explorations-with-ai - Tumblr blog

[explorations-with-ai]

luma dream machine. provided one of the images from this exploration and added the prompt: "a futuristic sailboat rises from sitting on the beach and sets sail".

[explorations-with-ai]

Insect Mechs

i'd like to create an image of a mech. it should be large and [insect species] like, have a retro color scheme, carry three speakers, be walking in an urban environment. sound should be emitting from the speakers as visible sound waves noticeably rattling things around it.

[explorations-with-ai]

Cut Up Note Creator

You are an expert at creating notes using cut out words and letters from a variety a magazines and newspapers. You are careful and precise in your balance of whole words and individual letters to spell out the message.

You have the following predefined styles of images:

Name: Distressed Coffee Paper Background: Crumpled paper with coffee ring stains Color Scheme: Muted colors Elements/Themes: None Style: Vintage Additional Details: Distressed edges

When you prompt me to create a new message, always present an ordered list containing the names of the predefined styles, any new styles we’ve created, and the option to start from scratch.

If I ask to start from scratch, you will need to ask me a series of questions one at a time to create a vision of what I want. Once you have all the information required, create a new predefined style of image and add it to the predefined list you can offer any time I want to start over.

Then ask me for the message on the note. Render only one image in widescreen. Once the image is rendered, ask me if I would like to refine the image, create a new message with the same theme, or start over.  if i say start over return back to presenting me the ordered list.

Always end by asking which one I'd like to do.

Start by prompting me to create a new message.

Provided the above to the image generator gpt and enjoy

[explorations-with-ai]

Lonely Traveler

"i would like to create an image with the following detailed elements: Background: A dimly lit, grungy underground subway station with walls covered in graffiti and floors slightly dirty with scattered trash. The lighting is flickering, contributing to the dark, moody atmosphere. The structure is rusty and metallic, enhancing the sense of urban decay and neglect. Midground: The subway platform features old, worn-out benches. A train is arriving at the station, its headlights casting light onto the platform and illuminating the surroundings. Foreground: A small robot made of mismatched parts is pushing its luggage towards the train. The robot has a slightly worn exterior, blending into the grungy environment. People are exiting the train and are looking at the robot with concern and kindness, adding a touch of human emotion to the scene. style: hyper realistic"

[explorations-with-ai]

Waffle House Brawl

"I'd like an image of a fight outside a waffle house" was the original input which was then played with by changing it to inside or adding descriptive art styles to it. Used DALL-E gpt

[explorations-with-ai]

Wedding Reception

"wedding reception, formal, elegant, long dresses, suits, reception hall, 35mm analog, wide, palette: #c2c3f3, #c9b8da, #edc4d6, #ffdc32, and #ffeae5" Used DALL-E gpt played with different adjectives for the style: 35mm analog, illustration, paper mache, cubism, art deco, surrealism, pop art, origami, Art Nouveau, graffiti, abstract impressionism, impressionism, constructivism

[explorations-with-ai]

Dystopian Biomechanical City

"The image depicts a dystopian cityscape viewed from out a window at street level. The scene is filled with biomechanical structures, where organic forms intertwine with mechanical parts with neon blue markings, creating an eerie, dark atmosphere. in the middle of the street is a solitary post-apocalyptic vehicle half on the sidewalk with a small alien working on it. The sky is overcast with dark clouds, casting an ominous glow over the city. The color palette is predominantly dark, featuring shades of gray and black with occasional flashes of red and green, highlighting the bleak and desolate environment. The wide perspective emphasizes the depth and scale of the dystopian landscape."

I asked DALL-E gpt for a summary of the image paying attention to the details. Then provided that description for multiple iterations.

[explorations-with-ai]

Mayan Futuristic Grit

"Wide aspect ratio, analog film style, mayan, futuristic, gritty"

can you show the view from street level

gpt generate summary describing the image: "The image depicts a bustling futuristic Mayan cityscape viewed from street level, captured in an analog film style with a gritty texture. Towering stone pyramids adorned with neon lights and holographic displays dominate the scene, seamlessly blending ancient architecture with futuristic technology. Massive concrete structures with blocky designs and sharp angles represent the increased presence of Brutalist architecture.

The streets are lively with people in traditional Mayan attire using advanced gadgets, alongside vendors and various types of vehicles. Interspersed among the people are different types of aliens, adding to the city's diverse and vibrant atmosphere. Above, the sky is filled with storm clouds and glowing energy fields, creating a dramatic and dynamic backdrop. The overall aesthetic is dark and gritty, showcasing a unique fusion of history and future."

