Later On

A blog written for those whose interests more or less match mine.

Archive for December 15th, 2015

Astonishing meme-formations: YouTube and copyright law and a personal crusade

leave a comment »

Quite an intricate and fascinating path is trodden by Jason Koebler in Motherboard in “Weezer’s Bizarre Copyright Crackdown on ‘Hash Pipe’“. As you might guess, it was somewhat difficult for me to follow, never having so much as heard of “Weezer.” But the formations into which memes can quickly evolve are amazing. (Memes are the meta-lifeforms that live in and through our cultural creations, subject to the same Darwinian principles as normal lifeforms, and thus constantly evolving, only on a much faster timescale. More in The Selfish Gene, by Richard Dawkins, and in other publications sense.)

Formations of the mineral variety persist for years, gradually shaped by erosion, glaciers, etc. Formations of animals are either very slow moving (coral reefs, for example) or quite fluid (a large herd, flock, or school, for example) or even hyperfluid (a political party, a body of law, a cultural practice, a fashion). So meme-formations are best thought of as 4-dimensional shapes, so that you have to examine the entity over time.

Written by LeisureGuy

15 December 2015 at 2:03 pm

An example of the Dunning-Kruger Effect: The common understanding of how Medicare and private insurance interact

leave a comment »

Absolutely fascinating—and carefully argued—article. It is a must-read because it requires must-think. “The Experts Were Wrong About the Best Places for Better and Cheaper Health Care,”
By Kevin Quealy and Margot Sanger-Katz, in the NY Times.

Written by LeisureGuy

15 December 2015 at 11:29 am

When the State Department Tries to Choose Muslim Thought Leaders to Win “Hearts and Minds”

leave a comment »

Glenn Greenwald has a good article on the boneheaded PR programs from the State Department:

Few things produce darker and more warped comedy than when the U.S. Government launches new propaganda campaigns to “win the hearts and minds of Muslims.” Remember when George W. Bush dispatched his long-time political aide, Texas’ Karen Hughes, to the Middle East as a State Department official to change Muslim perceptions of the U.S. and that promptly (and predictably) resulted, as Slate put it, in a “jaw-dropping display of ignorance and malapropism that made her the laughing stock of the region”?

In fairness to Hughes and the State Department, she was vested with an impossible task. How do you convince the people of that region to like you when you’ve spent decades bombing, invading and droning them, arming and propping up the tyrants who suppress them, lavishing Israel with the weapons, money and UN cover used to occupy and brutalize Palestinians, and just generally treating their countries like your own private plaything for war and profit?

As a 2004 Rumsfeld-commissioned study about the causes of Terrorism put it: “Muslims do not ‘hate our freedom,’ but rather, they hate our policies;” in particular, “American direct intervention in the Muslim world,” our “one sided support in favor of Israel”; support for Islamic tyrannies in places like Egypt and Saudi Arabia; and “the American occupation of Iraq and Afghanistan.” As a result, trying to change Muslim perceptions of the U.S. without changing U.S. policies of imperialism and militarism is the ultimate act of futility.

Destined though they are to fail, the propaganda efforts don’t have to be quite so comically bad: the government could, for instance, put people in charge of these campaigns who actually know something about the part of the world they’re trying to propagandize. Yet these efforts seem only to get worse. One of the most embarrassing tactics is when the U.S. government (and its media allies) select people whom they hold out to the Muslim world as the people they ought to follow; invariably, the U.S.’s selected “leaders” spout views and engage in conduct more anathema to the overwhelming majority of Muslims than the U.S. government itself is.

Last year, the State Department announced with great fanfare a new social media campaign to counter ISIS’ online messaging. They called it “Think Again, Turn Away,” and created Twitter and Facebook accounts in that name. Its self-described purpose on Facebook: “Our mission is to expose the facts about terrorists and their propaganda. Don’t be misled by those who break up families and destroy their true heritage.”

It was a massive comedic failure from the start. And that failure continues. Yesterday, Think Again, Turn Away’s Twitter account promoted and hailedsomeone they think will serve as an inspiring thought leader for Muslims around the world:

Is Ayaan Hirsi Ali likely to be the effective messenger to the Muslim world which the State Department envisions her to be? Last year, she revealed her choice for who should win the Nobel Peace Prize: Benjamin Netanyahu; “he does what is best for the people of Israel, he does his duty,” she said. “I really think he should get the Nobel Peace Prize. In a fair world he would get it.”

Earlier this year, she told a gathering hosted by Israeli Consul General that she previously tried to convert to Judaism and hoped one day to try again. She has spouted some of the most virulent anti-Muslim bigotry, the worst of which may have been her 2007 interview with Reason, where she said she rejects the notion that “we” are at war only with radical Islam but instead are at war with Islam generally. Behold the State Department’s chosen Ambassador to the Muslim world: . .

