I am Stanislav Starchenko, a photographer and researcher of artificial intelligence (AI) tools and neural networks. I actively use these technologies in design, photography, and film production.
My experience with neural networks is a year and a half. During this time, I have worked with large companies and created products for them using AI (images and videos). Practice has allowed me to draw certain conclusions.
I would like to point out at once that the issues of AI-related authorship have not yet been legally regulated. The very notion of AI is not yet autonomous enough, as the algorithms of neural networks are written by humans. They are trained and educated on the basis of previously created videos, images, and texts.
AI does not create in the human sense of the word but compiles information, which is certainly impressive. For example, it is possible to faithfully recreate clothes that were in fashion in the 1970s.
Neural networks come in both open-source (allowing you to control the process of creating images or videos) and closed-source (where you can only enter a text prompt). Each type has its own advantages and disadvantages. Closed-code neural networks are no less effective, and sometimes better in quality, and require a less expensive period of training a person to work with neural networks. But it should be noted that when creating images, character poses and other elements in the frame can be unpredictable.
Open-source neural networks take longer to learn, while closed-source neural networks produce faster results.
In no way is it meant to replace humans with neural networks and AI, especially in design or visualisation, making videos, or writing books. But as an additional tool, they are good helpers in finding inspiration or solving parts of tasks. For instance, I have designed presentations and visualised a film script.
There are experiments in the film industry, such as bringing actors who have passed away back to life or using AI to age young actors. Technology can be used to create short stories and quests in both video and images.
AI can automate the process of finding models or film actors according to the requirements of the script.
Neural networks have become indispensable assistants for photographers; when processing photos, they can, for example, clean up the background of photos, do basic retouching, and remove creases in clothing. The same can be said about AI-enhancers of the appearance of actors or fictional characters. But it is still very early to talk about the complete replacement of the work of artists and videographers by artificial intelligence.
Artificial intelligence and neural networks will inevitably become part of our lives. We need to learn how to work with them in school. Already now, there are ethical dilemmas and difficult-to-solve legal disputes related to these technologies. The issue of regulating AI and protecting it from malicious use is also relevant.
Source of graphic AI works: Stanislav Starchenko’s works by means of neural networks.
