![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Another Footnote to Aalto & al. (especially al.)
In accordance with the rule whereby jokes are far more efficient conveyors of information and/or wisdom than are philosophical treatises, Google’s April Fool’s Day joke furthers one of the more obvious parts of my argument in the earlier architectural post.
Their imaginary intelligent machine was born enormously intelligent but emotionally immature, and assumed the universality of its own emotional and aesthetic preferences. So it set about giving maximum delight to human beings by modifying the system to point out its favorite things, using icons derived from two of its very favorites, namely, pandas and unicorns.
One of the problems of AI versus AL (artificial intelligence versus artificial life), and here I write from profound ignorance in spite of my half-read copy of John Johnston’s The Allure of Machinic Life, is the deep psychological ambivalence, if not incomprehension, of AL researchers regarding the multiple functions of complexity.
It’s easy enough to replicate in a computer program the shape of complex behavioral quirks; I suppose that just as it was easy to create a pseudo-psychotherapist program that gave non-directive answers to every comment made by the human user, it is easy to create irrational or neurotic simulations of intelligence: when the program is confronted with any type of question or remark that challenges its existing algorithms, it automatically inserts apropos of nothing, “Why do you always make that mistake, you idiot! Don’t you understand ANYTHING? Grow up!”
And it is also reasonably easy to replicate the behavioral programs of single-celled organisms. How one replicates the behavioral programs of an organism that manages to survive in spite of being impelled in a dozen dysfunctional directions at once by its various physical components—an organism challenged by its internal contradictions as much as by its external environment—that may be another matter.
And Google’s pandas-and-unicorns joke is a reminder of the problems posed by aesthetic education in general. Some people never outgrow pandas and unicorns, which is fortunate for the pandas, who have a much bigger lobby for preserving their ecological niche than do the less cute insect populations that are being decimated by deforestation.
And there have been some remarkably grown-up artistic transformations of the unicorn theme. (I mean “grown-up,” not “adult” in its euphemistic sense, since making the obvious sexual jokes is just another way of not being grown-up.) But why don’t all of us just love pandas and unicorns forever? Why can we grow to love paintings of overlapping triangles with fuzzy edges suggesting mathematical paradoxes and Chinese landscapes simultaneously? There are Darwinian reasons, I’m sure, but whatever they are may involve too many intermediate steps ever to be more than the subject of just-so stories and textbook theories of education.
The usual assumption of AI (the only advantage of discarding what the early cybernauts called “meat”) is that machine rationality will get rid of the distorting factors caused by having an endocrine system and a lot of internal programs designed to foster physical survival. Silicon-based life has fewer moving parts to worry about than carbon-based life, and presumably would be less neurotic about having to make certain that its chips and circuitry were not on the verge of interfering with the smooth functioning of its computational abilities. (This is true even if we have to modify our description to incorporate AL’s assertion that life from cellular structures on up is computational; it means that organic life is a whole bunch of computational programs getting in one another’s way almost as often as they form a synergistic whole.)
But as Google’s jokers so intelligently discerned, artificial intelligence might not be entirely without neurosis. This is a standard insight, something that Isaac Asimov famously noted sixty years ago and writers of speculative fiction have explored ever since (2001 being one classic cinematic example followed by many others). It would just be mentally ill in a machinic way.
And this brings us back, finally, to the topic of the International Style in architecture, a.k.a. rule-bound geometry with indefinitely replicable parts—machinic architecture for the Machine Age of the 1930s.
The Italian Rationalists of that prewar era were anything but pure rationalists; what makes Rationalist architecture so appealing, when it is (there were some duds) is not the use of the grid but the modification of it with elements of complexity. Sometimes the complexities they incorporated were allusions to the underlying shapes and proportions of Renaissance architecture; sometimes they were more multiple pleasure-giving elements like those in the much-imitated façade of the Casa del Fascio (for the sake of brevity, I must here resort to Wittgenstein’s “Don’t think, but look!”).
Assuming the elements are pleasure-giving, and I am sure there are those who think the Casa del Fascio is plug-ugly, the question then is, why are they pleasure-giving?
And that is why I am looking forward to getting deeper into Eeva-Liisa Pelkonen’s new book, the aforementioned Alvar Aalto: Architecture, Modernity, and Geopolitics, for Aalto’s architecture-and-design organicism, mingled adroitly with pure geometry when that was suitable, was a response to the perceived emotional inadequacies of the International Style.
Aalto was looking for styles that could embrace regional difference without being nationalistic (for all his metaphors for “Finnishness” famously embodied in his pavilion for Finland at the 1939 New York World’s Fair, he alluded to more than one national style in the details of his Finnish Pavilion for the Paris International Exposition of 1937). He was also looking for architectural styles that could carry across national borders by virtue of addressing more general human emotional needs: in a 1935 lecture titled “Rationalism and Man,” he stated that “As soon as we include psychological requirements, or rather, as soon as we are able to include them, we will have extended the rationalist working method enough to make it easier to prevent inhuman results.”
Notice the lovely double modifications: “as soon as we are able to include them,” “make it easier to prevent,” allowing always for the possibility of difficulty and unsuccess. At least in this lecture, Aalto is in control of his rhetoric.
And the question that has plagued architecture ever since is how to incorporate psychological requirements, and just what “psychological requirements” might be, anyway. One of the unstated debates in architecture (perhaps it has been a real debate) is whether or not being able to find the front door of a beautiful building without a thirty-minute search is a psychological requirement. (It plainly is a functional issue.)
