There’s no question that some aspects of artificial intelligence (AI) offer, in both camera capture and software processing, enhanced creativity that can drive deeper consumer and professional engagement in imaging.
Admittedly, on first blush, it all seems quite impressive. Scene recognition, subject tracking, “smart” selection processes in software and features such as “super resolution” algorithms aim to produce images of superior quality. And foster the creation of never-before-seen photographs and videos.
Who could have imagined technology such as “motion blend retouch” (think Duchamp’s Nude Descending a Staircase captured in one press of the shutter button) would become standard fare. Or cameras would integrate “human pose estimation.” (That’s the ability to distinguish between multiple people and maintain focus on the targeted subject.) While some photographers respond to this with excitement and interest, others are wondering just where it all might lead.
Artificial Intelligence: Promise or Threat?
There is indeed a current rush toward incorporating AI—an acronym that marries artificial with “intelligence”—into image capture and creation. Moreover, commercial photographers, both still and video, in the fashion, architectural, product and even fine arts fields seem increasingly concerned about their future role. As well as, alarmingly, their ability to continue to earn a livelihood.
It seems a whole new discipline required of pros is coming. Most worrisome for some is a potential shift away from years of practice of their trade and the accompanying creative lifestyle that may have attracted them to photography in the first place.
Candidly, proponents of artificial intelligencee take a somewhat Darwinian stance about the future prospects for today’s pros. The attitude is “adapt or fade away.” There will come a day when, at some level, computers and their algorithms and generative imaging programs may at first slowly, then quite rapidly, displace the classic studio and location photo venues of today.
Want to do a fashion shoot on the island of Bimini? Forget about schlepping all that gear, all those techs and models. Just get some stock shots of palm trees and sand, scour the stock photo catalogs for models, then clone the clothes on, set the lights in software, and up pops a two-page spread.
It’s the Newspeak (check Orwell’s 1984) that freaks some people out. It’s the clinical and rationalized phraseology that’s so in vogue these days. “Generative AI” apps are a prime example. They’re dubbed “advanced natural language machine learning algorithms” that interpret verbal cues to create an image. While the first examples were fairly primitive and mundane, a whole new slew of them has upped the game.
AI Image Generators
If you haven’t already done so, you should check them out by googling “AI Image Generators” and take the tour. Some allow you to upload your own image. Others rely on text input alone.
Basically, you type a descriptive phrase into a text box. Try something silly at first like “penguins sipping pina coladas poolside.” Then move on to the meatier stuff like “a man with empty pockets stares at a crescent moon.” Once you do that you may be prompted to choose a style, from old lithographs to modern art to spooky stuff and steampunk.
Many sites also have sensors that act as “woke” censors, so nothing too political or racy gets through. Although, taking the Internet’s proclivity for such subject matter into account, that may not be a problem in the future.
Further, most of the current flock of popular apps are aimed at smartphone users and only allow for quite low-res final products. However, I am aware of scalability research going forward and companies that already promise 4K resolution and upscaled pixel output.
In addition, while amateurs may find this amusing and something to do while waiting for their delayed plane, pros are investigating its potential. However, it is worthy of note that stock libraries like Getty Images, no small operation, are putting the brakes on by not allowing any “generated” images into their stacks. They cite concerns about copyright infringement and other matters that keep the lawyers busy. But that doesn’t mean that commercial photographers and illustrators won’t see a need to, as mentioned above, “adapt.”
Waters Already Tested
In a sense, these content creators have long engaged in artificial intelligence imaging, if you will. They work with Adobe Photoshop, ON1, DxO Photolab and similar advanced image-processing programs. The refinement of masking (selections and isolation of very discrete areas in a scene, like sky behind a pine tree, or modeling lighting in very discrete ways); selective color and tonal control (for saturation and enhancement); as well as the incredible number of presets (“looks” and emulations of toning and a myriad of push-button effect filters) have prepped anyone who works with camera and software processing to deal with coming artificial intelligence offerings.
Cameras = Computers with Lenses
Indeed, the modern camera has evolved into a computer with a lens. This started back with picture styles and scene modes. It has pushed new heights with the specter of the smartphone always lurking in the wings.
Of course, reviewing all the amazing tech we’ve seen in cameras in the last decade would take more pages than allowed here. However, one recent development, so-called IST “intelligent subject tracking,” makes the point.
Rather than relying on the instincts of seasoned pros to capture the peak moment in a sports match, for example, photographers can engage “real-time tracking” on a basketball point guard and bang away for 300 frames at 10 frames per second. Voila! You’ve got the cover shot. What’s more, ISTs can track humans, animals, birds, insects(!) and even planes, trains and automobiles.
The replacement of mechanical with electronic shutters also has resulted in a previously unimaginable parsing of time. Nature photographers can now ensure that the eye of the tiger is always sharp as it roams the forest or that of an eagle as it soars through the sky. And videographers, using all the tools now available in hybrid cameras, gain equal facility.
On one level this is quite exciting stuff. It provides unheard-of benefits in image capture. Additionally, it hands over classic techniques to aids such as eye detection AF, HDR, focus mapping and things such as Google Clips, which with AI recognition of your significant others remembers who you frequently photograph.
It gets kind of odd; you have to wonder where it’s all going. According to a recent press release from one manufacturer: “The subject-detection function is now capable of recognizing insects and drones. Set ‘subject detection setting’ to ‘bird’ when you want to detect insects and to ‘airplane’ when you want to detect drones.”
Art & Craft
All this is developing fast, and the learning curve is steep. The question remains how it might appeal to enthusiast photographers who have other livelihoods but also a love of photography and imaging.
The increasing attraction of “analog” imaging—film and darkroom—may be one manifestation of bucking the trend. These analog advocates see AI in cameras and software as alien to their concept of what photography is all about.
This ain’t your father’s photography folks. It’s digital photography. Like the steamroller that reset the photography world at the turn of the century, I think the AI cat is out of the bag. It will continue to have a growing influence among camera makers, their customers and the print creation end of the business.
There’s much discussion of what all this means for the industry and the photographer. There is a lot of hype going on about AI. Expectations may be met with eagerness or perhaps disappointment. However, one thing is clear: there are many companies busily engaged in developing algorithms and software that will undoubtedly initiate profound changes for everyone involved.
We have already seen artificial intelligence affect all aspects of our society—and this will continue. The worry is that AI tools may allow, to an extent, the replacement of designers, writers, editors and, yes, photographers. Yet, it also offers opportunities for businesses large and small.
It is disruptive, no doubt. How it will affect every level of the creative chain of image making, distribution and publication remains to be seen. It may end up like a new toy under the Christmas tree that’s played with enthusiastically for a while but ultimately ends up in the closet. Or it could open entirely new avenues of creativity and business creation.