Frank Gehry passed away late last year and in this post I want to acknowledge the work of the great architect. My interest in architecture goes back as far as I can remember. At an early age I gave my father (an engineer) a bit of a shock by announcing that I wanted to become an architect—one of the worst choices I could make in his mind. I gained a sense of his perspective years later when I started my engineering degree at Melbourne University (yes, my dad persuaded me to change my ways). In a civil engineering 101 lecture, our lecturer chose the following image to illustrate the difference between architecture and engineering:
We all recognised the picture as a stylized rendition of the Sydney Opera House, and he proceeded to add the mechanism that an architect would, in his telling, use to hold up such a structure: the skyhook.
Perhaps it is civil engineers who hold the greatest contempt for architects, on the basis that they have to solve the challenges posed by the architect’s vision for a structure. I’d argue that the Sydney Opera House, like many iconic buildings, is a triumph of both architecture and engineering, and we’re fortunate that Jørn Utzon’s bold vision turned out to be implementable. The engineering firm Arup solved the construction challenges without resorting to skyhooks. There is a great podcast about the drama around the opera house in the Cautionary Tales series.
I went on to become an electrical engineer and computer scientist, but I have continued to have a strong interest in architecture, both of the physical kind and of the sort that computer scientists get to work on, that is, system architecture. In fact, “system architect” was my first job title at Cisco, a place that wasn’t entirely pro-architect, to the extent that one of my colleagues called himself an artichoke rather than use the other a-word. I’m not sure I ever shared my title with my father.
As I discussed in my SIGCOMM keynote, I have a longstanding interest in bridges, which also capture for me the intersection between architecture and engineering (as well as being a favorite networking metaphor). But my interest in architecture goes beyond bridges, and I’ve enjoyed visiting iconic structures since at least 1985, when I made my first visit to the Centre Pompidou, right before I arrived in Edinburgh to start my PhD.
Fast forward to 2005 and I was at an IETF meeting in Vancouver where I noticed a lot of interesting modern architecture along the waterfront near the conference hotel. I decided to go and buy a disposable camera to capture some photos (ah, the pre-iPhone era) and this eventually led to my starting a blog with an architectural bent to it. To finish up the roll of film I then took a few photos of cool architecture in Cambridge, MA, where I lived at the time. Exhibit A was the Frank Gehry-designed Stata Center, which had opened the previous year as the home of CSAIL.
Gehry was selected as the architect for the Stata center not long after he became a household name for designing the Guggenheim Bilbao museum. You don’t have to spend much time looking at Gehry’ s work to wonder if he would be the ideal choice for a building comprising mostly offices and lecture theaters. According to a reliable source I know at MIT, the faculty who were destined to be housed in his building were justifiably a bit nervous before meeting the famous architect. Gehry allegedly walked into the meeting with them and said “I bet you’re all scared sh*tless. But in fact I can design buildings that people work in.” A couple of years later I ended up with a visiting lecturer position at MIT and my own (shared) office inside the Stata center. I can say that for me the building worked pretty well, aside from one infamous meeting room (see below). There were, however, a number of problems with the building, including claims of sick building syndrome and ice falling onto sidewalks. In the resulting lawsuit, Gehry blamed “value engineering” (i.e., cost cutting in the implementation). Some things never change.
A Perennial Tension
The tension between architects and implementers is a perennial challenge. Fellow author Jim Kurose recommended a great book on the subject when we sat together on a flight home from SIGCOMM 2006. A Place of My Own by Michael Pollan is one of several books that touch on this conflict, and we discuss some others here.
So the one Gehry-designed feature that bothered me at the Stata center was a meeting room on the ground floor—you can see it in many exterior photos as a multi-story yellow cone. On the inside, there were wood panels arranged in such a way that lines between the panels led up the walls to a skylight, with none of the lines being quite vertical. For many people (including me), sitting in a room surrounded by lines that are not quite vertical induced seasickness. At one point large, vertical cardboard tubes were placed in the room and as long as I focussed on them I was OK. Apparently Gehry was brought into the room at one point, became sick and said “Well, I never made that mistake before!”
Just down the street from the Stata Center, the MIT Media Lab had its headquarters in a rather less flashy building designed by I. M. Pei. Pei won the Pritzker prize a few years earlier than Gehry, and is perhaps best known for the pyramid at the Louvre. There is, I believe, a bit of competition between the Media Lab and the computer science department of MIT, and when it came to extending the Media Lab buildings, a third Pritzker prize winner was selected for the job: Fumihiko Maki. According to a story I heard at a Media Lab event, during the pre-building consultation between Lab leadership and Maki, the nearby work of Frank Gehry came up. Maki apparently scrunched up a napkin and threw it onto the table, saying: “That’s a Gehry building. His buildings are explosive on the outside; my building will be explosive on the inside.” I will certainly say that Maki’s Media Lab building is lovely on the inside (and the outside) and it never made me seasick.
Obviously the negative view of architects that I heard expressed by my father and my university lecturer hasn’t prevented me from enjoying architecture throughout my life, whether as an amateur appreciating the buildings around me or as a computer scientist and engineer trying to shape network architecture over my career. For many years I could walk around my neighborhood and admire the work of three Pritzker prize winners. Now that we are hard at work on our 7th edition textbook, my challenge is to figure out how to teach an appreciation of network architecture to the next generation.
Speaking of MIT, I am a big fan of Rodney Brooks, one of the former directors of CSAIL (housed in the Stata Center) and a noted roboticist. In 2018 he made a number of very specific predictions about the future of technology, especially in the fields of AI and ML, and he has been reporting back on his accuracy ever since. He’s very good at this and I wish more people would listen to his views, especially on such topics as humanoid robots, where he probably knows more than almost anyone else. It’s also nice to see him pointing out how many hyped claims made by others in the field have failed to play out.
Interestingly, Brooks has now added quantum computing to the list of areas where he is making predictions, and while I tend to agree with his (rather modest) predictions, he is notably less sanguine than the quantum computing expert whom I trust the most: Scott Aaronson. Aaronson thinks we might be very close to breaking traditional cryptography with quantum computers.