Continue reading.

Written by LeisureGuy

15 December 2015 at 10:24 am

How the state’s attorney for Chicago refuses to prosecute cops for criminal behavior

leave a comment »

State’s Attorney Anita Alvarez sees one of her primary responsibilities as being to kill any investigations or prosecution of police officers whose misconduct amounts to criminal behavior. You’ll recall that she refused to indict the cop who murdered Laquan McDonald until a judge forced the release of the video of the murder—and even then she waited until the last minute, and has made absolutely no move ot indict the cops whos perjured testimony protected the killer. And Daniel Denvir in Salon gives another example of why Alvarez should be immediately removed from her position as state’s attorney: she simply refuses to fulfill the obligations of her office.

In a meeting on July 24, 2012, Chicago Police Officer Allyson Bogdalek broke down and cried as she admitted to prosecutors the obvious: She had lied under oath in the case of a man accused of robbing a Back of the Yards liquor store and shooting the owner in the leg.

The victim of the shooting had picked the suspect, Ranceallen Hankerson, out of a lineup. But Officer Bogdalek lied on the stand during an April 13, 2011, hearing when she denied that the victim had been shown photographs of possible suspects prior to Hankerson’s arrest. In fact, the victim had been shown photos, and he had failed to pick Hankerson out—evidence that would have proven beneficial to the defense.

Prosecutors opened an investigation, and recommend indicting Bogdalek for perjury and other felonies, according to Cook County State’s Attorney’s Office files provided to Salon. In February 2014, however, the process came to a screeching halt: State’s Attorney Anita Alvarez overruled her subordinates and instructed them that no charges would be filed. The case, which until now has escaped much public notice, provides evidence to back charges that Alvarez, currently under fire for her handling of the fatal police shooting of Laquan McDonald, protects officers accused of misconduct.

“It’s a powerful example of State’s Attorney Alvarez’s refusal to address systemic perjury by Chicago police,” says Craig Futterman, a civil rights attorney and professor at University of Chicago Law School who reviewed the case at Salon’s request.

Bogdalek’s lie had become clear at the 2011 hearing. After she testified that they had never shown the victim photos of possible suspects, Hankerson’s defense attorney played a recording from Bogdalek’s squad car, in which she can be heard asking a sergeant whether they should take Hankerson into custody given that the victim had failed to identify him in a photo array, meaning a group of photographs of potential suspects shown to a witness. [This 4-minute video is included in the Salon article. – LG]

Bogdalek finally came clean more than a year later, after prosecutors asked her about the discrepancy as the case was about to go to trial. Not only had she lied, she stated, but police detectives, multiple superiors and her partner, Officer Dominick Catinella, had encouraged her to do so. She said that she had wanted to inventory the photo array but Catinella “wanted her to forget about it because it hurt the case,” according to the prosecutors’ summary.

Bogdalek and Catinella could not be reached for comment.

Futterman says that the case demonstrates that the “State’s Attorney has prioritized convictions over justice” and “numbers over truth,” a mind-set that deprives defendants of their rights and encourages the conviction of innocent people.

“Police perjury is so common here in Chicago that we call it testilying,” says Futterman. “The state’s attorney has relied on those very lies to win convictions.”

In a statement released to Salon, the Cook County State’s Attorney’s Office blamed judges and juries, saying that they decided not to prosecute Bogdalek because it is simply too hard to win convictions against police officers.

“This case underwent extensive legal analysis at all appropriate levels of the State’s Attorney’s Office,” emailed spokesperson Sally Daly. “That analysis included significant evaluation of both the strengths and weaknesses of this case and how those strengths and weaknesses would play out at trial.

We face a reality here in Cook County, and around the country, that it is extremely difficult to convince judges or juries to convict police officers of misconduct in the line of duty. The ultimate analysis in this case led to the determination that the State would not have been able to meet the legal standard that is required, which is proving guilt beyond a reasonable doubt.”

* * *

Bogdalek, according to the prosecutors’ account, was having a crisis of conscience over the lie—namely that it might get her into trouble. She asked two or three detectives (whom she claimed she could not identify) what to do. She claims they told her to “forget about it.” Bogdalek approached a supervisor, Sgt. John Ward, and asked him how they should handle the situation. She claimed that Ward called Catinella in, and asked if they had shown the victim a photo array. Catinella said they had not. She claimed that Sgt. Ward then announced that “the argument was settled,” and the photo array was not inventoried. . .

Continue reading.

