Later On

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

Archive for August 13th, 2018

Every pol needs an EP..

leave a comment »

From Brian Stelter’s newsletter:

That is, every politician in the TV age needs an executive producer. Dylan Byerswrote about Bill Shine’s role as Trump’s EP in Monday’s PACIFIC. “Trump’s 2020 challengers will need their own EP’s who know how to counter-program…” (PACIFIC)

Written by Leisureguy

13 August 2018 at 8:06 pm

Posted in Media, Memes, Politics

Universal Method to Sort Complex Information Found

leave a comment »

There are some heavy-hitting adjectives in that title, which names an article by Kevin Harnett in Quanta:

If you were opening a coffee shop, there’s a question you’d want answered: Where’s the next closest cafe? This information would help you understand your competition.

This scenario is an example of a type of problem widely studied in computer science called “nearest neighbor” search. It asks, given a data set and a new data point, which point in your existing data is closest to your new point? It’s a question that comes up in many everyday situations in areas such as genomics research, image searches and Spotify recommendations.

And unlike the coffee shop example, nearest neighbor questions are often very hard to answer. Over the past few decades, top minds in computer science have applied themselves to finding a better way to solve the problem. In particular, they’ve tried to address complications that arise because different data sets can use very different definitions of what it means for two points to be “close” to one another.

Now, a team of computer scientists has come up with a radically new way of solving nearest neighbor problems. In a pair of papers (one posted online in April, the other forthcoming), five computer scientists have elaborated the first general-purpose method of solving nearest neighbor questions for complex data.

“This is the first result that captures a rich collection of spaces using a single algorithmic technique,” said Piotr Indyk, a computer scientist at the Massachusetts Institute of Technology and influential figure in the development of nearest neighbor search.

Distance Difference

We’re so thoroughly accustomed to one way of defining distance that it’s easy to miss that there could be others. We generally measure distance using “Euclidean” distance, which draws a straight line between two points. But there are situations in which other definitions of distance make more sense. For example, “Manhattan” distance forces you to make 90-degree turns, as if you were walking on a street grid. Using Manhattan distance, a point 5 miles away as the crow flies might require you to go across town for 3 miles and then uptown another 4 miles.

It’s also possible to think of distance in completely nongeographical terms. What is the distance between two people on Facebook, or two movies, or two genomes? In these examples, “distance” means how similar the two things are.

There exist dozens of distance metrics, each suited to a particular kind of problem. Take two genomes, for example. Biologists compare them using “edit distance.” Using edit distance, the distance between two genetic sequences is the number of additions, deletions, insertions and replacements required to convert one into the other.

Edit distance and Euclidean distance are two completely different notions of distance — there’s no way to reduce one to the other. This incommensurability is true for many pairs of distance metrics, and it poses a challenge for computer scientists trying to develop nearest neighbor algorithms. It means that an algorithm that works for one type of distance won’t work for another — that is, until this new way of searching came along.

Squaring the Circle

To find a nearest neighbor, the standard approach is to partition your existing data into subgroups. Imagine, for instance, your data is the location of cows in a pasture. Draw circles around groups of cows. Now place a new cow in the pasture and ask, which circle does it fall in? Chances are good — or even guaranteed — that your new cow’s nearest neighbor is also in that circle.

Then repeat the process. Partition your circle into subcircles, partition those partitions, and so on. Eventually, you’ll end up with a partition that contains just two points: an existing point and your new point. And that existing point is your new point’s nearest neighbor.

Algorithms draw these partitions, and good algorithm will draw them quickly and well — with “well” meaning that you’re not likely to end up in a situation where your new cow falls in one circle but its nearest neighbor stands in another. “From these partitions we want close points to end up in the same disc often and far points to end up in the same disc rarely,” said Ilya Razenshteyn, a computer scientist at Microsoft Research and coauthor of the new work along with Alexandr Andoni of Columbia University, Assaf Naor of Princeton University, Aleksandar Nikolov of the University of Toronto and Erik Waingarten of Columbia University.

Over the years, computer scientists have come up with various algorithms for drawing these partitions. For low-dimensional data — where each point is defined by only a few values, like the locations of cows in a pasture — algorithms create what are called “Voronoi diagrams,” which solve the nearest neighbor question exactly.

