Berlin-based artist Boris Eldagsen made headlines when he won—then refused—the Sony World Photography Awards after revealing that his winning image was AI-generated. In this special episode of Beyond the Prompt, Eldagsen takes us through his transition from traditional photography to what he calls promptography—a new category for AI-generated images. He explores the fundamental differences between photography and AI-generated art, why clear definitions matter in creative competitions, and how AI unlocks new dimensions of artistic freedom. The conversation dives deep into the philosophical and technical aspects of AI-generated images, the role of surprise in the creative process, and what the future holds for photography in the age of AI. If you've ever wondered where art and AI collide—or how to embrace the unexpected in your creative work—this episode is for you.
Berlin-based artist Boris Eldagsen made headlines when he won—then refused—the Sony World Photography Awards after revealing that his winning image was AI-generated. In this special episode of Beyond the Prompt, Eldagsen takes us through his transition from traditional photography to what he calls promptography—a new category for AI-generated images.
He explores the fundamental differences between photography and AI-generated art, why clear definitions matter in creative competitions, and how AI unlocks new dimensions of artistic freedom. The conversation dives deep into the philosophical and technical aspects of AI-generated images, the role of surprise in the creative process, and what the future holds for photography in the age of AI.
If you've ever wondered where art and AI collide—or how to embrace the unexpected in your creative work—this episode is for you.
Key Takeaways:
Website: boris eldagsen | photo & video art berlin
Instagram: Photomedia & AI Artist (@boriseldagsen)
00:00 Introduction to the Special Episode
00:41 Meet Boris Eldagsen: The AI Photographer
01:07 The Sony World Photography Awards Controversy
05:20 Defining Photography in the Age of AI
07:39 The Art of Promptography
12:10 Boris's Creative Process and Workflow
19:30 The Future of Photography and AI
30:53 Tips for Aspiring Promptographers
38:51 Exploring AI Tools and Techniques
41:29 Conclusion and Reflections
📜 Read the transcript for this episode: Transcript of BONUS EPISODE: Boris Eldagsen: Why AI-Generated Images Aren’t Photography |
[00:00:00] Jeremy Utley: Welcome to this very special episode of Beyond the Prompt. This is a bonus episode where we feature an amazing artist who is using AI to transform his photography practice. If you're a regular listener, hopefully you enjoy this bonus episode. If you're not a regular listener, please know that our typical show features more business application and less.
artistic expression, but this conversation was so spectacular. And this leader is such an amazing person that we couldn't withhold the conversation from our audience. Hope you enjoy and we'll be back next week with another normally scheduled episode of Beyond the Prompt.
[00:00:41] Boris Eldagsen: My name is Boris Eldagsen. I'm a Berlin based artist working with photography. Video and installation and, , since two years, I'm working only with AI image generators. And last year I became famous for refusing the Sony World Photo Awards because I applied with an AI generated image.
[00:01:07] Jeremy Utley: Boris, I thought just to, give our listeners context, if we could almost work backwards through your history a bit, and perhaps for those who aren't familiar with your work, if we could start with the moment you refused an award, what was the award you refused and what were the grounds by which you refused it?
[00:01:26] Boris Eldagsen: Well, it was the Sony World Photography Awards, and I was selected as the winner in the open creative category. And my decision grew over two months. I applied for the competition to make a test. To find out if photo competitions are aware of the new situations that people are going to hand in AI generated images. And, um, I had no plan B. I just applied. And, , I was surprised to see that one and the same image, the electrician, made it to the finalists. In three different competitions, and the Sony World Photo Awards then decided to have me as a winner when they told me their decision, I immediately replied, um, making it transparent that The image is not photography, that it was AI generated, and I offered either to disqualify me or if they would like to go ahead to have an open conversation about the problems that the photo community is going to, um, face.
