Tim Tadder is widely revered as a bold and visionary photography icon. With numerous accolades to his name and a client list ranging from Amazon to Nike, he is recognized as a leader in advertising photography and beyond. Since late January, though, he has turned into a highly controversial artist in the industry with his work in AI. I sat down to chat with him about the polarizing pieces he has been creating.
I was ready and excited with a list of questions. I had a cozy introduction planned and some softball questions to start the conversation. But it was not more than a few sentences in that Tadder jumped into what he really wanted to say: he is completely fed up of being attacked for his explorations in AI. So, I went with it and jumped right into the flames.
Attacks From the Art Community
January 23rd, Tadder began posting work created with Midjourney, an AI image generator. Tadder described himself in this way:
[ I am an] established artist who has been a paid creator for 25 years at the top level. I am sharing my experience using this new tool. I am […] exploring the most powerful creation tool ever made.
For him, it’s a way to exercise his creative mind and explore self-expression. Over the decades, he has transitioned from film to digital, to CGI, to Photoshop composites. With the development of a new creative medium, is it any surprise that a photographer heralded for his boundary-pushing work is exploring it? Though many photographers are praising his other-worldly visual explorations, the backlash from the photography community has been merciless, according to Tadder.
Because I’m doing any AI, I’m being labeled, boxed, and attacked because I am being an artist and I am exploring creativity. Just because I am exploring a powerful tool, sharing my knowledge, and sharing my insight from having lived through many mediums, if that is justifiable to hate, to discount, and to completely characterize me as a monster, then I don’t want to be a part of that community.”
He continues: "Being attacked is exhausting, and it happens every single day.” He later adds: “Why are you hating on an artist who is creating?”
The truth is, although I didn’t write any of my thoughts on his posted images, my initial reaction to the AI pieces was the same guttural response as many: “What? That’s cheating!” I was vexed.
The Image-making Process
I first came across his work when we were both featured photographers in the 2021 Sports issue of Lens Magazine. His work was completely unique and jumped off the pages. I loved it instantly, and I have followed his thought-provoking photography since.
After the New Year, though he maintained a recognizable cohesion, his work changed somewhat. I started reading the comments below more carefully. He was posting AI-generated images. As one more beautiful and imaginative piece rolled out after the other, I got increasingly irked: “That’s cheating! How can he do that?” One night, irritated and curious, I sat down and tried Midjourney. I was painstakingly shooting a campaign that week, and after seeing one of his pieces, I thought: “Well, maybe I’ll cheat, just to see what happens.” I suddenly understood his perspective. I must have sat there for two hours. It was not giving me what I wanted. Each AI-generated image gave me parts of what I wanted, but it never got it quite right. I had expected to sit down, type what I wanted, and get it. I had thought similarly to a follower who attacked him on an Instagram post: “It takes absolutely 0 skill to type words into a machine until it spits out what you’ve typed… you’ve run out of talent.”
After my failed trial, I went back and looked at his AI pieces. They were cohesive. The palette, the look of the “models,” the backgrounds, and the airiness. How did he do that? My mind shifted at that moment about his AI work, which led me to pursue this interview.
He described the specificity and laboriousness of his process:
It’s a cohesive choice that I’m directing this tool with my vision and my sensibility. AI didn’t sit down and spit out images for me. I sat down and labored over the images for hours and hours and hours, and then I curated it, and then I fed it back into AI, and then I pushed it further, and then I mixed this image and this image in Photosho and put it back into AI and had it regenerate it. I went through a process.
The concept that making intentional cohesive pieces is as simple as stringing a few letters together on a keyboard and pressing enter is not how the process works in Tadder’s pieces. The pieces require direction, re-direction, and compositing. It is a very intentional and time-consuming creation process.
