Month: March 2016

I am watching the Netflix series Halt and Catch Fire about computer entrepreneurs in the 1980s. I very much appreciate the fact that the heroes are mostly computer programmers or hardware hackers.

The technology is all absolutely spot-on. Every detail, no matter how arcane or nerdy, is completely correct and chronologically accurate. Clearly somebody on the writing or advisory team was actually there.

But what really intrigues me is that feeling of heady possibility, of creating an astonishing future that you know is just around the corner. It’s exactly what being in computer graphics felt like to me when I was just starting out.

And it’s exactly what it feels like now.

I participated yesterday in a workshop filled with extremely smart and exceptional people. In general the entire experience was wonderful and inspiring, and I learned a lot from everyone. But there was one odd moment.

One the talks, you see, involved a bit of back and forth. From time to time the speaker would show something on the screen and solicit a response from the room. At one point showed an image of the Mona Lisa sporting a mustache. Next to this he showed a photograph of a man’s face. He then asked “Who is the man in the photograph?”

I shouted out the obvious answer, expecting that a chorus of us would give the same answer: “Marcel Duchamp!” Yet in that entire room, only one other person spoke up. I realized then that nobody else knew about Duchamp’s iconic work L.H.O.O.Q.. Either that, or they had suddenly all become strangely shy.

I’m certainly no art historian, and my knowledge of 20th Century art has huge gaps. But it seems to me that some things, like iconic works by pioneering artists, should be part of the general knowledge base of our populace. Yet clearly it is not, which tells me that something is screwy with the way education works in this country.

OK, maybe this isn’t the most important problem with our education system. After all, our high schools also manage to carefully avoid teaching mathematics, or even letting kids know how amazingly creative and fun math is. Instead they mostly teach a sequence of rote exercises and formulae that they mislabel as “mathematics”. Believe it or not, in most parts of this country you can get all the way through high school without ever learning the beauty of Euclid’s proof of the infinity of prime numbers.

So maybe in a way our education system is indeed teaching absurdism to our children. Except instead of painting a silly mustache on Leonardo da Vinci paintings, they are painting a silly mustache on rational thought itself. I wonder if many kids get the joke.

I saw a wonderful talk today about machine learning. Most of the time when people talk about machine learning they deal in abstractions. They write down some math, they wave their hands, they mutter vaguely about neural networks, and in general they say things that are completely mysterious to most of the populace.

But the talk today, by Chris Olah, was anything but mysterious. He pretty much laid it out for us, in terms that anybody could understand.

Essentially, machine learning algorithms are like Silly Putty. They take the space of all of the variables that go into whatever an algorithm is trying to recognize, and they stretch and distort that space in all sorts of interesting ways.

After all that distortion, whatever it is the algorithm is supposed to recognize ends up on one side of some plane, and everything else ends up on the other side. For example, if the machine learning algorithm is trying to recognize pictures with dogs in them, then after all the Silly Putty distortion, all the pictures containing dogs will end up on one side of the plane, and all of the pictures without dogs will end up on the other side.

Then it’s just a matter of using a mathematical knife to cut through that plane. On one side will be all the dog pictures, on the other side will be the non-dog pictures.

And that my friends, in a nutshell, is what machine learning is all about. I had no idea, until today, that Silly Putty could be so useful.

All in a circle
We talk about the future
And so — create it

Today I went to a large event at the MIT Media Lab in honor of Marvin Minsky. This was very different from the much smaller event I went to shortly after he passed away, which was just for family and a few friends.

Many wonderful things were said, and I took notes. Looking at those notes now, one if my favorites is from Pat Winston, who summed up Marvin’s contribution to A.I. like this: “Alan Turing told us we could make computers intelligent, and Marvin Minsky told us how to do it.”

Another is from Brian Silverman. He was explaining how when Marvin was working with Seymour Papert on developing programming languages for kids in the early 1970s, and there was no computer that could do what Seymour needed, Marvin just designed a new kind of computer and built it himself.

I particularly like the way Brian said it: “Research required a particular thing. If that thing didn’t exist, Marvin just invented it.”

They also passed out fortune cookies at the event, each with a quote from Marvin. At the end of the evening I saw somebody carrying out a large bag of left-over fortune cookies. “Careful,” I told her, “if you eat too many of those at once, everything will start to make sense.”

Humans are very good at picking up on subtle non-verbal cues. We can generally tell when somebody is nervous, or excited, or joyful or confused, without needing to hear a single word.

There is, reasonably enough, much worry about whether these sorts of important interpersonal cues will be preserved when people are having extended conversations in a shared virtual world. But I am not worried. In fact, quite the opposite.

