Science, Technology & Health: February 2005 Archives
I've been working like mad on my PhD while I've been looking for a job, and now I'm writing a chapter about the properties a communication simulation needs to have in order to be analogous to the real world. I don't want to give too many examples -- because I don't want to taint your suggestions -- but I'm thinking that simulations must deal with things like:
- Time. No cause and effect without time, so it's pretty fundamental. Plus, I doubt the human brain can even conceive of a world without time.
- Space. Most useful simulations will have something analogous to physical space, for objects to exist in and move through.
- Tasks. Simulations must have tasks to be completed, or they're useless.
- Objects. Something has to do the acting and moving and learning.
And then there are concepts with more specificity:
- Resources. Types of objects that are required to complete a task. Are they scarce? Are there an infinite number?
- Feedback loops. In real life, success leads to more success.
- Aging, death, birth.
- Types of knowledge. Where things are. The states of other agents. How tasks are completed.
- Indirection. John says that Jane said that Tim said that John is dead.
- Truthfulness. Some people lie, and some are just mistaken, but either way not all signals are correct.
And so forth. I've got a huge list of things to be considered, but I'd like to hear your suggestions. What properties of real life are important for a simulation of Distributed Autonomous Communicators?
Every human being has a selfish personal agenda that they're attempting to advance through their professional life. In scientific fields, the system is supposed to work in such a way that the scientist reaps the most personal benefit if he follows a certain set of rules and always acts honestly and without philisophical bias. Unfortunately -- as in all human systems -- that lofty ideal is rarely met, and in some cases reality falls far short indeed. As an example, take the case of professor Reiner Protsch von Zieten, who apparently lied about the age of skulls he discovered in Northern Europe.
Reiner Protsch von Zieten, a Frankfurt university panel ruled, lied about the age of human skulls, dating them tens of thousands of years old, even though they were much younger, reports Deutsche Welle."The commission finds that Prof. Protsch has forged and manipulated scientific facts over the past 30 years," the university said of the widely recognized expert in carbon data in a prepared statement. ...
Among their findings was an age of only 3,300 years for the female "Bischof-Speyer" skeleton, found with unusually good teeth in Northern Germany, that Protsch dated to 21,300 years.
Another dating error was identified for a skull found near Paderborn, Germany, that Protsch dated at 27,400 years old. It was believed to be the oldest human remain found in the region until the Oxford investigations indicated it belonged to an elderly man who died in 1750.
These weren't just mistakes, they were purposeful lies, and these lies led to a whole host of subsidiary mistakes when other researchers relied on his results. Professor Protsch thought he could beat the system and advance his career without playing by the rules, and he succeeded for decades, only getting caught by chance. Anyone who thinks this example is singular within the academic community is sadly delusional.
Why? Because there isn't much acclaim to be gained by going back over the work of others and seriously checking it for errors. No one wants to be seen as a backstabber, especially in tight-knit, incestuous scientific circles. Plus, it's hard to find lies, even when they're there, because even the liars have a lot of specialized knowledge that makes it difficult for others to replicate their work. See also, the downfall of Michael Bellesiles.
Update:
Raina points to another potential fraud from last year regarding Bernard Kettlewell and his peppered moths. Such examples likely abound, just waiting to be discovered.
A reader pointed me to this article about robot toddler that learns to walk like a human child. This achievement is an excellent example of how artificial intelligence techniques can be used to reap efficiency gains that cannot be easily engineered by human hands.
The machines use what the researchers called a "passive-dynamic design" that closely mimics the way humans walk. Earlier robots required powerful machines to stroll, with each leg, knee and ankle requiring motorized assistance. The effort requires a lot of energy.The passive dynamic design uses gravity, along with muscle-like springs and motors. The energy required is just a fraction of that needed by other walking robots, said Andy Ruina, a Cornell University researcher.
Ruina said the walking robots move like humans, falling and catching themselves as they move forward. This essentially is the same movement people use, a motion toddlers must master to walk.
"We let the machines take care of a lot of the motion," he said. In contrast, most walking robots, such as Asimo, developed by the Honda Motor Co. (HMC), require a motor to power every motion.
Could a control system for this kind of movement be designed by hand? Probably, but it would be incredibly hard, particularly if it were to be as adaptable as this learning model.
"It can learn to walk in 20 minutes," Tedrake said. "Once it learns to walk, then it adapts its gait to new terrain."He said the sensors take measurements at the rate of 200 times a second and constantly send new instructions to the motors that control the tilt and motion. The sensors also direct actuators that control the tension on springs in the robot ankles. This helps the machine push forward with each stride.
"Every time it takes a step, it changes the parameters a little bit, based on its experience," Tedrake said. "It will walk on any surface and adjust the way it walks."
In effect, the robot changes its stride just as humans do when moving from sand to grass to pavement.
He said the machine even has learned to walk on a treadmill, making adjustments as the surface tilts or speeds up. The robot can start on its own and even walk backward.
I have a little experience with robotics, and these results are pretty impressive. Artificial intelligence is very well suited for developing control systems like these walkers, and there's more promise in this direction than towards the kind of AI you see in science fiction books and movies.
I flipped on the radio this morning on the way to lunch and heard Rush Limbaugh slinging Zicam, the homeopathic cold remedy, and I thought to myself, Why, I've proclaimed the benefits of Zicam on my site, and I've never been paid thousands of dollars to do it! Yeah sure, a few people claim to have lost all sense of smell and taste after using the product, but for most people it works great and drastically lessens the suffering caused by the common cold. I even own stock in Matrixx Initiatives, because I believe in the product. So where's my payola?
An article about a particularly fast-acting strain of HIV gives some informative numbers for the rate of typical HIV progression.
The normal time of progression from infection to full-blown AIDS in an untreated patient is about nine years, with death following within 18 months, said Carly Stanton, a spokeswoman for the U.S. Centers for Disease Control in Atlanta. For someone treated with anti-viral drugs, the average progression to disease from infection is 11 years, with death occurring within an average six years, Stanton said.
For much more detailed information about every stage of the disease, see this HIV/AIDS primer.
If you're like me and you enjoy buying metal items (such as knives), you've probably wondered what the differences are between the various types and grades of steel. Check out the research page at Metal Suppliers Online and you'll no longer have to remember why kitchen blades are often made from stainless steel 420, while stainless steel 440 C is used for superior pocket knives.