Looking back, my own outrage came from a combination of being uneducated about the process and, if I was honest, a fear of being obsolete after investing so many years of honing my craft. It seemed that all my learning, practicing, and years of experience were being threatened by a computer that needed nothing more than a few words strung together to out-create me. When confessing this, Tadder replied:
People are worried about being obsolete like you said, so am I! So am I.
A follower on his Instagram added a similar sentiment:
The real reason we people are reacting in such a negative way to AI is because we are scared. People are simply afraid of anything new, especially if they see it as a threat.
Yet another commented: “Resenting a new technology will not halt its progress. - Marshall McLuhan"
AI Is Fundamentally Bad for Commercial Photography and for Humanity
My biggest surprise from our talk was Tadder’s straightforward moral characterization of AI:
AI is fundamentally not a good thing for humanity. It is fundamentally a bad thing for humanity.
He even continued by adding:
“It can replace photography, and it will replace photography in advertising campaigns, in commercial campaigns, in magazine articles.”
I never got to ask the obvious follow-up question, “if it’s fundamentally bad for humanity, why are you using it,” as he quickly followed up:
Me not using it is not going to stop it.
Pandora’s box has been opened, and the technology, he reasons, is now part of our ecosystem. It’s out, regardless of how we feel about it. His general sentiment seemed to be that we might as well get used to it and get on board.
AI images are generated from learned information, not collages.
He also addressed some of the misconceptions about how images are created. Tadder explains that the platform has scanned images and created mathematical representations of them. They turn images into numbers. The technology then generates mathematical algorithms for different characteristics. For the “in the style of” for example, Tadder explains that the software attributes mathematical representation for an artist’s use of color, of lighting, of model choices, casting, and location. It then generates images from scratch in that style. It makes its images from learned information, not from collages. That’s a big misunderstanding.
AI Suits and Copyright
Another point of contention that he touched on is copyright: “If we are going to sue people over style emulation, then every single photographer who ever created is guilty.” Art was built on emulation and expansion. Would there be a Picasso, were there not a Braque? A Monet without a Pissarro? He also pointed out that copyrights are only as practical as their enforceability, recounting many stories of people around the world using and profiting from his work. Tadder believes that, over time, artists will be able to have the copyright to their AI pieces, and artists will be paid to create AI works:
After the anger fades and the hate fades, people are going to be like, ‘There’s actually something to this.’
Closing Thoughts
In closing, I think some questions are worth being asked.
- Why is the art community responding so aggressively towards an artist exploring a new medium and sharing his experience?
- What is it about AI that makes artists so upset? Is it really the copyright, the perceived easiness of the medium, or is it something deeper?
- If we disagree with the concept of AI, are we then morally bound to boycott its use? Or, should we accept it and adapt, as so many other mediums introduced throughout the history of art?
I want to thank Tim for agreeing to take the interview. Initially, he declined the interview due to being burnt out by incessant attacks and criticism. I pointed out that novelty in art has never been accepted. Van Gogh only sold one painting before his permanent abandonment into despair at age 37. Monet’s paintings were so snubbed by the art community during his first two decades that his poverty often led him to starvation, even resorting to eating the fruit in his still-life setups. Digital photography was snubbed by film photographers, then Photoshop was vilified as cheating. Artists historically don’t respond well to change. I suppose something in there made him come back around, and I’m grateful. I applaud him for continuing to make art despite the pushback. Despite the many pundits, there are even more of us who continue to appreciate his bold pioneering. I’ll close with a quote I have on my desk, paraphrased from Warhol:
Make art, and while everyone is deciding whether they like it or not, make more art.
My conversation with Tadder was recorded, with permission, mostly as a reference for the quotes in the article. I am far from adept at video and YouTube, but if you would like to watch the full conversation, you can do so above.
The heart of the debate is Realism vs Idealism.
