Science, Technology & Health: October 2003 Archives

I posted a bunch of links to space elevator information many months ago, and I'm quite a fan of the idea. SDB tries to rain on my parade, but I've read some papers that address his main concern.

So you got your elevator crawling up the kevlar ribbon, and like the elevator in the building, it is not only forcing itself and its passengers "up", but also "sideways". To reach the classic geosynchronous orbital altitude, you not only have to pick up a lot of altitude, but also several kilometers per second of lateral velocity. ...

The lateral force on the lower end of the ribbon is a design problem for those who design the anchorage, but it's not really too bad. They can deal with it. The anchorage will be designed to transmit that force into the earth itself, where it can be ignored.

Unfortunately, the lateral force this applies to the counterweight is a pernicious problem which is not so easily solved. Remember our weight hanging from a string? Try pushing slightly on the string. The weight starts swinging, right? That's what's going to happen here.

Yes, but there are ways to deal with that problem. I'm not a rocket scientist, but I play one on TV, and my buddy Robert Cassanova (sounds like a soap opera name, doesn't it?) assures me that the physics is a-ok.

"Technically it's feasible," said Robert Cassanova, director of the NASA Institute for Advanced Concepts. "There's nothing wrong with the physics."

In fact, the biggest problem is most peoples' minds isn't balancing the counter-weight, it's building the cable, as SDB mentions near the end of his post.

Which is all well and good, except that I've given it a lot of thought, and I can't see how you could even get one ribbon connected. I can't figure out any way you can actually build this system.

You can build an anchorage. You can put the counterweight into space. How do you connect the ribbon between them?

In the next article I'll explain why that's a tough problem.

I look forward to reading his analysis, but I've seen plans that sound plausible to me. He may shred these ideas tomorrow, but I may as well tell you what I've seen so far.

The cable won't be made from kevlar; the material-of-choice will be carbon-fiber nanotubes. The cable won't be strung (or hung?) all at once, but will be laid incrementally.

Getting the first space elevator off the ground, factually, would use two space shuttle flights. Twenty tons of cable and reel would be kicked up to geosynchronous altitude by an upper stage motor. The cable is then snaked to Earth and attached to an ocean-based anchor station, situated within the equatorial Pacific. That platform would be similar to the structure used for the Sea Launch expendable rocket program.

Once secure, a platform-based free-electron laser system is used to beam energy to photocell-laden "climbers". These are automated devices that ride the initial ribbon skyward. Each climber adds more and more ribbon to the first, thereby increasing the cable's overall strength. Some two-and-a-half years later, and using nearly 300 climbers, a first space elevator capable of supporting over 20-tons (20,000-kilograms) is ready for service.

"If budget estimates are correct, we could do it for under $10 billion. The first cable could launch multi-ton payloads every 3 days. Cargo hoisted by laser-powered climbers, be it fragile payloads such as radio dishes, complex planetary probes, solar power satellites, or human-carrying modules could be dropped off in geosynchronous orbit in a week's travel time," Edwards said.

Using a laser beam to boost the climbers into space is doable, said Harold Bennett, president of Bennett Optical Research, Inc. of Ridgecrest, California. "If you do it right, you can take out 96 percent of the effect of the atmosphere on the laser beam through adaptive optics," he said. The strength of the pulsed laser beam is less than the intensity of the Sun, so birds, airplanes, or human eyes wouldn't be affected, he said.

A lot of great (kooky?) minds have worked on this concept, and so I'm hopeful that SDB's skepticism is misplaced. Wikipedia has more.

Over the decades, the United States has developed dozens of different types of nuclear warheads, for many different uses. Some are dropped as bombs, some are fitted to inter-continental ballistic missiles, some are designed for artillery shells, and some are even used on air-to-air rockets. For a (complete?) listing, check out "Designations Of U.S. Nuclear Weapons" and "UNITED STATES: Nuclear warheads and applications".

