By Dr Greg McMillan GMAPS, EFIAP EPSA BPSA

Photography as a Serious Hobby

Throughout the 20th century, photography evolved with the introduction of film, colour photography, and digital photography. Today photography has become an integral part of our daily lives, with high-quality portable cameras (smartphones) allowing people to capture and share moments instantly.

Whether the discussion is about photography for social use, for exhibition or competition purposes, as part of an intent to sell photographs and images, or as part of photography services, Artificial Intelligence (AI) and its impact on photography is an issue generating significant opinions and comments.

Broadly, Artificial Intelligence (AI) aims to enable machines to perform tasks that typically require human intelligence, such as decision-making, pattern recognition, and natural language processing. In this context, AI has enabled photographers to improve their techniques, enhance their creativity, and streamline their workflow. This allows photographers to push the boundaries and produce unique and captivating images. However, at what point is that ‘photograph’ no longer a photograph?

Arguably there are three key elements to this issue: the source, authorship and post-production. A discussion on these elements may significantly vary if it is a photograph used as a hobby, for entering a competition or exhibition, or to generate income as a professional photographer. There is also an issue of equipment but that is for another article.

What is a Photograph?

The term ‘photograph’ traditionally refers to an image originating from the action of light or other forms of radiant energy upon a photographic medium or device. For example a film or digital camera or digital scanner. AI images are generated by artificial intelligence algorithms based on input data or instructions. The challenge is that language evolves and changes and the term ‘photograph’ now seems to be often used to describe what may otherwise be referred to as an AI generated image.

The second key element to consider is authorship. The author is the person who produced the work and is often also the owner. There will be circumstances where the ownership of a photograph (digital or print) is by another entity and this will often be due to circumstances where a commercial transaction or engagement is enacted.

The third element and perhaps the most contentious, is in post-production. The general use of the term AI in post-processing has been arguably more used as a marketing term than a true reflection of what is ‘AI’ and what it can do. However, recent advancements of software that provides generative fill (or similar) and text prompted actions certainly fits within an AI definition. One of the key issues here is ‘control’. How much does the photographer make decisions about the changes made to their image and at what point does the AI action make decisions on behalf of the author.

For example:

• Removing an object which has been replaced by pixels around the object does change the image; however, the image is still fundamentally that taken by the photographer.
• Compiling several images into one composite image where all elements of the image have been taken by the photographer remains the work of the photographer.
• Using stock images as part of a composite image depends on the context: if that was a competition or exhibition then it is unlikely that the composite image would be meet the rules or conditions.
• If using stock images was to create an image for commercial sale, provided the author purchased the stock images as well as the copyright to use the image in this way, then that would be ok.

However:

• If the image (or part thereof) was generated within a software environment from random elements in the web, then this is in part or whole, an AI generated.
• Whether this AI image is eligible for entry into a competition or is saleable would depend on the legal (and moral) entitlement of the author to use or sell the created image and any entry conditions for a competition or exhibition.

Can we Categorise Photographs

In the book world there are accepted categories such as fiction, non-fiction, crime, history, science fiction etc that provides some insight about the content. This is useful for selecting a book to read or share with others or in judging the authenticity, quality and relevance of the contents. For example judging the historical accuracy of a fiction book based on history is different to judging the accuracy of the book if it is categorised as non-fiction.

Photography does not have these accepted categories so when an image is displayed there is no established protocol to indicate that the image may be from: in-camera export only, a single image with limited post-production, a composite image or AI generated. This transparency is important when the image is being used, for example, in photojournalism as a representation of what occurred, in a competition or exhibition, or if it is being used by a person to generate sales. It is also arguably important in less formal areas (social media) where an AI images may be presented as being ‘real’ rather than as a digital art creation.

The Challenge

AI is having a profound impact on the field of photography. On one hand using AI technology allows photographers to push the boundaries of their creativity and produce innovative and exciting work that can appeal to a broad range of viewers. On the other hand, there is an important issue on how we share or disclose information on our photographs or our images. It is certainly an issue that will continue to excite, concern, and challenge us.

In the interests of transparency and relevance: only 2% of this article is similar to online sources; therefore, I can argue that while the article used some information accessible in the public domain, the production of the article and the presentation of the ideas are overwhelmingly mine.