When you and I speak on the phone, we don’t feel that our inability to see each other visually destroys our ability to communicate. Instead, we both understand quite well that the only channel we have is voice, so we pay more attention to vocal cues. When communicating with each other, people are very good at sussing out where the good quality information is, and focusing their attention accordingly.

I think something similar will happen for face to face communication in virtual worlds. At first, the body cues will be a strict subset of those in real life. We will be able to see each others’ head movements, and then perhaps hand movements, but it will take a little while longer to transmit all of the subtleties of full body motion.

At every step of this evolution, we will instinctively know where the quality information is coming from, because that’s what people are good at. Once we are used to any particular mode of future face to face commmunication, we won’t think of it as odd, or off-putting, any more than we currently think that way about talking to each other on the phone.

But what if it’s even better than that? What if it turns out that our brains are more evolved and capable at supporting face to face communication than our bodies are?

If that is so, then there might come a point when computationally enhanced body language actually lets us convey and apprehend subtle cues of body language that are not possible in physical reality. When that happens, we might find that body language and facial expression, suitably enhanced by computer intermediation, will allow us to communicate with each other more effectively than was ever before possible in the history of humanity.

After a generation or two of living with such advanced support for non-verbal communication, people might wonder how the human race ever got along without it.

Today at our lab we held part of our weekly production meeting in VR. We were all in the same physical room, and we could have seen each other in person, but we opted to put on wireless headsets, using our Holojam technology, and hold our conversation “in world”.

Because we could move around the room freely, it felt as though we had all been transported together to that alternate world. We could continue to talk with each other, but while inhabiting alien bodies on another planet.

This was just an early experiment. We still haven’t added enough things to do on that planet to make it a place we would prefer to spend a lot of time hanging out together. But we are going to keep adding. Every week, our alternate meeting room on a friendly alien world is going to become an ever more interesting place to hang out.

Still, to quote the last line of a great movie, there’s no place like home.

Happy Ides of March! 🙂

Today I was interviewed about a virtual reality dance performance we recently did. The interview questions were really good, and I was enjoying the process. And then something unexpected happened.

I was trying to explain the idea of giving a dance performance in virtual reality. So I said “It’s like a non-fictional version of the Holodeck.”

At this last statement, the interviewer looked confused. “What,” he asked, “is the Holodeck?”

I realized, with a mix of astonishment and sneaking delight, that he really didn’t know. So I explained about Star Trek the Next Generation, and that first episode in 1987 when the world was introduced to this wondrous fictional plot device.

Yes, the Holodeck is a fabulous (in every sense of that word) literary invention. But there is also something fabulous, and oddly refreshing, about meeting somebody who has never heard of it.

Happy PI day, everybody! 🙂

I had an on-camera conversation recently with a filmmaker who is putting together a documentary about how evolving technologies might, in the long term, change the human condition. He was particularly interested in the sorts of conjectural views that center around everybody uploading their minds into a computer.

As the discussion evolved, it became pretty clear that from his perspective, I represent the relatively staid voice of old-fashioned humanism. In fact, I am not really interested in those sorts of questions, and I consider them a distraction.

Of course I understand why people are so fascinated by these possibilities. We all have a strong instinct to want to not die, and the idea of uploading yourself into the Cloud, and thereby becoming some sort of immortal being, can be very compelling.

When I told him that I wasn’t really interested in those sorts of conjectures, he was surprised. “Wouldn’t you want to live forever,” he asked, “if you had the chance?”

I told thim that I thought he may be missing the point. Sure, if you lived forever there would be many things to reconsider. But that’s not the important question, is it?

“What’s the important question?” he asked.

The most essential quality of the self, I suggested, is not our mortality, but rather our uniqueness.

“If you were to meet another you,” I asked him, “or ten thousand other versions of you, which one is you? For example, if for some reason only one of you could survive, which of you should get to live?”

He didn’t seem to have an answer for that.

Today, because I needed an interactive animated avatar for our latest social VR research, I decided to repurpose a character I had created years ago. That was back when I first started programming for the Web in Java.

As soon as I saw the newly reconstituted character (now translated into Javascript), I realized that I had created the original exactly twenty years ago. Strangely, it did not feel at all as though twenty years had passed. In fact, it felt like just a moment.

Looking at the character, and the code I had written to model and animate it, my mind slipped back to the exact moment of first creation, and into the exact thoughts I had had in that moment. The feeling was as though I had put down my tools for just a second, and then picked them right up again after a brief pause.

I wonder whether there is a separate category of time, which might be called “artistic time”. In the world around us, years might elapse, nations may rise and fall, and entire generation of children can be born and grow up to become young women and men.

But when we are in artistic time, all of those things can feel like a distant dream. For within that creative place inside us, we know that it was all just a moment.