If anyone is interested, here are some satellite photos of the California wildfires. All links go to photos with resolutions of 1 kilometer per pixel, but from those pages there are links to photos of 4x and 16x the resolution. (The 250m/pixel pictures are spectacular!) The fires and smoke are clearly visible; very neat.

- 10/28/03 -- True color.
- 10/27/03 -- True color.
- 10/26/03 -- False color.
- 10/26/03 -- True color.
- 10/25/03 -- True color.
- 10/23/03 -- True color, 500m/pixel.

Note all the smoke pouring west into Los Angeles proper, ugh.

Update:
Now Drudge has up satellite photos too! Copycat.

SDB writes a bit about using Bayesian networks to fight spam. I use the same tool for filtering my email that he does -- POPFile -- and it does work incredibly well. Furthermore, you may have noticed that I installed an anti-spam filter for the comments section here that is also built around a Bayesian network. It doesn't work perfectly yet, and some comments are hidden (not lost) due to the software mistakenly thinking they're spam, but it's getting better.

Go read more about Bayesian networks, and rejoice in one of the first truly widespread applications of artificial intelligence in everyday life.

I just installed an anti-spam filter for the comments on my site. It was written by James Seng, and uses a Bayesian network to detect probable spam -- just like many popular email filters. We'll see how it works. I don't get a lot of spam here, so it may take some time to train it. Meanwhile, I won't have it actually filter anything, so no comments will be lost.

No, I cannot resist adding artificial intelligence to everything, whenever possible.

Here's a really spiffy government-run National Atlas of the United States. I love maps, and I could spend all day on this.

Partial-birth abortions will be illegal, as soon as Bush signs the bill that the Senate passed 64-34 yesterday (the House passed it a month ago). In theory, this law will prevent up to 5,000 abortions of convenience each year (since, as Bill Hobbs notes, the AMA says the procedure is almost never medically necessary; Donald Sensing says that physicians have testified for years that the procedure is never medically necessary).

Considering that I view abortions of convenience as murder, I would have preferred if the federal government had stayed out of it and left it to the states (which generally prosecute murderers), but my affection for federalism is outweighed by the thousands of lives that will potentially be saved. Furthermore, many similar state laws have been struck down:

The measure is similar to, but somewhat more detailed than, a Nebraska state law that the Supreme Court struck down by a 5-4 vote three years ago. That ruling had the practical effect of nullifying 30 state laws. Up to that time, Congress had been trying unsuccessfully for five years to enact a similar proposal at the federal level.

My lamentable Senator, Barbara Boxer, has this to say:

California Democratic Sen. Barbara Boxer, who helped lead opposition to the bill, called it "a very sad day for the women of America."

But, of course, she's only considering women who are already out of the womb, and has little concern for the women who will now not have their brains sucked out by vacuums and their bodies dismembered.

Along the same lines,

But an abortion rights supporter said the ban "will bring an end to providing the best and safest health care for women."

It will bring an end to the mass-murder of thousands of children. Physicians all seem to agree that this procedure was never medically necessary, so it certainly can't be required for the best and safest health care for anyone. Congress concurs:

In drafting the new national measure that has now passed, Congress wrote lengthy findings that contradict the Supreme Court's conclusion that abortions using the procedure banned by the bill are sometimes medically necessary to protect a woman's health. "Congress finds that partial-birth abortion is never medically indicated to preserve the health of the mother," the bill's preamble says.

The problem is that the abortion-rights people don't seem to understand that they're arguing a different point than most people are conerned with.

"This dangerous ban prevents women, in consultation with their families and trusted doctors, from making decisions about their own health," said Gloria Feldt, president of the Planned Parenthood Federation of America.

Most Americans don't see abortion as a private issue that only affects the mother, no matter how much abortion-rights advocates want to spin it that way. They purposefully misstate the pro-life position, which is that an unborn baby is a human being, and that as such the medical privacy concerns of the mother are inconsequential compared to the right of that other human being to live. As I wrote in that earlier post:

2. Michelman states that the position of pro-life advocates is that the government should be involved in people's private medical decisions, when that isn't the crux of the matter at all. To an opponent of abortion, the critical issue is that a fetus is a human being, and as such should not be killed without a cause more substantial than mere convenience. It has nothing to do with a lack of respect for the privacy of the mother, or with a desire to interfere with her private medical decisions. To a pro-lifer, the decision to have an abortion isn't private, because it necessarily involves another person: the unborn baby.