In the 1800's, Impressionists and Historical painters in France hated each other over the same exact debate. The Impressionists were concerned with Realism and they painted the fleeting moments of actual life. Historical painters depicted scenes from the past that may or may not have been accurate so they were basically creating a fantasy that didn't necessarily correspond to reality. Photography was invented in the same period so it was immediately seen as part of the debate. By the end of the century, the Realists in photography shot "Straight" photography while the Idealists used hand tinting and mixed in painterly techniques to create "Pictorialism." Stieglitz eventually figured this out and famously switched from Pictorialism to Straight photography.
Digital photography has confused matters significantly for photographers. On the one hand, it is dependent on light and optics to give data to a sensor. The light is real and the optics are focusing on real objects. But something strange happens when the original data is manipulated in post because the real images created by light can be altered and manipulated into something that doesn't correspond with reality at all. Digital photography could roughtly be called 50% real plus 50% fake. In fact, some digital cameras capture images where as much as half of the original data is noise in the first place. This leads to a situation where digital photography has one foot in Realism and one foot in Idealism. It is inevitable that this unsteady relationship between two polar opposites would eventually cause a crisis within the medium.
Ai generated images are pure fantasy. There is nothing real about them except the appearance of being real. Ai images are pure forms of Idealism and they're going make the post processing and manipulated aspects of digital photography look pathetic in comparison.
This is my public service message to you:
Photography is Realism and it is what happens before capture. Anybody that thinks they can use post processing in photography is going to lose to Ai because that's like jumping into the water to fight a shark. Idealism is Ai's territory now. This is also a good time to think about film in a new way because maybe analog's direct connection to the object is the best way to compete with Ai's lack of connection to any object at all. This could be a good time to follow Steiglitz's earlier example and switch sides.
Prediction:
Expect more knock-down drag-out debates on the matter. Expect people to hate each other because that's exactly how France was in the 19th century. Also, rainbow stormtroopers are just the beginning. The way that Ai generators claim to be creative and compete with reality is by re-combining reality in ways that don't exist in actuality. Ai artists act predictably in unpredictable ways. We shouldn't be too surprised when an Ai early adaptor sends in rainbow stormtroopers to claim digital photography for the Ai empire.
Good point about film, analog may be the way to prove your photography is "photography" and not digital artwork. Let the flame wars begin!
Amen brother!
Thank you. Perhaps the outrage comes from the proximity of the finished pieces? Clearly, there are no linoleum print makers attacking charcoal drawers on tic tok with agressive remarks. Why is it that photographers attack AI artists? It's art. Its a different art. Its not photography it's AI art. Why the agression? Are artists not free to create with any tool they want? What is the real issue here?
You are absolutely right. Not seeing this is pure obfuscation.
Let's not forget even analog photography can easily by altered in the darkroom. So, unless you want to do an exhibition solely based on negative strips or contact sheets...
People seem to forget (or maybe never knew) that the top commercial photographers delivered chrome to clients during the last decades of the film era. Slides couldn't be manipulated easily and clients saw that as some proof that the photographer knew what he was doing before capture.
Reliance on darkroom work (post processing) was reserved for the lowest level photography like photojournalism, editorial, portrait, events, and photo students etc. There were a few big fashion shooters in NYC that owed their careers to the printers that bailed them out in the darkroom. Also, nobody called themselves a "professional" photographer back then except for amateurs. If somebody said "I'm a professional photographer" then it was a dead giveaway that he went to camera club meetings. Some things never change because post processing in the film era was for the people that needed their asses bailed while most of the people that called themselves professionals were just faking it.
Bailing pros in the dark room(s) is an understatement. Just name it, E6, black and white, even people using VPS, one of the worst emulsions ever in my opinion, for product adverting. What a jungle it was. Drum scanning Kodachrome, a real gamble to separate. Real pros would bring their 8x10 sheets and knew where to go to maximize their use of a good E6 line. Those rarely showed up at labs, a driver would drop off and pick up because they only relied on their own light tables, therefore had little use chatting all day at labs.
Omg VPS...I actually forgot that ever even existed. What a trip down memory lane :-)
You're so right about light tables because I only trust using my own. Every photographer that I respected was the same way about light tables and some of the guys didn't trust labs so they ran their own Jobos for e6.
