help, someone's been theorizing my work!!! averyreview.com/issues/42/wewo…
I feel compelled to respond. To start, the technical understanding of our paper is fairly flawed and puts words in our mouth (we never made any claims about "design skills" anywhere, for instance), and there are lots of incorrect statements about @WeWork.
Given the timing it's also hard not to see it as a bit of clickbait with an academic gloss. Which is a shame because some of the core questions — about what automation means for the architecture industry — are really important!
And what this piece obscures, in the interest of sensationalizing and painting the company as a big scary dehumanizing monster, is the extent to which our philosophy about automation has always come from a place of sensitivity to, and respect for, the labor of designers.
I think this is better expressed in our more recent paper, "Humanizing Architectural Automation," which it seems the author has not read. If he'd reached out at all I'd have been happy to send a copy.
From the introduction: "We hypothesize that the adoption of advanced automation processes in practice requires focusing on improving the experience for human designers as much as it requires focusing on improving the performance of algorithms."
We specifically propose "putting the designer at the center of the automation process (both in developing the tool and in using the tool)"
And this is something I talk about all the time. What do we think is the alternative to automation? Just Not Automating Stuff is fundamentally not an option.
Architects and designers must take ownership over this process and drive it, or we'll wind up in the nubureaucratic dystopia foretold by the article.
In our work we've always started by asking designers the questions: which parts of your work suck the most? Which parts of a design process feel like tedium and drudgery? And those are what we build automation routines around.
We also explicitly focus on AUGMENTING rather than automating design. Let's give designers better tools to do their work; let's build on the trajectory from drafting to cad to BIM to reach even higher levels of conceptual abstraction and control over design outcomes.
Another key finding in my work, which I also talk about publicly a lot, is that AUTOMATING THE WORK OF DESIGNERS IS WAY HARDER THAN IT SOUNDS.
Design is hard, and complicated. Even small, seemingly straightforward decisions involve a vast amount of clever problem solving and creativity, thinking at multiple scales at once, juggling conflicting demands and constraints, trying and inventing multiple strategies at a time.
And I see this as a good and reassuring thing for the profession. Robot architects are not taking your job away anytime soon.
Even though I disagree with most of its specific assertions, interpretations, and conclusions, I appreciate this article for trying to take the issue of automated architecture head on. Not enough people are even beginning to think about what it might mean.
But I think it's unfair to paint the work of my colleagues and me in this simplistic light when we've spent so much effort trying to do it right — in a way that fundamentally respects the process of design and the agency of designers.
I would love to see more people (designers, theorists, technologists, and whoever else) engage critically with this issue!
