Between Artificial Intelligence and Consciousness: The Art of Daniel Ambrosi
Explorations of AI Art — Episode 18
"Beyond pondering the true essence of the natural phenomena that trigger our sensations and perceptions is the more enigmatic question concerning the nature of consciousness itself. Working with artificial intelligence busts that line of inquiry wide open." Daniel Ambrosi
[Fig.1] Dunboy Castle, County Cork, Ireland. Dreamscapes, courtesy of Daniel Ambrosi
In the recent years the field of machine learning has advanced enormously and nowadays computers can generate realistic images that do not actually exist, identify people and objects and get very close to mimic human behavior for some tasks. However, artificial intelligence still remains primitive with respect to humans.
In the artistic field, many have started using AI as a tool, medium, artistic collaborator and subject of study. Living in a world permeated by technology this was foreseeable and, after all, inevitable: in our existence, in fact, the boundaries between real and digital are increasingly fainting.
While on the one hand many discussions on artificial intelligence, based on a dystopian scenario, focus on its potential destructive threat and divert attention from other important issues, on the other hand the introduction of this technology in the artistic field opens a potential that goes beyond the imaginable.
In our interviews we have often talked about how artificial intelligence and art can together open a glimpse into blind spots and push us to question our individual and social limits. However, there are many other aspects of our existence that can be investigated and presented by this union of (only apparently) opposite worlds. We believe that the artist can play an effective role in contributing to ethical and philosophical discussions often overlooked by the scientific and institutional world, and at the same time act as a communicator and creator of strong and shared experiences that well represent our times.
This month I had the pleasure of speaking with Daniel Ambrosi, one of the founding members of the AI artistic movement and well-known artist for the nuanced balance achieved in the human-AI hybrid art. He is based in Half Moon Bay, California, and has been exploring novel methods of visual presentation for almost 40 years, after earning a Bachelor of Architecture degree and a Masters in 3D Graphics at Cornell University in 1984.
After developing a unique form of computational photography that creates immersive landscape images, in the last few years the artist has devoted his effort to develop large scale artworks that display exquisite sensitivity and intricacy. These artworks, called Dreamscapes, are built upon his previous experiments by adding an enhanced version of DeepDream expressly modified for him by two brilliant software engineers - Joseph Smarr (Google) and Chris Lamb (NVIDIA) - to operate successfully on his huge images.
Ambrosi's AI-augmented artworks and grand format landscape images have been exhibited at international conferences, art fairs, and gallery shows, as well as installed in tech offices and featured in multiple publications. He has been defined by art dealer Paul Fisher as "a groundbreaking visionary artist straddling the worlds of fine art, science, and nature."
When I started to put together this interview, I thought the focus was going to be around AI and photography. To my surprise, I discovered a completely different world that amazed me; indeed, it turned out to be a conversation full of food for thought far from photography. I learned, first of all, that Ambrosi's art has more to do with landscape painting than with landscape photography, being greatly influenced by movements such as Romanticism and Impressionism. Then, that the application of AI in his work is a great generator of deep conversation and shared experiences, where both the artist and the viewers are encouraged to question the concepts of consciousness and perception of reality.
The use of technology is optimistically viewed as a means to expand creativity and expressiveness. At the same time the artistic production, enriched by a different aesthetic, becomes a co-creation with the public which generates a unique experience. All this leads to a deeper question of the nature and mystery that surrounds our human life: seeing becomes a subjective creative act that takes place mostly in our mind.
[Fig. 2] Ithaca Falls, Ithaca, NY. Dreamscapes, courtesy of Daniel Ambrosi
What is a possible role for computational artists in our society? Is it about advancing the art field in which they work or is there something more?
I suppose one role computational artists can play is to show by example that a desire to express oneself, to advance ideas, and to create can be greatly facilitated by thoughtful use of computational tools. In my own case, I was driven to find a more effective way to capture and convey the experiences I was having in the presence of special places, especially amidst great landscapes. In the process of investigating this, I became utterly captivated by the 400-year tradition of landscape painting in the western world. The remarkable achievements of the highly skilled artists of the Hudson River School, and of the European Romanticists before them, and the Impressionists who followed, deeply inspired and influenced my own artistic journey. In fact, while my landscape art starts with a camera, I consider my work more closely aligned with the tradition of landscape painting than with landscape photography. This is somewhat ironic since I have no training or practice in painting. Computational tools in general, and artificial intelligence in particular, enabled me to participate in this fine art tradition and--with great reverence and respect--take it to new places. So, yes, for me at least, advancing this art field has become my primary mission.
