With the increasing buzz around AI, many creatives have expressed concerns about its potential impact on their professions, writes Sydney photographer Robert Edwards.
As a commercial photographer, I have been using AI in various capacities for years, and now the spotlight is on Generative AI like ChatGPT and DALL-E from OpenAI. These tools claim to enable anyone to become proficient in writing or visual arts with just a few prompts. Well, supposedly.
Creatives feel threatened that their livelihoods will be replaced by artificial intelligence. Generative AI is trained on content published on the internet by thousands of creatives, much of it copyrighted. At time of publishing this article Hollywood writers are on strike and part of their dispute includes studios potentially using Generative AI to replace them. These AI generated scripts will be based on work that the writers previously created.
Visual creatives, including photographers, are worried Generative AI is making images based on their work and that it’s being used commercially. While the latter is most certainly happening, it’s not clear if your work is being appropriated. Unless Generative AI accidentally includes your watermark as Getty Images discovered.
While commercial copyright is important for business so too are our moral rights. Creatives have legal rights to be financially compensated and credited for their work.
When I’ve tried using Generative AI to extend images or add backgrounds the results have a painterly effect and don’t blend too well with my original images. Now that’s partly due to my inexperience with using it and the AI failing. The more I use AI the better I’m getting at providing the right examples and prompts.
I have far better success using AI within Photoshop to create masks. I can create masks to separate skin tone for retouching hundreds of images quickly. Masking was time consuming but it’s just the beginning of the process. Like any retouching I prefer being subtle so the viewer isn’t aware of it.
For many years my use of AI has been more Machine Learning, as used by hardware and software manufacturers. How my camera analyses light and what it chooses to autofocus on is based thousands of images. To this I add some human intelligence to know when to override the settings to capture what my client wants.
Generative AI will disrupt professional photography but it’s not like we haven’t been disrupted before. The internet changed how we interact with clients. Digital photography opened the profession to more people. Generative AI will be a tool we use to create images and do business. It’s not something I fear but rather embrace, sensibly, to future-proof my career.
– Robert Edwards is a Sydney-based commercial photographer who has been in the business for over two decades. So long that he secured the enviable website domain, Photographer.com.au. This article originally appeared here.
My takeaway from this piece is that soon the “photos” of creators such as the writer will not actually be photos but computer created digital art. No thanks.
Gil that’s probably why I’m a professional photographer and not a writer!
My point was that AI isn’t going to be the death of photography. Just as digital didn’t kill off photography. Professionals do need to keep an eye on image rights while using the AI for the betterment of their art and livelihood.
Hi Robert,
I reckon you made your point pretty clear. It remains to be seen how Generative AI will be useful for photographers, but I can’t see how it will replace them.
One area in which photographers are already being replaced by smart machines is architecture. It is now possible to generate high quality internal and external images of a new building. Before it is built! Based on data from the architects and engineers design work.
Hi Peter. Unfortunately the same happened with car advertising. Some building contractors like electrical, painting, stone masons, etc, need images that feature their work. Others want lifestyle however CAD + AI may eventually get humanoids right.