In conversation with AI artist: Chad Nelson

Chad Nelson stands as a testament to the transformative potential of AI, beckoning artists and innovators alike to embrace the boundless possibilities that lie at the intersection of technology and visual art. With a career spanning over two decades, Chad’s creative journey led him to the doorstep of OpenAI, and this was only the start.

Agnieszka Celej: Can you describe your profession? What’s the story behind you getting so closely engaged with OpenAI? 

Chad Nelson: My career has always been about taking new technologies and turning them into entertainment or content that people can really engage with. If I look at my career of roughly 25 years, in the 90s, in the 2000s, and then now into the 2020s, OpenAI in a way is a continuation of that journey.

In April of 2022, a closed beta version of DALL-E was released, which was the first publicly released generative AI tool. I knew I had to try this. I thought this might change the landscape of creativity and so I found a way to one of the individuals who was granting access through their artist program. And within my first moments of using these tools, it was almost as if I went from the dark to the light. 

Long story short, I played with their tools for roughly eight hours straight after I got access. When I finished, I created a collage of all these little creatures that I generated and I sent them a copy as a thank you note. 

Someone wrote me back from OpenAI asking, ‘how much of this was DALL-E, or how much of this was Photoshop?’ I confirmed that it was all DALL-E and they wrote me back the next day, and they said, We want to meet you. We don’t know how you got those results.’ So in a way, it was like they knew it was powerful, but they’d never seen someone do what I had done, create with that much visual clarity and personality. Once I met with them and their research team, the conversation just kept growing and growing. 

A: How long did it take you to master the prompting skills?  

C: I had been a programmer at one point, so I knew how to code. I wasn’t an expert, but I knew the basics in how to think structurally like a computer program. What I found was this structuring of prompts was key. It needed to be more logical and also follow an order that the computer could decipher and understand exactly what you want. So in a way, you need to understand your visual priorities. You start with the first priority, then go to the next and the next. It’s like a funnel. 

The other key, don’t bounce around and describe visual ideas like it’s a bowl of spaghetti. You have to be very logical in your descriptions. I found that in the early days of DALL-E, this was essential. However, now prompting with GPT-4 is so advanced that some of those early problematic examples don’t even hold up anymore. But in the early days it was definitely more sensitive, more primitive. 

A: A couple of months later your joint work with photographer Carlijn Jacobs, supermodel Bella Hadid and stylist Imruh Asha lands on the Vogue Italia cover. How does it feel?

C: I’ll be honest, the cover of Vogue was never something I had on my list. 

Wyświetl ten post na Instagramie

Post udostępniony przez Chad Nelson (@dailydall.e)

There is something magical that I think we in our digital society sometimes forget, which is the beauty of having a physical thing. Being able to turn a page, to see it not as projected light, but reflected light. Just that little difference. I have all the images digitally, but to see the actual final file printed on a cover – it was very special. 

A: Can you describe the project and walk us a little bit more through the process? 

C: Vogue Italia and Condé Nast contacted OpenAI to do an experiment with AI, and I was the one that was selected to work on the project. Photographer Carlijn Jacobs had the vision and provided much of the creative director for the shoot. And then obviously Bella and Imruh had all sorts of their own ideas, as well as Vogue and Condé Nast. It was a very collaborative effort and an incredible opportunity. In the end, I think, it became a 16-page spread inside, all featuring Bella, who is clearly one of the top supermodels of the world. So it was a great honor to work with the whole team. 

In a way it shows as a real testament to how AI will evolve creativity and creative opportunities. I’m kind of living that, you know?

A: Do you have your favorite piece from the shoot?  

C: Carlijn had some very strong imagery in her head for what she hoped the AI could deliver – obviously, she had already photographed Bella. And in some cases, it was her vision that we had to visualize. And that was one type of challenge. On the flip side, there were others where she would give me abstract descriptions and then I could visualize it.

Wyświetl ten post na Instagramie

Post udostępniony przez Vogue Italia (@vogueitalia)

One of my favorites is the zebra striped one. I really like the keyboard dress. I also love the one where she has the peacock on her knee. If you actually saw the original photo – she’s just in an empty studio in New York. There’s no bed. There’s not even a pillow under her head. All was created by AI, except for Bella’s designer outfit. 

A: Can you shed more light on the advantages and limitations of using AI in this specific creative process?

C: The opportunity to fulfill whatever creative vision that you imagine is far more possible with AI, as opposed to when you’re dealing with the realities of financial, time and logistic limitations. So to a certain degree, the advantage is that cost is not the big barrier anymore. If you don’t like a concept, say a location, cost is not the factor in redoing it.

How many times in the life of a filmmaker, painter, or novelist were they under time and budget pressure? Happens all the time. I think AI allows you to creatively explore far more than ever possible, within those limitations.

The weakness, though, is that AI doesn’t know everything. The AI is only good at what it’s been taught well. There’s a lamp on the table within the peacock shot I’ve mentioned. We rendered hundreds of lamps, hundreds, because we were trying to find this one Italian designer or a certain style of lamp. And it just didn’t know how to handle it. 