I love this question because it leads to what I enjoy most about viewers’ reactions to my artworks: the surprise, wonderment, and delight they experience when the details of my Dreamscapes reveal themselves to be entirely different from what they thought they were seeing. That experience clearly shows them that seeing is a subjective creative act that mostly takes place in their mind and is not just objectively about the photons that gather on their retina. That experience--and the realization that follows--does indeed make the viewer an integral part of my artworks.
[Fig. 3] Ithaca Falls (detail), Ithaca, NY. Dreamscapes, courtesy of Daniel Ambrosi
This segues nicely from your prior question. Beyond pondering the true essence of the natural phenomena that trigger our sensations and perceptions is the more enigmatic question concerning the nature of consciousness itself. Working with artificial intelligence busts that line of inquiry wide open. This fascinates me to no end. To paraphrase author/philosopher/neuroscientist, Sam Harris, “Consciousness is the only thing we can be sure of, yet it’s the thing we know the least about!” While I have no illusion that I’ll ever learn the definitive answer, I love that collaborating with artificial intelligence compels me to think deeply about this grand enigma and inspires stimulating conversations with my viewers.
[Fig. 4] Double Arch Alcove, Zion National Park, UT. Dreamscapes, courtesy of Daniel Ambrosi
To answer this question, I think it’s important to first consider what we’ve seen thus far with AI art. In my view, AI art comprises a spectrum that ranges from works that are primarily AI-generated to those that are better characterized as AI-augmented, or “human-AI hybrid” art. I place my efforts far on the AI-augmented end of the spectrum: AI has enabled me to achieve a specific artistic intent that preceded the arrival of the tool itself. That said, while I can set the direction of this tool, I can’t control the details. What’s especially amazing to me about these algorithmically-generated details is that they are contextual; my AI responds to the character of the specific source photography it is analyzing. Like all AI artists, I can and do carefully curate the results of my AI’s semi-autonomous efforts. This combination of human-led direction/curation with AI-based responsive manipulation makes this an artistic tool like no other. To me, it truly feels like a collaboration.
[Fig. 5] Fitzgerald Marine Reserve, Montara, CA. Dreamscapes, courtesy of Daniel Ambrosi
There’s no doubt that my AI has enabled me to create artworks with a level of intricacy, mystery, and grace that I could never achieve on my own. In this sense, it has taken me beyond my initial ambitions and unlocked a superpower for me. At the same time, my AI would not have risen to this occasion if I hadn’t convinced a couple brilliant engineers (Joseph Smarr from Google and Chris Lamb from NVIDIA) to modify it for my purposes. As originally released by Google, the “DeepDream” open source software that my AI is built upon, was simply not designed to work on multi-hundred megapixel images. Fortunately for me after months of effort Joseph and Chris succeeded in enhancing DeepDream way beyond where its original developers had envisioned. This dance between man and machine has been taking place for millennia and shows no signs of ending or slowing down; quite the opposite, it is accelerating. This brings both promise and peril as so many notable writers and speakers have pointed out in recent years. I choose to focus on the promise and to use artificial intelligence for purely creative pursuits.
[Fig. 6] Japanese Tea Garden, San Francisco, CA. Dreamscapes, courtesy of Daniel Ambrosi
I’m happy to report that I’ve had a bit of a creative surge while sheltering in place these last few months. My latest project, Infinite Dreams, is an exploration of Cubism-inspired “refracted” dreaming (see: http://www.danielambrosi.com/Infinite-Dreams). This seemed like a natural direction to take since my landscape art has taken an arc somewhat parallel to the tradition itself, which started with representational paintings, evolved through Impressionism, and then into Cubist approaches.
I also remain committed to manifesting my vision for an immersive exhibition centered around the theme of climate change. Over the course of my photo expeditions of the last decade I’ve witnessed alarming amounts of forest loss due to wildfire and tree diseases exacerbated by global warming. With my Swan Song project (see http://swansong.ai), I hope to encourage participants to contemplate what is at risk of being lost and, in doing so, inspire them to redirect their grief and anxiety toward involvement and action.
Finally, I continue to collaborate with developers to create virtual reality experiences with my Dreamscapes. I’ve had the good fortune of experiencing this myself in multiple VR labs, including at Facebook and Google, and I can tell you it’s a wonderful way to experience these images. Thus far these have only been experiments and demonstrations, but I’m hopeful that someday these immersive experiences will be more widely available.
To follow Daniel Ambrosi on social media:Twitter: @danoramas Instagram: @danoramas
References and Resources
About the author: Beth Jochim is the Creative AI Lead at Libre AI, and Director and Co-Founder at Cueva Gallery. She works at the intersection of technology and arts. She is actively involved in different activities that aim to democratize the field of Artificial Intelligence and Machine Learning, bringing the benefits of AI/ML to a larger audience. Connect whit Beth in LinkedIn or Twitter.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.