Ken’s Blog

I had a far ranging discussion today with some students about the future possibilities of virtual reality. The discussion went to some surprising places.

One of the tropes of VR is that everything has the possibility to seem incredibly real, because of the sense complete immersion. But one of the more interesting topics in our conversation this evening centered around just the opposite.

When you go to see a movie, or a play, you are not expecting to see a literal transcription of life. In fact, if a filmmaker or dramaturge were to attempt to show you such a think, you’d probably be very bored indeed.

Rather, each medium, as it has matured, has developed its own particular set of stylizations, uniquely suited to that medium. In any developing medium, it takes a while to work out what kinds of innovations really work, and which do not (talkies: good; color: not bad; smellovision: meh).

All of which is to say that when VR finally becomes a mature storytelling medium, I suspect that its successful innovations won’t be those that mimic literal reality. They will be those that deviate from that reality in interesting and powerful ways.

This morning, in Dublin, I was conversing with a woman over breakfast. She had the loveliest Irish accent, and she posed me a provocative challenge, which she had apparently heard on a radio show. It seems that your answer to this challenge reveals a lot about what is currently going on in your life. Here is what she said:

“If you were walking along a road in a forest, and you were to come upon a bear, answer the following two questions: (1) Describe the bear, and (2) When you find yourself faced with the bear, what do you do?”

My immediate response was “I don’t know. What’s the answer?”

She would have none of this. “Oh no, you can’t get out of it that easily. You need to answer the questions.” I was so entranced by her lilting Irish accent that I accepted the challenge.

I thought about it for a bit, and then I gave my response.

“In answer to the first question,” I said, “it’s a teddy bear. In answer to the second question, I would pick it up and give it as a gift to a child.”

I was worried that she would think that I had somehow cheated. On the contrary, she was completely delighted by my reply. “It’s clear,” she said, “that you are leading an untroubled life.”

I was very happy to hear that.

Late last night I had a very inspiring conversation with a colleague, after which I excitedly wrote down a note to myself. I was sure this would be a wonderful subject for a blog post.

When I looked at the note sometime later, I saw that it simply said “Bartender == Film Editor”. That’s it, just three words and an equals sign.

I’ve tried to reconstruct in my mind what this could possibly have meant. I vaguely recall that it had something to do with the way a good bartender can seem to read peoples’ minds. A talented and perceptive bartender can tell whether you are happy or blue, and knows at a glance whether you are in need of a stiff drink or just someone to talk to.

Presumably, at the time, I saw how all of this tied together with the art of film editing. But I must admit, the connection now eludes me entirely.

And so, alas, the brilliant and pithy hypothetical blog post I might have written on the subject is not to be. I shall not be expounding today on the subtle yet powerful connection between the talents of bartenders and the talents of film editors.

Yet all is not lost. For it seems that I have indeed found a topic for today’s post.

It occurred to me today that it would be an interesting programming challenge to generate alien speech. I mean the kind of speech you hear spoken by extraterrestrials in science fiction films.

You wouldn’t need to generate actual functioning speech, just the feeling of it. In other words, when an ET opens up its mouth (or functionally equivalent bodily organ) to speak, what comes out conveys to the audience not only the sense of intelligent speech, but also some sort of coherent culture.

This effect was created very well, of course, for Klingon. But in that case, an actual language was designed. I’m suggesting something much less than that, but in a way also more. Our program could simply generate faux speech, as was done for The Sims. But it would need to be plausible and self-consistent faux speech.

For example, we may have an alien race with particular characteristics. Say, tall and thin, argumentative yet pacifist, aerially winged, living in a low gravity planet with a thin atmosphere, etc. We should be able to dial those characteristics into our speech generation software.

Then the software can proceed to generate plausible speech for such a creature. I am envisioning a process somewhat akin to generation of procedural texture — something we now see in movies all the time.

