Bookmark: I’m not convinced ethical generative AI currently exists

“Some devs are working on approaches to fairly compensate people when their work is used to train AI models, but these projects remain fairly niche alternatives to the mainstream behemoths.” I’m Not Convinced Ethical Generative AI Currently Exists

The article explores the ethical dilemmas associated with generative AI, noting two main concerns: the opaque acquisition of vast datasets and the substantial environmental footprint of these technologies. The major players in the AI field often disregard the need for consent from content creators whose works fuel these AI models, arguing that the scale required would stifle innovation. Even with existing licensing, these agreements cover only a fraction of the necessary data. Although some developers aim to fairly compensate creators used in AI training, these efforts remain marginal compared to mainstream practices.

Furthermore, the energy demands of generative AI are significantly higher than non-generative technologies, exacerbating environmental concerns. While emerging models like DeepSeek offer some efficiency improvements, leading AI companies remain focused on rapid progress over ecological considerations. Moreover, reshaping AI to be ethical involves rethinking developer practices and user interactions rather than attempting to make AI inherently “wiser” or “ethical.” The challenge lies in the human elements—intentions, biases, and development ethics—that underlie AI creation and deployment.