ISSUE 5. Artificial Intelligence. Interview of Gabriel Esquivel by Alfonso Arias
Alfonso:
Gabriel, given the fact that our current issue of AGENCIA is dedicated to artificial intelligence, and you have been working on it for a while. I thought it was important to talk to you about what your point of view and experience has been on this topic. Why did you become interested in AI?
Gabriel:
I have been researching and teaching digital fabrication for more than a decade, trying different materials, methods, experimenting with Robotics and some other things. I was looking for something different to do and suddenly the opportunity of starting to work with AI was there. I was working on a research project with a group of people along the idea of “Architectural Intelligence” and Serlio. So that ignited the possibilities to start working with AI. I’ve created a new office as well with the purpose of developing these type of projects, it is called (BE)² with a great collaborator, his name is Shane Bugni, he has been a critical part for this research to take place. (photo) https://aiarchitects.org/portfolio/esquivel_bugni/
The next thing I did was create an advanced research lab. The purpose of this type of research lab was to focus on the transforming potential of artificial intelligence robots and automation in architecture. The purpose was to open the discussion about the multitude ways of looking at for the problem of robots, AR, VR and Artificial Intelligence and Architecture. Not only as a toolset to optimize specific processes and practices but rather to emphasize architecture’s ability to serve as a cultural marker, challenging the idea of computational methodologies as tools of expedience and efficiency but seen as mediums which architects work from.
Alfonso:
In AGENCIA, we have talked about all kinds of problems in regard to the architecture studio and its pedagogy. Have you developed an application of AI for the architecture studio?
Gabriel:
Well, we haven't really worked on an application for an architecture studio per se, that is the reason why I created the advanced research lab. We needed to start researching the idea of what architectural intelligence is first. This exploration as I was saying earlier was around the figure of the Renaissance architect Serlio and his treatise. We were considering his books as documents on architectural intelligence, and we used his images to create several data sets. After we have been working on this for a while, we started to think about the importance of using artificial intelligence as a way to support the architecture studio. In other words, it is not meant to substitute in the architect. The traditional methods in the architecture studio whether analog or digital are still valid but AI could extend the possibilities of development within the studio, this could be in terms of performance, simulation or form making. Our main focus of investigation in AI is about form, specifically the translation between 2D and 3D using a workflow of neural networks, as well as parametric and modeling softwares.
Alfonso:
I know this is a typical question but for you what is artificial intelligence?
Gabriel:
I think there is not one particular definition accepted by all experts of what artificial intelligence is. First, because it is a relatively new, constantly changing, and it is an experimental science. And second, because we can't even define exactly what human intelligence is. I would say that basically, AI is the attempt to mimic human intelligence using a robot, or software. But this is a very vague concept because there are many ramifications. In 2009 Stuart Russell and Peter Norvig differentiated four types of AI, systems that think like humans, like neural networks. Systems that act like humans, like robots. Systems that use rational logic, such as expert systems, and systems that act rationally, such as intelligent agents.
AI is capable of ingesting vast volumes of data, recognizing patterns and making decisions at a scale beyond human abilities. Of course, there is a lot of discussion about it at conferences, in the media, academic circles, in the corporate world and government. But these discussions focus on a high-level viewpoint and application, that of the workforce and government. As executives, policy makers and academics discuss the broad implications of AI and robots in the workplace for example, little attention is paid to the specifics, the individual people and what will happen to them? But the real rise of artificial intelligence, on a practical level, I think happened when powerful and inexpensive computers began to appear, capable of experimenting with AI on a global and everyday level. There are many types of AI, some of them are still on experimental stages. Examples of a type of AI are those used in computing, mobile phones, Internet services, and other areas close to ordinary users. Concepts such as machine learning, neural networks, and other technologies that we often hear about, but we do not really know how they work.
Alfonso:
One of the issues that most architects worry about is if AI is going to substitute the architect?
Gabriel:
I don't think so, I think there is a long way to go before that really happens. I think that is a myth among different people and different ideas. We are at the beginning of the exploration of AI in terms of architecture. I believe that a complete substitution of the architect is very difficult. If we as architects start working with AI now and start experimenting and demystifying it, is a good start. To learn how it can help us and by doing so we will be prepared for the future, gaining agency with this particular technology. I think as architects we already use artificial intelligence, but we don’t think about it.
In general, there is great excitement about the incredible things that AI, robots, and other forms of automation can do. As technology advances, AI and robots are moving from curiosity and entertainment to practical tools capable of working, analyzing, and creating. We have seen robots and automation take jobs in manufacturing plants and some office jobs like secretaries, and telephone operators. But today we're seeing AI make strides in claims processing, analytics, sales, marketing, and more, where higher levels of education and skills for this particular field are needed. So, I believe is only natural to implement more aspects of artificial intelligence in architecture schools and the practice as extensions of the discipline.
