Saturday, December 20, 2003

Lessons of Libya: "[Bush] was right"

That's something you will almost never hear from the New York Times. Bush's decision to not let up sanctions until Libya admitted and gave up WMD programs (though it admitted governmental involvement with Pan Am flight 103 to satisfy the UN) was right.

Calling a spade a spade pays off. I just hope we don't forever dance around Chinese suppression of Taiwan's independence & democracy, or N. Korea's blatant threats to its neighbors.

(0) comments

Friday, December 19, 2003

Libya to give up weapons programs

Watch Bush's address regarding Libya's commitment to dismantle any weapons programs & invite UN inspectors into the country.
Two important motivations:
1) Countries which do not act in accordance to the peace and prosperity of the world will be punished. Iraq is learning this the hard way.
2) Countries which cooperate with the world authority will be rewarded.

I am certain Libya will benefit greatly from this with trade deals, movements for internal reform, and greater stability.

This is a very clear example of how America need not invade every country that stands against its security interests. In fact, I think it could be argued rather convincingly that only there needs to be only one demonstration of our willingness to enforce our demands, before other countries will act without one.

This might seem like America is taking a very aggressive stance. It IS, but the positive point here is that we are aggressively defending peace, prosperity, and freedom for the world, rather than simply dominating it. It is no coincidence that the countries that threaten us most with WMD development, also suffer economically even compared to other countries in the region, in addition to having autocratic oppressive governments.

It has been argued (notably by Thomas Friedman), that we are entering an age where there are two sides in a divided world: the world of order verses the world of disorder. We need to wake up from history and move everyone in the world to this world of order. This is the process of globalization, and it is inevitable.

The only question how many people will die in the resistance to such forces.

(0) comments
CNN.com - Chinese Authorities Return Biological Weapons To Civilian

He plans to use them in an operation called "Red Moon"

(0) comments

Wednesday, December 10, 2003

Media on Media on Politics

From this article:Ted Koppel, Anchor Provocateur (washingtonpost.com), you can read about the planning of ABC News & Ted Koppel on how to stir up debate and get better coverage of the Democratic presidential candidates.

It seems rather odd to me that I can read in a newspaper a story about the machinations of television journalism. The layering onion is unmistakable, and it all seems a bit in-bred.

I remember lamenting that there should be an objective source of information that gives a description of any politician's views, actions, and how both change over time. Then I remembered that this is the supposed function of the media, which clearly fails miserably because no source is objective, and all coverage is filtered.

This is one reason why the Internet is a important and blogs especially, if they manage to achieve a critical mass. With enough 1st person account of events, Big Media becomes less needed (e.g. imbedded journalists during combat in Iraq from dozens of news agencies are still not as interesting as the accounts of the soldiers themselves and their many blogs). With enough analysts taking on these events in a meta-level, the official views of any one source can be subjective, and the system still works.

One final ingredient would have to be the involvement of the electorate on a far more massive scale. I suppose this is a general problem with democracy, but hopefully the supply of information and its exponential growth rate can spur demand.

(0) comments

Monday, December 08, 2003

Flypaper ::

Andrew Sullivan has an interesting take on a more global strategy for the war on terror. Rather than viewing the insurgents in Iraq as an unfortunate counter force to forces of democratization and stabilization, view it as an opportunity.

Every foreign fighter that comes to Iraq is fighting on our terms, against our best, instead of civilians fighting & fleeing in New York or London.

The article quotes U.S. Army Gen. Ricardo Sanchez, commander of U.S. ground forces in Iraq::

[Sanchez] just opined on CNN that attacks against U.S. forces have increased in "sophistication, especially in the improvised explosive devices that they are using, and we're working to learn from that and to be able to counter them." He went on, critically: "This is what I would call a terrorist magnet, where America, being present here in Iraq, creates a target of opportunity... But this is exactly where we want to fight them. ...This will prevent the American people from having to go through their attacks back in the United States." You won't find a better description of the "flytrap" strategy anywhere - or from a more authoritative source.