The problem with JOBOS and small amount mix kits was that E6 required very large bath in order to control the chemicals balanced day after day. But what they were facing was most likely not having local labs that processed enough rolls and sheets to maintain that constant. Dip and dunk used nitrogen for agitation so you never had air mixed in during the process. Also, the racks would move at extremely precise time, but one end of your film would get processed may be 5 seconds more than the other because it would go in first and come up last from the chemical. That time frame was however a constant and tiny compared to hand mix where you empty and refill with the next chemical. You couldn't do test cuts reliably with small tank process for example.
Yeah that's exactly how I remember it. The hardest part is being busy enough to justify mixing a lot of chemistry at one time. I used to mix a big batch and it would last maybe two or two and a half weeks. There is no way to mix small batches of chemistry and keep a consistency IMHO I definitely prefer the dip and dunk at good labs because of the risk of scratching the film when loading it in the drums. I never noticed a difference in processing between the beginning and ending of a film roll, but it could have been there and maybe I just didn't notice it. Also, I never thought about the air vs nitrogen so I didn't know that dip and dunk had that advantage. Nowadays, I shoot 8x10 provia and fedex it to a lab for processing and scanning, so I don't get involved in any of the lab work at all anymore.
That's valid for black and white. For color, not so much. Just dodging or burning from a negative would create color filtration shifts. Photographic paper didn't react like curves in digital. And for chromes, you had few paper options. Printing chromes wasn't that easy, you had to have your color filtration perfectly nailed and that left no margin for much alterations of any sort. Your best bet was to separate on a scanner first but that came later for nearly 100% of photographers.
I'm glad you brought this up because I actually wanted to ask you about printing color negatives. I only did it a little bit of color neg printing and always got color shifts when burning/dodging. But I wasn't sure if there were better ways of doing it that I didn't know about.
I used to have a lab make cibachrome prints of my slides in school but they never looked good because my technique as a student wasn't up to par yet. Within about 1-2 years of graduation everbody was drum scanning and printing was something I didn't think about anymore. It's so sad we can't make cibas today because I'd really like to try it again.
As soon as you take a photo, digital or analogue, it has already been interpreted by the sensor or film stock. We know certain film stocks are prized by photographers for their particular colours (or b&w contrast) and grain. Digital sensors don't reproduce a scene exactly as you saw it either both with exposure and colour. Also lenses will distort the scene too, whether wide angle or telephoto compression, depth of field, for example. Anyone can call post production fake but imo it's the human artists interpretation of the photographs (colour, exposure) that I find really interesting and what I think is the whole point of photography as art - the historical painters as you put it. AI is disrupting photography in a big way but it can't replace human made art in the same way. If (a big if) people start to become disinterested in art (AI will affect all forms of art) and purchasing human made art, that's when it will look most under threat.
Excellent thoughts here! I liked your point about how cameras themselves are interpretations of what the eyes see. Your comment "Ai is disrupting photography in a big way" made me wonder- is the reason which the outrage has been so noteable because the finished pieces of AI look so similar to the finished pieces of photography? I made a pivot in my own mind, "this is not photography- there is no camera. This is Ai art"
I have followed Tim Tadders on IG for a long time now, and his AI images are some of my favorite images out of his entire body of work. They're beautiful. I never attacked him nor any other AI artist. In fact I follow quite a few. That said, I just wanted to raise a point here about the realism vs idealism debate: I'm one of those "captured in camera" or "practical effects" type of photographer. For us it isn't about realism. But more about the process of capture or creation. We aren't trying to represent reality perfectly. It's just about the joy and technical achievement that comes out of trying to do everything in camera, using practical effects and optical illusions, doing everything through photographic and optical processes only, and not photoshop montages or AI. I've put real models on top of cars in rush hour traffic for a photo shoot, because that was my process and I had fun doing that. I wouldn't do it any other way, even though a montage would have been 10x easier and more simple to do.