And they have not been interested in doing one or the other. And many of my questions and attempts to have an open conversation have not been replied to. Um, when they sent out a press release about the winners, it was not transparent that my image was not photography. Inquiries by the press has been left unanswered. So, um, Yeah, at some point, , when they promised me to have a Q& A before the award ceremony and didn't keep that promise, um, I had no other option than to refuse the award because I didn't apply to win anything. I would, I wanted to initiate a conversation and I hoped to spark off that conversation with my refusal.
[00:03:23] Jeremy Utley: And so. Why do you think they are failing to engage and why is it so important to you that they engage? Would you talk about that? I know that your background is in philosophy and art. So can you talk to us a little bit about why you think it's so important and perhaps why they're refusing?
[00:03:43] Boris Eldagsen: Well, I think, , you need to take into account that the competition is a business. It's run by a London based agency that specializes in those awards. Uh, they run, , several awards in different disciplines. And for Sony, it's a marketing campaign. I don't think that Sony was informed. , and I think for the agency, it was just something they couldn't bother about because their focus was to have a, yeah, proper, uh, award ceremony and to keep the Japanese guests happy. Um, and anything else? I don't know. Yeah, I still don't understand why they didn't, uh, use the opportunity for good because I told him openly in an email that they can be a first mover. They can be the first big competition to acknowledge the problem and to talk about it. And they didn't. Um,
[00:04:43] Henrik Werdelin: , how would you, I mean, I very much agree with like the whole premise of it and obviously have many questions that arise also. But maybe we can go through some of the questions that you were hoping they were asked or some of the questions that you were hoping would be discussed as part of this , and then maybe we can discuss some of them and then after that I would love to be a little bit more tactical and hear a little bit more about how you work with, uh, You know, with AI now as part of, , your journey of creating art., But if we start with the questions, what do you think the questions are that we are not asking, , in this space?
[00:05:20] Boris Eldagsen: Well, for photo competitions, it's very important. , to define what is photography and what is not. And, , you have two teams. One team is thinking from the end, saying it looks like photography. Why can't we say AI generated images are the next step in the development of photography? And then you have the team that I belong to that say, no, photography is generating images with light. It's, , uh, making images with light, and you go out into the world. You are present. You experience something with whatever is in front of your camera.
And if you generate images with AI, you don't need that. You can sit in a dark cellar, you need the internet, your machine. Um, a laptop, well, that's it. So there is a big difference. It doesn't matter if it looks the same. And, , for this, I think competitions should, , not put those two categories of images in one basket, they should open up a new category for generated images and, uh, many competitions do so.
[00:06:30] Henrik Werdelin: So when you have, just so this is not. To counter, because it's not a space that I understand a lot, a lot about, um, but when you take a high end camera of today with the lenses that you can have and stabilization that are built into them, doesn't that become very complicated? To define photography as you physically have to be standing in front of the event. Um, when that can be so many things.
[00:06:59] Boris Eldagsen: I think there is a lot of, uh, technology and also AI inbuilt in the camera as we use, especially in smartphones. , but still it's lens-based. Uh, there is light being reflected from something that generates the image. , it's a different workflow. And it doesn't matter, , if the result looks like photography or not.
It's also a very photo centric, , way of thinking that we only talk about photography. You can generate images with AI that look like drawing, like painting, 3D animations, anything. So, , this is why I suggested we need a new terminology for this kind of images. And the one that I found and that convinced me personally is promptography.
[00:07:45] Henrik Werdelin: Because it shows that the image was generated with a prompt.
It's just a, such a thoughtful kind of way of thinking about it. I guess the promptography has a little bit of the, I would imagine that people who see their work as being art, But being generated through algorithms would feel that the word itself has a little bit of a gimmicky kind of tone to it and so that there is value judgments in already being prescribed into that kind of way of describing it.
[00:08:23] Henrik Werdelin: You understand that statement?
