Later On

For some reason, although I’ve had this Altesse soap for a good while, I’ve never tried it until today. I think I just didn’t like the little plastic shell for some reason. Today I used it with the little Omega brush. Not only is the plastic shell actually rather nice—put the lid aside, cup the lower part in the non-dominant hand (the submissive hand?) and brush briskly with the wet brush, letting lather overflow as it will into the sink.

The result was a brush packed with a terrific slick, dense lather. Really quite a good soap, as it turns out, and the brush is not bad, either. I enjoyed this part quite a bit, and then the Slant holding a previously used Kai blade smoothly exorcised every trace of stubble—one of the smoothest and easiest shaves I’ve had. The XTG pass was pretty much BBS, but I did ATG anyway to a perfectly wonderful shave.

A splash of Saint Charles Shave Woods aftershave, and I am feeling quite on top of the world. I like that soap. And the brush… the whole shave, in fact.

It suddenly occurred to me why cats instantly learn that food is associated with a particular sound—a can opening (and soon the sound of the drawer holding the can opener), a rustle of a kitty-food bag or kitty-treat bag, whatever: they hear the sound, get the food, and by God they know that link, right now.

It’s because cats hunt by hearing: sound is their primary tool in finding food. It’s a very natural and easy link for them to make because they’re primed for it: hear the rustle of a leaf, pounce on it, and lunch—and forever after know that sound means a possible lunch.

I like the part of Sunday when all the chores are done and the rest of the day is mine to have. Since I have only two Sunday chores normally (count out meds and bring up recycling bins), it is generally easy to work them in, but OTOH, I find both sort of boring so I postpone them (always) until after lunch, and then the afternoon starts to slip away. But today, I had them done by 2:00, so lots of time for reading and relaxation.

Hoping you are the same.

20 quadrillion floating-point operations per second: now that’s computing power. The Sister passes along this story by Dan Lyons in the Daily Beast on China’s breaking into the top supercomputing ranks:

Lawrence Livermore National Laboratory is one of the great symbols of America’s scientific and military prowess. For six decades, here on this tranquil campus tucked away in the hill country east of San Francisco, where scientists stroll along leafy paths and zip to meetings on bicycles, huge breakthroughs have been made, like the discovery of a half-dozen elements on the periodic table and the detection of a key component of dark matter.

Livermore’s biggest claim to fame involves designing the world’s most advanced nuclear warheads—this was the mission of the lab when it was created in 1952 by Edward Teller, father of the hydrogen bomb. To do this, Livermore relies on powerful machines called supercomputers, which hum away inside top-secret, heavily guarded buildings. The U.S. has long dominated the industry. Which is what made the news that Bruce Goodwin, head of the lab’s weapons program, received last November all the more momentous: the Chinese had unveiled the world’s most powerful supercomputer, a machine five times more powerful than Livermore’s biggest computer.

To most of us, this might sound like no big deal, akin to Apple coming out with a faster smartphone than Microsoft. But to the scientists, industry titans, and world leaders who understand how delicate America’s position as a global superpower really is, this was a Sputnik moment. Only this time, it wasn’t Russia trouncing the U.S. in the space race, but China surging ahead in one of the most vital areas of national security. By running thousands of processors in parallel, supercomputers not only help design weapons systems, they also model climate change, crack codes, and help develop new and life-changing drugs. Cranking out 500 trillion operations per second, just one of Livermore’s supercomputers throws off so much heat that if the air-conditioning system were to fail, the computer would start to melt within minutes.

Globally, high-performance computing is a $25.6 billion industry, and whoever holds the lead in the field gains huge economic and military advantages. Or put another way, if the U.S. falls behind in supercomputing it could quickly lose its edge in all areas of science, in industries like oil and gas exploration and pharmaceutical research, and in security and military fields. In the race to develop the most powerful computers, both our economic prosperity and our national security are on the line.

When China flipped the switch on the Tianhe-1A, also called the “Milky Way” supercomputer, last fall, it placed itself at the top of the technology world with a stunning demonstration of its newfound engineering prowess. The Chinese grip on the top spot turned out to be short-lived, since six months later, a team in Japan announced an even bigger supercomputer that bumped Tianhe-1A into second place. Nevertheless, . . .

Continue reading. I do wonde how computer-assisted software development (or even computer-directed software development) is progressing.