I agree and have a similar approach in my photography to you. I use a 40mm lens which is close to what the eye sees but even that lens has it's own qualities that influence the way the photos turn out. I try to keep the software post processing minimal, mainly just playing with the exposure and black and white conversion. I don't crop or clone out objects but like an 'it is what it is and that's what I saw' approach when going through photos I've taken.
I agree with your point here Raphael Vieira. Although Tadder's AI work is some of my favorite work of his as well, I don't care to create art using AI work myself. I like using a camera more than a computer. That doesn't make the work not valid, or "cheating". As you said, you didn't attack him just because it's not the kind of art you feel drawn to making yourself. What do you think is at the root of photographers attacking his artwork?
That's a good point about different film stocks and thow they already interpret according to different color palettes and tonal curves. Lenses have different levels of contrast and some lean to warmer or cooler color balances. It's true that interpretation has always been happening in both film and digital and that post processing is just an interpretation too.
However, the difference between Realism and Idealism is not about interpretation but rather about the "object" of interpretation. In Realism, there is always an actual object that exists in the physical world that is to be interpreted as an image. In the case of Idealism, there is an imaginary object that is being interpreted as an image. Film photography is always connected to an actual object even if that object is only light. Ai imagery has no connection to an actual object and relies purely on imagination. Digital photography sometimes has a connection and sometimes doesn't have a connection to an actual object and that's why I see it as being the most threatened by the other two mediums.
We should also be careful when comparing human made to non-human made art because painters used to claim that photography wasn't an art since a machine (the camera) did all of the work.
This is a very intelligent, cohesive, historical comment. Thank you for taking the time to compose it. Your thought your seperation of realism vs idealism was especially insightful and exactly right. Here is a follow up question, why are some, as you call it, realism artists aggressive and attacking towards idealism artists? It's a different form of art. Does the ceramist attack the photographer for using a camera instead of a wheel? Of course not. Are artists not inherentely allowed to make art with any tool that they want? Since you clearly are a smart man, I'm curious what your evalution is on *why* many photographers have responded with such aggressiveness towards this.
Thanks Michelle, you wrote a great article and people are really responding well to it.
"...why are some ..... realism artists aggressive and attacking towards idealism artists? It's a different form of art."
You're correct that it is a different form of art. I think that photographers get aggressive when the boundaries are blurred between the mediums. Clement Greenburg, the famous art critic, coined the phrase "medium specificity" to deal with this complex issue. His idea is that the qualities associated with certain art forms are directly related to the mediums that are used to create those art forms.
For example, a sculptor might use stone and a chisel to create a 3 dimensional image while a painter may use a brush and canvas to create a 2 dimensional image. The painter and the sculptor may both create images of the Virgin Mary, but the sculptor's would exist in relief and the painter's would be on a flat surface. The basic difference between 2D and 3D are the direct result of the difference between the mediums of painting and sculpture itself. Therefore, part of the medium specificity of painting is that it is always 2D. Likewise, part of the medium specificity of sculpture is that it is always 3D.
The example I used is a simple one but things get really complicated when new mediums are introduced that can emulate a previous medium. When photography was first invented, it was immediately associated with painting and was actually named "drawing with light." This caused a crisis within painting itself because the painters had to compete with photography in a way that photography couldn't imitate painting. One of the solutions they eventually came up with was Impressionism. All successive movements in painting, from Cubism to Fauvism to Abstract Expressionism, shared the notion of claiming a territory for painting that photography couldn't easily compete against.
The problem that photographers have with Ai is that it can emulate photography and this is definitely going to cause a crisis within the medium of photography itself. Photographers are now in the same position with Ai that painters were in when photography began. The only solution is to start thinking about what is part of the medium specificity of photography that is different from Ai. Then, photographers can cede territory to Ai while still maintaining a realm of their own. This is not an easy process and I think that the photography of today is going to look drastically different from the photography of tomorrow.