As Matias Del Campo suggests, That we need to generate conversations about the nature of AI, and its possible impact. The term AI is quite a generalist term and is used to describe several different approaches. Matias Del Campo also suggests that this opens up questions about the nature of creativity, the methods to evaluate this, and the nature of creativity at large. These are questions that we need to address, because there is an enormous amount of fear of losing human agency in design. In most cases a fear that is not based on fact, and we need to address this. The idea that AI can creatively generate a sensibility, but also from a profoundly ethical point of view.
Alfonso:
After we heard the opinions of some of the people that have contributed to AGENCIA magazine, what do you think are the possibilities to implement this kind of technology within Mexican academia?
Gabriel:
I know, as well as you do, that there are lots of issues about digital technology and academics in Mexico, which we have been discussing in several issues of the magazine. Most of the people that we have asked this question, particularly those working directly with digital technology, most of them see the future quite difficult. However, I want to think that there's a promising future out there. This week I was listening to a round table discussion about artificial intelligence on Facebook Live from the Colegio Nacional in Mexico which as you know is one the highest academic institutions in the country. The topic was specifically about “How to transform information into knowledge,” using AI. This event had truly remarkable Mexican scholars like Carlos Gershenson, and Saiph Savage talking about different applications in the workplace, in government research as well as academic research. These experts have studied and work binationally in Mexico and the United States, they do research and teach in both countries, therefore I think there is an enormous interest in this particular field. Now in terms of architecture, perhaps a little more difficult, because there's always, as we said before, a resistance from within the architecture academic community for the exploration of these particular topics of technology and how they could affect the architecture curriculum.
Architectural education in most Latin American countries is built mainly from the design studio, this educational modality where the teacher guides some type of design process. However, we rarely see in this classic studio setting an innovative or experimental methodological proposal, much less speculative. In the interview he gave to AGENCIA, Rodrigo Shiordia, director of the master’s degree in Digital Fabrication at the Universidad Anahuac Mexico, points out in his article about this educational modality where the teacher guides a design process. However, we rarely see in this studio an innovative or experimental methodological proposal, much less speculative. So, the chances of introducing AI in the studio environment is limited, however I think this particular research will start at the academic institutions with an architecture curriculum that includes masters and PhD programs.
Alfonso:
Do you think that it is through research the way to start with the implementation of AI?
Gabriel:
When people think about architecture and architects, they often think about architects building buildings, and that is basically the only thing that architects do. But we know that architects are also searching, looking for things that might be complementary, supplementary, to the production of buildings. For example, they could be researching issues related to the history of architecture, architectural technology, performance, or AI as well as the impact of architecture on society which might be economic, social, political. Research in architecture involves developing knowledge or expertise in a particular area. It is not about simply implementing new technologies, this is important, because it may not be highly effective to see research in these limited terms, sometimes the line between these activities and strategic research is blurred. My interest in technology is in combination with other aspects of the discipline for me personally theory and technology are quite important. That is why 10 years ago I developed the T4T Lab based on this principle. https://www.t4tlab.com/
Research is not simply about finding out about new materials, or the latest technical software for your latest project and then filing the information digitally for later reference. I would always push for a line of research that is aligned with its current strategies or with its intention to develop new areas of expertise. Bringing practicing professionals into the academic research realm is important. Research and development are central to any relationship, engagement or linkage between architecture academia, practitioners, and emerging businesses. There are two potential difficulties you need to think about. The first is that academics in Mexico need to consider a couple of things: their time is often hemmed in by teaching commitments and an overly regulated bureaucracy, but also a lot of the architecture faculty teaches and practices, this leaves very little time for research. The second we need to understand the dynamics of practice or business protocols. But drawing professionals into the world of our research can be truly beneficial integrating these worlds and begin to dissolve this problematic dialectic between academia and practice that seems that it will never go away.
Alfonso:
How do you see the potential of AI going into the architecture practice?
Gabriel:
I think AI is something that is here to stay, and it is going to change lots of things. We don't know exactly how. There are lots of people as we know that they have been working on this issue for a long time. I'm kind of new at this. I've been doing this for two years. During this time, we have been able to discover and explore different ideas. A lot of these things are potentially helpful in the understanding of architecture not only in terms of design and form making but in performative issues, behavioral issues, as well as all kinds of simulations that are already being use by Mexican architects like the ones that appear in our current issue. Maybe it can start with AI in planning of urban issues as it is indicated by Juan Ponce Briseño in regard to his field of interest. Or in the work of Laura Spinadel about a virtual campus.
In architecture there is a lot to be explored in terms of the image and representation using things like style Gans, the potential for a new language for architecture and by extension the development of a 21st century aesthetic regime that comes from these explorations. This is one of the projects that I'm currently working on. The connection of graffiti, language (semiotics) and architecture, specifically in the study of the facade and the architectural surface. So, I think there's a lot of things that can happen in terms of practical uses of AI in the architect’s office.