Mental Maps: How Iregular and a Montreal Museum Visualised the City in the 19th Century through the Eyes of AI

Can we travel back to a world that no longer exists? 

When Montreal’s McCord Stewart Museum set out to create their ‘Becoming Montreal’ exhibition based on the works of James Duncan, this was the question they asked. The little-known Irish artist, who painted primarily in watercolours, was a chronicler of his era. Between 1830 and 1880, his work came to define Montreal as he documented its transformation into a cultural and industrial giant in the 19th century. In a sense, Duncan became one of Canada’s first photojournalists - in a time before photos were popularised. 

And so, any exhibition would need to be worthy of his work and legacy. As part of the plan for ‘Becoming Montreal’, the Museum turned to Iregular - the pioneering digital art studio - for assistance. The result of their collaboration was MENTAL MAPS, an immersive video installation concluding the exhibition, which plunges audiences in an imagined universe based on James Duncan’s body of work. 

In order to make MENTAL MAPS a reality, the team at Iregular leveraged the cutting edge of visual technology. Artificial intelligence and a 3D image generation technique known as NeRF (neural radiance fields) were essential to the creation of the installation, in which attendees step inside an AI’s interpretation of reality and explore layers of landforms and topographies, becoming themselves part of the scenery. 

To find out more about the groundbreaking project and how it’s offering new perspectives on some of Montreal’s significant art, LBB spoke to the founder of Iregular Daniel Iregui, alongside the McCord Stewart Museum’s head of digital outreach Stéphanie Poisson and project manager for exhibitions Catherine K. Laflamme. 

Above: Daniel breaks down MENTAL MAPS in this promotional video provided by the McCord Stewart Museum. 

LBB> Stéphanie, can you talk to us about your initial vision for this exhibition, and why you wanted to partner with Iregular to help realise it? 

Stéphanie> The attentive and constant gaze brought by James Duncan on his city makes it possible to imagine the evolution of Montreal in the 19th century. The artist circumscribes it as a whole, with a beautifully detailed focus on daily life. We wanted an exhibition to define the different views worn by Duncan which, when combined, provide a vivid impression of Montreal and the reality of those who lived there. 

For several years, we have been closely monitoring the work of Daniel and Iregular and have been waiting for the perfect project to work with them. From the start, we had the idea of integrating a project that would create a strong contrast with Duncan's work.

We liked the idea that their very linear and graphic visual signature, often monochrome and data-driven, would create an interesting contrast with the watercolours of James Duncan. We were delighted to know that they wanted to dive into the adventure, and create a work for the exhibition. 

LBB> And Daniel, why was this an exciting project for you and Iregular? 

Daniel> To Stéphanie’s point, it’s absolutely true that the painterly, figurative, and watercolour world of James Duncan is at the opposite end of the artistic spectrum from our usual style. Our work is geometric, digital, and monochromatic. I loved the idea of a meeting between these two differing artistic universes. 

How could we apply our approach to something which was, up until this point, outside of our wheelhouse? For us, that was a very exciting question. 

LBB> Can you tell us how MENTAL MAPS has evolved - to what extent do your initial visions align with the finished exhibition? 

Daniel> When we started to work on MENTAL MAPS a year ago, the intention was to visualise how the artistic decisions of James Duncan and his resulting body of work deeply influenced the depiction of Montreal in people’s imagination. 

Fast forward to the present, with the rise of artificial intelligence and machine learning systems becoming commonplace, the value of MENTAL MAPS now resides in how it’s helping us better understand the power of AI, and the role we humans carry in a world where such systems possess the capability to generate intricate images from simple prompts on a computer.

Catherine> Iregular's initial proposal consisted of projections that used lenticular prints, which immediately appealed to the Museum team. It was easy for us to understand the project and where it was headed towards. However, the concept was completely transformed when the team decided to adopt artificial intelligence and remove the lenticular images.  

We cannot hide that this turnaround raised questions within the Museum team - but we were still convinced that Iregular was the right partner to explore this new avenue. This was a novel approach for the Museum, and it called for a totally fresh kind of expertise. From our perspective, we were embarking into the unknown since it was our first experience with AI.

Ultimately, it was a bold proposal that would bring a new look to Duncan's work and anchor it in the 21st century. We could not miss this opportunity to take an entirely fresh approach to Duncan's work. In addition to this, we felt that Iregular’s proposal was likely to attract a new kind of audience, which was another of the Museum’s objectives. 

And finally, we felt inspired by the team’s drive and motivation to explore this new path. The time was right to do something different! 

LBB> So how exactly does MENTAL MAPS utilise AI? 

Daniel> To answer that, I need to explain why the concept of ‘point of view’ was so important to our team. It became apparent early in our research that James Duncan’s point of view was incredibly influential in the history, culture, and common understanding of Montreal as a city. His work gave visual life to the city in a world before cameras. 

So we wanted to give weight to the perspective of the exhibition’s attendees. We wanted them to feel as though their point of view mattered, transforming the world around them in some meaningful way. The idea we settled on relied upon AI, as well as a groundbreaking filming technique called NeRF. 

NeRF can build 3D digital worlds from static images - but in order to do so, you need to provide the AI algorithm with multiple images from different angles. So, we fed the algorithm with all of James Duncan’s work, giving it the best possible sense of the Montreal the pioneering artist captured in the 1800s. Then we asked the software to reveal the world it imagined. The result was extraordinary - like the AI is dreaming about James Duncan’s paintings and you are stepping inside of that dream. 

LBB> What do you think James Duncan himself would say if he saw MENTAL MAPS? 

Daniel> Well, I’m not sure what his taste was! But I think what he would appreciate is that the data he created - the worlds he drew as he saw them - has informed our cultural perception of an entire city. I think this installation is a testimony to just how influential his work has been. I hope that would make him feel proud. 

LBB> Catherine, what were the challenges associated with this project from the Museum’s perspective, and how did you overcome them? 

Catherine> The big difference with what we usually do when we integrate digital into our exhibitions was the relationship with our partner, in this case Iregular. It was a different kind of rapport, since they virtually had carte blanche. 

Given the nature of the technology being used, there were a lot of unknowns involved which, as a museum institution, we are not used to. We are accustomed to controlling everything that goes into an exhibition and knowing well in advance what is going to happen. 

The deadline was tight and we had no idea of the visual before the very end, which could  sometimes lead to uncomfortable situations. But we had confidence, which required a certain amount of trust and letting go. 

We’re glad that we did. These kinds of experiences generally appeal to a younger audience - one that the Museum wants to attract. 

LBB> And Daniel, did you learn anything over the course of this project which you think could be applied to future installations or exhibitions? 

Daniel> Oh, absolutely. I think this experience has been very energising for our team, getting to grips with technology and learning whilst doing. There’s a lot which can be applied to future projects. 

When you work in a Museum, you’re looking to tell a story with the material you’ve been given. And I think what we did shine an entirely new light on James Duncan’s work through these unique and ethereal dreamscapes. I think that’s something which could be applied to plenty of other exhibitions. 

For us, however, we’d want to be even more ambitious in any upcoming projects. This technology feels like a great new opportunity for art and for immersive experiences.