|
I've seen more than a few folks on my Mastodon timeline express enthusiasm about Glaze, a system whose goal is to protect artists against style mimicry by generative models. If you're using this work, or otherwise have heard of it and thought "wow this is neat I'm happy people are doing it", I strongly recommend you read this blog post by security & privacy researchers Nicholas Carlini & Florian Tramèr about evaluating Glaze's claims: https://spylab.ai/blog/glaze/ 4 comments
@tedted "So we should maybe just be transparent that [schemes for protecting against machine learning] are mostly for show." @tedted yeah I'm p sure I caught a bunch of blocks for demonstrating a bypass to Glaze 1.0 w/ GIMP for lineart and flats (not by the glaze team but just periphery ppl) bc it fundamentally cannot work for those, and the Glaze peeps didn't really acknowledge that at all until their Nightshade paper π¬  @tedted@hachyderm.io security by obscurity will never stop being hilarious |
Gregory's Server
@tedted yeah. I was starting to have some doubts when I learned that there wasn't any public source code released, which I found odd from a research perspective.
(Although the thing that kept me from using it ultimately was the lack of a linux release π)