What a well thought out and clearly articulated comment. Thanks so much for that insight! You've shown me a different way of thinking about this issue, and it helps me make sense of it all.
Thanks Tom :-)
One thing AI cannot do is replace real life events such as a sports game or more specifically a wedding. You cannot nor would want AI to virtually reproduce photos of your big day. AI will drastically change art in many ways but not completely replace it. The biggest threat is to commercial photography such as advertising and we are already seeing companies recreate their products virtually rather than hire a photographer to take photos of the real thing.
That's been happening with CGI for quite a few years, however it's not as simple as it seem. Small manufacturers for example can't afford having all their products turned to CGI. There is a time issue where CGI has stages of meetings, creation, approval and reviews and possible corrections that complicate and slow down the process and that cost has to be discussed upfront in contracts so every one knows what the limit of their cost to produce or to have produced will be. If there are no real need for a full 360 in regard to place an item in any scene, then it most likely can be a waist of money and time. Even with the best tools, trying to make many products looking real can be challenging. Try to have an artist in Eastern Europe recreate wood type and wood grain without a good sample for example. Next you have to ship 40-100lbs by air of fabrics, metal finish plastics, wood, paper... over seas. It can be expensive and slow and the cost depends entirely on actual needs. Not something to get involved with without proper understanding of the technology.
Yes it's limited to bigger companies like IKEA who can afford it. Commercial photography, on the whole will switch to AI generated imagery when they can save money and time from hiring a photographer. Think stock photography for one example and that is already very difficult for photographers to make decent money from anyway.
I agree but did you see the charity campaign that used AI. Give money to save - fake kids?
It's a fascinating interview. Any work of art that has evoked an emotional response, whether positive or negative, has worked as art. That's its point. People were similarly outraged by Marcel Duchamp's Fountain, and it was that anger that proved it was art.
I've worked in an industry where I got replaced by technology, and it will probably happen again with a portion of my photography work as AI takes it away. For the businesses selling products, whether the image is created by a human or a robot is irrelevant. However, that will not stop humans from wanting to create and nor will it stop intelligent viewers from seeking the meaning in the art that others have created, meaning that isn't there in solely machine-generated images. Tim Tadder's work is not only created by AI, but he is using it as a tool, so there is still human meaning there.
Nevertheless, there is the copyright issue of whether the AI is reproducing elements of other people's work. Parallels can be drawn to musicians sampling other people's music without permission. That's down to the lawyers and judges to argue about.
Great feedback as always Ivor Rackham . I loved your reference to Duchamp. Very relevant. I agree completely that legal has not been able to keep up with the incredibly rapidly developing technology. There are alot of great questions surrounding these issues and it will be interesting to see how the legal side of things unfold.
My objection to AI is based on my concept of what it is and how images are created.
I have an idea of what AI is, but I am not sure that my idea is correct. Here is what I think the AI process is, when it comes to creating images:
A person puts keywords into a program or app, and the program then creates images based on the words that are given to it and the myriad of millions of images that it has access to. The person puts keywords in, hits "enter", then sees what images the program produces. If they want something about the images to change, they change a keyword or two and then hit "enter" again and see what comes out. They do this over and over until the program gives them something they really like.
Is that how AI works? If so, then I have a problem with saying that someone "creates" AI images, because it is the software creating the images, not the human. One doesn't need to be a photographer at all to "create" AI images, because the images are based on the images that other people have already taken. So someone could be an "AI photographer" without ever taking a photo. So if that is the case then "photographer" is a complete misnomer.
I realize that my conception of AI may be very wrong. If it is, and if the person actually does create the images (as opposed to a program creating the images), then my apologies to those who I may have offended with my disdain.