The US government seems to be a prime enabler to the big drug gangs. The CIA’s involvement in the narcotics underworld during the Vietnam war continues to be rumored, and we know that the BATF send thousands of firearms to Mexican drug cartels, with deadly effective. Though some US officers were killed, none were punished. And now we read about the DEA laundering drug profits for the cartels. (The DEA is the same agency that opposes legal and constructive approaches to drug control, which would undermine the DEA’s size, power, and authority: keep all drugs illegal, and the DEA will always have plenty of money and plenty of work.)

And, of course, the FBI seems to be the go-to agency for initial help in getting a terrorist operation underway—indeed, the FBI seems to take the initiative in that sort of thing fairly often.

As I’ve mentioned, I tend to work in cycles: I have studied Esperanto intensively, for example, 4 or 5 different times, each time learning more and going further, then drifting away and putting it aside. I’ve learned to save the project materials for those kinds of things, because I’m pretty sure the interest will in time revive.

Other such interests: Go, chess (though I believe I’ve permanently abandoned that), letterwriting (pens, papers, and inks), and so on. One reading project is to read one excellent biography of each US president in order of their time in office, the idea being that the time-span overlap of the various biographies will give me multiple takes on significant events and developments, which should reinforce the learning as well as provide diverse perspectives.  (I’ll worry about Grover Cleveland if I ever get that far.)

And, of course, there’s the Shakespeare project: to read and ponder the works of The Bard, whoever s/he was. This article makes me think it’s time to get back to that one.

I blogged earlier on the development of increasingly sophisticated pattern recognition and data-linking software as businesses and government agencies begin to dig through the enormous store of on-line information and the continuing flow of on-line activity (that is, looking at the information from the perspective both of content and of links).

I just posted an update:

NPR has a two-part story on this topic. The first part discusses the uses of “Big Data.” The second part tells people who can help make sense of big data are located and groomed for the job.

Three things I noted: First, the articles make no mention of what is undoubtedly heavy government involvement (think DHS and NSA) in the entire enterprise: finding and grooming and recruiting talent, and developing applications to trawl big data looking for patterns.

Second, I would think it would be effective to use big data to find the people who can more effectively use big data: with school records, test scores, social networks, Amazon purchase history, movie-viewing history, and so on now available on-line (with a certain amount of hacking), finding people who can better find more people would be a logical direction of development.

Third, I find it interesting that the volume of digital data is driving development: new things and approaches being developed simply because the data are there.

I don’t, and I don’t see buying one for this recipe, but it does sound good. Found it via Kafeneio and Steve, who discusses his impressions of the cake here.

When I left after my cataract surgery, feeling fine (anesthetic) and perhaps a little woozy, they handed me a handful of papers with follow-up instructions and the like, including a little card that identified the manufacturer, make, model, serial number, and strength of my intraocular lens implant. Interesting idea, and so I dutifully recorded the information and then tossed the card onto the revolving bookcase beside the chair and went back to regular life.

The Sister, though, emailed me about a card I should have received (the very card) and explained that I must carry that always in my wallet: in case of an accident, the trauma unit would find it and, if eye surgery were needed, they would know what I had.

The thought had not even occurred to me—I’m an optimistic sort of guy—but now I get it. So that’s why it’s credit-card size—I had sort of wondered about that, but thought it had to do more with the packaging of the product than the purpose of the card.

So now the card is safely in my wallet, and I’m not thinking about the circumstances in which it would be useful.

Know what makes my jaw drop? The spectacle of the leading contenders for the GOP nominee for the highest office of the most powerful and probably still most affluent nation on earth. It really is on the level of reality TV—not very good reality TV—of the usual “survivor” format, with contestants being voted/polled off.

James Fallows simply points out what is actually happening. It goes deeper than self-parody, somehow.

Interesting, eh? “Be careful what you wish for.”

This rapid-response continual-refinement approach would have saved the Newton: that poor thing had to live with its glitches until a new version release could be arranged on a finite revision cycle—not like Siri.

But even Siri’s development is not quite where I was talking. The refinements are on-going, but they are still handmade, so far as I can tell from the story—a kind of one-off approach to fixes, bird by bird, as it were.