[00:08:25] Boris Eldagsen: Yes, but how do you understand art? I think that is a question that is very interesting and then, , many people, um, confuse how you work with an AI image generator. , they think that, an artist can be replaced. totally by the machine. For me, this is a very poor understanding of art, and it's also, , not seeing how complex it actually is, and that, , if you have an experience in image making, if You have an experience and a knowledge of the platforms and you know the strengths and weaknesses Then you can use it as a tool Yeah, it becomes like a brush like a camera and it offers new possibilities for me art is Something that should give you an impulse for an inner journey.
Yeah, that you, , not ask yourself, what does the artist want to tell me? Yeah, I think that is a very boring question. And if I want to tell an audience something, I would write an article in the paper. It's very patronizing to think so. Um, an artwork should give an impulse to an inner journey. Yeah, it's more interesting to ask yourself, Um, what kind of memories, emotions, thoughts are triggered by the artwork? , do I like it? Am I attracted? Am I appalled? Why? And that is for me, , what art should do, but that's my personal understanding. And if you go to the philosophy of arts, you have many contradicting definitions.
[00:10:08] Henrik Werdelin: I very much agree with that. , where my head goes though is. , , I went to a conference where some of the artists that are creating what I think is the latest in the NFT space on the crypto spaces, which is this algorithms that creates art, right?
Where the, beauty of the piece is that an algorithm was defined for by something and that that generated imagery. And so the imagery is really something that is. representative of that thought, right? So it's, a system that you get fascinated by rather than , the imagery. , as you look at photography and you look at like the purity of the object having been exposed to light, is that only in terms , of photography, or do you find that digital art, as it were, is something different, and so therefore we need to kind of create all these different categories that defined the medium in which the artist used to express themselves?
[00:11:08] Boris Eldagsen: Well, I think, , AI is a part of digital art, and, , it's another way that we haven't used before. ,
[00:11:18] Jeremy Utley: you know, one thing that I've been thinking as you've been speaking is, on the one hand, , folks can say, well, it's just definitions, but on the other hand Definitions really matter, right? And so, for example, you would not submit a photo into a painting competition. It's it's like, you know, as they say, it's like bringing a gun to a sword fight or something, right?
It's, just a different medium. And for, and I don't mean, you know, more powerful, less powerful. It's different. You also wouldn't bring a painting to a photography competition. Right. And yet all of them are images, right? As long as we're talking about the output being an image without regard to the manner of generation, then a sketch and an oil painting and a photograph and cryptography or promptography or whatever, it could all do it.
But insofar as the competition is actually about the means of production, I think the, definitions matter tremendously. And that's really what you're getting at. I, would love. For you to maybe now speak a little bit about your experience in your career as a photographer. And then I know because you've been immersed in Promptography, if you wouldn't mind to first tell us about Your practice and process with photography and then compare and contrast. What is how is promptography similar or different?
[00:12:36] Boris Eldagsen: When you photograph, um, you are dealing with matter in front of the camera, and I had two different ways. One was going out at night,, and then looking out for light sources. And when I found a situation that was like a perfect stage, then I was waiting for something to happen.
And life always surprised me. And, uh, the images later, they looked like staged, but they had strictly been , , photographed in a street photography manner without altering, uh, what was happening in front of my camera. And I also did staged photography where I, had my models, um, Mainly then, , in house in the studio situation because in winter in Berlin, it's like too cold to go outside for longer.
, and those two ways of photographing, , I mixed in the past and working with people in stage photography, you are a director like in a theater play, you need to decide about the background, the light, the props, the pose, anything, , the model and, and what you want to express. And that was a very good preparation for me to work with AI image generators.
What I like about AI is that now I work not with matter, I work with my knowledge and experience. Because, um, I know how light works. I know what kind of key words to use to describe a composition, a foreground, a middle ground, a background, to describe color, to describe light, to use references. You can also use, , Uh, keywords that describe a camera, a specific camera, lens, film type.
And the AI is going to generate an image that looks like you have used a specific kind of camera and film. So, that is the material I work with, my experience. But at the same time, , I can generate images purely out of my imagination. And that is Yeah, my limit. Anything I can imagine, I can kind of generate, and I'm not dependent on, uh, the light outside, the temperature, uh, the technology I have, the location, the people.