I should note that I have no problem whatsoever with people using AI to produce images from photos. The thing I have a problem with are some of the words that are used, because they may not be correct and may be being used in a way that is not consistent with their literal definition. For example, calling someone a "photographer" if they are not taking photos. Photographer means someone who takes photos, not someone who works with photos. Another example is saying that someone "created" an AI image - if the program created it, then we should not say that the person created it. Typing keywords and hitting "enter" is not creating.
Semantics really are everything - as long as all of the terminology used is correct and literal, then have at it! I would love to see more stunning AI creations, as long as the language used to describe the process is worded correctly.
Great thoughts here Tom Reichner . Although the description of the process is a little bit elementary and in complex pieces such as Tadder's there is compositing and other techiniques- I think your note on the semantics is insightful. Should it be called photography if it makes no use of the camera? We don't call digital art or CGI photography no matter how realsitic it looks. We call it what it is. CGI. Digital art. Perharps we need a new lexicon for the new form of art. AI art. Period. Drop the association with photography, since as you noted, it is not using cameras. What do you think of this?
I wholeheartedly agree with not using "photography" in the name or description of AI generated images, other than to note that true photos are used as input for the program to learn from.
I suppose that to me, whether AI images are art or not art depends on how much a person controls and directs the process. No, more than that, how much actual creating the person does. If the program is doing 99% of the actual creating, then I can't think of the person directing the process as an artist. In such cases, the program itself is the artist, and the person is more or less a data entry tool.
But regardless of what semantics are ultimately decided upon, the resultant imagery is absolutely stunning! I love the works that the program has produced for Mr. Tadder. Whether or not he is actually doing the creating or not, his artistic vision and direction are superlative!
Great feedback here Tom Reichberg. It seems logical to me too that the level of creating that the artist does would seem to play into the equation. Thanks as always for your input.
Yeah, it sounds more like they should be called an AI image director instead!
ohhhhh I like that title alot. Good input
Then somebody who takes a lot of photos to get what they want is not a photographer?
Or is the RAW converter or development bath creating the photo...?
If you use anything automated to shoot (proximity trap, laser barrier, an assistant), are you not the photographer or creator?
Anyways, what you are describing is the same as snapshot vs photograph.
Intent matters, not the physical technique.
I have no idea how you arrived at the thing you say in the first line of your comment. No one said or implied any such thing.
If someone is taking the photos that are used in the final creation, then the person is most definitely a photographer.
If someone just uses AI and the images that the AI program has in its enormous database (photos that others have taken), then the person is not a photographer. They can have fantastic artistic vision and direction, and be excellent at curating what the AI gives them, but a photographer they are not, if they are not taking any of the photos used in the AI creations.
You implied it because you focussed so much on the triviality of hitting enter. Also because you deny people being the creator - "Typing keywords and hitting "enter" is not creating."
What if you add your teach your own concepts/styles to existing AI models? Using your own photos? Or licensed ones?
Also technically, the AI doesn't have a database of photos. The model is only a few gigabytes big.
You are mistaken as to what I meant.
Typing keywords and hitting enter is NOT creating.
Adding ones' own photos is absolutely creating.
I never said or implied that someone taking their own photos is not a creator. In my previous comment, the one you replied to, I actually said directly that someone taking their own photos is a creator, so how in the world can you possibly accuse me of having the opposite stance?
It would behoove you to read more carefully and more thoroughly before commenting.
You think that I implied something that I didn't imply because you took something I said and expanded it to mean a little more than what I actually said. Please go only by my actual words, and do not expand what I say to mean any more than I actually said.
If hitting enter is not creating, why is clicking a button on a camera?
Taking a photo entails much more than clicking a button. One first decides what to shoot, then they decide how much context to include in the frame, then they decide what angle to shoot from, and what distance to shoot from. They move around and kneel down and stand tippy toe and shift from left to right to get everything in the frame to align a certain way. They determine what type of light they want on the scene and what direction to shoot from, relative to the light source(s). They determine what exposure to use to ensure that it correlates with their artistic vision and post processing parameters. Doing all those things with a camera to produce an image is creating.