For a really excellent scientific explanation of why unborn babies (from conception) are alive, and are "real" human beings (without any reference to religion), I highly recommend "Life: Defining the Beginning from the End".

And finally, "Who, after all, could consider a fetus as life unworthy of living, once they've held its hand?", asks Sydney Smith, a family physician, and author of MedPundit.

Yay, I finally got some nifty PNG graphic libraries to compile into the Quake 2 source code, and now I can generate PNG images as output files for my PhD dissertation project. The text maps I was using previously were a real pain to read, and the new color-coded pictures look really cool.

Want to see? Well, ok! The image is pretty small (32x32); each color represents one of 4 tribes that can control a sector of territory, and its resources; a black sector is not controlled by anyone.

This image was taken from the very early stages of a simulation run, and the control patterns are basically random. However, after the tribes have some time to learn and grow, I hope to see territorial boundaries emerge around food concentrations, just as would happen with tribes of real humans. The tribes will learn to use simple language constructs to coordinate their members' activities (such as attack, defense, and mating), and will then compete for resources in the simulated world.

Time for bed!

It took the FBI 5 weeks to find some box cutters that a man hid aboard two Southwest Airlines jets (link removed), despite an email he sent to federal authorities alerting them to his actions. Aside from the obvious security concerns, this brings to my mind the really nifty concept of attractors and strange attractors, and is an excellent example of an attractor in real life. What's an attractor?

The basic idea behind an attractor is that a dynamic system will tend toward certain states as time goes on. The simplest form of an attractor is the point attractor. Consider a normal pendulum, it doesn't matter where you release it from, it will always come to rest in the same position, perpindicular to the ground. This state is the attractor for the system.

From another story, it appears that the box cutters were simply hidden in a compartment in the airplane bathrooms. We've all seen the compartments, I imagine; there are several panels in those bathrooms that all look removable. But we've never opened them, and have probably not even wondered what they're for. The man told the authorities exactly which planes to search, but it still took them 5 weeks to find the knives because they were hidden outside the areas the searchers were attracted to, for whatever reason.

Strange attractors are like normal attractors, except that they're chaotic. Chaotic systems never revisit a point they've been to previously, so you may wonder how such a system could have an attractor at all -- well, it can't, but it can have a strange attractor. A chaotic system may never return to an exact previous position, but it can go to a position that's similar to a previous position (and much depends on how you define "similar", which in turn depends on the system in question).

Human behavior is (arguably) chaotic -- it's immensely dependent on initial conditions, and tiny changes in our inputs can yield drastically different outputs. If I happen to get a piece of dust in my eye, I may twitch, stub my toe, decide not to walk to lunch, and avoid getting hit by a car. Nevertheless, humans often behave in ways that are very similar to their past behavior, and we are often quite predictable. Our patterns, movements, thoughts, and life can be seen together as a giant strange attractor that represents the most likely state of our being and that describes our operational progression through time.

Consider your movement patterns through your house. There are probably several areas where you spend the vast majority of your time -- such as the bed, the couch, the bathroom, the computer -- and the rest of your house may be rather sparsely visited. How often do you peer into the crawlspace under the floor, or go up into the attic? How often does the crevice behind the fridge see the light of day? How often do you open that cabinet over the oven? Maybe once a year, or maybe less than that.

If you were to draw a map of your house and trace your movements over the course of a month, you'd probably see that 75% of the floorspace was completely untouched, and that 95% of the volume enclosed by your house did not ever contain a human being. We look at the corners of our rooms from time to time, but we never go up into them. You see the ceiling every day, but when was the last time you touched it? Even when we lose something and we say we've looked everywhere, we know that's just a turn of phrase. We haven't looked under the carpet, or behind the shelves, or inside the TV. But we shouldn't have to, because our car keys are not going to be inside the TV -- that location isn't a high-frequency part of the strange attractor that represents the movement of our keys.

