Go to Squarespace.com for a complimentary trial, and when you're all set to release, go to save 10% off your very first purchase of a website or domain!
AI has definitely been the buzzword in numerous circles recently with services such as ChatGPT. Imaging AI is quickly being established too with the ability to create photographs that do not exist. AI even has the capability to copy photographers – what does this mean for the future of photography? This is definitely a hot subject and I would love to hear your thoughts on this also.
CBS Sunday Morning AI function
Enjoy more Ted Talks
Join my email neighborhood
My Adobe Lightroom and Capture One Presets
AoP T-Shirts
Required a website?
On my channel you will discover videos about photography, cinematography, post processing tutorials for Capture One, Lightroom and Photoshop, image projects that YOU can participate in, the Artist Series and more. The Artist Series is an ongoing set of videos I produce as documentaries on living professional photographers. I am very passionate about photography and video and my objective in making these videos is to share my enthusiasm and enthusiasm with you! Do not forget to subscribe and ensure to hit the like button and share this video if you enjoyed it!
Ted Forbes
The Art of Photography
2830 S. Hulen, Studio 133
Fort Worth, TX 76109
United States of A.
-> Get Paid To Take Photos <-
I dont care if other photographers take better pictures then me, I also don’t care if AI can make better pictures then me. I enjoy the process of making photos for me.
I loved your post. I think the same, I don’t care if other people are better than me. I just love taking pictures. And I love the pictures I take. For me, that’s enough.
I agree, however, I would object strongly if AI software could trawl the web and cloud-based storage such as Adobe CC and use / profit from our photos. We buy the gear, travel to the location, buy / lease the post-production software, etc. To have some AI ‘artist’ type in a few commands and their software gain free access to our photos and get paid for their efforts seems totally unacceptable to me. This seems as much a debate about ethics and copyright as it is about whether AI will do away with photography as we know it.
@Paul Jenkin This is true, but change is inevitable. Technologies put weavers working from home out of business, but we still have people who do it. Some people only want hand made products. There will always be a need for ‘hand made’ photography. What form that will take for everyone I don’t know. I already only use methods lots of people think of as obsolete for my own work, which I don’t put online.
For enthusiasts like you and me, that’s all well and good. The issue Ted is discussing is whether AI advancements will affect pros: those who shoot commercially, art photographers who sell their prints, etc. For them, potential clients who use AI instead of hiring a real photographer or buying a real photograph represent lost income. It could be a big deal down the road for some pros.
@John Drummond they’ll have to adapt, just as they did with digital and when stock photography became a thing. In my youth there were so many pro’s with studios and physical shop fronts in my town it was unfeasible for me to start my own company without being driven out of business. Within 15 years most of those were gone and I’d moved away from thoughts of doing it as a way of making money. AI can’t do weddings or family portraits, some companies will still want to have real photographs, but that market is going to shrink. Whatever caused the event it was bound to happen anyway.
Been a beta tester for AI art tools for over a year. I’m very excited for the future of media and communication aided by AI. Many people will find new forms and ways of expression, enjoyment and contemplation. However, it actually makes me more interested in my own photography.
Photography, like playing an instrument, is an experience. AI cannot replace the total experience. It may offer something new but it is not a complete replacement for the photographic experience. That said, eventually certain types of personal photography and commercial photography may be replaced by AI tools.
Challenge yourself, push the boundaries of your creativity. Or simply do not worry, and just enjoy taking photos.
Thank you for a sane take on AI. Anyone who has worked as a commercial photographer knows that clients want images of specific objects, presented in a specific style, and to have input into the process (that’s what they’re paying for!). AI fails on all counts. In terms of personal work AI has no insights, makes no statements or observations, and adds nothing of any meaning to images. Again it fails. By it’s very nature it is always going to fail at these things.
That said, it can be fun to plug some prog rock lyrics into it and see what you get. I’ve wasted more hours than I’d like to admit doing that.
Despite all this talk about photography being art, I have always believed that, at its core, photography is rooted in documentation. This is not to say that photography is not artistic. But the one thing that no other creative medium can’t do is show what things looked like, what happened at a certain point in time. This is the one thing AI can never do. Of course, it will be able to simulate (in other words: fake) things, but then it is not documentation anymore. And I think this is why photography will always be able to justify its existence, regardless of how good AI will become at creating images. And, of course, everything @Usuallyroamingrob wrote 45 minutes ago.
