So much has been written about Artificial Intelligence (some of it by humans), that I am not sure there is much original left to say. AI will be a boon to architects as a virtual assistant and quality controller, researching codes and checking plans for compliance and collisions. AI will be a bane to the architecture profession by combing vast data sets to serve up average solutions for new problems, and then it will feed those same solutions into its data mill to resolve the next problem. AI offers answers based loosely on the average of all solutions in its data set. As the dataset itself is populated with average AI generated solutions, the dataset reverts to a mean and true creativity suffers. That’s one theory, anyway.
For architects, a more interesting and worrisome AI might be Artificial Intuition. Intuition – we flatter ourselves to believe – is one of those uniquely human qualities that will keep us relevant and safe when the robots take over.
Hah!
Generic House Form – Minimalist 3D House Design with Sleek Lines and Modern Aesthetic for Real Estate Architecture AI Image: Adobe Stock/Fajr
What is intuition?
For that question, I refer to Daniel Kahneman’s 2011 seminal book, Thinking, Fast and Slow (I only finished it this year because I was thinking slow). In his book, he cites psychologist Herbert Simon’s explanation of intuition, “(A) situation has provided a cue; this cue has given the expert access to information stored in memory, and the information provides the answer. Intuition is nothing more and nothing less than recognition.”
The reason this recognition feels almost magical rather than slow and analytical is the incredible speed at which it takes place. It is too fast to consciously articulate what is happening, so we assign this phenomenon to subconscious. This is part of our thinking that Kahneman calls our System 1, “a machine for jumping to conclusions.” System 2 is the analytical part of our brain that kicks in when System 1 cannot figure out what is going on. You’re using System 2 right now to check the time and figure out where the hell I am going with this. When we encounter a situation, our minds use System 1 to automatically connect it to past experiences, patterns, and knowledge stored in our memory, leading to very quick, unconscious, judgments or responses. System 1 looks only at what is immediately apparent and filters out extraneous or conflicting information in exchange for the capacity to make almost instantaneous – possibly lifesaving – decisions on how to react to anything unfamiliar. This feels like intuition.
So, if intuition is recognition as Herbert Simon states, and what is being recognized are “associative memories” according to Daniel Kahneman, then intuition must be something personal, dependent upon, and unique to, each of us. With intuition (and System 1), not just the memories are personal but the filters as well, formed by feelings, biases, fears and self-consciousness.
In contrast, Artificial Intuition must use generic datasets instead of individual personal memories and filters defined by algorithms to suggest surprising patterns and insights at blazing speeds that may impress but are never connected to personal experience. If Artificial Intuition returns biased results like a human, it’s due to poor programming rather than lived experience. This implies that “Artificial Intuition” may never exist for reasons of semantics more than philosophy or technology.
So, do I still have a job?
Like any creative profession – architects can take some small comfort for now that true invention based on personal intuition safely remains a “human” activity. For now, we exploit the AI’s extraordinary powers of search, synthesis and analysis, use AI for potential productivity gains and quality control, but we remain wary of the gristmill of banality when using AI for inspiration and actual design. As AI becomes more ubiquitous and draws from ever larger datasets, will we subordinate our personal associative memories to a sort of collective associate memory of ChatGPT, Gemini and others? Will we recognize the difference? Will we be seduced by the ease and inventiveness of this new “simulated intuition”? Even if everyone recognizes the difference between simulated and real intuition, if it is cheap and “good enough” for a less discerning market, the greatest economic risk for architects is not to be directly replaced by AI, but squeezed out.
I don’t think there is a road map for any of this, or even a road, but I am reminded of a famous film by Charles and Ray Eames called “Powers of Ten”. Progressively stepping the camera away from its target (picnickers in a park) by a factor of ten, the film soon takes us to the scale of the universe. Then, just as quickly, the camera divides by a factor of ten until we pass beyond the picnickers until reaching the nucleus of an atom. I think AI is the voyage upward into the expanding universe where we are awestruck by the vastness of available knowledge. I think the trip inward to sub particles represents our personal selves that can be aggregated but not eliminated. The macro and the micro are co-dependent and never mutually exclusive. AI may wash over us but never wash us away.
And that, dear reader, is my personal intuition.