There is no bar in downtown Las Vegas with the moniker “After the Apocalypse”, yet. There should be. A speakeasy. Below ground. With a thick roof and solid metal doors. The design of the card below just happened. Playing around with AI in illustrator. You turn on your computer in the morning and there is an email or something. Firefly now exists in Illustrator and boom. Seven hours later, you have a new thing. Just goofing around.
The way I ended up making the thing after several hours of trial and error involved using Firefly to produce an AI image of an octopus in Firefly on the web. This would also work in Adobe Express. The new thing with these programs, or maybe they are called services, is that you can now upload an image and direct the AI how to use the image in the creation of the AI image that is developed. You can then generate all sorts of images that Firefly creates, picking some as models for others. That is the way it has been.
The new thing, the uploaded image, is a huge thing. Before I could put a white collie that looks like my Luna along the Seine. Generating lots of options would eventually arrive at a similar dog. Now I can upload a actual image of Luna. And boom there she is in Paris. The real Luna.
So I uploaded an image of an octopus. I specified a bunch of stuff and went through many options and found my octopus. I then popped it into illustrator, used the old-school image trace feature to get an illustration style that suited my imaginary speakeasy, and went to work with the new AI recoloring feature. I decided to go with acid-wash-octopus sort of look. Because, well, after the apocalypse…
Next I made a new layer, put it behind my octopus, and filled it with a text-based AI image of a dirty Las Vegas alleyway. I colored this along the same lines of my octopus using the same recoloring AI tool.
I added the text using old-school illustrator methods. I then realized my speakeasy needed a logo. So I hit up Firefly again in search of a text-based atomic mushroom cloud. This was very difficult. I have no idea why. Maybe mushroom clouds are off-limits in Adobe AI.
I think it came out pretty well for a day’s effort all while learning how to use it along the way.
This can be compared with the above early in-progress version of an illustration I made in illustrator for a family project early in January. It took me probably two and half weeks. The red-rock background is from Valley of Fire State Park and the desert big horn sheep are from Boulder City, both in Southern Nevada. They are from photographs taken by myself. The background photo was modified somewhat using a new Photoshop AI feature which extends the photo to larger boundaries.
I then followed my old-school illustration methods where I composite the photos on different layers in Photoshop. Then I tried a new AI method in Photoshop to bring the the subject photo in sync with the lighting of the background photo. I can’t remember if I used that or not.
I popped the photos on different layers in illustrator, image-traced them at different levels of detail. Then on top of the illustrator generated vector images I hand drew my own take on the way the background and the sheep should look. This includes my own particular styling. I’ve done this sort of thing for a long time.
As I progress I hope to find a really nifty place which involves old-school and new-school, imposition of my own personal style, and a much more rapid turn-around made available with AI.
Over the past couple of weekends I’ve hit up a few events that I’ve done before. The Vintage Home Tour in Las Vegas, the Boulder City Garden Tour, and the Utah Tulip Festival at Thanksgiving Point in Lehi.
Much fun was had by all and I finished up with around 650 new images. These are located in my photography galleries in southern nevada usa 15 and 16, las vegas nevada usa 4, 5 and 6, and northern utah usa 9 and 10. I expect to be doing these events on and off going forward.
In particular, the Vegas Vintage Home Show done by the Las Vegas Home and Heritage Foundation is a wonderful thing. After a while you can develop a negative vibe on the roadways of the neighborhoods of Las Vegas. But every time I do one of these shows, I get a very good feeling about the direction of the town. There are so many good people doing the right thing in restoring these old houses. One house was the band-house for the group Fans of Jimmy Century. That and all of the others were very cool. My favorite was an artist’s house with a lovely mix of art and mid-century modern.
In editing these photos I’ve started removing things and people using the AI method in Photoshop of outlining the thing using the lasso tool and using generative fill without any text specified. It works phenomenally.
Of the other AI methods I’ve used in the last little bit, I’ve found the incorporation of uploaded images in Firefly to work very well. This was also the case with the recoloring feature in Illustrator and the image expansion feature in Photoshop. I found the text-based image generation feature in Illustrator to be good for basic illustrations but not up to prime-time for detailed illustrations. It didn’t work at all for nuclear mushroom clouds as was the case with rattlesnake fangs. Go figure.
Another AI feature that has suddenly happened upon me is the new Samsung keyboard which allows you to download language packs and converse with others in real time in any language. So far I’m yacking away in French, Spanish, English and Hindi. To be clear I can’t read a single word in Hindi. But I can send and receive real time translated texts in Hindi. I did check out the Hindi texting with a native Hindi speaker and I suspect it could be a little dangerous if you don’t know Hindi at all because it isn’t always accurate. The French and Spanish are pretty accurate. Brave new world. Seriously this could be such a great thing. Next stop, writing html, css and javascript with Github Copilot and generating layout designs with Adobe AI tools. I’ve mainly written python with Coplilot so far. So. Well. Why not. Could be fun.