For higher-dimensional data, where each point can be defined by hundreds or thousands of values, Voronoi diagrams become too computationally intensive. So instead, computer scientists draw partitions using a technique called “locality sensitive hashing (LSH)” that was first defined by Indyk and Rajeev Motwani in 1998. LSH algorithms draw partitions randomly. This makes them faster to run but also less accurate — instead of finding a point’s exact nearest neighbor, they guarantee you’ll find a point that’s within some fixed distance of the actual nearest neighbor. (You can think of this as being like Netflix giving you a movie recommendation that’s good enough, rather than the very best.)

Since the late 1990s, computer scientists have come up with LSH algorithms that give approximate solutions to the nearest neighbor problem for specific distance metrics. These algorithms have tended to be very specialized, meaning an algorithm developed for one distance metric couldn’t be applied to another.

“You could get a very efficient algorithm for Euclidean distance, or Manhattan distance, for some very specific important cases. But we didn’t have an algorithmic technique that worked on a large class of distances,” said Indyk.

Because algorithms developed for one distance metric couldn’t be used in another, computer scientists developed a workaround strategy. Through a process called “embedding,” they’d overlay a distance metric for which they didn’t have a good algorithm on a distance metric for which they did. But the fit between metrics was usually imprecise — a square peg in a round hole type of situation. In some cases, embeddings weren’t possible at all. What was needed instead was an all-purpose way of answering nearest neighbor questions.

A Surprise Result

In this new work, the computer scientists began by stepping back from the pursuit of specific nearest neighbor algorithms. Instead, they asked a broader question: What prevents a good nearest neighbor algorithm from existing for a distance metric? . . .

Continue reading.

Written by Leisureguy

13 August 2018 at 6:07 pm

“Stephen Miller Is an Immigration Hypocrite. I Know Because I’m His Uncle.”

leave a comment »

David Glosser writes in Politico:

Let me tell you a story about Stephen Miller and chain migration.

It begins at the turn of the 20th century in a dirt-floor shack in the village of Antopol, a shtetl of subsistence farmers in what is now Belarus. Beset by violent anti-Jewish pogroms and forced childhood conscription in the Czar’s army, the patriarch of the shack, Wolf-Leib Glosser, fled a village where his forebears had lived for centuries and took his chances in America.

He set foot on Ellis Island on January 7, 1903, with $8 to his name. Though fluent in Polish, Russian, and Yiddish he understood no English. An elder son, Nathan, soon followed. By street corner peddling and sweat-shop toil Wolf-Leib and Nathan sent enough money home to pay off debts and buy the immediate family’s passage to America in 1906. That group included young Sam Glosser, who with his family settled in the western Pennsylvania city of Johnstown, a booming coal and steel town that was a magnet for other hard-working immigrants. The Glosser family quickly progressed from selling goods from a horse and wagon to owning a haberdashery in Johnstown run by Nathan and Wolf-Leib to a chain of supermarkets and discount department stores run by my grandfather, Sam, and the next generation of Glossers, including my dad, Izzy. It was big enough to be listed on the AMEX stock exchange and employed thousands of people over time. In the span of some 80 years and five decades, this family emerged from poverty in a hostile country to become a prosperous, educated clan of merchants, scholars, professionals, and, most important, American citizens.

What does this classically American tale have to do with Stephen Miller? Well, Izzy Glosser is his maternal grandfather, and Stephen’s mother, Miriam, is my sister.

I have watched with dismay and increasing horror as my nephew, who is an educated man and well aware of his heritage, has become the architect of immigration policies that repudiate the very foundation of our family’s life in this country.

I shudder at the thought of what would have become of the Glossers had the same policies Stephen so coolly espouses— the travel ban, the radical decrease in refugees, the separation of children from their parents, and even talk of limiting citizenship for legal immigrants— been in effect when Wolf-Leib made his desperate bid for freedom. The Glossers came to the U.S. just a few years before the fear and prejudice of the “America First” nativists of the day closed U.S. borders to Jewish refugees. Had Wolf-Leib waited, his family would likely have been murdered by the Nazis along with all but seven of the 2,000 Jews who remained in Antopol. I would encourage Stephen to ask himself if the chanting, torch-bearing Nazis of Charlottesville, whose support his boss seems to court so cavalierly, do not envision a similar fate for him.

