I took the image on the left in the 1980s and submitted it to the Tony Stone Photo library (later Getty Images). The images on the right are generated on my computer using the Stable Diffusion program with the text string “big ben clockface high resolution golden light dark sky dslr medium format highly detailed”
The image on the left was captured on film on location in Parliament square, the images on the right were generated by a program that used an algorithm to process such “real life images” and extract from them a sense of what the clock face is and without copying pixels recreate new images. I don’t know if my picture was one of the millions of web images processed in order to train the AI but I doubt that permission to use images in this way was granted by the photographers.
Getty Images announced that they will not represent images generated by AI on its sales platforms for this very reason, neither the source images were likely licensed and the output from these AI programs likely copyrightable. Getty clients are of course free to generate their own AI images but may run into trouble if their work then becomes public domain and a new message attached.
Some of the images generated were more like my photograph but they always came out kind of “strange”, not quite right. What I found more interesting was the mistakes (like the above images), the results of a fever dream or using a consciousness altering drug.
I’m reminded of a scene in the 1980 film “Fame”, in which the tutor (Shorofsky) and his student have this conversation:
Shorofsky : One man is not an orchestra.
Bruno Martelli : Who needs orchestras? You can do it all with a keyboard, an amp and enough power.
Shorofsky : You going to play all by yourself?
Bruno Martelli : You don’t need anybody else.
Shorofsky : That’s not music, Martelli. That’s masturbation.
What I think Shorofsky is getting at here is that a piece of classical music played by an orchestra is made up of humans responding to each other. Each player and their instrument are following and reacting with the conductor and others to create something unique and subtle, a synth can only play samples and does not have this human dynamic.
I would equate ethical photography to playing in an orchestra, an individual photographer interacts with the scene in front of them, they respond to the subject and the subject to them, they move around a scene and frame to find the light and the moment that comes together to make the image meaningful.
Ethical photography depends on honesty, introduce or remove elements (either in real life or in photoshop) and suddenly that nature is tainted, introduce AI generated elements and the integrity is compromised completely.
The counter view:
Use the same synthesiser and create new kinds of music (Kraftwerk / Brian Eno) and suddenly something new and interesting is created. I like the way AI images give us a fresh take on things, I like the “mistakes” stable diffusion and other AI systems make. Is it Photography? – no, not in a million years. Does it have a use? perhaps!