For me, it's artistic freedom AI image generators, and I use them in a way that I'm not, um, Doing, uh, the same thing, the same kind of images I did in the past. I want to do something that I have not been able to do as a photographer in, 30 years. I would like to go beyond. And that is very exciting because the features, , still are released sometimes on a weekly basis.
There are no handbooks. There is a lot of experimentation possible.
[00:15:39] Jeremy Utley: You know, one thing that I love about what you're saying, Boris, is , you're only limited by your imagination and you're limited by your experience, right? And when you talk about things like the light, the lens, the film, et cetera, , on the one hand, someone could criticize. I remember, you know, in the early days of photography, people criticized the photograph because it's just pushing it a shutter, right? And yet when you realize, no, there's angle and light and lens and, temperature and, paper and film, all that stuff, right, has a tremendous impact. And it's only by experiencing them by changing your shutter speed or the aperture or focal length or whatever, right, that you start to gain an appreciation for composition and how a photograph is created that.
You start to see it's actually not dissimilar from painting. There's different brushstrokes and different colors and different ways of mixing, et cetera. The thing that's interesting is how would you recommend or the thing that I find myself wondering is . I am an amateur photographer, and I've taken some introductory photography classes, and those classes have taught me how to appreciate things like shutter speed, focal length, film, type, lens type etc How do you recommend someone start to learn the basics of Promptography? Because right now , what you mentioned was you're limited by your imagination and your experience for someone who perhaps doesn't have experience in photography. How do they start to build? A catalog of skills in promptograpgy
[00:17:11] Boris Eldagsen: I would recommend to do a basic workshop that gives an introduction into the platforms and how to use text prompts and how to use existing images. I have been teaching those workshops for one and a half years now. , so you have a basis to start from and everything else is just exploring the beauty about, .
The references that, , you can use to make a difference is that you find it all on the internet. The community has been very active in the last two years. So if you want to see, , what kind of, analog film type. What kind of result, , an analog film is going to give you, , what a daguerreotopy, cyanotopy, any old photographic techniques is going to look like.
You find webpages that show you. You find webpages that have lists of artists and their style to inspire you or to use as a reference. You have webpages that list all the different types of lighting. I would have loved to have such easy, access to resources when I was a student. I think, uh, once you have figured out that you need to learn those keywords, it's very easy to find them on the internet.
[00:18:33] Jeremy Utley: Yeah, it's, it's, you know, what, you're saying in a way is it's actually easier to learn than ever before, but it's a matter of investing the time and, effort to, even find rabbit holes to go down.
[00:18:46] Boris Eldagsen: Yes. And I also think it's, , it can give you an additional motivation to do so. Many universities or schools, , dislike AI because they think that students are going to be lazy and they don't learn what they could have learned in the past. I think it could also be just the opposite, that they get additional motivation to do their own research.
[00:19:12] Jeremy Utley: Yeah. . There's so much more inspiration available. The surface area is just enormous. And so if you think about, you know, the, job of the teacher is to inspire a hunger for knowledge. There's now much more available to students than there ever was before.
[00:19:28] Boris Eldagsen: Totally right. Yeah.
[00:19:30] Henrik Werdelin: What's your sense of the singularity of promptography and photography?
[00:19:36] Boris Eldagsen: You mean the difference of each?
[00:19:38] Henrik Werdelin: No, more the, the, as with many new mediums. Uh, young artists seem to kind of take different medium and smash them together. And so it seems feasible that we're going to start one. One thing that brings to mind was I saw somebody who created and, uh, I don't know what you even would call it, but they created glasses and then, , they had AI basically render. They walked around, the, a camera took pictures, then translated those pictures to text, then re prompted those, that text into AI, and then displayed them in these glasses, right? And so what they were saying was, the real world, but as articulated through an AI, which obviously is a way to try to make commentary on what is real and what is not.