A tangent issue is the manipulation of the archival image to give intentional misreprentations.
I agree with you, the journey to a photograph in which you plan travel and see the photograph and you have a camera to capture that image that is why I take photographs.
Consider an image created by AI compared to an image taken by a photographer and then drastically edited. Both result in what was wanted not what existed.
Hi Ted. I’m a mostly retired attorney as well as a life-long avid photographer. On the AI question, my first thought is that any commercial use of AI-manipulated art is ripe for legal challenges and that, as you pointed out, AI is necessarily based on all the information fed into the system (including existing, “real” artwork, i.e., copyright protected material). In that sense, it is not dissimilar to the current problem of people posting unauthorized copies of photos on the web. One of my workshop instructors some years back mentioned that his attorneys collect instances of unauthorized use of his work and make an annual sweep, sending demand letters to the wrongful users, resulting in a nice amount of passive income each year. On a related issue, the editorial use of computer generated images is similar to the issues surrounding the early use of digitally manipulated images in news stories. You probably remember that Newsweek published a cover photo from which one person’s image had been removed. Newsweek included a note that the cover was meant as a photo illustration rather than an actual record of what occurred. I think that any AI-generated “news” images should (at least ethically) be similarly noted, which means they would not likely supplant photographers’ real images.
As for the intellectual property matter (and I am not am IP lawyer), it is important to keep in mind that statutes are only one part of the matter. As you noted, the facts of s specific matter, as well as how they are presented (how the case is framed), the jurisdiction in which it is brought, and the previous case law are largely determinative of the outcome. An unsatisfactory result in one instance does not mean that the overarching issue is resolved. Unless the courts somehow decide that copyright protection no longer exists, it is hard to imagine that a company could freely publish AI artwork that was demonstrably based on protected material.
It’s all fascinating, for a technical, gee whiz standpoint as well as the legal ramifications, and, like most of us, I’m routing for the human side of the equation.
The fascination with Cartier-Bresson is not that you couldn’t arrange items in a frame artificially like a painter. Neither is the fascination that AI can cut down the time a painter needed to milliseconds. The fascination is that he managed to find these arrangements in the real world and record them in fractions of a second with (often) interesting subject matter to top it all off. The only problem with AI is that it will become more difficult to tell fact from fiction. Other than that the fascination with real and good photography will remain the same.
It would be really helpful if there was software available to identify whether AI had been used to create the imagery. If it becomes so difficult to distinguish reality from fake, politicians and criminals could use it to pursue their own deviant agendas.
Thank you for the stimulating comments. I’ve drawn a parallel between AI in photography an CGI for movies. I think the comparison is apt. CGI has been around for awhile. Cinema historians will know how long. And even now I can still differentiate between CGI and video imagery. I think.
I was eager to know what you had to say about this and believe I’ve been amply rewarded. Thanks again.
Many years ago when I was a student in Computer Science, I wrote a program that would put words together to form a sentence and sentences to form a paragraph, with amusing results. As time went on and computer graphics improved to the point where they are almost photo realistic it occurred to me that new movies could be made with Humphrey Bogart or Marilyn Monroe. All well and good. Until unscrupulous people use this technology for their own benefit. And yes, I’m thinking about politics. And that’s what worries me.
2:50 Being a professionel photographer for 30+ years I’ve always supported technical evolution. On the other side Via my secondary activity as photography teacher for the past 15 years I saw the evolution of visual content. As we all know there’s a huuuuge load of daily visual content being spread around the globe and between people. It’s sad, but at the same tile stunning to see the younger generation X and Z use more visuals than words to express themselves. The need for photographic collages and montages was increasing immensely… hence the evolution in super easy/quick masking and selecting in all photo and video editing software. For the moment I don’t think the AI generated imaging will be a threat for photography. 10 years ago CGI was the photography killer number 1. But instead of killing photography, it pushed itself out of contention because product photography for instance got back to proper photography instead of CGI thanks to major photographic developments in editing software. What this AI generated imagery will cause is an extended way of expressing ideas and feelings that can be created “on the spot” and shared instantly. It will only benefit photography because of the obvious big quality difference. I could go on for a bit longer, but it’s an excellent item to share especially te subject on copyrights… who owns the AI generated image for instance if it’s created with visuals of other people, artists and visions 🤔🤔🤔 💪🏻💪🏻💪🏻 subject
To this question I would pose this additional question: which famous photographer’s body of work would likely be easier to replicate via AI, Ansel Adams or William Eggleston? Now think about why. What lesson can derived from the answer?
