Editor's note: The Reframe Column is a series of essays and interviews aimed at examining new media and visual storytelling. For this three-part interview series, Clary Estes sat down with New York-based British editor and business manager Stephen Mayes to discuss the state of photojournalism. For Part I, Mayes argues that the photographic images today are not, at times, photography.
Part II of this interview series will be published tomorrow, July 27th, 2019.
Part I: This is Not a Photo
Stephen Mayes is a New York City-based British editor and business manager working with institutions and individuals to develop effective visual storytelling since 1987. Currently, he is the Executive Director of the Tim Hetherington Trust.
Across twenty-five years, Mayes has directed several organizations and managed the work and careers of top-level photographers and artists in the diverse areas of art, fashion, photojournalism, and commercial photography. As Creative Director of eyestorm.com, he worked with artists such as Damien Hirst, Ed Ruscha, and Richard Misrach. He also represented the archives of luminary photographers, including Robert Mapplethorpe, Steven Meisel, and David LaChapelle during his time with Art + Commerce as Director of the Image Archive. From 1985 to 1994, he worked as Director of Network Photographers in London.
In 1995, Stephen Mayes became part of the founding management team at Getty Images. His work as SVP and Group Creative Director helped launch the company as the world’s most successful image supplier. From 2004 to 2012, he was Secretary to the World Press Photo Awards Competition in Amsterdam. From 2008 to 2013, he was the CEO of VII Agency in New York.
Often described as a “futurist” Stephen has broadcast, taught and written extensively about the ethics and practice of photography and visual communication. With a focus on creative management, including project design, execution, distribution strategies, and business structures, Mayes has been at the forefront of operations for American, Asian and European imaging companies.
Don't Panic. Is image evidence?
The Oxford dictionary defines a photograph as “A picture made using a camera, in which an image is focused on to light-sensitive material and then made visible and permanent by chemical treatment, or stored digitally.” And indeed, given the above definition, Stephen Mayes, a creative photography director, CEO and described ‘futurist’ of the photography industry, can convincingly argue that the photographic images that saturate our visual landscape today are not, at times, photography at all.
As Stephen Mayes explains: “Digital photography [in the mainstream media today] looks much the same as photography as we’ve known it for 170 years and we approach it with similar expectations. However, the indexical medium of photography has been replaced with computational processes that have thrown into question all our assumptions about the image as evidence.” He points out that the level of manipulation that can be achieved in a photographic image could easily fall under the category of science fiction.
For instance, the website, thispersondoesnotexist.com, is a Blade Runner-esque experiment in just how coequal a photograph is from a photographic image, visually. When one visits the site the viewer is immediately presented with an image of a person - one who does not exist, meanwhile, the lower right-hand corner of the frame stating, “Don’t Panic.”
The thispersondoesnotexist.com website was created by LyrnAI, a tech company developing Style-based Generative Adversarial Networks (GANs), which according to their website “synthesize artificial samples, such as images, that are indistinguishable from authentic images.” Different from Computer-generated imagery (CGI), which is the application of computer graphics to create or contribute to images dictated by the input of an outside creator, GAN technology is the computational process of an unsupervised machine deep learning technique that creates new content based on artificial intelligence. In LyrnAI’s Generating and Tuning Realistic Artificial Faces version of GAN technology, upon each refreshing of the webpage, a new “person” appears. At times, these GAN images are easy to recognize as a result of a ‘person’ having too few or too many teeth, inconsistent jewelry choices, equally odd clothing cuts, psychedelic backgrounds, or small pixelated indicators, etc., pointing towards the programs computational roots.
However, the longer the site stays up and the technology improves, the harder it becomes to recognize those criteria with the human eye.
“AI is not designed to follow orders; it is designed to think.
And from here onwards we cannot even imagine what the consequences will be."
The implications of generating images. Are we still in control of the images we use?
While Artificial Intelligence (AI) technologies have a variety of uses outside of photojournalism, within its context, the implications of these generating images have people like Mayes pondering.
Mayes reflects: “My take is that our immediate experience of GAN imagery is crude and expresses the general belief that we are still in control of the imagery we use. People think of these technologies as super-slick expansions of Photoshop, where the artist sets some parameters and the machine obeys. The element that we have yet to come to grips with is that we are now dealing with intelligence."
He continues: "AI is not designed to follow orders; it is designed to think. And from here onwards we cannot even imagine what the consequences will be. Some (for example, Stephen Hawking and others) predict doom, others predict growth. We don’t know [what will happen], but it’s becoming clear that humans won’t be the only force in the driving seat. I’ve heard it said several times now that we don’t understand artificial intelligence any better than we understand human intelligence - ask three psychologists about how we think and you’ll hear four theories or more. We have no idea how machines think but very soon we’ll begin to see the visual expressions of that process, and unlike photography, we won’t have the benefit of shared experience to interpret the image.
At this [point] I can’t anticipate the consequences; all I can say is that the sooner we come to grips with the fact that we’re not in Kansas anymore the better we’ll be able to deal with what’s ahead. Much of the imagery will doubtless look like photography as we once knew it, but it is not at all the same.”
Stephen Mayes has observed how image-based storytelling is organically moving to different technologies and offers an example of how similar creative technologies are already accessible to the media and photojournalism industry.
AI expands into deep fake.
In July of 2017, a similar “deep fake” computational process was used to create a false, yet realistic video of former President Barack Obama giving a presentation on a school shooting that in reality never happened. As Business Insider reported in July of 2017, “Researchers at the University of Washington designed a program that can take a video of someone talking and replace their audio with something else. The result is surprisingly realistic — and completely fake.” In a TED Talk presented on July 25, 2018, computer scientist Supasorn Suwajanakorn shows how, as a grad student, he used AI and 3D modeling to create photorealistic fake videos of people synced to audio. The presentation discusses both the ethical implications and the creative possibilities of this technology as well as highlights the steps being taken to fight against its misuse.
The only certainty is change.
Mayes analyzes these new advancements and their potential effects in his essay titled, “The only certainty is change; Instability as a catalyst for creativity,” published on Viewbook in 2017 and subsequently republished in a book of collected essays called “Transformations”.
He writes: “For 170 years, photographers found roles in commerce, journalism, art and a myriad of niche specialties, yet whatever our market or form of creative expression we all accepted an extreme limitation on our function that restricted our status to that of a supplier… our business was subject to the whims of our clients who determined what they wanted, in what quantity, when and at what price. Whether selling prints to collectors, accepting commercial assignments or traveling internationally for the press our only real control was the word “no”...
Mayes sees an opportunity in the new model of image-making. He perceives a world in which these developing technologies create new industry spaces for image-makers to wrench control from traditional power structures, which have come to dictate every level of distribution and publication, and instead reorient that control inwards by becoming publishers themselves and completely shifting the publishing pyramid to a model that could be more sustainable for the creator.
In Part 2 of this interview, we will discuss Stephen Mayes’ perspective on some of the results of introducing new technological developments for image-making and publishing in the photojournalism industry. We will also reflect on Mayes’ assertion that new image-making protocols open the door of possibilities to new methods of storytelling and subject-creator-audience engagement.
Written by Clary Estes
Editor: Adriana Teresa Letorney
Published July 26, 2019