Art Tagging

By Ben Smith | July 23, 2025

In an effort to be transparent, I have started tagging all my artwork with information about each art piece. I didn't start tagging my art pieces until fairly recently. But, I've done my best by going back through my files and notes to make these tags as accurate as possible, but being human, I may have missed a few. This page is for describing what tags I use.

Artwork Tags

The things I try to document for all my artwork:

• Application(s) that I used to create the artwork
• Date of the art piece
• Whether or not stock imagery was used
• Whether or not any kind of generative image model was used

Application

When I was learning to create digital artwork and trying to develop my skills, I often found myself trying to figure out what applications other people used to create their artwork. It was often a fraught exercise, so I wanted to make it easy for myself and others to know which application(s) I used to create each art piece.

Generally speaking I create almost everything in Adobe Illustrator, but there have been times where I have also used Adobe Photoshop and or Blender. I keep telling myself that I will spend time learning applications like Blender, but history has shown that I haven't done very much of that.

Date

For a very long time, I was incredibly rigid with my starting and finishing each art piece between midnights everyday. Practically speaking, I work graveyards and I generally stay up really late, so I do create an art piece for each day, but it is more based on when I end my day by going to sleep than when the calendar day actually ends.

My philosophy for this comes from Wii Fit of all places. As a kid, I remember playing this game a lot and really enjoying it. This game, like many, has a concept of daily streaks, and I was corrupted by one setting. Wii Fit has the option to set your 'new day' rollover point for streaks to 3 am instead of midnight. This has stuck in my brain for around 15 years at this point.

This is also compounded with the issues of daylight savings and traveling across timezones. As of writing this, I have been creating artwork for over 4 years and I have created something for each of those days.

Stock Imagery

On the occasions that I use stock imagery, I don't use anything that isn't licensed or freely available to use. The ways I use these stock photos varies, but the primary ways I will integrate stock photography are as follows:

• For texturing, often this is done with a transparent cloud image (ex. Day 1361)
• Silhouettes of people running/standing (ex. Day 1268)
• Faces/People, often. through Adobe Illustrator's Image Trace Vector tool (ex. Day 321)
• Direct implementation into the artwork (ex. Days 662 and 704)

No matter how little stock imagery I use in my artwork, I have tagged them to show that they have used stock imagery. I personally don't think there is an objective measure of whether using stock imagery is positive or negative when it comes to artwork. I have found in my own design process it can be both a useful tool when used correctly and a crutch when used poorly.

It is important for me, especially in the era of generative imagery, that I be upfront about what I use in my artwork.

Generative Models

I have experimented with various generative models over the course of this project. I know feelings about the use of such models is complicated and often very polarizing. I've tried to take special care to annotate each image that I have created to include that they were made with assistance with generative models and also which models were used.

I've included some information about each model I have used. I've put them in the order in which I used them, which roughly corresponds to how capable these models are and how polarizing each of them can be.

Adobe Photoshop's Random Tree Generator

I have used this only a few times, and I don't know if it necessarily counts as a generative model, but included, if only to be thorough. At least when I used it, it could generate a couple styles of trees based on a random seed and you could use them however you wanted.

This Person/Chemical Does Not Exist

thispersondoesnotexist.com and thischemicaldoesnotexist.com both predate what generally is accepted as modern generative models. I have also included these since they aren't creations of my own, but don't qualify as stock photography. These sites have collections of images that have been generated (I can only assume as a pre-generated collection), that gives you a random one each time you reload the page. I have used the one for faces twice (Days 236 and 270) and once for the chemical (Day 391)

Stable Diffusion Via DiffusionBee

I, like many in the Apple community, played around with DiffusionBee when it first came out. Not many of the things I created in DiffusionBee worked their way into my published art, I personally think each of them was tastefully created and weren't overly reliant on the models from an artistic perspective (Days 392, 572, 585)

Adobe's Firefly Model

This is the model I have definitely have used the most and undoubtedly I have the most conflicted opinions on. By my count, when writing this, I have 171 of my daily art pieces that were created with some or significant usage of this model.

This was done exclusively from within Adobe Illustrator with the native integration of the Firefly model that enables text-to-vector image generation. Of all of the generative image models that I have used, this one by far has the most variance in the quality of artwork I have produced.

As I look back on the artwork I created using this, my favorites are ones that I still put a lot of effort into (Days 991, 993, 1055, and 1068) and the ones I like the least are ones that had very little added to them beyond a base image (Days 1070, 1107, 1139, 1161, 1226, 1227, etc.). As I worked with this tool more, I liked it less. I found myself getting lazier with the creation process as I used it more.

This, like stock photography isn't inherently bad in my eyes. But it is a tool to be wary of. When I used it well, it has helped me to create some of my all time favorite artwork, but on the flip side, many of my personally most despised and disliked pieces I have been created using this model.

The last art piece I made using any kind of generative image model was on Day 1248 nearly a year ago. I won't say that I will never use them again, the future holds a lot of mysteries, but for the foreseeable future, I don't have any plans to use these models other than out of curiosity and exploration.