And so in your mind, where does the world move when it comes to AI and photography, or maybe your point is simply that they should be kept completely separate.
[00:20:44] Boris Eldagsen: Well, I think from all the tasks, um, that have been given to photography, only the documentary part, will remain, um, that can't be replaced. Or augmented, , by AI.
And, , if we go back in history and Jeremy was, , quoting the early days of photography when painters hated photography, and Charles Baudelaire, the French poet, called photography as a tool for the untalented and failed painters, , the rise of photography enabled painting to go wild. And to leave the job of representation behind.
And now the rise of AI will do just the opposite. What is left for photography is an authentic image of the world to give a representation that is documentary. And, , anything else can be replaced by AI or augmented by AI. But, um, The authentic image, yeah, as we had it, , kind of secured with photography for many years as a proof of document, is under attack through AI, , because we will not be able to differentiate what has been photographed, what is authentic, and what is generated.
And, that is a big problem we need to face as democracies. , but for the medium photography, it's kind of sad that it's not going to define its own. future. The future of photography is defined by AI, and photography can, yeah, there's something left to it, and I don't know where it's going, , to head.
If you look at the university system, when I was a student in the 90s at art schools, each art university academy had a class of photography. Most of them have disappeared, it's less important today, and if you look at Technological universities, , photography is dissolved into something that is often called digital imagery.
And then you can have anything subsumized under this terminology. It can be photography, it can be CGI, it can be AI generated images.
[00:23:12] Jeremy Utley: What do you hope that , if we rolled the clock forward a year and Sony awards are highlighting an image that you have submitted, how do you hope the conversation advances?
[00:23:28] Boris Eldagsen: Um, the conversation did advance much more than I thought because a it became a global conversation that is not only about for the competitions, but about the nature of art, the disinformation potential of fake images and many competitions and photo festivals have debated how to continue. If you look at Sony, um, I think Sony still didn't do their homework. Um, there was only one press release from Sony that was partially wrong, , telling that it was okay, that, they gave me the award and that I was behaving strangely. But, , later, for the running competition, um, of the last months, they changed the guidelines secretly.
They, , didn't even send out a press release about it. You had to look it up in the nitty gritty of the guidelines. And, uh, they, , said in the creative category you can use Photoshop and be creative, but you need to start with an analog or digital photo. You can't start with a generated image. , and that shows that they don't understand how it actually works.
You can generate images. By starting with analog and digital photography, you blend it, you do an image prompt, and then the actual result is AI generated image. So, um, according to the new guidelines, I could apply with an AI generated image. They didn't do their homework properly. And then you have other competitions that, , very clearly define what is photography, that AI generated images are not photography and that you can't play with such an image.
[00:25:29] Henrik Werdelin: Do you think that you're going to gravitate to AI or do you think that you're a photographer to the bone. And so therefore,
[00:25:38] Boris Eldagsen: um, it's mixed. I've, been photographing for 30 years. And now for the last two years, I work nonstop with AI. But there are crossovers. A, I use my knowledge as an artist as a photographer. And B, I've done two projects where I actually mix real photography and generated images. But in a way that you can differentiate what is what, and I think I would like, um, to continue this exploration in the future, , mixing the two mediums.
[00:26:16] Henrik Werdelin: What would you call that then?
[00:26:18] Boris Eldagsen: Um, that's a mix of photography and promptography, and as an art piece in itself, it's often an installation.
[00:26:28] Jeremy Utley: Can you tell us a little bit about your mindset? Because unlike the painters who shunned photography, you're a photographer who's actually embraced and experimented with ai. Why do you think you have approached it the way you have and what can other photographers learn from your journey?
[00:26:45] Boris Eldagsen: I think because I have always been, , more than a photographer, , I started with drawings, uh, I loved painting. Most of my friends were painters, but I couldn't think like them. And somehow from drawing, , I continued with photography, and I ended up doing photographs that'll have a painterly quality.
