bruno caldas vianna<p>I was really happy for having this article accepted for the Evomusart conference. It talks about all the aesthetic problems in the training of the first stable diffusion models - only SD1.5, since after that no completely open model came out. In short, the taste that got imprinted in the model represent a very small slice of society, in particular some white, male, rich techbros from the West.<br><a href="https://works.hcommons.org/records/g917m-xrj41" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">works.hcommons.org/records/g91</span><span class="invisible">7m-xrj41</span></a><br><a href="https://mastodon.social/tags/stablediffusion" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>stablediffusion</span></a> <a href="https://mastodon.social/tags/generativeai" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>generativeai</span></a> <a href="https://mastodon.social/tags/aesthetics" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>aesthetics</span></a> <a href="https://mastodon.social/tags/ethics" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ethics</span></a> <a href="https://mastodon.social/tags/diversity" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>diversity</span></a> <a href="https://mastodon.social/tags/opacity" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>opacity</span></a></p>