I’m thinking of an AI program making Bayesian inferences from the on-going data stream. The customer asks a question, and Siri responds with (in effect), “I don’t understand, please rephrase,” and the customer rephrases, and Siri gets it. So now the AI program has more information about the probable meaning of the first phrase—even though the meaning is an inference about the meaning of a set of phoenemes that will vary from customer to customer. It will take a program some time to suss out what exactly are the factors to watch for to determine the meaning of the phrase. With AI refinement, those factors can turn out to be very weird, not the kind of thing a person would think of at all. When people do the tuning by hand, as it were, they look for things that make sense. So they don’t spot weird connections even when those weird connections are more efficient. AI and evolutionary alogorithms do.

I’ve been told that the way inventory is arranged when a warehouse has complete automation and software control of the operation, making its own decisions for optimal performance, parts get stored in weird locations—like a pile of stuff out in the middle of the floor, which location turns out to be optimized for some cost/efficiency metric, and all the (automated, centrally controlled) machines running around on the floor “know” exactly where the pile is and what is in it. But the distribution is not as any human would do it—though it is provably optimal—and the machine’s movements also do not follow humans’ unconscious expectations, as would movements of human-controlled vehicles—which is why humans are not allowed in those areas. The machines move fast, in unpredictable (to human) directions, and with small clearances, since all the machines always know where each is with respect to the others.

That’s the sort of connections and efficiency that would result (I would think) from an AI doing the continuous on-going adjustment/learning/monitoring of Siri’s responses and “understanding”: very fast learning, but based on weird and (to humans) incomprehensible rules developed by software for software.

New Scientist has an interesting article on the health benefits of saunas, which turn out to be substantial. Most of the article is behind a subscription wall (but I hope more and more of you are subscribers), but the gist is visible:

People with chronic heart failure who took saunas five times a week for three weeks improved their heart function and the amount of exercise they could do. Meanwhile, neurons that release the “happiness molecule” serotonin respond to increases in body temperature, perhaps explaining the sauna’s pleasurable effects.

Heart failure occurs when the heart is unable to supply enough blood to the body, resulting in shortness of breath and difficulty exercising. Previous studies have hinted that saunas might boost health. To investigate, Takashi Ohori at the University of Toyama in Japan and colleagues asked 41 volunteers with heart failure to take 15-minute saunas five times per week, using a blanket for 30 minutes afterwards to keep their body temperature about 1 °C higher than normal. …

That, of course, immediately led me to look for public saunas, which I quickly decided were public microbe depots, so I looked at portable 1-person saunas, a very niche category and devoted to infrared saunas—well, infrared via electric heaters rather than steam, and with saunas available in near-infrared and far-infrared as well. Google searches can tell you more; the far-infrared is the one more intriguing to me.

If you have knowledge and/or experience, please share.

Humanity’s first word? “Duh.” We don’t know what it meant, but there’s a good chance this was the first word sound we made.

As this article in New Scientist reports, proto-humans still possessed the air sac that allows primates their booming jungle calls—and messes up speaking (articulation, clear distinction between various vowels and consonants) like nobody’s business.

So once language began, the survival value in being able to communicate efficiently resulted in quick natural selection for smaller and smaller air sacs: a good example of how our memes shape us.

I observed a few posts back (in an update) that human groups are by and large created by memes, and our personalities draw heavily upon groups for basic templates. It occurs to me that our personalities and identities and inextricably interwoven with memes: we are meme creatures, at a deep level, a symbiosis of a physical lifeform and a culture consisting of memes. And no wonder: our very evolution has been shaped by memes ever since we created them.

Why? Because the companies can make money, and companies have that as their only “principle”. Paul Marks reports for New Scientist:

The ease with which totalitarian regimes can buy western technology to intercept and store every electronic communication made by their citizens has been revealed in a joint document release by Wikileaks, the pressure group Privacy International and several media organisations.

Posted online today, the tranche of 287 documents details the wide choice of cellphone and internet surveillance technologies on offer from 160 intelligence contractors – and show, in part, that dissidents using common tools like Google’s gmail service, or devices like Apple’s iPhone and RIM’s BlackBerry, stand little chance of hiding their missives from authoritarian regimes – unless they know how to use the Tor anonymising network.

Many of the surveillance firms appear to be operating in two distinct ways: offering technologies that adhere to legal surveillance norms in their home markets – but when it comes to selling to other nations, they adopt an “anything goes” approach to interception functionality. Some of the them, says Wikileaks founder Julian Assange, will even tap the global network of undersea telecommunications cables to harvest traffic going into and leaving a nation.