This article is linked to in a more recent post that makes good use of Sun Tzu in describing why the recent flight of combatants from Afganistan to Iraq is good for the U.S.::

If [your enemy] is in superior strength, evade him.
Sending al Qaeda foot soldiers to attack mechanized infantry divisions is not an economical use of their rather limited resources.

Attack him where he is unprepared, appear where you are not expected.
We are expecting terrorist attacks in Iraq, and prepare for them constantly.

You can be sure of succeeding in your attacks if you only attack places which are undefended.
The World Trade Center was undefended. The US military in Iraq is a hard target.

The clever combatant imposes his will on the enemy, but does not allow the enemy's will to be imposed on him.
Iraq is a far better place for the Coalition to fight al Qaeda than Afghanistan. We have tremendous military strength in country; the terrain is favorable for modern mechanized warfare; a substantial portion of the population is friendly to us and hostile to terrorists; and our intelligence capabilities are growing stronger every day.

(0) comments

Friday, December 05, 2003

Steel Tariffs Taken Down

The economist perspective (you might need a subscription, email me for one)

It contains the best use of a latin phrase I've seen in a while:

All of this [steel industry became strong in US] has happened since the tariffs were introduced, but did any of it happen because they were introduced? Gary Clyde Hufbauer of the Institute for International Economics (IIE), a Washington think-tank, warns against the post hoc ergo propter hoc fallacy

Any restriction on free trade (including subsidies from foreign governments to aid their own steel industries) should be eliminated, with the exception of provisions to ensure domestic law isn't violated (such as sweat-shops spinning silk in India).
I see no reason why commercial interests should usurp labor laws. This is an issue with NAFTA and other free trade organizations that still needs to be resolved. In the case of Mexico, regardless of how your Cathy-Lee sweater was made, we are obligated by law to import it without restriction. Because what goes into a product is often hidden from the final consumer, there must be some other level of enforcement, rather than letting supply/demand take care of it (as I would usually suggest).

(0) comments
Bush in Iraq::

Interesting blog with an eye-witness account of bush's visit over thanksgiving...

it's pretty crazy that he would go to iraq now...but those troops must really have loved it.
(0) comments

Thursday, December 04, 2003

Human Visual Memory Representation

So I just had a meeting with my advisor, and there were some fairly interesting ideas discussed.

One goal in robotics is to be able to use vision as a tool in navigation, recognition and a number of other areas. Clearly they are related: "I've seen this place before, I know where I am". But algorithmically, things get very hairy, very fast. For instance this guy uses features extracted from a test image, measures parameters of each feature, and matches those feature parameters to training images, where the same feature extraction was used. He can then successfully detect an object in a scene that has been seem before in training data.

But the problem becomes harder when you are dealing with navigation, because the number of features seen grows so large. I mean, human vision operates at around 60Hz, and lets say we find 100 features in each image (very low estimate). From the time you woke up just today, you have seen ~129,600,000 features, with plenty of overlap. Now you find an arbitrary seen. Have you seen it before? What if it was 30 years ago?

I am amazed we even function! How can I possibly remember places, and I mean literally the contours of the walls and shape of the environment, that I haven't seen for so long? And as far as I know, it is still a mystery how humans represent visual imagery. Is it holographic, like Gibson suggests in 'Neuromancer', 3-D, 2-D, or 10^10-D ? Or everything but on different scales.

My advisor told a story: when his daughter was 3, he stopped at a phone-booth when they were lost. A year later, they were driving and his daughter says, "we've been on this street before, you stopped to make a phone call right THERE" and points to where the phone booth USED to be , but was removed after some construction or something. They hadn't been back to that point since. She must have taken in enough information from the scene to identify it without actually seeing the primary item, the phone booth. Simply amazing.

