Thursday, June 26, 2008
Wednesday, June 25, 2008
Annals of University Bureaucracy
The complete text of an email—that included no links—from the CSULA administration to the entire campus.
from Office of the Vice President of Administration and Finance and CFO
to
date Tue, Jun 24, 2008 at 11:59 AM
subject Administrative Procedure 430
mailed-by cslanet.calstatela.edu
Campus-wide email – Do not reply
The following Interim Administrative Procedure has been uploaded to the CSULA web site:
AP 430, Interim University Regulatory Training Requirements, effective date 06/19/08
Please update your department manuals accordingly.
Monday, June 23, 2008
Jig-saw puzzle: to recognize vs. to do
In the Meno, Socrates spoke of the apparent paradox that we can recognize something that we have a harder time creating. In computer science this is often thought of as the P =? NP problem.
There are many problems for which one can tell in Polynomial time whether a proposed solution is actually a solution but for which no one knows of a way to solve a typical instance of the problem in less than exponential (non-deterministic polynomial) time. The prototypical example is to find a set of values for Boolean variables that cause a Boolean expression to evaluate to true. Given a proposed set of values, it is trivial to determine whether the expression evaluates to true. But in general it is very time-consuming to find such a set of values or to determine that no such set exists.
That was brought to mind when a common-place intuitive example of this sort of problem occurred to me: a jig-saw puzzle. It's easy to tell whether a given configuration of jig-saw puzzle pieces is the correct one. But given a box full of pieces it may be a very time-consuming task to put them together correctly.
That was brought to mind when a common-place intuitive example of this sort of problem occurred to me: a jig-saw puzzle. It's easy to tell whether a given configuration of jig-saw puzzle pieces is the correct one. But given a box full of pieces it may be a very time-consuming task to put them together correctly.
Friday, June 20, 2008
Yahoo! ! diversity
A New York Times story included the following paragraph listing executives who recently left Yahoo!
"Today’s news is that three more of Yahoo’s best executives are leaving the company: Qi Lu, Brad Garlinghouse and Vish Makhijani. That follows the departure of two executive vice presidents, Jeff Weiner and Usama Fayyad. …I find the diversity of the names heartwarming—although it's too bad only one woman was mentioned. There are some things this country does well—and benefits from it.
Hilary Schneider is expected to take some of Mr. Weiner’s responsibility.
Thursday, June 19, 2008
The wheels of bureaucracy …
Sen Carl Levin (D. Mich) is the chairman of the Senate Armed Services Committee. His opening statement at the June 17, 2008 hearing describes the history of the use of interrogation techniques by the US DoD. It's quite amazing.
One particularly striking feature is how bureaucratic it all was. The proposed techniques were passed up and down a number of chains of command. It's a good thing that so many people were involved in reviewing the policies. Many had the courage to object.
The account makes it clear, though, how far the military moved away from treating people like people. It was only at the end of Levin's statement when he quoted a statement by Gen. Petraeus that there was any mention at all of treating people humanely. Most of the discussion was about legal issues, as if the concern was not whether it's right to inflict pain but only whether it's legal. There was no discussion at all about the subjective experience of the prisoners, no attempt to understand what it must feel like to be treated this way, only whether the techniques would be legal and effective. It was like listening to two guards discussing the relative merits of different makes of thumb screws. Should we buy Brand A because it's cheaper or Brand B because it's more sturdy? A study is needed to assess the comparative ROIs and make the right purchase decision.
One particularly striking feature is how bureaucratic it all was. The proposed techniques were passed up and down a number of chains of command. It's a good thing that so many people were involved in reviewing the policies. Many had the courage to object.
The account makes it clear, though, how far the military moved away from treating people like people. It was only at the end of Levin's statement when he quoted a statement by Gen. Petraeus that there was any mention at all of treating people humanely. Most of the discussion was about legal issues, as if the concern was not whether it's right to inflict pain but only whether it's legal. There was no discussion at all about the subjective experience of the prisoners, no attempt to understand what it must feel like to be treated this way, only whether the techniques would be legal and effective. It was like listening to two guards discussing the relative merits of different makes of thumb screws. Should we buy Brand A because it's cheaper or Brand B because it's more sturdy? A study is needed to assess the comparative ROIs and make the right purchase decision.
Friday, June 13, 2008
Political Play: No show hands of hands, please
Great AP story.