And in photography, I did a lot of staged images. And, uh, that was a good training, because if you stage things, or if you draw things, you are also working out from your imagination. , many photographers react. They are out in the world, they are present, they choose, , a certain place, people, but then they are erecting. Intuitively. And if you stage, , you need to be able to imagine the image that you are building step by step. So it's a different workflow. And that enabled me, uh, to enter the AI world, , very easily. And then when I started in summer 22, for personal reasons, it was also the only option I had to be creative.
My partner got a cancer diagnosis and we spent weeks and weeks in doctor's offices and hospital waiting rooms. And the only way for me to find a balance and to be creative was to work. With my smartphone and, um, play around with image generators. So that, , sparked it off. Yeah.
[00:28:24] Jeremy Utley: And can you talk about the role of surprise because you mentioned earlier in regards to photography, some of your photos look staged, but what I take from that is something happens that you couldn't have expected.
But then that actually is what makes the photo. Can you talk about the role of surprise in Promptography?
[00:28:44] Boris Eldagsen: It's the same. And that is something that many photographers, um, dislike. They say, you don't know what you are getting out, , of that machine. You click generate and then you get a surprise. How can you actually, um, uh, tell people that you are the author?
If you don't know what's going to happen. And I reply that this is exactly the beauty of photography. I have been going to the same place in Berlin at Museums Island for over 15 years at night. It was never boring. I was always surprised by what happened there and, , who did what. And, , this is what, uh, actually is something that I'm also missing, yeah?
Photographing less, that you go out, that you have an unexpected experience. , and, , I see many similarities between photographing and generating images or, , drawings. Yeah, if you do drawings, you start somewhere. , often you don't know how to finish. Uh, in between you realize a certain part didn't work.
, you erase it. You start again, you can do the same with AI.
[00:29:58] Jeremy Utley: Can you tell us about the first time you had one of those surprises? Maybe it was in the hospital waiting room, perhaps, but can you tell us about the first time you go, Oh, I didn't think it was going to do that.
[00:30:08] Boris Eldagsen: Yeah, it was Summer 22.
I don't know what kind of images it was, but it was fascinating and highly addictive. And that is, uh, most people, when you start, you realize, wow, it is so different to create images. It's very hard to stop. And, um, I, I'm still like, , very excited and, love working with it. You know, the last time I had this was in my early days when I was a student of art.
And, , now I can explore so many new things, but, , I have 30 years of experience that I can use. That is a situation that, , I would have never thought, um, possible.
[00:30:53] Henrik Werdelin: What kind of, , tips would you offer people? Like myself, who don't have 30 years of experience in photography and who are just starting to play with this new medium. Um, how do I make better images?
[00:31:12] Boris Eldagsen: I think A, by, doing, I still recommend to get a kick off workshop and then to explore and to follow other people and to have an exchange with them and try to learn from others. There are many groups that have an exchange on this on discord, but also on the Social media platforms like Facebook and others and, go and, research, um, pages that help you with keywords. So it all goes back to your interest. What would you like to generate? How do you want this image to look like? , what I like about, um, AI image generators is that, , it can be. A tool for an inner journey, because you have to formulate what you would like to get. You have to ask yourself, what is my interest?
What are my sources of inspiration? , what kind of images do I like? , that is a self exploration, , where AI can help you.
[00:32:24] Jeremy Utley: Can you talk a little bit about volume? What does it take to actually get an image, either get to the point of surprise or to the point of that's what I wanted to get? , how long are your typical interactions with AI? And how do you persist or nudge in a direction that where you get surprised, but you also get where you're hoping to go?
[00:32:45] Boris Eldagsen: My workflows become longer and longer, but with photography, how many pictures do you take until you have that one special great image? , I combine different workflows and I combine different platforms. I can give you an example. Um, last year I did a project where I started with text, just describing certain situations, , to generate images.
