Originally, artists and architects esteemed imitation. Historians know this from a variety of Greek and Roman sources such as, for example, the writings of the Roman educator Quintilian (b. AD 35) who in his Institutio Oratoria (Education of an Orator) states: “there can be no doubt that in art, no small portion of our task lies in imitation, since, although invention came first and is all-important, it is expedient to imitate whatever has been invented with success, and it is a universal rule of life that we should wish to copy what we approve in others.”
In our time--the “art world” having since around 1910 widely rejected the principle and practice of imitation in art on the basis that it can only ever engender derivation (“derivative” perhaps being the most soul crushing word ever uttered in modern art criticism)--we witness not artists but scientists, specifically computer scientists who after having experienced some success in the development of artificial intelligences have now met up with some roadblocks and to overcome them have with some sense of urgency (if not panic) renewed an interest in learning how humans learn through—of all things-- imitation.
AI developers have acknowledged that what we have thought of as artificial intelligence until now is not so much intelligence as it is an illusion of it. Computers can mimic (even seemingly originate) speech and imagery based on techniques of pattern recognition enabled by algorithms fueled by massive amounts of data (techniques theorized in the 1980s and manifested only in the last decade or two with the advent of the internet). Under the right circumstances (chess, translation) it makes for great theater, but machines have yet to master even the seemingly simplest tasks of which human 2-year-olds are capable.
The rigidity and fragility of the brute-force pattern recognition approach to AI has become so evident (even a minor glitch in the data or change of algorithmic goal can discombobulate the machine) that scientists and engineers now want to understand how it is that humans learn to then attempt to reverse engineer it with the hope that this will generate more flexible and resilient machine learning capability. They have turned to child psychologists, those who have studied the youngest of us-- infants and toddlers-- for whom, besides curiosity and exploration, imitation (of first the mother, then later others) is central to how we learn. “A crucial factor that sets children apart from AI”, says child psychologist and UC Berkeley professor Alison Gopnik “is that they learn socially, from other people. Culture is our nature, and it makes learning particularly powerful. Each new generation of children can take advantage of everything that earlier generations have discovered.”
The evidence is apparently incontrovertible, and scientists of the mind generally agree that it is more efficient to extend the bandwidth of our thinking by taking advantage of how others think-- grafting their thinking onto ours, bootstrapping ours onto theirs. The external scaffolding of what others know and can do enables efficient and effective learning, and it is this that interests the developers of machine learning (although Gopnik is skeptical that machines will ever master anything resembling real human learning except in specifically programmed task targeted ways). Whether our current art mythologies choose to recognize this or not the evidence in the history of art is also incontrovertible. There would be no Thelonius Monk without Claude Debussy, nor Rolling Stones without Muddy Waters.
The cult of originality into which we have been kidnapped has had deleterious consequences for both the practice and the education of the architect ranging from celebrity fetishization to how our schools are set up. Architecture schools are mostly situated within universities though they are both descendants of and would be strangers to the guilds and academies from which they emerged. The first academies were established to formalize apprenticeship, learning through training (imitation) whereas our schools are founded on the fiction of spontaneous inspiration (originality). We believe that somehow students are with little to no training supposed to come up with creations on their own. Instead, the so called “design studio” and the “jury” in architecture schools yield little more than students adept at performance art—the kind of mimicry pretending to be originality that AIs perform.
This kind of theater cannot develop the flexibility and resilience required for the lifelong learning experience that is the practice of architecture. It yields instead incompetence and alienation—artificial intelligence-- schools now secreting “graduates” uninterested in architecture because they know so little about it and disillusioned by the impotence of a “profession” (not to mention faculties) occupied by so many who know so little about it. If we were to relax our obsession with what are by now discredited creation myths in art, we might set ourselves up for less anxiety as students (and teachers) and eventually more fulfillment and creativity (and success) as practitioners.