Frequently Asked Questions

The Pixelthesia Creative Process

What makes these images special?

Creating meaningful generative images is an alchemical process that combines technical expertise with artistic vision. Like a photographer who understands both the mechanics of their camera and the qualities of light, I work with advanced generative tools to create conditions where beauty can emerge.

The essence of the work lies in careful observation - reviewing thousands of generations to find those pieces that offer genuine moments of reflection. It's about recognizing the subtle combination of elements that allows an image to transcend its technical origins and invite contemplation of beauty and wonder.

What was the prompt you used for (this image)?

I don't usually share prompts for specific images because my prompts are usually not direct descriptions of the result. Instead, I'm usually prompting the generative engine in a way that makes it use its in-built chaos and creativity, in an exploration of the latent space of images.

I do share prompting techniques and sometimes specific prompts in the Pixelthesia Workshop tier.

To get a prompt that might create similar images, you can ask one of the vision-enabled LLMs, such as ChatGPT or Claude, to create an image prompt that would create a similar image. You can also ask the LLMs for other details such as artistic elements analysis (composition, use of color, lighting, perspective, and technique) or about the subject of the image.

You should definitively use generative tools yourself!

As for why you might subscribe, while anyone can type a prompt and generate an image, consistently creating thoughtful and beautiful work requires deep understanding and dedication. Think of it like exploring an vast new territory - while anyone can enter, there's immense value in learning from someone who has already mapped the landscape and knows where beauty tends to emerge.

My work involves:

Mastering the technical aspects of generative systems
Creating the conditions for beautiful images to emerge
Developing an eye for the subtle qualities that invite reflection
Investing the time to hunt through thousands of generations
Understanding how to refine and perfect promising pieces

Technical Details

What size are the images and how are they delivered?

Newsletter images: Medium resolution (~1 megapixel), optimized for screens and email delivery
Paid subscribers may request high-resolution versions (~36 megapixels), suitable for printing

What tools do you use?

My alchemist's toolkit includes:

Midjourney as my primary generative engine
Claude and ChatGPT for prompt exploration and refinement
Custom-built tools for prompt manipulation; some of these are open source
Adobe Photoshop for precise adjustments and enhancements
Topaz Photo AI and Midjourney upscaling for high-resolution versions

Rights and Usage

What about copyright and generative imagery?

I take intellectual property rights seriously. Here's my approach:

All shared images are generated from scratch using advanced generative tools
I take care to review outputs to avoid trademark or copyright issues
I respect artists' rights and avoid using living artists' names in an attempt to copy their style
Images are provided under CC-BY-NC license (see below)
Commercial licensing is available upon request

The debate around generative AI training is complex and evolving. Artists, photographers, and other creators publish their images, whether copyrighted or not, for the purpose of being viewed. I believe that when an AI system views and forms an impression of an image, this process is fundamentally similar to a human doing the same - it just happens much faster and with many more images.

Current AI tools will sometimes produce an image that closely resembles a copyrighted work, or one that might trigger trademark violations regarding commercial logos, trade dress, or characters. This is analogous to using a camera to photograph a copyrighted work - the act of photographing or generating a similar image doesn't negate the copyright. In these cases, users of generative tools should not distribute such images, and I personally do not knowingly infringe on copyrights or trademarks.

I support fair compensation for artists and proper attribution of artistic styles. Furthermore, I believe it's unethical to use living artists' names as keywords to generate images mimicking their styles. AI companies that enabled this practice should implement opt-out mechanisms for artists by default as soon as possible, and should compensate artists for prior infringements, including damages. While "style" has historically not been protected by copyright or trademark - which has created this gray area - I believe the legal system should address this gap while carefully avoiding undue restrictions on artistic expression.

How can I use Pixelthesia images?

Images you receive through Pixelthesia are provided under a CC-BY-NC license, which means you can:

Use them personally
Share them with others
Create derivative works

With these requirements:

Credit me as the originator
Don't use them commercially

For commercial licensing, please contact me. For complete details, see the Creative Commons CC BY-NC 4.0 Deed.

Still have questions?

Feel free to email me anytime. I'm always happy to help!