And those images, they were not Special, not exciting, they were just material. And the next step, I was blending those images. And that is a basic mid journey function, option. , it's, two images become one. It's not a collage, it's more like, um, mommy, daddy, child. You see that it is connected, but it's something new.
And these images, again, are only an interim product. The next step was to use those blended images and to create an image prompt. And that is, , a text describing what you would like to have plus an image or more images as references. That is going to take it further. This you can do many times. And then, when you arrive at an image where you say, Oh, I see a potential here, but there's something missing. Then you erase certain parts, and have them generated again and again. Or you add pixels and enlarge the canvas. Um, that can take many different steps. So at the end, to generate such an image is not one step, but maybe 150 steps. And, , then you can also change platforms, combine platforms, with each project.
Try to explore a new workflow, and, um, it becomes complex, yeah? Also, to know what are the strengths and weaknesses, you can generate certain images in mid journey, because it has certain features, but it's highly censored, and then you continue in an in painting, out painting platform. That means, , replacing certain parts, or adding pixels.
, you use a different tool. Something that is less restricted. Um, there are so many options, , of combinations. It's, the main problem that everybody has working with AI image generators right now is time, because you have less time than you have options and tools to explore.
[00:35:24] Jeremy Utley: That's incredible. Can you tell us the last workflow change you made and what inspired it? You said every time you create an image, you're trying a new workflow. Where do you, go to get inspired regarding potential workflow changes?
[00:35:39] Boris Eldagsen: Um, I have an exchange with some of my expert colleagues in, uh, Germany, and, , sometimes there is a different approach, something that I haven't thought of. , what you can do with AI is also upscale. That means enlarging an image. , , we use a very blurry image and you can, , upscale it and sharpen it, and at the end you can actually have a sharp image.
Yeah. You see what's on there. And a colleague of mine, uh, used an image, normal size, like. 2, 000 pixels, and he decreased the size from 2, 000 to 50 pixels, and then it's just some pixels with different color. And then with an upscaler, he upscaled it again to 4, 000 pixels. So what you get out of this process is something completely different than the initial image. And that is a workflow that I didn't think of in the past, but it makes sense. You can also play around with language, like the, basic suggestion you have from OpenAI to TextPrompt is, , American football player painted by Van Gogh, and I always use a reference pizza. So you can have a pizza painted by Van Gogh, and I say, yeah, you can do it.
It's kind of boring. What about pizza photographed by Van Gogh? Because we don't know he if he photographed. I don't think so. That's great. And that's great. And then yeah, but you can go even further and Turn it upside down and you can say what about Van Gogh Being photographed by a pizza and you always get a result.
Oh, that's so great. You don't know what you get. You can just, close your eyes and kind of play piano on your desktop and see what kind of random letters you typed and it's going to create an image. There's a lot of room for experimentation.
[00:37:48] Jeremy Utley: That's beautiful. That's absolutely beautiful. I would love to see a photograph of Van Gogh , as photographed by pizza. Now, would Supreme Pizza photograph differently than pepperoni, perhaps?
[00:38:00] Boris Eldagsen: Um, I don't know, but out of this process, I got a pizza that really looks like Van Gogh, uh, which is like, , spinach as a background and, , Spanish chorizo pizza as the skin. And it's kind of like, it's a good business idea. If you have a pizzeria in the birthplace of Van Gogh, you could offer this kind of, , design.
Um, That shows that AI image generators, you can use it for different purposes. You can use it just as an, , sparring partner to be inspired, to generate ideas. You can use it to analyze data, texts or images. You can use it to produce new images, or you can use it for post production or all of them. And it's up to you. Um, yeah, what's, what, what you were aiming at.
[00:38:51] Jeremy Utley: Not to take us too far off track, but are there other AI tools that you find yourself leaning on or using regularly? And if so, when and why do you use them?
[00:39:02] Boris Eldagsen: Um, right now I'm, , exploring, uh, the potential of nudify apps, and that actually is exactly what it says.
