Artificial Intelligence and Architecture: Matias del Campo - Part I

In this lecture, Matias del Campo provides some thoughts on the ontology of Artificial Neural Networks and their relationship to architectural production. Oscillating between aspects of wicked problems (aesthetics, agency, authorship, inspiration, creativity) and tamed problems (analysis, feature recognition, prediction), this lecture paints a picture as of how architecture might operate in a posthuman design environment.
Dr. Matias del Campo is a registered architect, designer, and educator. He is an Associate Professor at Taubman College of Architecture and Urban Planning, University of Michigan, director of the AR2IL - The Architecture and Artificial Intelligence Laboratory at UofM, and affiliate faculty member of Michigan Robotics and MIDAS, the Michigan Institute of Data Science. Most recently, he published the AD Machine Hallucinations - Architecture and Artificial Intelligence with Wiley, and his book Neural Architecture - Design and Artificial Intelligence with ORO editions.

[@pickerrs]

He's wrong on Wittgenstein, fwiw. The proposition 'The limits of my language are the limits of my world', from the Tractatus, refers to the boundary condition of propositions that 'have sense' and are sensible/coherent. Whatever propositional statement about the world that is formed in language constitutes the sum of my 'picturing' of its attributes. It is a semantic process that is theoretically infinite, but individually only ever partial. The ML he speaks of is able to generate far more complex descriptive propositions than we usually employ, and that constitutes the recognisable quantum of an image when we experience the 'Verfremdungseffekt' or 'estrangement effect' he speaks of.

[@matiasdelcampo5085]

Interesting thought. However, I don't think my proposition is very different from what you just described in an alternative way. Let me be more precise: Wittgenstein's assertion suggests that how we can understand and describe the world is fundamentally tied to the language we use. In other words, the structure and limits of our language determine what we can meaningfully say about the world and what we can think about it. To that end, Wittgenstein believed that language is not just a tool for expressing our thoughts but also a medium through which we represent reality. According to him, when we use language to make statements or propositions, those statements are meaningful only to the extent that they accurately reflect the structure of the world. If our language lacks the necessary structure or vocabulary to express something, then we are limited in our ability to even conceive of that thing. Thus, "the limits of my language" refers to the boundaries of what we can effectively communicate or think about using our language. If our language doesn't have words or logical constructs to represent specific concepts or phenomena, then those concepts are beyond the limits of our language, and we cannot discuss or comprehend them. By asserting that the limits of language are also the limits of one's world, Wittgenstein is making a broader epistemological claim, don't you think? He suggests that what we can know or understand is constrained by the linguistic tools we have at our disposal. If a concept cannot be expressed in our language, it effectively falls outside the realm of our knowledge or understanding. This leads directly to the end sentence of the Tractatus: "Whereof one cannot speak, thereof one must be silent" (Worüber man nicht sprechen kann, darüber muss man schweigen) - which I always considered not an order to remain silent, but to invent or discover the necessary language to explain the phenomenon, thus constructing a reality around it. Probably another point where we both disagree on Wittgenstein.

[@pickerrs]

@@matiasdelcampo5085 Hi Matias, thanks for your response which is far more courteous than my gruff comment - which I apologise for. I think the point of difference is the idea that language 'represents' reality. In my reading of Wittgenstein, as an enthusiastic amateur, I understand him to mean that language is not some imperfect thing that has the task of interpreting reality, it is the basis of all sense claims. So we agree about the limits of my language being the limits of the world - as you say. And yes, you're right about the final Tractatus proposition - but I think its more about the exercise of philosophical thinking. I think he is saying, once you have thought through a question in philosophy it no longer exists as a problem, because what else is the purpose of philosophy than to clarify confusion. But can we discover a language? Or discover the nuances of effects in other languages - much like the nuances of effect in the famous example of Inuit words for 'snow'. Or of languages with complex senses of state and time, or gender. In terms of language prompts for AI, the estrangement issue is really the point of the exploration since it is inherently monstrous, or outside of propositional expectations, which is what I was trying to say above.
Enjoyed your book by the way!