Written by LeisureGuy

15 December 2015 at 10:09 am

Landmark Algorithm Breaks 30-Year Impasse

leave a comment »

Very interesting article in Quanta by Erica Klarreich—headed (at the link) by a dynamite GIF showing various transformations of a network into equivalent (but very different-looking) networks. The article begins:

A theoretical computer scientist has presented an algorithm that is being hailed as a breakthrough in mapping the obscure terrain of complexity theory, which explores how hard computational problems are to solve. Last month, László Babai, of the University of Chicago, announced that he had come up with a new algorithm for the “graph isomorphism” problem, one of the most tantalizing mysteries in computer science. This new algorithm, Babai says, is vastly more efficient than the previous best algorithm, which had held the record for more than 30 years. His paper became available today on the scientific preprint site arxiv.org, and he has also submitted it to the Association for Computing Machinery’s 48th Symposium on Theory of Computing.

For decades, the graph isomorphism problem has held a special status within complexity theory. While thousands of other computational problems have meekly succumbed to categorization as either hard or easy, graph isomorphism has defied classification. It seems easier than the hard problems, but harder than the easy problems, occupying a sort of no man’s land between these two domains. It is one of the two most famous problems in this strange gray area, said Scott Aaronson, a complexity theorist at the Massachusetts Institute of Technology. Now, he said, “it looks as if one of the two may have fallen.”

Babai’s announcement has electrified the theoretical computer science community. If his work proves correct, it will be “one of the big results of the decade, if not the last several decades,” said Joshua Grochow, a computer scientist at the Santa Fe Institute.

Computer scientists use the word “graph” to refer to a network of nodes with edges connecting some of the nodes. The graph isomorphism question simply asks when two graphs are really the same graph in disguise because there’s a one-to-one correspondence (an “isomorphism”) between their nodes that preserves the ways the nodes are connected. The problem is easy to state, but tricky to solve, since even small graphs can be made to look very different just by moving their nodes around.

Now, Babai has taken what appears to be a major step forward in pinning down the problem’s difficulty level, by setting forth what he asserts is a “quasi-polynomial-time” algorithm to solve it. As Aaronson describes it, the algorithm places the problem within “the greater metropolitan area” of P, the class of problems that can be solved efficiently. While this new work is not the final word on how hard the graph isomorphism problem is, researchers see it as a game changer. “Before his announcement, I don’t think anyone, except maybe for him, thought this result was going to happen in the next 10 years, or perhaps even ever,” Grochow said.

In recent weeks, Babai gave four talks outlining his algorithm. It will take time for his new paper to be thoroughly vetted by other experts, however. Babai has declined to speak to the press, writing in an email: “The integrity of science requires that new results be subjected to thorough review by expert colleagues before the results are publicized in the media.”

Other researchers are cautiously hopeful that his proof will pan out. “Babai has a sterling record,” Aaronson said. “He’s as trustworthy as they come.”

“I think people are pretty optimistic,” said Luca Trevisan, a computer scientist at the University of California, Berkeley, after Babai’s second talk. Assuming the proof is correct, he said, “it’s a landmark result.”

A Blind Taste Test

Given two graphs, one way to check whether they are isomorphic is simply to consider every possible way to match up the nodes in one graph with the nodes in the other. But for graphs with n nodes, the number of different matchings is n factorial (1 * 2 * 3 * … *n), which is so much larger than n that this brute-force approach is hopelessly impractical for all but the smallest graphs. For graphs with just 10 nodes, for instance, there are already more than 3 million possible matchings to check. And for graphs with 100 nodes, the number of possible matchings far exceeds the estimated number of atoms in the observable universe.

Computer scientists generally consider an algorithm to be efficient if its running time can be expressed not as a factorial but as a polynomial, such as n2 or n3; polynomials grow much more slowly than factorials or exponential functions such as 2n. Problems that have a polynomial-time algorithm are said to be in the class P, and over the decades since this class was first proposed, thousands of natural problems in all areas of science and engineering have been shown to belong to it.

Computer scientists think of the problems in P as relatively easy, and they think of the thousands of problems in another category, “NP-complete,” as hard. No one has ever found an efficient algorithm for an NP-complete problem, and most computer scientists believe no one ever will. The question of whether the NP-complete problems are truly harder than the ones in P is the million-dollar P versus NP problem, widely regarded as one of the most important open questions in mathematics.

The graph isomorphism problem is neither known to be in P nor known to be NP-complete; instead, it seems to hover between the two categories. It is one of only a tiny handful of natural problems that occupy this limbo; the only other such problem that’s as well-known as graph isomorphism is the problem of factoring a number into primes. “Lots of people have spent time working on graph isomorphism, because it’s a very natural, simple-to-state problem, but it’s also so mysterious,” Grochow said.