The data has been gathered by researchers like Eric King at London-based Privacy International, who has been examining the wares on offer at arms fairs and surveillance industry conferences – many of which have been platforms for selling the kind of bulk email interception technology shown to have been used by the dictators recently ousted in Libya, Tunisia and Egypt.

A French company called Amesys, for example, was found to have helped Libya’s late dictator, Muammar Gaddafi, monitor dissidents’ webmail accounts – because . . .

Continue reading.

I just had an insight that will go into the weight-loss book. Yesterday I mailed the Xmas packages and, it being noon and the Red House Café being next door the the PG PO, I decided to have lunch there. I had  been craving a grilled cheese sandwich, and it came with a pasta salad, and I succumbed to the chocolate pie: a carbohydrate blow-out, and this morning my weight is up 3 lbs. (Puffy fat, I’m sure: too quick a gain to be real.)

So I’m back on sensible eating, and I had a little insight. As a type 2 diabetic, I take a lot of pills, morning and evening. I have a 7-day, 14-compartment (morning and evening) pill box, so it’s easy to see if I forget and miss one morning or evening. And it is inevitable that I do miss one from time to time: perfection does not reign in this household.

When I miss, I do what anyone would do: say, “Oh, crap,” and then proceed with the next scheduled dose. In other words, no grieving, no regrets, just recognition and moving on: Exactly as one should do with a carb-blowout like the one I did: I ate unwisely, now move on.

That’s all pretty standard, but I realized that with the pills, I have one more component: I hate seeing an unemptied compartment, so when I fill the pillbox again, I make a focused effort to have a perfect week the next week: not one missed dose.

Somehow, I can carry this resolution through for a week: it’s not that long, and I just think each morning and evening, “This is the week: no misses.” So I get a week of good compliance on my meds and also build a little foundation toward an improved habit of pill-taking.

The same thing, I realized, can work with meals. Okay, I went overboard. Now to see if I can do a perfect week: breakfast, two fruit snacks, and a sensible lunch and dinner following the meal template I’ve worked out (3-4 oz protein, scant serving carb, at most 2 tsp fat, and vegetables must include greens)—all that for one week, no exceptions, no failures.

I’m sure I can: a week? That’s nothing. So now I have a perfect week underway, and in the course of that I’ll work to re-establish good food habits.

Interesting article in Bloomberg BusinessWeek by Alexandra Wolfe on how two hedge-fund analysts have turned their attention to philanthropies:

When bonus season came around, instead of deciding whether to buy a new suit, car, or a Caribbean vacation, hedge fund analysts Holden Karnofsky and Elie Hassenfeld, both then 25, agonized over which charities to donate to. They put so much time and effort into figuring it out that they eventually left their jobs at Bridgewater Associates, a hedge fund with more than $125 billion under management, to start GiveWell, a nonprofit that evaluates charities with hedge-fund–level rigor. “We were both interested in markets, but we were more interested in how to reduce diarrhea in infants in the developing world,” says Karnofsky, who graduated from Harvard in 2003.

He and Hassenfeld, a Columbia graduate, quit in 2007 and, with $300,000 raised from their friends and former co-workers, developed a grant program in which they would award charities up to $140,000. GiveWell now ranks organizations according to five criteria—evidence of effectiveness, cost-effectiveness, need for funding, transparency, and self-monitoring—and posts the results on its website. GiveWell recommends 2 percent of the charities it reviews, most of them in the developing world, where a dollar goes further toward saving lives.

Though the two now make salaries that are less than half what they earned at Bridgewater, donations to charities they recommend, such as Village Reach and Against Malaria Foundation, have more than tripled in the last year, totaling $1,606,274. Next, they are launching a project called GiveWell Labs, targeting high-net-worth donors. Hassenfeld says they hope to ultimately “change the rules of the game.” . . .

Continue reading.

Our various surveillance mechanisms—closed circuit TV, the NSA collection of everything that goes through telecoms (Internet, phone conversations, text messages, and so on), drone video footage, and on and on—pile up massive amounts of information, and I have remarked from time to time on the increasingly advanced AI and pattern-recognition software, doubtless refining itself through genetic algorithms and perhaps even neural nets, required to use the data and doubtless under constant development.

