Month: October 2013

Spending time in a village with buildings that have been around for six hundred years, I am of course charmed by the storybook quality of everything. These old stone walls and cobblestone streets feel like something out of a fairy tale, and one cannot help but be swept up in the magic of it all.

But somewhere in the back of my mind I am wondering — in another six hundred years what will people make of the buildings from my own time, buildings that today are considered modern? Will future generations walk around in such places and feel that they have entered a storybook?

Or perhaps all of our modern buildings will be gone by then, swept away by the merciless sands of time. In contrast, I suspect this sturdy old village will still be around. Future generations will wander through its cobblestone streets marveling at the quaintness and charm of twelve hundred year old buildings made of stone.

And as they wander these streets, those people may very well be thinking to themselves — in another twelve hundred years, what will people make of the buildings from my own time?

Here at the UIST conference I saw two talks today that both illustrated the same principle, in very different ways.

One was a talk about gaze tracking (sort of). It turns out to be very difficult for a computer to tell where a person is looking (which is frustrating, because this happens to be something our human brains do very well). There have been many approaches to the problem, most involving expensive and/or invasive hardware.

This paper started by redefining the problem to a simpler one: Is a person looking into the camera or not? It turns out that this much simpler question is quite solvable, with no special hardware required. The solution they showed even works with old photos or movies. And it happens to be good enough for a huge percentage of the applications for which you might think “gaze tracking” is needed.

The other paper showed a way to track someone’s hand position, using three simple and inexpensive photosensors instead of a fancy and expensive video camera. It can’t track as well as a camera. For example, you don’t get individual fingers and other finer details. But it turns out that those niceties are often not needed.

So here we have two ingenious papers, both illustrating the same principle: If you don’t know how to solve a hard problem, then solve an easier (but really useful) one.

Here in St. Andrews I had a very funny conversation with a colleague. I don’t know how much you know about haggis (you could look it up), but it’s the quintessential example of a food where you really really don’t want to know where it comes from.

I mentioned to my colleague that three years ago, when I first visited Dundee, a friend had taken me to a restaurant where they served vegan haggis. My colleague was astonished by this revelation.

Getting into the spirit of things, he asked me “Which part of the plant do they use, exactly.”

Warming up to the theme, I replied “You really don’t want to know.”

Then I pretended to hesitate, out of a sense of delicacy. “Actually, it’s the part of the plant that, well you know…” and here I feigned an embarrassed silence, a delicate understanding that to say any more would require me to cross a line of discretion that must not be crossed.

We both laughed, and then I had a thought.

“You know,” I said to him. “The way that we just had a fake conversation, because we were talking about plants?”

He nodded.

“Well,” I continued, “the way that it would have been real if we weren’t talking about plants. That’s why I’m vegan.”

He totally got it.

Walking around St. Andrews today, I saw all sorts of beautiful sights, almost too many to take in at once. I love the energy of a university built from a centuries old history and tradition, yet offering up to the minute research and education.

Strolling along Market Street I saw a small sign on the door of a beautiful old stone building: “Fluid Gravity Engineering and Applied Electromagnetics.” It was wonderful to see a sign like that a mere five minutes walk from the ruins of a thirteenth century cathedral.

I found myself wondering exactly what they studied in a place like that. What sort of innovations would deserve such an evocative appellation? I knew I could look it up on the internet, but it was much more fun to imagine.

I formed an image in my mind of a young scientist in a lab coat running excitedly out of the research laboratory. As everyone turns to look at him, he holds an odd looking electromagnet triumphantly aloft in one hand.

“It works!” the young man exclaims, as he flips on the switch, and promptly floats gently to the ceiling.

Just because your city is already filled with beautiful buildings hundreds of years old does not mean you can stop building. Populations grow and shift, technological needs change, and sooner or later you need to build. But how do you build new buildings in a way that respects and works with the old ones?

Walking around Edinburgh I saw two fundamentally different approaches to this problem. One was to create what might be called “neoclassical Scottish” architecture. These are buildings that were clearly built in the late 20th or early 21st century, and belong to the modern era, but that somehow reflect ideas from the older buildings around them. Perhaps a choice of stone color, or a general shape and form, or window placement, or a roof that hints of architectural details in the ancient rooftops around it.

The other approach is to go completely modern — glass edifices so different from the ancient structures around them that they seem to be from a different world (which, in a sense, they are). These are invariably built in a plain and unadorned style. We’re not talking about Frank Gehry here, or anything even remotely suggesting a lipstick case. Just straightforward glass boxes accented by simple dark frames.

