Comparing Glaze and Nightshade, from Sand Lab at the university of Chicago the distortion is notably different. Glaze offers less destructive manipulation, while Nightshade, with its mosaic approach, provides robust protection. We prefer Glaze for displaying art above 3 inches and for additional protection for smaller-sized social media content use nightshade. When the photographs are printed, they are reproduced from the original unprotected image.

Looking at the examples, Nightshade performed the best due to the distortion. While the Flatiron Building in NYC is quite famous, GPT had a hard time recreating something that would be plausibly usable upon further inspection above 2 inches. We have not tested this with advanced A.I. generators, but seeing where the distortion is created, it's safe to say this is a feasible way to protect your works online as of 07/09/2024.

With deregulation, there is a risk of fraudulent activities, not necessarily through artificial intelligence or large language models themselves, but through the malicious intent of users who design them for unethical purposes.

As culture adapts to deregulation and or slow regulation with rapid technological advances, we are presented with exciting opportunities and challenges. Artificial intelligence and large language models are powerful tools that enhance creativity and efficiency. While these technologies can inadvertently replicate existing work, they also drive innovation and open new possibilities for collaboration and creativity. Experiencing photography in person is more rewarding; the physical presence of photography as intended can evoke emotions and connections that are difficult to replicate online.

Protecting our work from copyright infringement is essential, it inspires us to find new ways to share our photography in person or at scale, where the print's true essence gets communicated. The internet is a powerful tool for sharing and discovering art. However, it is only one part of the artistic journey. By focusing on in-person experiences, we can create more meaningful connections with our audience and communities to truly celebrate the unique impact of our shared visual communications.

Specifications and requirements for successful cloaking of digital works posted online from Sand Lab at the university of Chicago:

Image specific: The cloaks needed to prevent AI from stealing the style are different for each image. glazed or nightshade cloaking tools, run locally on your computer, and will "calculate" the cloak needed given the original image and the target style (e.g. Van Gogh), you specify.

Effective against different AI models: Once you add a cloak to an image, the same cloak can prevent different AI models (e.g., Midjourney, Stable Diffusion, etc.) from stealing the style of the cloaked image. This is the property known as transferability. While it is difficult to predict performance on new or proprietary models, we have tested and validated the performance of our protection against multiple AI models.

Robust against removal: These cloaks cannot be easily removed from the artwork (e.g., sharpening, blurring, denoising, downsampling, stripping of metadata, etc.).

Stronger cloak leads to stronger protection: We can control how much the cloak modifies the original artwork, from introducing completely imperceptible changes to making slightly more visible modifications. Larger modifications provide stronger protection against AI's ability to steal the style."

https://nightshade.cs.uchicago.edu

We have yet to test the glazing effects on prints, but looking at the artifacts, they might be better as an overlay etched in the glass or acrylic display in the frame rather than in the print itself. 

our tests settings were:

1. glazed: default+slowest

2. nightshade: default+slowest

CHatGPT prompt:

• Copy this image exactly as it is.

All photography posted and sold on www.kevinlandes.com cannot be duplicated, manipulated, or modified for any intended use without written consent from Kevin Andrew Landes.

©kevinlandesphotogrpahy 2009-2024

• Generated images used for testing glaze and nightshade were made with OpenAi Chatgpt40.