Ken's Blog

At the moment I am in Chicago. I’m here for a SIGGRAPH conference jury meeting, which is a lot of work but also great fun.

Last night we all went out to dinner at a very traditional Chicago pizza restaurant. I thought that I was going to be out of luck, being vegan.

But that didn’t seem to be a problem at all. The waiter had them cook for me a really delicious vegetable Chicago deep-dish pizza.

Being from New York City, I think of Chicago deep-dish pizza as something exotic and almost otherworldly (come to think of it, it does look sort of like a UFO). So I was very happy with this option.

The meal turned out to be delicious and very satisfying. Before eating it, I snapped a photo with my SmartPhone, which I sent to a friend.

My friend, looking at the photo, was quite taken aback. She said that in her opinion, all deep-dish pizzas are unnaturally giant and oddly intimidating.

I replied that the phrase “unnaturally giant and oddly intimidating” seemed like a great description of Chicago architecture. We agreed that it’s probably a Chicago thing.

Not that there’s anything wrong with that.

For the last few days I’ve been binging on Scott Walker’s songs and videos, from his early teen idol phase through his Jacques Brel / Mort Shuman period and on through his radical later reinventions creating music that can perhaps be best described as “Bowie meets Stockhausen”.

I can’t even think of anyone else to compare to Scott Walker. He was truly unique, absolutely fearless in his approach to musical exploration, and impossible to categorize, with a voice that was simply sublime.

Sadly, Scott passed away earlier this week. But I am sure that his incomparable music will live on, to shine its radiant light upon the world for centuries to come.

Sometimes the solution to a mathematical problem seems so obvious that you don’t think you should even bother writing it down. I often make that mistake.

But every once in a while I think about something that seems obvious on the face of it, and I take the trouble to write it down. And then suddenly I realize I’d had it all wrong.

I discover that the thing that had seemed to consist of just one simple step actually consists of three different steps. And each of those steps can be unpacked to contain sub-problems that are all interesting in their own right.

Before I know it, I’m generalizing the problem. I’m connecting it to other problems I’ve seen, and getting a whole new set of insights.

That’s why I try, on my better days, not to take mathematical problems for granted. And also to remember to carry a pen.

This evening was the third time I have seen Lanford Wilson’s play Burn This on Broadway. The first time was the original production with John Malkovich in the leading role of Pale.

The second time, that role was played by Ed Norton. This evening, the same role was played by Adam Driver.

It amazes me how different these productions are. Each time the same character, each time the same words, yet each time an entirely different outcome.

It is great to have the luxury to compare. And I must say, having now seen the interpretations by Norton and Driver, both of whom are excellent stage actors, I am far more appreciative of the original.

Even across the memory of all these years, I can still feel the elemental power and subtlety of Malkovich’s performance. His portrayal digs deep into a character whom it is all too easy to render as a caricature.

It’s a shame there is no good way to capture for posterity a live theater performance, while keeping all of its subtlety and sense of presence intact. Fortunately, we’re working on that. 🙂

In a rare move, I took some days off this past week. It was a revelation.

Not only did I have a wonderful time, but I also — to my great surprise — came up with a lot of new ideas for research. Although I suppose it shouldn’t have been a surprise.

After all, when you stop filling your head with day to day problems, your mind becomes more free. New ideas that pop into your head are embraced, rather than dismissed as a distraction.

This raises a question for people who do research. I wonder what would be the ideal ratio R of days spent “doing research” to days spent “taking a break from doing research”.

If the goal is to maximize the number of good research ideas, then wouldn’t it then make sense to spend less time just trying to get things done? If the ratio R is 2 to 1, that would mean two days working for every one day not working.

That sounds plausible, but wouldn’t it be funny if it turns out we have it all wrong? Maybe the optimum value for R will turn out to be something more like ½.

There are numerous examples throughout history of somebody creating something we end up labeling as “art”, which were never intended to be experienced by other people. One canonical example of this is the diary of Anne Frank.

Another famous example is found in the many manuscripts of Franz Kafka, most of which he instructed his friend and literary executor Max Brod to burn upon Kafka’s death. Fortunately for us, Brod did not burn anything.

Which leaves us with a question: If somebody never intends a work — be it painting, sculpture, poem, song, novel or something else — to be experienced by others, is it still art?

Or does art require that most basic contract between creator and audience: That there be an intended audience?

My own take is that we need a different word for such works. The word “art” doesn’t not quite encompass the antisocial provenance of such creations, because the creation of art is fundamentally an act of intentional communication.

But what would be a good word?

Well, tonight I finally got to the end of Murder on the Orient Express. Such an amazing movie.

And sure enough, it ended exactly the same way it ended the first time I saw it.

I wonder whether movies will always do that. Or will there come a day when a movie can have a different ending every time you watch it?

With all of the new digital technologies coming down the pike, you never know.

When I was six years old, and my brother Mark was eight, our family was staying in a summer cottage up in Mountaindale New York. One afternoon our mom handed us each an empty Maxwell House coffee can, and our dad took us out into the woods where wild blueberries grew.

We were each supposed to fill our coffee can with blueberries. Which was not so difficult a task, since blueberries grew wild and plentiful in those woods.

At the end of the afternoon my brother’s coffee can was filled to the brim with fresh juicy blueberries. My can was pretty much empty.

That evening our mother took all of the blueberries and made a blueberry pie. My brother enjoyed it very much. Alas, I had a tummy ache, and didn’t feel hungry at all.

I think there’s a lesson in here somewhere.

This evening I started watching Murder on the Orient Express on Amazon Prime. I am a big fan of director Sidney Lumet, so this was an extra special treat.

The old saying is really true: They don’t make movies like they used to. This is a classic Hollywood film in the grand style, combining a vast and sweeping canvas with an unerring attention to the finest detail.

Albert Finney’s performance is, all by itself, a priceless cinematic treasure. And the other performances aren’t too shabby either. Just seeing Lauren Bacall, Ingrid Bergman and Vanessa Redgrave playing a scene together is worth the price of admission.

Alas, the film is more than two hours long, and today has been a long day. So tomorrow I shall watch the rest: Murder on the Orient Express, part 2.

Psst: Don’t tell me how it ends. 😉

Let’s back down half a dimension from yesterday’s post and talk about 3.5D. What, you may well ask, is 3.5D? Actually it’s a term I made up, but it refers to something very specific.

This evening I went out to the movies to see one of the big science fiction flicks currently on offer. It was filled with great computer graphics, and thrilling action scenes.

I had made the deliberate choice to see it in 2D, even though I could have seen it in 3D. The reason? 3D in movies, the way that term is currently defined, is highly problematic.

In real life, when you are sitting across a table from somebody you’re talking to, and you move your head — even a little — the entire room behind your conversant appears to move. This is simply motion parallax: When you move your head from side to side, nearby things appear to shift by a greater amount than far away things.

But in a 3D movie nearby and far away things move together when you move your head. In fact, they always move together.

In my case, this takes me clear out of the movie. All I can think about in such moments are the deficiencies of today’s stereo 3D movies, and those thoughts take away from my enjoyment of the film.

In contrast, the collective VR technique our lab is developing for immersive cinema solves this problem. Just as in real life, when you move your head from side to side, near things appear to shift more than far away things do.

I call that 3.5D. And I’m sticking with the name until somebody comes up with a better name.