Can a Machine Be an Artist?
In 2022, a new wave of Artificial Intelligence (AI) tools emerged that could generate stunningly detailed images from simple text prompts. Users could type "a photorealistic astronaut riding a horse on Mars" and receive a unique, complex image in seconds. These AI art generators, such as Midjourney and DALL-E 2, have sparked a fierce debate in the art world. Proponents hail them as revolutionary tools for creativity, allowing anyone to visualise their ideas, regardless of their technical drawing skill. They argue that AI is simply a new paintbrush, a sophisticated instrument that opens up new artistic possibilities.
However, many human artists and designers are deeply concerned. A major point of contention is how these AI models are trained. They learn by analysing vast datasets containing millions of images scraped from the internet, many of which are copyrighted works by human artists. These artists argue that the AI is essentially "copying" their style and labour without their consent or compensation. For example, an AI can be prompted to create an image "in the style of Weta Workshop" or "like a Rita Angus painting," raising complex questions about artistic ownership and copyright law.
The central question is whether AI-generated images can truly be considered "art." Does art require intention, emotion, and lived experience, or is it purely about the final product? If an AI cannot feel joy or sorrow, can it create a genuinely joyful or sorrowful image? This debate challenges our very definition of creativity. As this technology becomes more sophisticated, it forces us to consider what we value most in art: the aesthetic beauty of the image itself, or the human story and skill behind its creation. There are no easy answers, and the conversation is set to reshape our understanding of art for years to come.