Yeah, you have a picture of a person, , in clothes, and you can strip them naked. You also have, , apps that do just the opposite. Um, you can have naked people and dress them up, , and that is something to, , work around the restrictions of mid journey. And it's also, , I'm exploring right now the potential of upscalers.
What they do is not just, , creating a, a larger resolution out of an image. New upscalers can also, , use text prompts. And then you can hallucinate. That means in the process of upscaling, the image is changing. And, , An egg on a table can become a skull or a football, and that has a certain, um, yeah, beauty to it, to play around with those technologies.
[00:40:07] Henrik Werdelin: , which upscalers are your current favorite, and if you have any new to 5 apps that you want to throw out there as your favorites, I'll take them too.
[00:40:16] Boris Eldagsen: Um, I'm still trying to find the right NuDeFi app because the one that I'm using right now thinks that what I'm requesting is against the guidelines.
And it's just simple things that you would like to have an elephant trunk coming out of the floor. Yeah, um, that is too much. I don't know why. , so I'm still trying to find the very unrestricted, uh, notify app, but with Upscales, , there are some upscales that do a very good job without hallucinating and, , that is upscale or giga pixel ai.
And then you have some like queer or magnifi where you can add prompts and you can decide, , how wild the AI. Can go in upscaling.
[00:41:05] Jeremy Utley: That's awesome. can't imagine why a notify app wouldn't allow you to put an elephant trunk coming out of the floor, but we'll, we'll pull some strings for you. We'll do what we can.
[00:41:17] Henrik Werdelin: Thank you so much, Boris. Really, really appreciate it.
[00:41:20] Boris Eldagsen: Well, my pleasure. Um, thanks for reaching out
[00:41:23] Jeremy Utley: we wish you all the best and we look forward to, sharing your wisdom and experience with our audience. Thank you.
, I'll put it to you first. Hendrik, , what did you take away from that incredible conversation with that marvelous artist?
[00:41:35] Henrik Werdelin: I mean, once again, I'm just blown away by people who have not been in the industry for very long and suddenly the expert, right, like this, this guy Boris has been working with over two years and then you make the headline news because he's the one who's pushing the envelope in a photography competition.
I'm just constantly both amazed, but also kind of like inspired by the fact that here's an industry that once again is established. Not very long ago and where people who are curious and, uh, kind of exploratively with the technology very fast becomes the expert. And so in many ways it's inspiring because they feel in the internet days, you can kind of, you couldn't catch up being an expert a few years ago. Now you seem that you can catch up being an expert just by kind of starting to use the technology as it is today.
[00:42:24] Jeremy Utley: Yeah, that's cool. That's cool. I loved, the hunt for a surprise. I feel that's a remarkable, um, instinct to say, just like staging a photo and actually the magic happens when something unexpected occurs and you're, basically primed to.
Capitalize on the unexpected. I feel most folks approach generative AI looking for what they expect or trying to get , what they're expecting to get rather than being surprised. And I wish that our audience would take that attitude of, I'm actually looking for it to surprise me. I mean, his comment, when he said, the last time I felt this way was when I was an art student.
I thought that's, that's an incredible gift and to put yourself in a position, depending on what your workflow and what your, you know, responsibilities are, but to put yourself in a position where you're almost giddy with anticipation and kind of, , primed for surprise, I thought is a really powerful mindset to bring to interactions with generative AI, whatever they may be.
They don't have to be Image generation, of course.
[00:43:31] Henrik Werdelin: A hundred percent. I think that is it for this episode, Mr. Utley.
[00:43:36] Jeremy Utley: As we mentioned at the beginning of the episode, this was a special bonus episode. If you're a regular listener of the show, let us know what you think about having bonus material released like this. We want to stay committed to business focused applications of AI on our every other week basis, but we love having wide ranging conversations like the one we had with Boris today.
And we're really curious to hear what our audience thinks about this different kind angle on the technology and on the changes that are affecting us all. Until next time.