The computational capacity is quite awe inspiring. I am certain it is related to the power of forgetting combined with our focus of attention. Right now, as a computer, it is cheap never to forget any sensor data, and you wouldn't even know what to throw away...
(0) comments

Wednesday, December 03, 2003

Back to the moon!? ::

When President Bush delivers a speech recognizing the centenary of heavier-than-air-powered flight December 17, it is expected that he will proffer a bold vision of renewed space flight, with at its center a return to the moon, perhaps even establishment of a permanent presence there.

This is excellent news if true. It would be a boom for the robotics industry and many others.

For those who think it isn't a cost effective effort, the main reason to go into space is to get effectively unlimited resources (both mineral from asteroids & energy from the sun) without polluting the earth. This is from now to 100 years from now. On the order of 1000s of years, the motivation changes, but the mandate is equally clear: to ensure that some fool with a virus, a lucky asteroid, or the nuclear powers don't whipe out the human race, we MUST eventually go into space. It is the only worthy insurance.

(0) comments
Personal Robot Future Form Factor

An interesting article I found on robotic trends.

It highlights the major difference between the Japanese and American approach to personal robots [i.e. robots used for entertainment & service].

While the Japanese might prefer to have humanoid robots because they are reluctant to accept other people to perform certain tasks, Americans are much more ambivalent, even pragmatic. Â?Here robots must be cheaper than other solutions before we can accept that it can do some work,Â? Lewis suggests. Â?The exceptions are search and rescue and military robots, where robots can potentially save human lives.Â?

It is important to realize the costs involved here. A humanoid robot will undoubtedly be more expensive than, say, the Roomba, if only because of the great mechanical and processing power needed to make a biped run (err ... or even to make it walk).

But when you have a humanoid, they should be able to do many things you can do, because everything in your home is designed for a human. Think toilet bowl cleaners, upright-vacuums, dusters, handles, door knobs, stairs, etc. Joseph Engelberger said the same thing when he visited CMU to inaugurate the new Robot Hall of Fame

I think the main issue is market penetration. Humanoid robots probably can't be affordable until they are made en mass. Maybe some rental scheme would be sufficient, where you pay on the order of $500 a month for a robotic butler/jester/rent-a-cop, and constantly receive upgrades of new editions. Then the older generations could be rented out for less, and the oldest generations could be sold at a low price for academic experimentation or battle-bots. I'm thinking of a generation as 6 months to a year for new computer hardware and/or mechanical body, and monthly/weekly for software.

This might seem expensive, but it is like leasing an expensive car. The difference with this purchase, is that this is so much more than a status symbol: your house is secure and clean, and there is an entertainment robot for the whole family that is much more immersive than a television. Imagine the mulitpurpose functionality: you call your robot to the kitchen to bring up a recipe from the internet on its LCD-screen enhanced chest, and then the robot goes back to your toddler to play teletubbies.

All of this is extrapolation based on both the current cost (Asimo runs around $100,000) and performance (Asimo needs to be remotely controlled), and couldn't be achieved in less than 10 years.

UPDATE: Reader Abby comments:
" but englerbert was pointing out that there are many aspects of a human that we *do not* need to replicate for various functions - ie, walking and running for a nurse robot, for example. So, again the focus should be on saving human lives (or making their lives better) not on just obsession with mimicry or creation in our own form. The priority should be on FUNCTION not FORM. "

I should have specified that i was referring to his statement that robots should be able to use our current tools.

I think he is wrong about not needing legs. Sure, maybe in a nursing home, but think about an apartment with a stoop. Would you install those lifters on the stairs outside, as he suggests? Of course not! It is a needless roadblock to the full immersion of robots into the home.

Also, you should know as an pro in HCI that the form follows function when you are interacting with a social/entertainment robot, i.e. the most natural interaction will most likely come from something that looks most like a human.

