How we created AI photos of poverty

without harming those we seek to help

Share this page to help us fight furniture poverty:
Facebook
Twitter
LinkedIn

A how-to guide for generating AI photos

AI-generated art is still very new and the designers and laypeople who are working with it are still in an experimental phase discovering together how far it can go. Though it doesn’t entirely replace the work of a traditional designer, it is another graphic design tool that helps to put ideas into visual statements quickly and economically, and is helping us at Furniture Bank disrupt the charity sector’s reliance on “poverty porn.”

With access to AI tools like MidJourney, Dall-E2, and stable diffusion, charities can compose a key phrase that encapsulates the reality of their clients. It takes tweaking, and practice, and to generate a well-defined concept you have to follow a creative process of digital creation and leverage skilled collaborators to make this possible.

Furniture Bank hired an  “AI architect” Pablo Pietropinto, who worked with us with the following work methodology:

  1. Reviewing our clients’ stories about living with furniture poverty, videos, and understanding the mission and mandate of the organization, along with the reality of what furniture poverty is.
    1. Worked to define style parameters for the images, including:
        • style
        • moods and feelings
        • illumination
        • ambience
        • scenery
        • inspiration
        • camera position
        • focus
    2. Translating the written impact stories into the words and phrases we needed to describe what furniture poverty looks like behind closed doors. Specific details were critical here.
    3. Generating sample images – trial and error is part of this process.
    4. Tweaking the AI-generated images, and adjusting parameters until the scene truly captured the story.
    5. Finally, the images are retouched and ready for the annual campaign.
  1.  

Like any form of art or skill, you can spend A LOT of time and expense learning the ins and outs of the process. We found hiring an AI Architect saved a lot of time, allowing us time to focus on fine-tuning the images to most accurately reflect our clients’ stories.

Stories drove the process

“As a university student, I really needed a desk to study at, a place to store my books. It's pretty hard to study on a bare floor, and I couldn't just get up and go to the library with a toddler.”

“Sleeping on the floor was painful. Without anything in my apartment, every little noise was amplified. Those blank white walls were black once the sun went down.”

“Right now, it still feels like I am homeless because I got an apartment but nothing in it—no bed, no furniture, no pots and pans. I got a key for the place but there’s nothing to sit on, just our clothes to sit on and sleep on.”

“When your place is empty, you feel empty. I’m grateful to be out of the shelter but it’s been sad here without furniture. These walls aren’t what makes a home.”

“Furniture Bank brought me a chance. This is my chance at a do-over. I have a place to live, and now, a place to sleep. With nothing in it, this room used to feel like a cell. Now I’m allowed to say I have a home.”

A great return on investment

What did 40 photos cost the charity?

The idea of doing this project any other way was an impossible task in terms of ethics, time and costs. 

In 25 years of operating, we’ve had two families invite us to see the realities of living in furniture poverty. The probability of finding 20-40 families that would be willing to allow professional photographers to demonstrate their reality behind the closed door is less than zero, not to mention the ethical and moral dilemma of asking those emerging from crisis to put their suffering on display.

And that’s not even factoring in the cost of a photographer for this complex months-long project—we would see the organization invest over $30,000 of funds to pay for photos. 

In using AI, this project was completed within days, and cost less than $900 with an AI architect. Furthermore, we’ve been trained how to do these ourselves, allowing us to focus on finding stories, not photos for years to come!

Final Upscaled Example

How to start using Midjourney

9 steps to get started

Make sure you have a Discord login. Midjourney works entirely on Discord, so you’ll need an account there to use it. Signing up for Discord is also free if you haven’t done it yet. Very popular with younger generations – ask around they can show you too!
Here, choose Join the beta. This will automatically take you to a Discord invite.
Accept the Discord invite to Midjourney. Choose to Continue to Discord.
Your Discord app will not automatically open. When it does, select the ship-like Midjourney icon on the left menu.
 
In the Midjourney channels, locate the Newcomer rooms. There will typically be a number of newcomer rooms open, with names like “newbies-108.” You can select any of these to begin.
 
Now you’re ready to begin creating AI art. Before you get started, note that you only have a certain number of prompt options available as part of your free trial. You can create around 25 free images. After that, you’ll have to purchase a $10/ month membership to continue. 
When ready, type “/imagine” in the Discord chat for your newbies room. This will create a prompt field where you can type the image description. The more precise that you can be with your description, the better the AI will be able to produce good results.
 
 
Be descriptive, and if there’s a particular style that you are looking for, include it in your description.
 
When finished, select Enter to send your prompt.
Give Midjourney a minute to generate your images. Typically, the AI will create several different versions based on your description. You now have a number of options to continue.
 
 
Look below the images, and you’ll see a section of U and V buttons labelled 1 through 4. The numbers correspond to the four images that Midjourney produced. Choosing U will upscale that particular image into a larger, more defined version. Choosing V will create an all-new image based on the present image that you choose. You will also see a refresh button to the side to request a new set of images. Keep in mind that each of these choices will use up some of your available free prompts, so only do it if you are sure you want to proceed.
If you find an image that you like and choose U to upscale it, you will be given a new set of options. You can choose to Make variations on that image, Upscale to max before you download, or choose a Light upscale redo if it’s not quite what you wanted, among other options.
When finished, you can DM the image directly to yourself by selecting the Envelope emoji and adding it to your chat. That allows you to save and download the image whenever you want.
If you plan on using a lot of Midjourney, you can use any bot channel in Midjourney’s Discord and type “/subscribe.” This will create a link that you can follow to pay for a subscription. Those who are really serious about using Midjourney in the long term will also want to take a look at the manual, which will provide you with a greater list of commands and some advice about how to create images.

How the process works

1.

/Imagine

Its a simple as imagining…

Type /imagine and describe in DETAIL what you want to see. Example here:

/imagine a child sitting on the floor, two books open on the floor, setting a empty room, urban home no furniture, intricate details, beam of sunlight through window

2.

Four pictures will be created

Based on your command, you will get 4 choices. You can run new versions, try alternatives to a particular composition or ask for a more detailed version!

3.

Then focus on one to finalize!

Upscale your image to get a final version.

Final Upscaled Example
Share this page to help us fight furniture poverty:
Facebook
Twitter
LinkedIn
AI generated image

Furniture Bank's 2022 postcard campaign

The picture isn’t real but the reality is.

Sample of AI Art used to show how to avoid poverty porn while showing the pain caused by furniture poverty.

Visit the AI photo gallery

Check out the 40 images we created using AI that imagine the reality behind closed doors.