The alien speech generator should handle prosody, an impression of vocabulary, a kind of sound that suggests a particular biological equivalent to throat, breath and vocal cords, and generally something that maps recognizably into nuanced emotion.

If nothing else, this sounds like a really fun programming project.

I am sitting in Penn Station, NY, NY, waiting for a train that will take me to the airport. And I am realizing that there is something about Penn Station which is both fascinating and disturbing.

Nobody is here because they want to be — everyone here is just passing through on their way to somewhere else. So there is an odd sort of energy, a kind of pervasive restless feeling.

It’s different from, say, Grand Central Station, which is an actual destination in its own right. Over there, people stop and look around, draw in a deep breath, spend some time taking in the magnificent vaulted ceiling, luxuriate in the sense of spacious grandeur.

Here it hasn’t been that way since 1963. Once one of the most magnificent buildings in the world, this place is now cramped and soulless, a place nobody wants to be.

But it’s a place that lots of people need to pass through, while waiting to get to somewhere else. So if nothing else, it’s a really awesome place for people watching.

I was telling a colleague yesterday about the oddest coincidence. You see, I have lots of friends who have had experience with contractors. Maybe they needed their roof fixed, or their kitchen redone. I also have some other friends who are contractors, who have had various experiences with clients.

As you probably know, before a contractor starts to work, a contract is written up. It says exactly what work will be done, with an estimate of how much it will cost. Just so there will be no surprises.

Yet it happens that every person I know who has worked with a contractor has horror stories to tell. The contractor didn’t actually do the work, or did the wrong work, or ended up needing to fix something that shouldn’t have needed fixing, and everything ended up costing vastly more than the estimate.

It also happens that every person I know who has worked as a contractor has their own horror stories to tell. The customer changed their minds in the middle, or complained about some phantom problem with the work, or refused to pay what they had agreed to pay.

So here is the amazing coincidence: Every person I know who has worked with a contractor turns out to be an honest person who was taken in by an unscrupulous contractor. Yet every person I know who is a contractor turns out to be an honest person who was taken in by an unscrupulous customer.

I ask you, what are the odds?

If we lost the Web,
Twitter, Facebook, would we learn
About the old ways?

Once you start playing with 4D puzzles, it’s difficult to stop. They are highly addicting.

After thinking about the 2x2x2x2 puzzle for a while, I started wondering how many ways there were to build a kind of 4D Soma puzzle. That is, little pieces all fitting together perfectly to create the 2x2x2x2 hypercube.

There are endless variations on such a puzzle, so I narrowed it down a bit. Consider that there are a total of 16 little hypercubes in a 2x2x2x2 hypercube. Suppose we only consider puzzles where each “piece” consists of exactly 4 little hypercubes.

It will take four such pieces to build the 2x2x2x2 hypercube (since 4×4 = 16). Suppose we restrict things even further: All four of those pieces need to be the same shape.

How many different 4D puzzles can we make, if we follow those rules? More tomorrow.

In some 4D world
Someone’s solving a puzzle
In five dimensions

Now that you’ve seen what it looks like to rotate a 2×2 block inside a three dimensional cube, let’s see what happens when we rotate a 2×2 block inside a four dimensional hypercube. It turns out that there are at least six different ways you can rotate the same 2×2 block within a 2x2x2x2 hypercube:

0 1 2 3 4 5 6 7 8 9 a b c d e f	0 1 2 3 4 5 6 7 8 9 a b c d e f	0 1 2 3 4 5 6 7 8 9 a b c d e f
0 1 2 3 4 5 6 7 8 9 a b c d e f	0 1 2 3 4 5 6 7 8 9 a b c d e f	0 1 2 3 4 5 6 7 8 9 a b c d e f

All six of the above cases represent the same 2×2 shape embedded in a 2x2x2x2 hypercube. The only difference is how they are rotated differently (in four dimensions, of course). The good news is that all those rotations provide lots of opportunities to make really interesting puzzles.