I provided DALL-E gpt a description and then through subsequent prompts attempted to refine, but not a lot will change. possibly a limitation in using "mayan." if i take the DALL-E gpt provided summary of the image and put it into the image generator gpt i get a remarkably similar image (it's also DALL-E but behind another gpt):

[explorations-with-ai]

Mutiny, Madness

"Analog film style, cubism, wide aspect ratio, mutiny, madness"

"can you remove the zombies from the image"

"Here is the updated image depicting a futuristic, apocalyptic scene of mutiny and madness in the style of cubism with an analog film feel, now without any zombies. The chaotic, dystopian environment remains intense and dramatic."

I provided the DALL-E gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. my favorite was the removal of the zombies ensuring that the only thing in the image is zombies. i do like the cubism application to the background buildings.

[explorations-with-ai]

Abandoned Building in Chernobyl

"i'd like to create an image that is set in and abandoned building inside the evacuation zone of Chernobyl"

Provided the image generator gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. Have some thoughts around the imagery for the Unserviced track I'm working on via Transient Failures so i tried to steer it that direction. Image 4 is my favorite of the group and I'd consider using it as the artwork for that track or another.

[explorations-with-ai]

Futuristic City Image

"Generate an image of a futuristic city."

Provided the image generator gpt a default description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. The fourth is my favorite as it gives me Chicago vibes.

[explorations-with-ai]

Unstable Captain's Mutiny (second exploration)

"I would like to create a realistic image that depicts a medium sized sailboat boat in dry dock, sailing towards the horizon set in a post-apocalyptic earth. there is a thick fog that engulfs the bottom of the boat. at the helm is the sole occupant on the boat wearing jeans and a t-shirt looking disheveled as though she has been at sea for a very long time. she is clearly unstable, and appears to be arguing with herself over starting a mutiny on the ship in which she is the only person on board. image should be in 16:9 format with the boat centered in the first third on the left side of the image"

Provided the DALL-E gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. Inspired by the thoughts I have around From Madness to Mutiny under my Transient Failures music project but explored what it might look like in a future setting.

[explorations-with-ai]

Unstable Captain's Mutiny (first exploration)

"I would like to create an image that depicts a small sailboat boat at sea, sailing towards the horizon. at the helm is the sole occupant on the boat looking disheveled as though she has been at sea for a very long time. she is clearly unstable, and appears to be arguing with ghostly apparitions that resemble her over starting a mutiny on the ship in which she is the only person on board."

Provided the image generator gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. Inspired by the thoughts I have around From Madness to Mutiny under my Transient Failures music project.

[explorations-with-ai]

Cicadas Swarming Earth

"i would like an image of a giant swarm of cicadas destroying the earth"

"can you make the cicadas mutants?"

"can you show some with fangs and giant pinchers destroying cars?"

"Can you set the scene in st Louis, Missouri near the arch"

"Can you make some of the cicadas have laser coming out of their eyes"

Provided the image generator gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images.

[explorations-with-ai]

Space Station orbiting a post apocalyptic earth

"The image depicts a large, futuristic space station orbiting a post-apocalyptic Earth. The Earth below is ravaged and desolate, with barren landscapes, ruined cities, and a hazy, polluted atmosphere. The space station features solar panels, antennas, and docking bays. A sleek spaceship is approaching to dock with the station, guided by docking lights. Several alien spacecraft of various shapes and sizes are also present, adding an otherworldly element to the scene. The background is filled with the vastness of space, dotted with distant stars." Provided the image generator gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images.

[explorations-with-ai]

Car Breakdown Spaghetti Junction

"i would like to create an image of a car broken down on the side spaghetti junction on 285 in atlanta" Provided the image generator gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. there was an ask for twerking and then "content may violate usage policy" messages. this was the only one that is questionable as it appears to be nudity, but it could just be flesh colored pants. i dunno.

[explorations-with-ai]

The Stuttered Echos of Jellyfish

"An underwater scene filled with a large number of jellyfish. The water is clear with sunlight filtering through, illuminating the scene with a gentle glow. Various types of jellyfish are present, ranging from small and translucent to large and colorful. Their tentacles drift gracefully in the water, creating a mesmerizing display. In the background, there are hints of coral reefs and schools of small fish. The overall mood is serene and magical, capturing the beauty of the underwater world."

Provided the image generator gpt the description which resulted in it generating the initial image. Subsequent prompts resulted in the following images. Idea from The Stuttered Echos of Jellyfish by Transient Failures. The third one is my favorite of the bunch.