Diego Trujillo Pisanty (Mexico City, 1986) is an artist working with information as both a concept and a material. His past work has explored how information is captured, stored, transmitted, interpreted, and destroyed. He is also interested in the role that information plays in shaping identity, politics and our understanding of nature. His work, Blind Camera, was the 2024 Lumen Prize Still Image Award Winner. Blind Camera creates images using sound instead of light while inquiring about bias in AI. The device uses a custom-made artificial neural network to convert sound into image. It was trained with data from Mexico City restricting its interpretations of the world to this locale. We spoke to him about the award winning Blind Camera, exploring working and training AI, capturing sound and the artistic opportunities of AI.
1. What are the core concepts behind your artistic practice?
I think that art is an activity that tries to make sense of the world and give it meaning. In this context I see my works as enablers of situations where this sense-making can happen. Many of my works are machines that somehow interact with the real-world, either because they are used “in the wild” or because they incorporate data found out there. In this context the artworks behave similarly to scientific experiments in that they create the conditions for things to happen.
This is where I think information becomes a raw material, not only in that it often drives my devices’ functionality as code but also in that these situations need to be interpreted both by myself and by the audience. Making and using my devices also forces me to see the social, political, and cultural implications of using and making technology. In the case of Blind Camera this was explored largely as a concern with bias and identity, but in other works I’ve looked at technology mediated power dynamics or how we ascribe value to artefacts.
Where did the idea for Blind Camera come from and what is your interest in linking sight and sound?
The idea originally came from an interest in the plasticity of digital materials. If everything is encoded as bits then it must be possible to represent the underlying data in any form. I think AI achieves this spectacularly and that was partly why I ended up working with an AI model.
In parallel I was interested in how we perceive sound. While a visual scene can be frozen and captured it seemed almost impossible to do that with sound and that fascinated me. I wanted to come up with a thing that captures a sound instant because that seemed very difficult to define. At the same time I like how sound is much more generic than vision in our mind. It’s very difficult to determine exactly what a sound is, for example, we can recognize a car engine but just by hearing it we know nothing about its colour, model or brand. I was really interested in making images that echoed that level of perception at a visual level.
3. What was your experience working with AI for Blind Camera, treating and programming AI for artistic purposes?
It was initially very daunting, there was just so much to understand in theory before I could write the code. There was however a point where I realized that all the information is out there and that I just needed to be disciplined and engage with it. Once I understood what I was meant to be doing it sort of just flowed.
More interestingly, the way I made decisions on how to write the underlying neural network and its training cycles was driven more by how they made sense from what I was trying to achieve as an artist, not only in the images' aesthetics but also in how the data was processed. At some point it felt like the inner workings of the device had to follow my artistic concepts, the device had to follow rules that matched the questions I was asking even if no one knew this but myself.
4. What interests you about working with new technologies and AI in your work?
There are two things overall: The first one is that new technologies need to be contextualised and given a place within our lives. This is often done by marketing departments within the companies that develop the tech. I think part of my role as an artist is to provide alternative readings of these technologies while highlighting their potential unintended consequences.
Secondly, working with new technologies offers the possibility to innovate within visual and conceptual discourses. I’m quite cautious when it comes to this as I try to make sure that the technology is still reflecting my own aesthetics and areas of interest as it can be tempting to let go and see what it can do when left to run wild.
5. Looking forward, what do you think the biggest opportunities of AI will be for artists?
Like any new tool it will provide visual possibilities that were not necessarily possible before. While it has already made image-making easier I think that it will force us to consider the intent and control behind these images more rigorously. For those artists willing to make or train their own AI systems I think the possibilities will be endless, not only in terms of the work made by AI but also in the rationales behind gathering data and what computers can extract from it.
I’m very excited to see a growing number of artists engaging with AI systems and appropriating them to do things that maybe computer scientists or engineers don’t find worthwhile.