I think that AI is very interesting and impressive technology. That said, I couldn’t be less interested in buying the software or even using the option to change backgrounds and skies in PS / LR, etc. I read somewhere that it’s already having a detrimental impact on commercial photography as it can produce acceptable results, to order, much quicker than commissioning a photographer to shoot it – and reduces the cost as it seldom requires travel and subsistence expenses or hiring actors, etc. However, I am really NOT in favour of AI ‘artists’ or software using photographs or any other form of imagery which they haven’t licensed or purchased. If AI is that great / powerful, get it to create original content. Apparently, Adobe automatically allows images it has has access to to be used by AI. Supposedly, there’s a setting where PS / LR users can opt out but I’ve not been able to track this down – yet. Personally, I’d prefer Adobe (and any others) to ask users if they’re happy to opt in and not assume agreement on our behalf. Not that my photographs are anything more than average – but they are mine and if anyone wants to use them, they should ask for permission and pay a fee. From what I can gather, there have been copyright court cases which have found against the original artist on the basis that the image had been significantly altered by the AI software but I think this is bad law. If the AI ‘artist’ has legitimately bought stock photos, etc. that’s different. Anyway, I’m not against AI but I believe we, as photographers, need to protect our ownership / authorship of our images.
I see AI image-creation technologies as another step in the direction of something I’ve been discussing for the last several years: images as literary works.
The history of photography is dominated by the use of the medium to document real people, places, things, or events. This background is the foundation for much of the angst people feel when they learn an image that everybody thought was “real” was actually Photoshopped or ‘shopped, as the saying goes. We assume a photo is of something or someone real. When we see an image that looks “like” a photo but isn’t of something real, we’re taken aback. We feel misled; cheated.
Of course, documentary photography, photojournalism, and other genres (landscapes, wildlife, portraiture) that document the real are not the only genres of photography. Editorial photography involves the creation of images that tell stories – not about the real people, places, things, or events in the photos – but about the audience consuming the image. Also, there are photographers who work in the medium to create pure works of art. And there are photographers (e.g. Richard Avedon) whose works transcend their intended genre to be seen as works of art.
In recent years as AI image processing and editing tools have become readily available in smartphone camera apps, more and more people use photography as a tool to create fictionalized images of themselves, the people they know, and the places they’ve been. This isn’t done out of a desire to deceive or mislead viewers. It’s done within the use of photography as a literary device. These images are fictional visual literary works.
In literature, there are writers who’ve created great biographical or historical works that tell the stories of real people and events. There are also great writers of fantasy, crime, spy, romance, horror, and other fictional genres. In fact, fictional works are far more popular than biographical or historical ones. Nobody who enjoys reading expects every writer to create only works about actual people, places, things, or events. Why should photography be any different?
AI processing and editing tools make it possible for “authors” to create visual works of “fiction”. Just as a writer who works with the written word may create a character in their own image or in the image of someone they know, photographers who use AI to manipulate selfies are creating fictional visual literary characters based on themselves. There’s nothing devious about this. The intent is not to tell lies. It is simply an application of new technology that extends the reach of one creative medium into the realm of another. Photographic images are manipulated to create visual literary works.
AI image creation is a continuation of this process. It’s not photography and, therefore, cannot replace what only photography can produce. AI image creation will never be able to document real people, places, things, or events with the authenticity of a photograph. A photograph is made by capturing light reflected off a real subject. AI image-making is different. That doesn’t mean AI image creation isn’t a legitimate medium. It just means that it isn’t a documentary medium.