Like other immigrants, our family’s welcome to the USA was not always a warm one, but we largely had the protection of the law, there was no state sponsored violence against us, no kidnapping of our male children, and we enjoyed good relations with our neighbors. True, Jews were excluded from many occupations, couldn’t buy homes in some towns, couldn’t join certain organizations or attend certain schools or universities, but life was good. As in past generations there were hate mongers who regarded the most recent groups of poor immigrants as scum, rapists, gangsters, drunks and terrorists, but largely the Glosser family was left alone to live our lives and build the American dream. Children were born, synagogues founded, and we thrived. This was the miracle of America.

Acting for so long in the theater of right wing politics, Stephen and Trump may have become numb to the resultant human tragedy and blind to the hypocrisy of their policy decisions. After all, Stephen’s is not the only family with a chain immigration story in the Trump administration. Trump’s grandfather is reported to have been a German migrant on the run from military conscription to a new life in the USA and his mother fled the poverty of rural Scotland for the economic possibilities of New York City. (Trump’s in-laws just became citizens on the strength of his wife’s own citizenship.)

These facts are important not only for their grim historical irony but because vulnerable people are being hurt. They are real people, not the ghoulish caricatures portrayed by Trump. When confronted by the deaths and suffering of thousands our senses are overwhelmed, and the victims become statistics rather than people. I meet these statistics one at a time through my volunteer service as a neuropsychologist for the Philadelphia affiliate of HIAS (formerly the Hebrew Immigrant Aid Society), the global non-profit agency that protects refugees and helped my family more than 100 years ago. I will share the story of one such man I have met in the hope that my nephew might recognize elements of our shared heritage.

In the early 2000s, Joseph (not his real name) was conscripted at the age of 14 to be a soldier in Eritrea and sent to a remote desert military camp. Officers there discovered a Bible under his pillow which aroused their suspicion that he might belong to a foreign evangelical sect that would claim his loyalty and sap his will to fight. Joseph was actually a member of the state-approved Coptic church but was nonetheless immediately subjected to torture. “They smashed my face into the ground, tied my hands and feet together behind my back, stomped on me, and hung me from a tree by my bonds while they beat me with batons for the others to see.”

Joseph was tortured for 20 consecutive days before being taken to a military prison and crammed into a dark unventilated cell with 36 other men, little food and no proper hygiene. Some died, and in time Joseph was stricken with dysentery. When he was too weak to stand he was taken to a civilian clinic where he was fed by the medical staff. Upon regaining his strength he escaped to a nearby road where a sympathetic driver took him north through the night to a camp in Sudan where he joined other refugees. Joseph was on the first leg of a journey that would cover thousands of miles and almost 10 years.

Before Donald Trump had started his political ascent promulgating the false story that Barack Obama was a foreign-born Muslim, while my nephew, Stephen, was famously recovering from the hardships of his high school cafeteria in Santa Monica, Joseph was a child on his own in Sudan in fear of being deported back to Eritrea to face execution for desertion. He worked any job he could get, saved his money and made his way through Sudan. He endured arrest and extortion in Libya. He returned to Sudan, then kept moving to Dubai, Brazil, and eventually to a southern border crossing into Texas, where he sought asylum. In all of the countries he traveled through during his ordeal, he was vulnerable, exploited and his status was “illegal.” But in the United States he had a chance to acquire the protection of a documented immigrant.

Today, at 30, Joseph lives in Pennsylvania and has a wife and child. He is a smart, warm, humble man of great character who is grateful for every day of his freedom and safety. He bears emotional scars from not seeing his parents or siblings since he was 14. He still trembles, cries and struggles for breath when describing his torture, and he bears physical scars as well. He hopes to become a citizen, return to work and make his contribution to America. His story, though unique in its particulars, is by no means unusual. I have met Central Americans fleeing corrupt governments, violence and criminal extortion; a Yemeni woman unable to return to her war-ravaged home country and fearing sexual mutilation if she goes back to her Saudi husband; and an escaped kidnap-bride from central Asia.