Similarly, thousands of people rode the Southwest jets over the 5 weeks the box cutters were hidden in the bathrooms, but no one found them because no one ever opens those compartments. They're 6 inches from your head when you wash your hands, but a million miles away conceptually.

Similar strange attractors can be found in almost everything, if you want to search them out. Consider the various ideologies that divide humanity, and that 90% of people believe in one of maybe a half-dozen religious systems. There are all sorts of reasons, but the system is so complicated and chaotic that it's impossible to fully describe. In all likelihood, no two people hold exactly identical religious beliefs, but the vast majority are similar enough that they can be easily clustered into just a few buckets.

Strange attractors are everywhere, and by recognizing and studying the attractors that describe our own behavior we can get a better understanding of how we are, and why we are.

Drudge points to a nifty report on battlefield laser weapons, and the report briefly mentions a very interesting question:

Are military computers and commanders ready for entirely automated weapons that deliver instant, lethal blasts of energy and can be retargeted in seconds? Lasers under testing for air defense already offer that capability. Fully automated firing on offensive targets is a short step behind.

"When you develop the capability to track, target and destroy something in a second, then the temptation to remove humans from the decision cycle becomes very great," said Loren Thompson, chief operating officer at the Lexington Institute, an Arlington, Va.-based defense think tank.

Our emerging use of Unmanned Aerial Vehicles controlled largely by computers brings up this same dilemma: how comfortable are we taking decisions out of the hands of humans, and putting them in the hands of computers? If we don't do it now, once our enemies start to let ultra-fast computers run their battles we may not have any choice.

One intermediate step may be the introduction of true neural control technology. Once the sensors and controls of a remote vehicle can be piped directly into a human's brain, the time required to execute operational decisions will be drastically reduced. A human will never be fast enough to target a laser weapon on an incoming artillery shell (a defensive use), but for offensive uses it might be possible to keep a human in the loop. Considering that laser weapons are line-of-sight, we're likely to have enough strategic warning of their use that we'll be able to plan ahead to counter them. However, once space-based lasers are in play that can destroy major targets (like buildings or city blocks), we'll probably be forced to turn all of our military systems over to tactical and strategic computer control.

I am very interested in the repetition of history, especially in the repetition of ideas. Two of my favorites are brought up every so often, one by the Right and one by the Left.
The first is the concept of the statistical bell curve as applied to human populations, normally in terms of IQ or some other measurement. Its most recent occurence was in the mid-1990's, reflected in books such as "The Bell Curve", and used to justify things like cuts in social spending, often with reasoning such as "they're poor because they are dumb, they are beyond help".
The second is typically a tool of the left, or of other environmentalists, and is sometimes referred to as simply the carrying capacity of the earth. A recent bestseller on this subject is "The Population Bomb". The concept here is typically doom and gloom - the earth is running out of stuff (food, ninja turtles, whatever) or there are too many people or both, and we're all gonna die for our arrogance.

I'm going to talk about the second one today, partly because I think my liberal leanings are wearing out my welcome on this blog, and partly because it's one of my all time favorite topics.

In 1798 (kind of before 1997), Thomas Malthus wrote his "on the Principle of Population". It is arguable if his ideas were new at the time (a Venetian named Ortes is often credited with preceding Malthus), but they have undoubtedly been repeated since. He postulated that population grows, typically exponentialy. But agricultural production grows linearly. So, you draw a snazzy graph, and the exponential curve eventually crosses the agricultural line and, well, people eat their feet or something. He obviously had a lot more supporting information, but this is the argument at it's heart.