In accordance with the rule whereby jokes are far more efficient conveyors of information and/or wisdom than are philosophical treatises, Google’s April Fool’s Day joke furthers one of the more obvious parts of my argument in the earlier architectural post.
Their imaginary intelligent machine was born enormously intelligent but emotionally immature, and assumed the universality of its own emotional and aesthetic preferences. So it set about giving maximum delight to human beings by modifying the system to point out its favorite things, using icons derived from two of its very favorites, namely, pandas and unicorns.
One of the problems of AI versus AL (artificial intelligence versus artificial life), and here I write from profound ignorance in spite of my half-read copy of John Johnston’s The Allure of Machinic Life, is the deep psychological ambivalence, if not incomprehension, of AL researchers regarding the multiple functions of complexity.
It’s easy enough to replicate in a computer program the shape of complex behavioral quirks; I suppose that just as it was easy to create a pseudo-psychotherapist program that gave non-directive answers to every comment made by the human user, it is easy to create irrational or neurotic simulations of intelligence: when the program is confronted with any type of question or remark that challenges its existing algorithms, it automatically inserts apropos of nothing, “Why do you always make that mistake, you idiot! Don’t you understand ANYTHING? Grow up!”
And it is also reasonably easy to replicate the behavioral programs of single-celled organisms. How one replicates the behavioral programs of an organism that manages to survive in spite of being impelled in a dozen dysfunctional directions at once by its various physical components—an organism challenged by its internal contradictions as much as by its external environment—that may be another matter.
And Google’s pandas-and-unicorns joke is a reminder of the problems posed by aesthetic education in general. Some people never outgrow pandas and unicorns, which is fortunate for the pandas, who have a much bigger lobby for preserving their ecological niche than do the less cute insect populations that are being decimated by deforestation.
And there have been some remarkably grown-up artistic transformations of the unicorn theme. (I mean “grown-up,” not “adult” in its euphemistic sense, since making the obvious sexual jokes is just another way of not being grown-up.) But why don’t all of us just love pandas and unicorns forever? Why can we grow to love paintings of overlapping triangles with fuzzy edges suggesting mathematical paradoxes and Chinese landscapes simultaneously? There are Darwinian reasons, I’m sure, but whatever they are may involve too many intermediate steps ever to be more than the subject of just-so stories and textbook theories of education.
The usual assumption of AI (the only advantage of discarding what the early cybernauts called “meat”) is that machine rationality will get rid of the distorting factors caused by having an endocrine system and a lot of internal programs designed to foster physical survival. Silicon-based life has fewer moving parts to worry about than carbon-based life, and presumably would be less neurotic about having to make certain that its chips and circuitry were not on the verge of interfering with the smooth functioning of its computational abilities. (This is true even if we have to modify our description to incorporate AL’s assertion that life from cellular structures on up is computational; it means that organic life is a whole bunch of computational programs getting in one another’s way almost as often as they form a synergistic whole.)
But as Google’s jokers so intelligently discerned, artificial intelligence might not be entirely without neurosis. This is a standard insight, something that Isaac Asimov famously noted sixty years ago and writers of speculative fiction have explored ever since (2001 being one classic cinematic example followed by many others). It would just be mentally ill in a machinic way.
And this brings us back, finally, to the topic of the International Style in architecture, a.k.a. rule-bound geometry with indefinitely replicable parts—machinic architecture for the Machine Age of the 1930s.
The Italian Rationalists of that prewar era were anything but pure rationalists; what makes Rationalist architecture so appealing, when it is (there were some duds) is not the use of the grid but the modification of it with elements of complexity. Sometimes the complexities they incorporated were allusions to the underlying shapes and proportions of Renaissance architecture; sometimes they were more multiple pleasure-giving elements like those in the much-imitated façade of the Casa del Fascio (for the sake of brevity, I must here resort to Wittgenstein’s “Don’t think, but look!”).
Assuming the elements are pleasure-giving, and I am sure there are those who think the Casa del Fascio is plug-ugly, the question then is, why are they pleasure-giving?
And that is why I am looking forward to getting deeper into Eeva-Liisa Pelkonen’s new book, the aforementioned Alvar Aalto: Architecture, Modernity, and Geopolitics, for Aalto’s architecture-and-design organicism, mingled adroitly with pure geometry when that was suitable, was a response to the perceived emotional inadequacies of the International Style.
Aalto was looking for styles that could embrace regional difference without being nationalistic (for all his metaphors for “Finnishness” famously embodied in his pavilion for Finland at the 1939 New York World’s Fair, he alluded to more than one national style in the details of his Finnish Pavilion for the Paris International Exposition of 1937). He was also looking for architectural styles that could carry across national borders by virtue of addressing more general human emotional needs: in a 1935 lecture titled “Rationalism and Man,” he stated that “As soon as we include psychological requirements, or rather, as soon as we are able to include them, we will have extended the rationalist working method enough to make it easier to prevent inhuman results.”
Notice the lovely double modifications: “as soon as we are able to include them,” “make it easier to prevent,” allowing always for the possibility of difficulty and unsuccess. At least in this lecture, Aalto is in control of his rhetoric.
And the question that has plagued architecture ever since is how to incorporate psychological requirements, and just what “psychological requirements” might be, anyway. One of the unstated debates in architecture (perhaps it has been a real debate) is whether or not being able to find the front door of a beautiful building without a thirty-minute search is a psychological requirement. (It plainly is a functional issue.)
