AI is used in many ways in everyday life (e.g. Google search, finding a route in map services, etc.). It can be used to generate new text, audio, images and even videos based on previously published material. AI models that can produce visual material are taught using a vast amount of images. The images have been collected from the internet, and especially from various image services available, such as Instagram and Pinterest, to which users have uploaded them. An AI can generate new images with the help of another AI that has been taught to identify what is depicted in an image (e.g. cats, clouds or flowers), but also the styles, executions and even moods in the images. An AI can also produce images based on a text description entered.
Images generated by an AI pose many challenges and ethical questions. The images in image material banks focus on western people and culture, and based on them, the AI generates very one-sided material. The images and audio edited with an AI also involve risks of abuse, such as making deepfake material (Explore the phenomenon: Deepfake). One problem that has arisen relates to copyright. An AI can adopt the style of visual expression from work published on the internet and generate similar images, which can endanger the livelihood of artists and illustrators.
AI also has positive uses in the creative industry: designers, illustrators and artists can use the images generated by an AI as examples and starting points for their own design work. It is also possible to generate 3D models and video footage with an AI, even though it is currently quite challenging.
Image editing can be tested using various internet services and applications. You do not have to upload your own images in them, and some of them are free of charge at least in the beginning. The current algorithms for image generation can one day be included in image processing software, for example.
In the future, AI should be developed but also used responsibly and sustainably. The use of AI algorithms is moving towards a more ethical direction. In some image sharing services, people can already deny the use of their images for teaching an AI. People can also delete their images from certain AI algorithms. There is also an ongoing discussion on how the image of humanity generated by AI models could be more equal.
Text: Tomi Knuutila
Image: Siru Tirronen