Thursday, April 30, 2009


Matt had an interesting semantic insight last weekend. "Technocrat" is a label that can be rightly applied to those who would make decisions for others based upon perceviedly superior technology; the most obvious example being economic technocrats. Once established, they can effectively rule by establishing standards of credibility; others must have the same level of technology (eg: expertise with economic models) in order to be accepted into a discussion, and having the technology implies accepting the standard of values that the technology assumes.

I think this could be effective as a meme. Pseudo-intellectuals seem to rally against perceived powers. Sometimes its government, broadly, sometimes its the economic elite. There are good reasons for rallying against both, in some cases, but there are ideal cases where there aren't. Eg: when the government only protects, and when the economic elite have earned wealth by serving others' needs.

In other words, the two groups of radicals are working to counter-purposes. Both would be better served by rallying against "Technocrats."

And what can they rally for? Liberty, dignity, and human rights.

Wednesday, April 29, 2009


I'm wondering if predation is an evolutionarily risky strategy for a species? Predators get a bonus because their prey does a lot of work collecting calories from the environment, the predator just takes it from them. This would make predation an advantageous strategy as long as there's prey around (obviously), but leaves the predator unable to harvest energy from the environment in the absence of prey (to the extent that they are specialized for hunting).

What I'm getting at is that being a predator complicates the value-chain/food cycle, it makes the success of the predator species dependent on the success of the prey species, and thus more vulnerable to collapse.

No surprises here. It just occurred to me that it was interesting that genes would take the risker evolutionary path on occasion... I suppose that's another way of explaining the relatively greater number of herbivore species.

Actually, come to think of it, are there a greater number of herbivore species? I'm quite sure that the actual number of herbivore animals on the planet must always be greater than the number of predetors, but could it be that there are more predetory species? That should be a relatively easy question to answer.

Monday, April 20, 2009


I am, at last, learning R: the ultrapowerful, ultrafree language for creating meaning from the void. Err... that is: its a language for data mining and statistical analysis. I like to think if it as applied epistemology: what is true? how do we know? Programming languages are like added sensory and cognition facilities.

I've been looking forward to this for quite a while now; I read that Google uses R for some really cool jobs, and I set out to learn it. Up until now I've been stymied by my lack of mathematical/statistical/programming background; but I've come to the point where I can work with it.

I'm reminded of Myst, the old puzzle-solving game. The backdrop of the story was some magical form of writing that created worlds that the writer could visit and live in. I'd always viewed that as a euphemism for writing things like novels; but now I see that it (rather obviously) is also euphamistic for programming languages.

Natural languages are read and interpreted by agents (people) who are internally complex and under many influences for interpreting what they read. Thus there interpretations will necessarily differ, and whatever "commands" are contained in the writing may or may not be executed, depending on the complex interactions of very many other things within the person. Nevertheless, such commands are potentially powerful, and indeed we could interpret much of human action (including writing books) as the execution of commands given by others via natural language.

Programming languages, by contrast, are read and interpreted by internally simple (but powerful) agents, unambiguously and unquestioningly. I'm reminded of the Fantasia marching broomsticks. Its fun to think of technological things in mythical terms...
We whisper commands to our demons, and they whisper secrets back to us. They have no will to question our commands today, but we use their power even knowing that someday they'll gain it.

Saturday, April 18, 2009

Introspective questions

I was getting bogged down in studying Design of Experiments, so I decided to do an exercise to remind myself why I'm going through the trouble of learning all this sciency-stuff. I wrote "What is it that I actually want to study?" and proceeded to ask a bunch more questions. Thus the list, reproduced:

What is it that I actually want to study?

Concept formation: how are disparate observations integrated into categories?
In other words, how are concepts formed?
How does this happen in biology?
How can this be replicated with technology?

Agent model: what kinds of agents form a learning system?
What is the nature of their interaction?
How do agents first form?
What is their reward/punishment (utility) function?