Once we wanted to do an image of an Italian sports car from 1968. It was a very specific style of a Ferrari where it had this almost space-age wedge design. The AI hadn’t been taught that car, so it didn’t know how to generate it. It did know that most Ferraris are red. Most Ferraris look like the ones you see on television. But that wasn’t our vision. We couldn’t do that. So with general knowledge and topics, you can do a lot with AI. But when it comes to the oddities, the obscure, that’s when you find the holes. 

A: In terms of that AI limitless variety of versions and concepts just at your fingerprints, the work seems to be almost never finished. You can always prompt further. Was it hard to say, ok, we’ve got it?

C: I think that was the beauty of having a Vogue deadline. It was going to print and we had to deliver the files before then.

A: What advice would you give to those interested in exploring the intersection of AI and visual art? 

C: Well, I think there are two things I would say. One, to existing artists. Learn. Don’t be afraid of these tools. You might not apply them to every aspect of your work. But it is amazing when you realize what benefits the AI might all of a sudden bring to your work. I like to describe these AI tools as giving me creative superpowers. 

It’s not that I believe they’re making me a better artist. I believe they’re letting me think through my ideas much faster and communicate them much quicker. In commercial art, you’re in the business of being a creative, and you have to present and sell ideas as quickly and effectively as you can. I have found that these AI tools allow me to do that better than I’ve ever done it before. 

The other thing I would say is to new artists, to young artists. You are starting your career at a time where you have the most incredible paintbrush ever invented. What that also means is your art education now should really be less about the techniques or the task. 

Now you need to learn much more about cultural history, art history… humanity! You need to experience life because the AI can do the task. So it’s really going to come down to what is it you have to say? What is it that you see about the world that’s unique? And if you haven’t given yourself that time to really learn those things and to understand history and where art has come from – you’re not giving yourself enough knowledge to best guide the AI. 

The AI will always be faster than you, in terms of recalling or replicating raw knowledge and information. So I think that is so essential for young artists to make sure you broaden your history as opposed to just learning a single craft or program.

A: What is your approach towards the risk behind the AI tools? 

C: Well, there are some issues that we’re unfortunately going to face. And to me, it reminds me of a story that my grandfather told me before he retired. He owned a letterpress print shop in LA. And that’s how he built his career. But computers came along in the 70s. 

And then suddenly by the 80s, his print shops were not getting the work because everyone could do it on their own computer with a laser printer. It will happen again in this generation. There will be careers that we’ve had based on a certain set of technology and tools and processes that will just somehow, you know, evaporate as careers. 

We’ve seen that throughout various industries and fields. So by that rationale, there will be jobs that are eliminated. And that is hard. 

If your job is very repetitive and it’s kind of a task that you know could be replicated, this is the warning sign. There’s definitely a yellow light that’s about to turn red, telling you your career or this particular job might be over. This is a good time to learn, to make a switch. 

The good thing is every single time, at least in my career, when new technology emerges, new jobs emerge as well. So there will be new jobs with this. I don’t know what they are yet. 

When I finished art school and entered my first jobs, there was no such thing as a web designer. No one was hiring a web designer when I finished art school. That job was invented. And what’s even crazier, now web design is not even a career anymore. We’ve already automated that entire field, for the most part!

A: Are there any specific projects you’re personally eager to explore in the nearest future? 

C: I want to continue doing more films with AI. In celebration of the one-year anniversary of the launch of OpenAI’s DALL-E, together with Native Foreign’s Nik Kleverov we’ve released an animated short, CRITTERZ, and it’s been received wonderfully. 

There are people in different industries that I think would be fascinated through the use of working with AI. So I would love to work with a select group of architects, designers, and I’d love to work with some real filmmakers. I’d love to pick some projects and partner with these individuals, and in a really collaborative way, much like I did with Carlijn and VOGUE, this is what we could do together utilizing AI.

A: Sky’s the limit! Thank you very much!


Chad Nelson is an award-winning creative director and technology strategist with over 25 years of expertise in designing cutting-edge interactive experiences and entertainment. His portfolio includes industry-leading video games, TV (both traditional and interactive), mobile apps, and 3D tools for digital artists. As co-founder of technology start-ups Eight Cylinder Studios and WGT Media, Nelson has collaborated with Fortune 100 companies such as Microsoft, Intel, Sony, and Virgin. Chad also serves as a creative collaborator with OpenAI.

His most recent creative venture, the animated short film CRITTERZ, is a groundbreaking achievement as the first film with visuals generated entirely by OpenAI’s DALL-E, then brought to life by incorporating classic Hollywood animation techniques. This past year, CRITTERZ has been screened at the Portland Festival of Cinema, Animation & Technology, Annecy International Animation Film Festival, Tribeca X, Cannes Lion, and was awarded “Best In Creative AI” at the Qld XR Festival 2023 in Brisbane, Australia.

Additionally, while under OpenAI’s creative program, Nelson worked alongside photographer Carlijn Jacobs for the April 2023 cover of VOGUE Italia. Modeled by Bella Hadid, the feature spread broke artistry barriers by expanding the possibilities of studio photographers through AI designed images. Discover more of Nelson’s AI visual creations and films on Instagram at @dailydall.e.


Continue reading