There are good reasons to suspect that graph isomorphism is not NP-complete. For example, it has a strange property that no NP-complete problem has ever been shown to have: It’s possible, in theory, for an all-knowing being (“Merlin”) to convince an ordinary person (“Arthur”) that two graphs are different without giving away any hints about where the differences lie.

This “zero-knowledge” proof is similar to the way Merlin could convince Arthur that Coke and Pepsi have different formulas even if Arthur can’t taste the difference between them. All Merlin would have to do is take repeated blind taste tests; if he can always correctly identify Coke and Pepsi, Arthur must accept that the two drinks are different.

Similarly, if Merlin told Arthur that two graphs are different, Arthur could test this assertion by putting the two graphs behind his back, moving their nodes around so that they looked very different from the way they started, and then showing them to Merlin and asking him which was which. If the two graphs are really isomorphic, there’s no way Merlin could tell. So if Merlin gets these questions right again and again, Arthur will eventually conclude that the two graphs must be different, even if he can’t spot the differences himself.

No one has ever found a blind-taste-test protocol for any NP-complete problem. For that and other reasons, there’s a fairly strong consensus among theoretical computer scientists that graph isomorphism is probably not NP-complete.

For the reverse question — whether graph isomorphism is in P — the evidence is more mixed. On the one hand, . . .

Continue reading.

Written by LeisureGuy

15 December 2015 at 10:00 am

Posted in Math, Software

The importance of freedom of religion varies a lot

leave a comment »

Seen on Facebook:

Screen Shot 2015-12-15 at 9.53.38 AM

Written by LeisureGuy

15 December 2015 at 9:55 am

Posted in Daily life, GOP, Religion

QED Mint Zest with Rooney Finest and the all-Maggard Comb-Guard Razor

leave a comment »

15 Dec 2015

Extremely fine shave today. You’ll note that after one use the label of the QEDman Mint Zest (spearmint, orange, and lavender) is more like an Impressionist painting than text: beautiful in its own way, but scant in the information department.

I did contact Charles at QEDusa.com and he said that this is deliberate. The tin in which the soap arrives is, in his view, merely a nicer wrapper than paper would be, and he expects the soap to be transferred to a mug that the purchaser already has. Since the tub/package is to be discarded, the label’s durability is irrelevant. Thus no waterproof label. He said that he didn’t expect the container to be used as a lathering bowl.

I think by “lathering bowl” he refers not to building the lather but simply loading the brush. A lathering bowl, as I understand the term, is a separate bowl, holding no soap, that is used to build the lather by working the loaded brush vigorously and adding water as needed. The lathering bowl (in that sense) is one of the three primary ways of building lather, the other two being palm-lathering (working the loaded brush vigorously on your open or cupped hand, adding water as needed) and face-lathering (taking the loaded brush directly to your beard and working up the lather there, adding water as needed). In all three methods, the brush must first be loaded, and that is done on the soap in whatever container holds it.

I pointed out that collectors of soap probably don’t have a lot of mugs around, and mugs have the disadvantage of lacking a label. (I have two mugs containing soaps, but I no longer recall what soaps are in them.) Thus it’s common for soaps sold in containers (tubs, bowls, or jars) to remain in those containers, with the brush (naturally) loaded on the soap.

That’s not how he sees it, and he seemed firm in his conviction that a waterproof label is simply not needed since the container will be discarded. So it goes. If you do get the soap and plan on keeping the container, I suggest using a strip of wide transparent packing tape to cover the label and protect the printing.

Once again I added small amounts of water as I loaded the brush, doing this several times. It’s my impression that soaps containing clay are particularly thirsty and require a fair amount of water to be added, the water worked well into the brush with each addition.

The resulting lather was extremely good: thick, slick, and having a very nice fragrance. And I do like this brush.

This is an all-Maggard razor this morning: the Maggard open-comb head and a Maggard handle. It is both extremely comfortable and extremely efficient—and I’ll add that it’s extremely cost effective: the head is $16 and the handle is $14. The total price of $30 is $1 more than for the Parker 24C, but for that extra $1 you get a better handle (stainless instead of plated brass) and standard threading on both head and handle, which makes swapping the handle or head simple. (Parker’s threading seems to have drifted off standard.) The Maggard combo would make a fine first (or additional) razor—and a great gift, IMO.

Carrying the citrus them along, a small splash of Mickey Lee Soapworks Italian Stallion aftershave milk (a discontinued product, alas) finished the job.

A great shave to start a new day.

Written by LeisureGuy

15 December 2015 at 9:45 am

Posted in Shaving

%d bloggers like this: