Disposable Images as
a Commodity

The advances in generative AI have truly made clichéd imagery a disposable commodity

Albeit one of very high quality. And with a lot of other things in the creative industry the price of those images is driven to the negligible cost of approaching almost zero.

We are already in a situation where images are reduced to being a disposable commodity because so many people and online services are creating images. A lot of them look almost identical or at least very similar. Just look at a typical stock photo service or on what people are posting to [[anti-social-media]] services like instagram, pinterest etc. It could be argued, that with so much [[sameification]] going there would be no need for so many similar looking images being used as a throwaway product. But many people are creating similar images as a creative exercise while still putting their own spin onto them. But a lot of those image are just used once. Created to never be seen again

Creating one-of images

Creating custom imagery for one-of purposes and one time uses can of course be of great benefit in certain situations. For example using these one-of images not directly as the finished work that everyone sees, but more as a way of creating concepts and finding a style for the image you actually want to create. In this way, the one-of image would serve the role of conceptualizing and generating ideas. In this stage I can see great potential for [[generative-ai]] and [[generative-design]] as useful tools.

Mass individualization

An interesting feature is the affordable creation of individualized images at scale. Something that was cost prohibitive (in a lot of cases) only a few years ago, now is actually feasible and can lead to interesting design decisions that would not be possible or feasible without [[generative-ai]]. Until now, most of the (digital) product personalization could only be text based. Today we can easily extend this to personalized images as well. Generative AI can also help us in finding or creating a style that will then be augmented with our own ideas or only parts of the image will be individualized. This can lead to feasible mass customization.

Mass customization

This mass customization is really what is interesting about this new approach. Creating large amounts of individualized images following certain rules so every viewer sees their “own” version of those images is really a powerful proposition.

On the other hand it has to be discussed if it’s really worthwhile to create a different images for each and everyone because if you do so what are the people really looking at were they to share those images with other users. For some campaigns this could be an interesting feature (i.e. hyper-individualized coke cans) to see what other people have gotten. But this can quickly lead down dangerous rabbit holes of misinformation. Just think of targeted advertisement using demographic information to further segregate the user base.

Save or delete?

In the case of one-of images a questions worth asking. How are the images used? One time only? Multiple times with one user? Multiple times with multiple users? How are the images created? In the artistic way by manual labor? Or procedurally by prompting [[generative-ai]]? In the latter case the images could be created over and over again every time they are needed. This of course would depend on the economic cost of generating those images, which at the moment is much higher than you would think.

Images as a commodity

Images fulfill varying needs in our [[contemporary-media-landscape]] and their uses are changing rapidly. But since we humans are visual images we succumb easily to beautiful images, no matter how they were created or how often we have already seen similar images. So in the meantime, I do not expect the generative AI craze to stop anytime soon.