a Paypal page, a donation by SMS system (text 3WORDSADAY to 53-555). I am not begging, I am explaining how it is important to continue manipulating images with AI for a good cause. This involves hours and hours of hard work, a good state of mind and a purpose. Art and documentaries are my purpose.
Click on the Google translate icon to find your language.