The obvious problem with his analysis is he completely ignores any exponential gains that can be made in agricultural production through technological innovation. Amazingly, Ehrlich does the same in "The Population Bomb". His argument is essentially the same. The thing that worrys me about all this is how often (and in what large numbers) people buy into these re-hashed or warmed over failed ideas. Heck, I worry myself about what bizarro truths I've taken hook, line and sinker; maybe the Fed really is run and owned by an international banking conspiracy!

There is one final aspect of the concepts created by Malthus and rewritten by Ehrlich that I'd like to touch on. Assuming the Universe, or at least the matter and energy in it, are finite, the idea of a limited carrying capacity is essentially true, and there really can be too many people. Unfortunately, people like Malthus or Ehrlich seem to always predict we'll hit the limit in the next week or so. What is that limit? If we can't move to other planets, how many people can really live here? Will we cram them in all underground? Eat a lot of Soylent green? Should we even take living conditions into account, is that even important? I bet if we all could photosynthesize we could cram a lot of people just on the surface.

Update:
Ehrlich even wagered on his predictions, and lost.

I have a hard time respecting environmentalists, but it's not because I want to pave the earth. I love clean air and water, fishies, owls, and all that sort of thing. What I don't love is environmental-mysticism masquerading as science. The Bay Institute, a Marin Country-based organization "Dedicated to the preservation and restoration of the San Francisco Bay Watershed, from the Sierra to the Sea" has recently "graded" the San Francisco Bay on various ecological criteria and found it wanting -- a result surprising to no-one, I suspect.

The San Francisco Bay is getting an ecological report card today -- and it doesn't look good. But there's hope behind the C's, D's and an F, according to the nonprofit environmental group that graded the bay's health. ...

The bay's grades -- one B, three C's, three D's and an F -- were based on historical conditions, environmental and public health standards, and restoration targets, according to the institute.

Ah, yes... why do I get the feeling that the Bay Institute itself determined these "restoration targets" based on their own agenda, and that these targets were in fact the driving force behind many of the bad grades? The Bay Institute itself admits that the bay is doing well in some areas:

"The destruction of San Francisco Bay's unique environment has in some cases been halted or even slightly reversed," said Grant Davis, Bay Institute executive director. ...

"Fish and wildlife populations that were crashing now appear to be stable. Many people are working to protect and restore habitat, improve water quality and use resources more efficiently. But progress is slow and needs to be accelerated," Davis said.

Why is that? Because otherwise their "restoration targets" cannot be reached!

So what's causing the problems? People want to use the fresh water that historically flowed into the bay for other purposes, like growing food.

Tina Swanson, fish biologist and member of the science team working with the Bay Institute, said decades of diversions of Sacramento and San Joaquin river hwater to growers and cities have taken a heavy toll on native aquatic species in the bay.

Historically, during rain and snowmelt, the fresh river waters have flowed into the bay and out the Golden Gate. The rush of the rivers creates a special mixing zone where the freshwater hits the salty ocean water, providing an important nursery for many estuary species, scientists say.

"The bay suffers from a permanent drought because so much water is diverted from its watershed," said Swanson. The species are not recovering from the steep decline they experienced over the last several decades, she said.

And so, some species of fish are becoming rare in the bay -- or are disappearing altogether -- because we humans need to use fresh water to grow food. Well, there's not much that can be done about that, is there? Water isn't free, and growers try to keep costs down by not using more than they need to. Water management in California is taken very seriously, and is a complex issue that I don't claim to understand perfectly, but I don't think that a great deal of water is being wasted.

So what does the Bay Institute hope to accomplish with this report card? I don't know, because they don't say; all I can infer is that they don't want us to grow so much food, and that they don't think there should be so many darn humans all over the place.

In a frightening trend, monkeys are learning to directly control robots with their minds. Seriously though, this is one of the coolest things I've ever read.

Monkeys with brain implants were trained to move a robot arm with their thoughts, a key advance by researchers who hope one day to allow paralyzed people to perform similar tasks.

A series of electrodes containing tiny wires were implanted about a millimeter deep into the brains of two monkeys. A computer then recorded signals produced by the monkeys' brains as they manipulated a joystick controlling the robotic arm in exchange for a reward -- sips of juice.