(0) comments

Tuesday, December 02, 2003

War Watch - November 23, 2003 - Zell Miller Tries to Save the Democratic Party - The Ornery American ::

Card comments on the democrats on lack of compromise, while praising Zell Miller as the exception:

Miller quotes John F. Kennedy: "Compromise need not mean cowardice. Indeed, it is frequently the compromisers and conciliators who are faced with the severest tests of political courage as they oppose the extremist views of their constituents."

Card always has interesting opinions. Many people predict that the democrats are risking failure as a mainstream political party. I really don't care either way as long as we don't go down a failing path of having the government provide for everything people don't have. That is simply not it's function...

For example:
Dean: A Socialist? ::

Is the socialist label unfair? I don't know. From Dean's website we get the following,

"The economic policies of the Bush Administration are misguided, unfair, and unsuccessful.
They fail to meet the basic standard of economic justice: decent, well-paying jobs for all who want them."

Certianly not the kind of pro-market position I'd like to see. A job's wage rate should be set by supply and demand and not some fuzzy headed notion of what is a "good wage".

(0) comments
Robotic Aide in walking / lifting / killing ::

The Pentagon wants the robotic suit to be able to vastly amplify the slightest motion of arms or legs - not enough that one could leap tall buildings with a single bound or dodge an incoming bullet - but enough assist to allow a soldier to carry large loads and a staggering amount of weaponry for up to 12 hours at more than seven miles an hour without tiring.

In an effort to make every soldier a tank, the military is heavily investing in robots aiding human mechanical tasks. Think of Ripley in the 'lifter' from Aliens, fighting the queen.

Of course, there are also civilian applications, like legs for the handicapped and any moving/lifting blue collar job. It would be like an active back-brace, and injuries would certainly be reduced.

More interesting might be applications in sports. I would image the pilots of these walkers would become jockey size, and would become extremely good at finessing the bot. I think contact sports would become much more interesting if human injuries and physical limitations were less of an issue. Also, the sports industry as a whole would actually aide innovation, instead of simply giving the plebeians their bread and circuses.


On a side note, this is an 80/20 problem. You see this trend a lot: a problem is too hard for robots, so a human is introduced to do the really hard part (like seeing and thinking) and the robot will do the rest. The last 20% is left for the human. Here, the robot doesn't need to perceive or think, only act and balance.

(0) comments

Monday, December 01, 2003

Robotic Trends uses news-feeds to get the latest news & reports in robotics. Reading this makes you think everything is right around the corner... which is where these things always tend to be.

Sports: Tilden and Wow Wee Toys Develop BEAM-based Humanoid Robot ::

Tilden is widely recognized as the father of BEAM robotics, and predictably the RoboSapien product is based on BEAM technology and principles. BEAM, an acronym for Biology, Electronics, Aesthetics and Mechanics, is a philosophy that espouses that the design and function of robots should be based on simple control circuitry instead of complex microprocessors and scores of transistors. This simple circuitry is coupled with sensor technology, and the machinery of the BEAM robots body itself, to produce elegant biomechanical robots that bear an uncanny resemblance to creatures found in the wild (especially insects).

"Conventional control systems are complex, power hungry, CPU intensive, and downright expensive," notes Tilden. "Our approach was to take advantage of wave patterns like those used in nature to solve the basic problems of upright walking, motion, and balance at minimal cost. It isn't simple, but is significantly faster to get cool results with, and toys are a natural market for cool."

One question I always ask: are such simple architectures scalable to, say, a cat-level intelligence (forget human)? Of course not. The perception problem alone shows the failings of this type of system, let alone the intelligent reactions to a complex environment one expects from interesting creatures. People don't keep pet insects except for ants, which are different because the emmergent behavior of the colony has a greater intelligence than the components. This means that the novelty of the robot is what is being purchased, and not the behavior.

BUT, this model is sufficient to make people more interested in robotics. That drive will make downright expensive methods feasible given innovation.

(0) comments
First Post!

I'm finally going to archive and publish my rants and opinions about things I find most interesting. Enjoy!
(0) comments

This page is powered by Blogger. Isn't yours?