"Glaze is a tool to help artists to prevent their artistic styles from being learned and mimicked by new AI-art models such as MidJourney, Stable Diffusion and their variants. It is a collaboration between the University of Chicago SAND Lab and members of the professional artist community..."

➡️ https://glaze.cs.uchicago.edu/

#ai #art #artists

@rasterweb I don't think this will work long-term.

#AI(s) are really good at learning complex patterns; this is just another complex, regular pattern.

There are really only two solutions that will work,
1. #cryptography
2. respecting #copyright restrictions via #metadata and #microformats

@rasterweb
#Cryptographic #AI #ImageProtection can work for a few reasons

1. declares an intent to protect
2. amplifies #copyright law around encrypted content
3. scrappers are usually simple, and this adds complexity at a different layer than the machine learning algorithm
4. #cryptography adds compute time to the process; how much depends on how dense the method is.

But #metadata and #microformats are also useful

@rasterweb
For a #Cryptographic #AI #ImageProtection solution in the browser,
Create an #OpenSource javascript library that can take an encrypted asset and then decrypt it. Then render the image in a #HTML5 <canvas> tag.

I see two possible implementations:

1. #openkey system - meant to add a decryption step; it won't stop anyone who uses the same library to decrypt it. which requires willful intent to steal the image.

2. use a private system, and use auth or payment to release a key