The joystick was later unplugged and the arm, which was in a separate room, was controlled directly by the brain signals coming from the implants. The monkeys eventually stopped using the joystick, as if they knew their brains were controlling the robot arm, Duke University researcher Miguel Nicolelis said. ...

"It really opens the possibilities, and it reduces the amount of time. Previously, I had thought it might be five to 10 years before we could apply this to humans. I'm getting more optimistic now, I think in a couple of years we may be doing the real clinical trials."

The implants remained in the Duke monkeys for 2 years showing they can be used for extended period. Over time, the monkeys' brains adapted to treat the robotic arm as if it was their own limb, Nicolelis said.

Thus, it is a virtual certainty that I personally will someday control some sort of robot [army] with my thoughts alone.

Although the methodology looks questionable to me, Harris Interactive conducted an "online" survey of teens and asked them what they thought about file sharing.

ROCHESTER, N.Y., Oct. 9 /PRNewswire/ -- Results of a new Harris Interactive® survey show that two-thirds (66%) of American teenagers (13-18 years old) oppose fining individuals who offer copyrighted music online for other people to download while about one in ten teens (13%) believe that people who offer copyrighted music on their computers for others to download should be fined. Half of teens (52%) strongly oppose such fines and two in ten teens (21%) neither support nor oppose the fines. ...

In addition, the poll found that most teens believe that sharing and downloading of copyrighted music should be legal. Three quarters (78%) of them feel that sharing (letting other people download music from them) should be legal. Additionally, 74% of teens said that downloading copyrighted music files from the Internet without paying for it should be legal.

Why am I skeptical of the results? Well, online surveys tend to be bogus, since the respondents are self-selected (only people who are interested tend to answer polls they come across online), but near the end of the article it says:

This Harris Interactive survey was conducted online within the United States between September 17 and 22, 2003 among a nationwide cross section of 642 respondents aged 13-18 years old. Figures for age, sex, race/ethnicity, education, urbanicity and region were weighted where necessary to bring them into line with their actual proportions in the population.

I don't know what this means; if the sample isn't self-selected, and the survey just happened to be done online with an actual random sampling of teenagers, then maybe the results are ok. It's not clear, however.

But, if the results are meaningful, then the modern concept of copyright is doomed, because these kids will be making policy in 20 years. I'm not saying this is good or bad, but it seems inevitable to me.

(Thanks MD.)

I hate linking to the BBC, but I can't find the story anywhere else.

So, China is expected to launch its first manned space flight within days. Thanks for the pointer, Matt.

There aren't many details, but in my opinion this is a momentous occasion. The space race between America and the USSR was driven by politics, and even though it's long over many of the technological breakthroughs of the past decades are the direct results of our massive space program. The USSR can't even maintain its committments to the International Space Station now, and with no real rival the United States' vision for space has stagnated.

China's entry into the space-faring community is political as well -- the space club is even more exclusive than the nuclear club, and China is aiming for the moon. Such programs capture the minds of the world, and attract scientists from all over who want to get involved in cutting-edge research. The immediate space projects will stay secretive for a while, but the peripheral work may be open to all comers; if China is smart, it will leverage the prestige of this endeavor counter America's dominance as the world's brain-drain.

What's more, I can only hope that if America's space superiority is seriously challenged we'll rise to the occasion and kick our butts back into gear. I want to go to Luna and Mars myself, and NASA sure as heck isn't going to get me there unless I stow away on board a rover. I have no doubt that private industry can out-compete China's communist party.

[You may also be interested in contributing to the X Prize (it's tax deductible).]

StrategyPage points to a page listing some excellent games and game mods created/paid for by the US military for training purposes. Some are available for free download, and the others all appear to be available for purchase. Note: these are all unclassified; there are plenty more classified games, particularly dealing with submarine warfare and the strategic use of nuclear weapons.

This is the kind of work I hope to get involved with once I finish my PhD in artificial intelligence.