CHICAGO (AP) — It's a cardinal rule in politics: Know your audience.Democrat Barack Obama, speaking in a high school gymnasium this week in Kaukauna, Wis., said nearly a quarter of Republican John McCain's proposed tax cuts would go "to households making over $2.8 million every year."
He asked for a show of hands from those in that category. Not one went up in the crowd of more than 1,000.
A few hours later, at a $28,500 per person fundraiser at the opulent Chicago home of Sara and James Star, Obama took the same jab at McCain's tax plan. This time, however, he wisely skipped a show of hands from his 60 donors.
Compiled by Charles Babington
'Herds' of wary cars could keep an eye out for thieves
From New Scientist Tech
What I like about it is that it relies on the overall goodness of strangers. It assumes that when one parks, most of the other cars will perform as expected. Society depends on that in general. Society doesn't work because the police keep us all in line. Society works before for the most part people are honest. If there are more honest people than dishonest people a society will work. Otherwise it won't — at least not in a way that we find comfortable. (There's a nice agent-based model about that, but I don't recall the exact reference.) A similar strategy is why some of the the wisdom of crowds effects work. It's also why it's better to be out in a crowd than out alone.
So this technique takes advantage of general honesty to let things (which are proxies for their owners) protect other things.
A new approach to car alarms gets vehicles to watch each others' backs like a herd of animals under threat from predators.This is a great idea. It relies on strangers to help accomplish a positive goal. The basic idea is that when parking a car, the car joins an ad hoc network of other parked cars. Any car that moves without authoritatively signaling that it's going to move causes the other cars to raise an alarm. See the details in the story or the original paper.
The security system relies on networks of cars constantly gossiping with their neighbours using concealed wireless transmitters. The cars raise the alarm when a thief tries to make a getaway with any of their number.
What I like about it is that it relies on the overall goodness of strangers. It assumes that when one parks, most of the other cars will perform as expected. Society depends on that in general. Society doesn't work because the police keep us all in line. Society works before for the most part people are honest. If there are more honest people than dishonest people a society will work. Otherwise it won't — at least not in a way that we find comfortable. (There's a nice agent-based model about that, but I don't recall the exact reference.) A similar strategy is why some of the the wisdom of crowds effects work. It's also why it's better to be out in a crowd than out alone.
So this technique takes advantage of general honesty to let things (which are proxies for their owners) protect other things.
Thursday, June 12, 2008
FightTheSmears
Obama has a new website
to fight the smears.
to fight the smears.
Labels:
politics
Tune-Deaf People May Hear a Sour Note Unconsciously
From NIH
Tune deafness [occurs] in an otherwise normal, uninjured brain,” said James F. Battey, Jr., M.D., Ph.D., director of the NIDCD. “[THis may make it possible to use] tune deafness … to study consciousness.”I didn't know there were brain signals for unexpected stimuli.
A person who is tune deaf is unable to perceive pitch, reproduce melodies, or identify deviations in a melody. According to geneticist Dennis Drayna, Ph.D., one of the study authors, not only is music not enjoyable for people with tune deafness, many of them don’t fully understand what music is. “For severely affected tune-deaf people, Yankee Doodle is no different than traffic noise in Manhattan. It’s fairly meaningless to them,” he said."
Volunteers [half tune deaf and half controls] listened to 102 familiar melodies, roughly half of which were correct, and half of which contained [an] errant last note. The researchers then sifted through the EEG data to isolate the brain’s response to a specific stimulus—in this case, the right or wrong note.
Of principal interest were two signals that brains normally generate when they are presented with a stimulus that doesn’t match what the brain expects to hear, such as the wrong note in a song. The first, the mismatch negativity (MMN), is a large negative signal that occurs roughly 200 milliseconds after the unexpected stimulus is heard; the second signal, the P300, is a large positive signal occurring roughly 300 milliseconds after the unexpected stimulus.
Because tune-deaf people consistently don’t recognize when a wrong note is played or sung, the researchers hypothesized that their brains would not generate the MMN or P300 signals, and as expected, this was true for the MMN signal. However, in the case of the P300 signal, tune-deaf volunteers were processing the wrong note in the same way as the normal participants, even though they weren’t consciously aware of the deviation. [emphasis added] Other brain signals demonstrated that correct notes were being processed equally well for both tune-deaf and normal volunteers.
Labels:
science,
subjective experience
Wednesday, June 11, 2008
Job creation?
Paul Krugman
has a nice graphic comparing job creation under Bush and Clinton.
Subscribe to:
Posts (Atom)