Saturday, January 05, 2008

Night Run in the Rain

I was at home watching My Big Fat Greek Wedding and lifting weights (I bought myself a weight set for Christmas) when the power cut out because of the storm we're having. Not to be discouraged, I figured it was a reasonable time (midnight) to go for a night run. It was fun! If it rained like this all the time in SoCal I might start running seriously again. Of course, I choose to live in SoCal preceisly b/c it does not rain like this all the time. Life is all about pros and cons. It seems the ankle injury I sustained last time I ran (a week ago) has healed itself.

Tuesday, December 18, 2007

Metaverse 0.0

When Linden's Second Life first came out I had some hope it was the first step toward a real Metaverse ala Neal Stephenson. However, further reflection convinced me unless Second Life changes its business model it will never morph into the Metaverse. The problem is that while Second Life's "browser" is open source its server code is proprietary. Linden is the only entity that can host a Second Life world.

In the world I envision I can create a little world on my own computer and invite one, two, or many friends to join me in it; up to some number of people (10?) peer to peer networking is all that's necessary to keep data flowing and sync events. Beyond that a central server is probably necessary, but there's no need for the server protocol to be proprietary. VRML (X3D) was a good step in the right direction.

I'm going to take a look at 3B (a 3D browser) and the Croquet Project (wiki it) to see if they do some of the things I imagine a Metaverse doing.

Gonna be a long week...

I have to put in ten hours at the UCLA zoo-archaeology lab this week, while still putting in 40+ hours for Raytheon. I stayed at work until midnight to get ahead on hours and prepare for my 10 am briefing tomorrow morning. The briefing's to some quite important mucky-mucks, but it's going to be an informal affair. I don't need polished slides. I just need to explain my ideas well. I think most people would be stressed by the week I'm about to have, but I'm used to it and, really, I'm just not the stressing type guy. Plus, I get to look forward to a whole week of paid vacation next week :-D.

Last weekend I started work on a Settlers of Catan web guide. I already reserved the domain www.catanguide.com. There's nothing there yet. I'll post when there's real content.

Monday, December 10, 2007

Google AdSense

Just because I can, I added Google AdSense to my blog. I tried to keep it unobtrusive. What do you think the first ad on my page was? It was for the Honda Accord hybrid. This is a major affront to my diginity, as I'd just finished explaining in my last post why hybrids are bad for the environment. At least I know they're trying to target my page with relevant ads.

Why Hybrid Cars are bad for the Environment

Hybrid cars are bad for the environment? How can that be? Let's say you are an equal opportunist enviornmentalist: that is, you are equally happy to clean up an African lake and thus provide drinking water to a poor village as you are to plant a tree to scrub carbon from the atmosphere. When you buy a hybrid car you pay more to transport yourself around than if you got an efficient conventional car. The money you would save by driving a conventional car can instead be used to clean up the African lake. The free market sets the price of a good based on the labor and resources necessary to produce the good. If one good does the same job as another, but does it cheaper, it frees resources and labor to be used in other endeavors, such as cleaning up the African lake.

Now, there's one objection that can be raised to this analysis. The environmental cost of pollution may not be accurately reflected in the price of gasoline (I happen to think it is). Basically, when you chose to buy a gallon of gasoline you are happy to pay, say, $3.50 for it. However, when you burn the gasoline it smudges your neighbors upholstery. He has to pay $0.05 to clean up his property every time you burn a gallon of gasoline. Since you don't bear the cost of the clean-up, you won't use the true price of the gasoline, $3.55. This is an example of a negative externality. The right way to handle this problem is to, as best you can, predict the price of the pollution and increase the cost of gasoline by that amount through taxation or other means, such as carbon trading. If a specific person or entity bears the cost of pollution they should be compensated directly. Now, I personally am not too worried about global warming, and thus I believe the current price of gasoline fairly accurately reflects the true price of production, transportation, and pollution. Allowing that I can make a strong claim that driving hybrid cars is bad for the environment.

Thursday, August 03, 2006

On Intelligence

I've been thinking a lot recently about how our brain works. Naturally I have more questions than answers, but the most compelling thing I've learned recently is that the cerebral cortex (the seat of most higher level functions, such as language and vision recognition) is very flexible. In fact, current research shows quite strongly that the brain matter governing one sense -- hearing, say -- is identical to that governing another sense -- vision, say. The difference in make-up and density of neurons is entirely a result of differing sensory inputs rather than speciliazation for the given task at hand. For example, researchers have rewired ferret brains so that the inputs from the optic nerve are go to where auditory senses normally enter (note that in humans, at least, optical input passes first through the thalamus before reaching the first layer of visual cerebral cortex, V1). I know, I know -- poor ferrets. But the ferrets see just fine . The cerebral cortex is just a giant, flexible pattern recognition "machine"! Notably, higher level intra-sense associative areas don't really care what the underlying sense are that they're correlating. They simply correlate!

OK, so ferrets cerebral cortexes are general pattern recognition machines. Is this true of humans, too. Yes! In one interesting study scientists placed a device on blind peoples' tongues that received inputs from a crude camera and turned the camera input into a grid of pressures on the tongue that indicated intensity of light by the amount of pressure. In a relatively short time subjects were able to play rock, paper, scissors, reach for doorknobs, and etc. Why? Because the asociative areas [see mosquito, feel mosquito, oh it's the same mosquito] of the brain don't care where the inputs come from. They just try to match patterns. This is true all up and down the cortical hierarchy.

Key point: many people, sadly, think of the brain in terms of functional areas. For example, scientists have identified certain areas of the brain (in the visual cortex) that trigger when we see images of movie stars (as I recall they used Angelina Jolie, among others), and news articles frequently report when new functional areas of the brain (language, vision, hunger) are mapped. This is because brain scans are very high resolution but not dynamic. That is, they give great spatial detail but terrible time sequence detail. This is unfortunate. Time delays and correlation are a (the?) key component of sensory association and higher level thought. Functional location is unimportant, as evidenced by the poor lobotomized ferrets.