A big part of an image’s potential for impact is its ability to transcend differences in culture and language to have an effect on people all over the world. Up till now, the photos that have been the most impactful have been images documenting actual people, places, things, and events. In the future, images that have a photo-realistic appearance but are of fictional subjects may become more popular and more widely shared than photographs that document the real. Will that mean photography has been replaced? I don’t think so. But just as in writing where fictional works are more popular and more widely read than biographical or historical works, it may mean that – within the emerging market for visual literature – fictional works are more popular than documentary ones.
Ted, love your breakdown of AI technology and its impacts.
In Hawaii, I own a major photographic firm that specializes in beach family photos, weddings, engagements, maternity, and fashion. In compared to Ai, a photographer’s ability to guide and utilize their nice nature to make the customer feel comfortable and calm in order to capture the correct image is incomparable. Unless a robot with all human sensibilities is created that you can no longer tell what’s human and what’s an Ai robot, but having sold that, I believe humans enjoy interacting with other humans, they enjoy the chemistry, energy, and warm personality that welcomes them to the island, and this would be difficult to replace. Also If our customers are aware that they are conversing with an Ai robot, they will not have the same experience, and I doubt they will appreciate the interaction or experience they will have during their picture session. Now, stock photography and advertising photography will be the most impacted of all the categories. My fear about Ai is that future generations will grow so used to it that human skills and abilities will become obsolete. I wish Ai, is supervised and sanctioned, such that there is a legal bar code or watermark on Ai that tells people if it is genuine or not. Humans will either have to step up to the plate and become more creative and quicker at inventing in the future, or they will just give up and become lazy, allowing Ai to totally take over. I recently utilized chat GPT on a blog post and instructed it to use relevant keywords and produce a 3000 word essay. It was created in 5 seconds and got to page one of Bing.com, however Google hasn’t even indexed it yet since their algorithm is so different. So, in conclusion, enjoy your photographs regardless of what Ai does to you. I believe humans will always have a special talent that will never be replaced by Ai; it will come close to taking over, but will fall short due to a lack of feeling, sensibilities, and love; yes, love and chemistry are in everything humans create, and we should be proud of this ability rather than fearing any takeover.
I almost have to wonder if this isn’t a more relevant question for illustrators than for photographers. I think very often what photographers sell pictures of is specific enough that Midjourney or any of the others can’t generate them (like a particular product or person or building) just like is the case with stock images. On the other hand, these AI image generators can fulfill open-ended requirements pretty well. Midjourney might be able to generate logos for your small business where you would’ve hired a graphic designer, but it won’t be able to generate a picture of the façade for the website (I think, tell me if I’m wrong).
A couple of thoughts. Since Ansel Adams, there has been movement away from photography (in camera image) to imaging out of camera. Most of us know Ansel was a photoshop geek far before photoshop ever existed. In one of the biographies about him, Ansel tells the author that he worked on one single print for three days, busy dodging and burning. There is more imaging and photography today that there ever was in the past and it’s only logical that massive changes in technology and procedures are going to occur. And these changes are going to affect some professionals in the marketplace. For example, a wedding photographer really can’t be replaced by AI, however AI might be used to enhance the looks of a groom and bride much like portrait AI is being used now and in the past. But I will invent software. So lets have seniors of high schools and universities take a photo of themselves from the nipple up against a plain background. The photo could be taken on a phone or camera by friend, family, lover, enemy; no need for any professional work. That image can then be forwarded to the imaging department of wherever and input into the AI software. The appropriate cap and gown added, the appropriate background is added and the photograph is good to go. No need for a professional photographer.
As developer and photographer, I constantly strip exif data from images as part of the performance optimization. We all do it. In fact some serves are configured to do this automatically as part of the request response. Case had no merit and photographer is overreaching with his media release.
On the Ai front, I’m super excited about it. Whenever new tech comes out I’m always thinking about how to integrate into what am doing. People that worry are operating on fear. I always thought that photography is personal–in every sense of the word, therefore no machine can replace that connection. Ai could be a crutch you lean on but will not replace photography.
I think AI might be great to create thumbnails and other artificial images like this for YouTube, social media or memes. Art is often connected with the process and the way how it was created. Telling a machine what to do isn’t that much of an impressive process after the first hype is over. Reportage style is also hard to do by telling the machine to create a picture of what has happened. Product or people need the precise image. Travel, landscape or wildlife can’t work. I don’t see a lot of applications.