AI has revolutionised numerous facets of our lives while deeply embedding itself into society. Addressing its inherent biases, which perpetuate inequality and discrimination, becomes increasingly crucial as AI's influence grows. While this article attempts to delve into the deep-seated existing inequalities, especially gender and racial biases, in AI generated art often embedded unconsciously during the development process, it barely scratches the surface of this profound issue. Studies like the one by the London Interdisciplinary School about how generative AI art makes its bias worse present very grim results regarding AI bias. The results of my own experiments on the subject through one generative AI tool, Leonardo.ai, provide further insights into the issue of bias in AI-generated art, which acts as an eye-opener by highlighting significant disparities.In the realm of AI-generated art, these issues of bias frequently come into play. The growing recognition of AI-generated art as a legitimate form highlights the need to address biases in this creative domain as well. Renowned artists are increasingly incorporating AI into their work, with examples such as architect Manas Bhatia using Midjourney to create surreal digital images for conceptual projects, and Australian illustrator Georgia Perry exploring new visual textures and environments. As AI art continues to evolve, ensuring that these creative processes are free from bias is crucial to fostering a more equitable and innovative artistic landscape.Continue reading at GenderIT.org. Cover image: AI generated art by Beatrix Kondo
  • Frank [he/him, he/him]
    ·
    18 hours ago

    AI has revolutionised numerous facets of our lives

    doubt

    The growing recognition of AI-generated art as a legitimate form

    doubt

    I know very few artists and art adjacent people who won't insta-block people who post ai slop so I see very little need to address it, biases or otherwise. But "we put reddit in to a blender then poured it in to a data center and now the data center is racist" is hardly surprising, and there's no way to "fix" it without somehow producing an anti-racist training set.

    AI models are neutral tools that reflect the data and instructions provided by their users.

    They absolutely are not. They were scratch built as weapons of class warfare and have little utility beyond eliminating workers, which inherently makes them class warfare on the race and gender and disability front.

    In essence, the responsibility lies heavily on developers and technology companies, who must implement robust ethical standards to mitigate the risks associated with biassed and explicit content generation.

    These things were explicitly built to create a lurching digital homonculus out of the rotting corpse of an artist's stolen dreams. They were built as weapons, they're being used as weapons, and the currently existing society will never use them as anything but weapons.

    This article is content free non-materialist liberal trash. "Ai" cannot be turned to any good purpose because the learning models and datasets were built by massive fascist tech firms to do fascist tech firm things.