I am not a particular fan of modern architecture. Yet I found the latter approach quite pleasing, whereas the former was rather monstrous in its effect. The glass boxes come across mainly as unassuming backdrops for the beauty of the ancient structures around them — backdrops that make no attempt to compete.

In contrast, the neoclassical approach creates a sort of frankenbuilding, something that does not have any of the rough beauty and integrity of the ancient buildings around it. These buildings bear roughly the same relationship to their forebears as a typical Hollywood RomCom bears to “Pride and Prejudice”.

There’s a lesson here, and I suspect it generalizes to all sorts of situations.

Spending a day wandering around Edinburgh was a great change of pace. Today I found myself marveling at the beautiful old buildings, and mentally contrasting the weight of history in these ancient stones with the lightness and energy of the young people darting about in their shadow.

It’s as though two species are coexisting at vastly different time scales. In many places the cityscape here changes very slowly if at all, and buildings can last for centuries. If they could observe us, perhaps we would seem very strange to them — energetic blurs flitting through life in the blink of an eye.

Today, mere hours before I am to get on an international flight, the trackpad on my MacBook suddenly has gone all haywire.

With care and patience I can get it to move to where I want (if you are reading this, then I will have scored a small victory in that direction), but it takes quite a while, and the cursor will often leap of its own accord to random places on the screen.

It occurs to me that this frustrating situation, and the immense frustration I feel in response, is a sort of electronic analog of physical dysfunction, a kind of technologically enabled Tourettes or Parkinsons.

The difference of course is that I can choose to walk away from my computer, or use another computer entirely.

Sometimes, when read a technical paper, I become curious about some work or other cited in the references. If that work is in my own field, then it is pretty easy for me to read the referenced work.

But often the cited paper will be in a field that I know next to nothing about. Even if it is the correct reference, this sort of cited paper does me little good. What is missing is some sort of graceful transition between the paper I am reading and the papers it cites.

In real life we see graceful transitions all around us: Trains arrive at platforms within train stations, oceans have beaches, cities have suburbs, houses have hallways leading to rooms. In the physical world, you rarely find yourself walking from an open field directly into a deep forest. Generally there is some sort of lightly wooded area between them, a transitional space where the density of foliage gradually increases.

Yet if I am reading a paper on computer graphics and I follow a reference, I might suddenly find myself plunged into a highly technical research discussion about biomechanics, or the physics of vorticity, or the quantum nature of light, written by and for experts in some other field. The problem here is that the cited paper has no concept of where I am coming from, and why I am reading it.

Obviously it would be impractical to manually create contextualized transitions for every cited reference. Yet perhaps tools could be developed to semi-automate the process of creating such transitional descriptions. If such a thing could be made to work, think of the benefits for scholars everywhere!

With any luck, Google is already working on it. 🙂

Usually you go through your day thinking that you are living in some sort of objective reality. Things are the way they are because that’s just the way the world is. All perfectly logical.

But every once in a while something happens that causes that comforting curtain to get pulled back just a little, and you realize that your day to day life is a social construct, a consensual illusion.

I recently got a small reminder of this. In my world — computer science — when you see the initials “ACM” you immediately think “Association for Computing Machinery”. It is, after all, the parent organization for nearly all computer science conferences, including the largest of all computer science conferences, the annual conference in computer graphics known as ACM/SIGGRAPH (which stands for Special Interest Group in Graphics, if you must know).

Earlier this week I was at a meeting at NYU that focused on a very different topic indeed — how to stop the roof from leaking in our building. After all, you’re not going to get much computer science done if your fancy computers end up getting rained on.

One of the building people, when discussing our plaster ceiling, kept talking about the ACM, and I was having a difficult time following. What conference was he referring to? Was there some algorithm involved in this roof fixing business that I had overlooked? Other people seemed to understand just fine, nodded their heads knowingly every time he said “ACM”.

I felt a little shy about announcing to everyone that I had no idea what this man was talking about. So I cheated: I opened up my notebook computer and surfed the web for “ACM”. And then, bingo!

It turns out that “ACM” also stands for “Asbestos Contaminated Material”. Supposedly all the asbestos was removed from NYU buildings years ago, but in some of the older floors there are still trace amounts, often under layers of old paint or plaster. Hence the worry about “ACM”s.

And I realized just how much this man, sitting right across the table from me, lived in a completely different world. And I was reminded just how much our reality — social, technological, existential — is a matter of context. And context is everything.

This perfect weather
May soon be gone, which only
Makes it more perfect