Wednesday, June 20, 2007

I see vacuous people....

Yoo-hoo
Oh, vacouous yoo
vacuously troo
Yoo knew;
trooth over survival
beauty over pain


You may have noticed the new graphic header for the blog.
It's a set of deformations of Solomonoff's Algorithmic Probability equation (Sholarpedia, MIT Lecture). ALP in essence takes the weighted sum of all the hypotheses about a set of observations. It has been shown that ALP learns to correctly predict any computable sequence with only the absolute minimum amount of data.
You can get it straight from the horse's mouth here, although, be forewarned, Ray is struggling with a missing front tooth which makes it hard for him to enunciate.

Something you probably don't know, is that Ray Solomonoff was in Steven Wright's short film, "One Soldier"!
In the film, there is a scene where Steven Wright's civil war character meets himself from 40 years in the future. His future self tells him not to worry about 40 years from now, he will will have been dead for 37 of them anyway. The man playing the 80 year old Wright is Ray Solomonoff. I saw his name in the credits and went back and checked, and sure enough it's the guy in the picture. As far as I know, no one else has noticed this. How deadpan is that?

So, vacuous you; have some fun with Ray and his beautiful and profound ideas about reality.





Tuesday, June 12, 2007

Word Jazz and SETI

In another life, as they say, before the invisible cities, before the invisible wars, my girlfriend and I got hooked on Ken Nordine's Word Jazz, which was on NPR. It was so profound and metaphysical, but with a jazz attitude. Very cool. After a while, it seemed like his shows had some sort of clue about our lives. We couldn't wait to get home from work on the nights his show was on and see what he had in store for us. It seemed like the them of our lives for that week was the them of the show! Sometimes they were very specific things, like something happening on a bus on his show that was weird but almost identical to something that happened to us. Yeah, I know, tropical fish look like they're dancing to the music, Ramsey theory says you always find patterns, etc. But, I've never known anyone who actually thought the fish were dancing, it just looks like they're dancing. Yes, we always find patterns, but some are more meaningful. For some reason we think it's more than a coincidence that when we read a book, we generally learn, or that in conversations, what the other person says, usually fits with what you were saying, (more searching for entailment).

The difference between superstition and science is, the amount of scrutiny the conclusions get. So, yes, I could think Ken was communicating clues to my girlfriend and I because of natural co-occurrence, Ramsey theory, Dada, mind as meaning maker, whatever. But once we ask ourselves, "Is it possible?" - things change.
Let's look at some of the ways this could occur.
  1. The message is deliberate, strictly personal and is equivalent to one given by an omniscient agent that knows the significant facts of who you are. This happens with no one else.
  2. Just like 1, except everyone gets a personal important message. In essence, KN is Santa Clause, knows every kid and what they need to do get the best Christmas ever. Everyone that listens will hear their personal important message.
  3. Just like 2, except ...

We see the pattern.
Basically we have {always personal and specific} {deliberate, not deliberate}, {only for you, for others{grouped, ungrouped}, for all others{grouped, ungrouped}}
Specific means, you might listening to the show and he would be talking about how greed brings grief with it, and you just realized that you had been greedy and hurt other people by putting yourself first.
So we want to see if there is a feasible interpretation over these possibilities.

Here's what I think: We live in a fractal reality, so everything is essentially the same thing. But there are "ways" of being the same thing, and some of them are death and some are life, some are ugly and some are beautiful. Everything true is personal and specific because the specifics are just the local distortion of the truth fractal. Actually, to be specific by attribution is so much less personal than by "speaking truth" if we are indeed truth incarnate. I'm pretty comfortable with that metaphysics.

So, what's this got to do with SETI. They have the same problem. Is this signal varying because it knows something we recognize, or is it merely the transcription of some deep truth. Is it intelligent? Does it care?

Ken Nordine speaks the cool truth, like Miles Davis did. I eventually moved away and never found it on radio again.
After all these years, I've never forgotten him or the excitement and mystery of listening to his shows and seeing how they seemed to fit with our daily lives. I ordered a couple of his Cd's and that was about it other than periodically surfing for him on the web. Well, he has a web site (of course) - and I just went there for the first time - there's no place like home, ah Oz....let's go see what he has to say tonight, shall we?


word jazz

Monday, June 11, 2007

Hell: a non-entailment problem?


"Hell is the impossibility of reason." - Oliver Stone's Platoon


I have been busy trying to find the right rectifiers for my nervous system and at the same time working pretty hard on getting Eclipse EMF to do some basic things, like represent XMI metamodels as ontologies and lesser model types like UML and Ecore . But it's like pulling teeth, even with all the fancy stuff. So, I have not been able to give you, p(t), anything new. I know you are currently the empty set, but that's ok, I don't discriminate. But what is time to the empty set and, to some future onlooker, what tree fell in a forest here before you arrived?

hELL
Why have I needed to put new "industrial strength" rectifiers between my nervous system and the real world? Well, there's reading Sartre's Nausea and there's being in it. Kind of like "Total Recall" with an existential twist.

eNTAILMENT
But this all has to do with entailment. If we postulate something like a valid culture (identity, perspective, value system, etc.), then shouldn't global (system wide) ethics entail every valid culture. Before globalization, tribal "micro-theories" could co-exist and since the interaction between tribes was minimal, tolerance (entailment of the other culture), was not an issue. Increased population, and advances in transportation and communications, make the tribal approach untenable. Tribalism was fine as a precursor to the current dynamic, but it wastes too much information to handle any real complexity. Global ethics is defined institutionally but has not been personalized by very many people.
Is it OK to stone your wife because she committed adultery? If we say yes for some cultures, is it because they are too crude to understand that this is cruelty, that other cultures have abandoned this practice and lived through it? We should advance a simple upper ontology of basic good that has two simple constraints:


  1. Don't hurt anyone
  2. Help as much as you can

Now of course, this entails quite a bit. You have to at least be honest enough with yourself to be able to tell if you are really doing those two things or not. Well, how do you know if you're being honest with yourself? That leads to more answers and as e.e. cummings used to say:


"Always the beautiful answer who asks a more beautiful question."