A Touch of Joy: Pretty Pictures with Pete

Welcome!

Hi, I am Peter Kaminski. I'm exploring the world of AI art with a photographer's eye and a digital technologist's background. I create thousands of images each month, and I want to share a few of my best with you.

This is my patronage page, where you can subscribe to my email newsletter with beautiful images, personal reflections on the images, and AI art tips and tricks. “Pretty Pictures with Pete” will focus on the images themselves, and keep the written part short. I'll send out a new email around 3 times per week.

An email archive and additional behind-the-scenes files will be available in your Gumroad library on the web.

A Variety of Amazing Images

I like to share objectively beautiful images with vibrant colors and careful composition, and I strive for variety and something a little different. I have decades of experience in digital images and photography, and I harness that experience to sift through thousands of raw images to curate and deliver you dozens of amazing images. You can see some samples in my Pixelthesia gallery - click images to enlarge or browse.

Images you receive through this membership are provided under CC-BY-NC license, which means you can use, share, or edit them, as long as you credit me as the originator, and you can't use them for commercial purposes. Contact me for a commercial license. For exact terms, see the Creative Commons CC BY-NC 4.0 Deed.

What you get: Inspiration and Education

Discover New Horizons: Subscribers get a front-row seat to the cutting-edge intersection of art and technology, exploring how AI tools can transform traditional creative processes.

Learn Through Experience: Gain insights from my personal journey, including trials, triumphs, and lessons learned while experimenting with AI art tools. It's like a guided tour through the evolving landscape of digital art.

Creativity and Enjoyment

Spark Your Creativity: Each issue serves as a muse, offering inspiration from AI-generated artworks and encouraging subscribers to think outside the box and experiment with their own creations.

A Touch of Joy: Enjoy a regular dose of beauty and wonder in your inbox, breaking up the monotony of daily routines with something delightful and surprising.

Money-back Guarantee

If you're ever not happy with your membership, you can cancel and get a full refund for the current month.

Frequently Asked Questions

Is it really art?

Traditional arts and AI art both stem from a wellspring of creativity.

Traditional arts demand years of study and the honing of skills, embodying a deep connection between the artist and their medium that is cultivated through dedication and practice. This rich heritage of craftsmanship and expression sets a high standard for what is often recognized as art or illustration.

Creating images with AI tools requires a unique blend of technical innovation, creative vision, and expert curation. AI art, much like photography or music played with digital synthesizers and other musical tools, represents a new frontier where the artist's role evolves towards setting conditions and refining outputs. This form of art challenges our traditional definitions, inviting a broader understanding of artistic expression in the digital age.

Are these images stolen from "real" artists? What about copyright?

To be very clear, I strive to ensure that any images I distribute do not infringe copyright, trademark, or moral rights of other artists or creators. Please let me know if you have any concerns!

Another thing: image generators don't just "cut and paste" sections of existing artwork to create something new. Instead, they generate an entirely new image from scratch. They analyze the text prompt provided, and then drawing on the "knowledge" gained from reviewing hundreds of millions of pairs of images and their corresponding text descriptions. The result is a unique, newly generated image that reflects the essence of the original prompt.

The issue of training AI on published images is complex and evolving. Artists and other creators publish their images, under copyright or not, so that they may be viewed. I think the process of having an AI view and form an impression of an image is similar to having a human view and form an impression of an image; it just goes a lot faster.

Current AI tools will sometimes produce an image that is very nearly the same as a copyrighted image. I believe the AI tools should have an internal process to discard that image without ever displaying it to the user. Nevertheless, it's up to me whether or not I redistribute an image like that, and I will not.

AI tools can also be used to generate images which if published might trigger trademark violations of commercial logos, trade dress, characters, etc. The responsibility there lies mostly in the hands of the user, and again, I don't knowingly infringe on trademarks.

Lastly, I think it's wrong to use the names of living artists as keywords that then produce images in the style of the artist. I think AI companies that allowed their tools to do that should now opt-out artists by default as soon as possible, and should compensate artists for prior infringements, including damages. Up until now, “style” has not been protected by copyright or trademark, which has allowed this gray area to open up; I think the legal system should close that hole, while not restricting expression of others unduly.

Anybody can make AI art, right? Why pay you when I can click a button and make my own?

Mostly, I ask for a small membership fee to foster a sense of shared value and commitment between us, to recognize that we're exchanging something of value.

It's true that making an AI image is just a matter of typing a prompt and pushing a button. Please, feel free to make your own! Don't feel obliged to subscribe. 🙂

However, while creating any one image is easy, coming up with imaginative prompts, going back and forth with the AI to get the perfect vibe, sorting through hundreds of janky images to get those few stellar shots, and paying for the services to create and share the images all takes time and money.

By the way, if you happen offer a similar subscription yourself, I'm interested in swapping reciprocal subscriptions. Just let me know!

What size (resolution) are the images you share?

The images I share are medium resolution, typically about 1 megapixel, as generated by the AI. They look great on screens -- web, desktop, phones, etc.

They are generated as PNG files, but a thing to note: when sent through email, the email software compresses them into WEBP format (with possibly more compression noise) and the filenames are changed to generic IDs.

Files I share in the Gumroad library will be either medium-resolution 1 megapixel PNG files (as generated, fewer compression artifacts than WEBP), or upscaled high-resolution PNG files, about 36 megapixels, which look great on screens AND as prints. The filenames will typical start with "peterkaminski".

If you need a medium- or high-resolution PNG file and it's not in the web library (i.e., I sent it via email only), ping me via email.

What tools do you use?

Currently I mostly use Midjourney, and sometimes DALL-E or other image generators. I use ChatGPT for creatively expanding prompts. I use Photoshop on some images to correct details, exposure, crop, or to do other small edits. AI upscaling is done with either Midjourney upscaling or Topaz Photo AI, or both.

Other questions?

If you have other questions, please send me an email! peter.kaminski@pathshiftpeople.com