President Trump wants to make us believe that these desperate migrants are an existential threat to the United States; the most powerful nation in world history and a nation made strong by immigrants. Trump and my nephew both know their immigrant and refugee roots. Yet, they repeat the insults and false accusations of earlier generations against these refugees to make them seem less than human. Trump publicly parades the  . . .

Continue reading.

Written by Leisureguy

13 August 2018 at 1:14 pm

Trump Foreign Policy Held Back by Struggle to Grasp Time Zones, Maps

leave a comment »

Jonathan Chait writes in New York:

Running an effective foreign policy for a global hyperpower is always tricky when the president happens to be a personally corrupt authoritarian bigot who is concealing shady ties to a strategic adversary. The problem gets even harder when the president is unable to grasp some of the basic facts and principles of diplomacy. Politico’s Daniel Lippman rounds up several harrowing new details of Trump’s attempts to interact with world leaders.

Trump’s shortcomings as a global strategist include, but are not limited to, the following areas:

Knowing all the countries. Maps indicate the world contains a bunch of countries whose existence Trump was never made aware of previously. “Trump appeared confused by Nepal and Bhutan, which lie sandwiched between India and China,” a person familiar with one meeting tells Lippman. “He didn’t know what those were. He thought it was all part of India. He was like, ‘What is this stuff in between and these other countries?’” One of the things they don’t tell you when you start running for president is, there are just so many countries.

Knowing how to read the names of the countries after seeing them. “In one case, Trump, while studying a briefer’s map of South Asia ahead of a 2017 meeting with India’s prime minister, mispronounced Nepal as ‘nipple’ and laughingly referred to Bhutan as ‘button,’ according to two sources with knowledge of the meeting.”

It’s like having Homer Simpson as president, but dumber:

Time zones work, how do they work? Trump reportedly gets the urge to dial up foreign leaders, and has trouble understanding that they may not be working or awake at that moment if they are located on the opposite side of the planet. “He wasn’t great with recognizing that the leader of a country might be 80 or 85 years old and isn’t going to be awake or in the right place at 10:30 or 11 p.m. their time,” a former Trump National Security Council official tells Lippman. “When he wants to call someone, he wants to call someone. He’s more impulsive that way. He doesn’t think about what time it is or who it is.”

A source tells Lippman the time zone problem comes up on “a constant basis.” Holding their daylight hours during inconvenient times is just another one of the ways all these foreign countries are ripping us off.

Which countries don’t like each other. Trump can intuitively grasp the concept that some countries will have better or worse relations with the United States. The idea that these countries may have different levels of relations with each other, independent of the United States, is a far trickier concept. During one meeting with Japanese prime minister Shinzo Abe, Trump repeatedly praised Chinese dictator Xi Jinping, reports a source, who notes that “everyone was cringing.”

Why world leaders have phone calls. Traditional world leaders have busy schedules, and use phone calls to focus on specific points of negotiations. Trump just calls world leaders for no reason. He especially likes to stalk French president Emmanuel Macron. Trump has developed “what one former Trump national security official calls a ‘bizarre’ fascination with calling French President Emmanuel Macron,” reports Lippman. “He wanted to talk to him constantly … Macron would be like: ‘Hey what are we talking about?’”

Most of your world leaders don’t have hours of “executive time” for watching television and rage-tweeting, so they need to use phone calls in order to advance a specific policy agenda.

Written by Leisureguy

13 August 2018 at 1:04 pm

Mühle silvertip, Tcheon Fung Sing, Fatip Testina Gentile, and Alt-Innsbruck

leave a comment »

Quite a pleasant shave today. The Mühle brush shown has quite a soft knot, which feels quite gentle on the face but creates lather very efficiently—and the lather this morning from Tcheon Fung Sing’s Tobacco Verde shaving soap—hard shaving soap, as they would have it—was excellent.

Three passes with my Fatip Testina Gentile left a smooth face,e to which I applied a good splash of Alt-Innsbruck.

I had a fine walk but forgot my pedometer, which is disconcerting. It reveals how focused I have become on tracking the walk rather than the walk itself. I need to refocus on the reality and not get caught up in memes. New route, 65 minutes, unknown number of steps. 😦

Written by Leisureguy

13 August 2018 at 10:07 am

Posted in Shaving

%d bloggers like this: