At the edge of a generative cultural precipice

TL;DR

This work interrogates the cultural implications of diffusion-based generative models trained on large-scale web-sourced art. Using data from Artstation, DeviantArt, and Danbooru, it analyzes how NFT pressures and model releases correlate with shifts in upload activity, notably a decline among senior artists and a rise among younger creators. It also discusses data-poisoning as a defensive response and frames the broader concern of model autophagy when training increasingly relies on synthetic data. The findings highlight potential erosion of diverse artistic voices and call for broader data access, rigorous analysis, and safeguards to preserve cultural richness in the face of rapid AI-enabled creation.

Abstract

Since NFTs and large generative models (such as DALLE2 and Stable Diffusion) have been publicly available, artists have seen their jobs threatened and stolen. While artists depend on sharing their art on online platforms such as Deviantart, Pixiv, and Artstation, many slowed down sharing their work or downright removed their past work therein, especially if these platforms fail to provide certain guarantees regarding the copyright of their uploaded work. Text-to-image (T2I) generative models are trained using human-produced content to better guide the style and themes they can produce. Still, if the trend continues where data found online is generated by a machine instead of a human, this will have vast repercussions in culture. Inspired by recent work in generative models, we wish to tell a cautionary tale and ask what will happen to the visual arts if generative models continue on the path to be (eventually) trained solely on generated content.

Figures (2)

• Figure 1: Uploads Artstation and DeviantArt, sample of 250 random artists with more than 2K followers
• Figure 2: Uploads for Danbooru2023