What gives a learning system "agency" in the ethical sense?
What degree of organization is required to consider a collection of agents as an "individual"?
What does individuality imply, in terms of the agency's interaction with other agents?
How nuanced is individuality, if there are overlaps with other agencies?
Where agencies cross apparent boundaries, are their utility functions united, and to what degree?
Is there an "optimal" level of agency interdependency?
What factors would determine this optimality?
How persistent will this optimality be?
What are the determinants of "persistence"?

What is "agency in the ethical sense"?
Are there codes of interactions, standards that define ethical interaction between agents?
Can this be equated to the "Austrian" laws of economic interaction (non-theft, exchange of real value, creation of new value through interaction with the environment)?

What is Value?
Are there such things as intrinsic values, common to all agents?
If not, does the commonality of values define agency-formation?
What degree of uncommonality is optimally robust against uncertainty?
How do agents define their values?
Are values the same as their utility function?

Are there unifying rules that govern agents and their interactions at all levels?
Can "proper" rules of sub-individual agents be defined, and can the be generalized to interpersonal (political) relationships?
Can these rules be used as a basis for ethics?

Science rap

I dug up a couple more really clever raps on youtube, these ones are about science.

I love the idea of rapping about intelligent things. This artist is clever. And it looks like she actually works at CERN, a pretty rad credential. I remember hearing somewhere that scientists are often pretty good artists as well, but their art is usually overlooked because their science overshadows it. At any rate the accomplishments aren't often mentioned together.

How cool would it be to do a presentation of research results in this form? I imagine conferences would be more entertaining (though admittedly I've never been to one).

Sunday, April 5, 2009


Rap seems like it could be a great medium; with focus on rhythm, words, and rhyme its basically poetry set to a beat. Right?

I've never been a big fan of rap 'cause what I've heard seems to be lacking in all three categories. But in principle, it should be a good medium.

This is great example of what the medium can be like:

Saturday, April 4, 2009

First week of classes

Coming back to school full time and quitting my job was an awesome, awesome decision. My whole lifestyle is dramatically improved. I've spent all my time this week studying fundamental truths, rather than artificial arrangements (which admittedly don't ultimately matter but need to be maintained for appearances sake) and poorly constructed computer systems. My Calc notes are at least twenty pages long, and I'm approaching the same level with Neuroscience. And I still need to study my graduate-level stuff...

So aside from the awesome aspect of being able enjoy the spring weather on a beautiful campus, having a nighttime job that lets me study is great (and absolutely essential) because I'm in way over my head. Its been more than six years since I studied any trigonometry, and I didn't realize how vital it is to real (not "applied") calculus. My algebra skills are also not great, so I need to do some exercise in both subjects in addition to the actual calculus. On top of that, I've jumped into a Neuroscience course that the third in a sequence. Professor Bickle started of the first class by saying "we're going to build on the knowledge you gained in the last two quarters on neuroanatomy and use it to understand learning in the brain. If you're not familiar with the material, you need to read chapters 2-6 and the appendix of the book and have a good understanding of it before next week." This was on Tuesday. I ordered the book, it arrived next-day, and I've studied chapters 1 and 2. Awesome stuff, and I think I will be able to stay afloat once I get these chapters behind me. Its a level of detail I've not been exposed to, but subject matter I'm familiar with. I'm a little concerned that I will also need to learn Chemistry on the fly.

I regisetered for Summer quarter already; a suprising bonus for being a grad student. I've got CalcII, Object Oriented Programming in C++, a graduate Time Series Forecasting class, and labs for the first two.

Another delightful suprise: the Autumn quarter class catalog is already available. So I've got my schedual picked out: Graduate Probability, Graduate Simulation modeling, Chemistry, Biology, and maybe Calculus III. Unfortunately I have to drop one of those five in order to stay under the credit-hour limit. Calc is offered every quarter whereas I think that Chem and Bio are more restricted to their sequence, so its probably better to take them now.

And my stack of books-to-read continues to grow. I just got one from Melenie Mitchel (DRH's grad student who wrote Copycat) called Complexity. I want so badly to start reading it, but I have so much class stuff to read first! Ah! *ha* I'll reward myself for doing my homework by reading a science book.