Hitting enter does not create an image. It prompts a computer program to create an image.
If one uses AI to generate images, and then puts the AI generated images into an editing program like Photoshop, and intentionally and purposefully manipulates the images, that is creating.
There is a great difference between making everything look the way one wants it to look and merely waiting to see what a computer program does with fodder that it has garnered from others.
Then your whole first paragraph is just a strawman based on a pretty superficial understanding. Just like somebody who would say "taking photos is just clicking the shutter".
You can intentionally and purposefully create the image before and after """hitting enter""" with AI as well.
But these semantics don't matter anyway. Digital photography will go the way of analog photography.
My apology for getting involved again, but what does AI create without actual photography input? Is AI going to make up sports photography, events, someone's new born baby, the photo of a new product that it does not have any data or model of? Are people going to be happy to receive a product that does not look like what they ordered. Shape, size, color, depth... How is AI ever going to make this up? I would love to understand why people see photography in such a doomed way.
I am totally lost with Christopher Giblin's thumb down. It says nothing.
You are right, documentary photography will still exist.
For product/fashion I imagine it will be a multiplier or enhancer. Maybe you don't need to build or shoot the complete set, but just 30% and have AI generate the rest. Just like in filmmaking, where green screens or OLED screens take over the actual stagecraft. You don't build the set, just the floor, put some props down and 3D render the rest on OLED screens.
Artistic photography - as tim tadder does - or even landscape might be 80% AI in the future. Landscapes are heavily photoshopped/composited anyway. Why not take like 5 photos of the mountain and outpoint or img2img the rest.
David, it's hard to argue with you and I anticipated that in my original post about how digital photography has one foot in Realism and one foot in Idealism. As soon as a good argument is made opposing Ai it backfires against digital photography because digital photography itself shares so many qualities with Ai. Benoit and Tom are 100% correct IMHO but they're arguing from a position of Realism (human vs non-human) and digital is only half Realism so it's a "catch 22." When a medium involves computation it becomes difficult to prove authorship.
You predicted that digital photography will go the way of analog photography. I actually think that Ai just made analog photography relevant again while changing the role of digital photography. I just looked at the front page of Petapixel and it's obvious that Ai has taken over. In my opinion, digital photography's new role is as a handmaiden to Ai by being a data entry point. Digital photography feeds reality to Ai as data so that Ai can process it into virtual reality. Suddenly, analog photography just became the best way to fight Ai and it's the perfect nemesis because it doesn't need computation at all.
Film photography = real photography = light
Digital photography = light transformed into data for computation
Ai = fake photography = computation
"Also technically, the AI doesn't have a database of photos. The model is only a few gigabytes"
It's a pass through but the images it uses are extensively broken down to the pixel which is an actual use of the image. AI can make models in a way we, as humans cannot do with our brains, however, those models are not born from nothing. The concept for these AI models comes from humans. Can you touch 1+1? No but you know it does something and you only know it because humans had to figure out a way to express it in a way they could take advantage of.
Contrary to the human "emulation" where the brain get stimulated by what it sees, AI has to actually "physically" possess the data in order to simulate emulation in the form of an output. People who are for and people who are not interest in AI imagery will use nuances to their advantage obviously. The way I look at it, emulation in an AI image is only possible through the human request when the person attempt to get the data to react in a way that person imagines. In the video, they talk about how many steps and thinking is necessary to get there as in an extensive amount of work. In a way they demonstrate that AI does not actually emulate but 100% process from data.
But no matter what, AI doesn't know what a photo is, it only understands data. So even if the model has filtered out any data it doesn't need, in regard to input and the model, photo or data is absolutely the same. It's pretty simple, if we talk specifically about data then AI cannot output an image, just data. The same way AI input can be cleverly called data so one could pretend not taking someone else's photo. Luckily things don't work that way.
Will you share the WHY on that. Curious about the reasoning