An interesting article about just one such enterprise can be found in the current issue of Bloomberg BusinessWeek, in a lengthy and informative article by Ashlee Vance and Brad Stone:

In October, a foreign national named Mike Fikri purchased a one-way plane ticket from Cairo to Miami, where he rented a condo. Over the previous few weeks, he’d made a number of large withdrawals from a Russian bank account and placed repeated calls to a few people in Syria. More recently, he rented a truck, drove to Orlando, and visited Walt Disney World by himself. As numerous security videos indicate, he did not frolic at the happiest place on earth. He spent his day taking pictures of crowded plazas and gate areas.

None of Fikri’s individual actions would raise suspicions. Lots of people rent trucks or have relations in Syria, and no doubt there are harmless eccentrics out there fascinated by amusement park infrastructure. Taken together, though, they suggested that Fikri was up to something. And yet, until about four years ago, his pre-attack prep work would have gone unnoticed. A CIA analyst might have flagged the plane ticket purchase; an FBI agent might have seen the bank transfers. But there was nothing to connect the two. Lucky for counterterror agents, not to mention tourists in Orlando, the government now has software made by Palantir Technologies, a Silicon Valley company that’s become the darling of the intelligence and law enforcement communities.

The day Fikri drives to Orlando, he gets a speeding ticket, which triggers an alert in the CIA’s Palantir system. An analyst types Fikri’s name into a search box and up pops a wealth of information pulled from every database at the government’s disposal. There’s fingerprint and DNA evidence for Fikri gathered by a CIA operative in Cairo; video of him going to an ATM in Miami; shots of his rental truck’s license plate at a tollbooth; phone records; and a map pinpointing his movements across the globe. All this information is then displayed on a clearly designed graphical interface that looks like something Tom Cruise would use in a Mission: Impossible movie.

As the CIA analyst starts poking around on Fikri’s file inside of Palantir, a story emerges. A mouse click shows that Fikri has wired money to the people he had been calling in Syria. Another click brings up CIA field reports on the Syrians and reveals they have been under investigation for suspicious behavior and meeting together every day over the past two weeks. Click: The Syrians bought plane tickets to Miami one day after receiving the money from Fikri. To aid even the dullest analyst, the software brings up a map that has a pulsing red light tracing the flow of money from Cairo and Syria to Fikri’s Miami condo. That provides local cops with the last piece of information they need to move in on their prey before he strikes.

Fikri isn’t real—he’s the John Doe example Palantir uses in product demonstrations that lay out such hypothetical examples. The demos let the company show off its technology without revealing the sensitive work of its clients. Since its founding in 2004, the company has quietly developed an indispensable tool employed by the U.S. intelligence community in the war on terrorism. Palantir technology essentially solves the Sept. 11 intelligence problem. The Digital Revolution dumped oceans of data on the law enforcement establishment but provided feeble ways to make sense of it. In the months leading up to the 2001 attacks, the government had all the necessary clues to stop the al Qaeda perpetrators: They were from countries known to harbor terrorists, who entered the U.S. on temporary visas, had trained to fly civilian airliners, and purchased one-way airplane tickets on that terrible day.

An organization like the CIA or FBI can have thousands of different databases, each with its own quirks: financial records, DNA samples, sound samples, video clips, maps, floor plans, human intelligence reports from all over the world. Gluing all that into a coherent whole can take years. Even if that system comes together, it will struggle to handle different types of data—sales records on a spreadsheet, say, plus video surveillance images. What Palantir (pronounced Pal-an-TEER) does, says Avivah Litan, an analyst at Gartner (IT), is “make it really easy to mine these big data sets.” The company’s software pulls off one of the great computer science feats of the era: It combs through all available databases, identifying related pieces of information, and puts everything together in one place. . .

Continue reading. Much more at the link—strange new world, to have governments track citizens so closely. Example from later in the article:

. . . Using Palantir technology, the FBI can now instantly compile thorough dossiers on U.S. citizens, tying together surveillance video outside a drugstore with credit-card transactions, cell-phone call records, e-mails, airplane travel records, and Web search information. Christopher Soghoian, a graduate fellow at the Center for Applied Cybersecurity in the School of Informatics and Computing at Indiana University, worries that Palantir will make these agencies ever hungrier consumers of every piece of personal data. “I don’t think Palantir the firm is evil,” he says. “I think their clients could be using it for evil things.”

Soghoian points out that Palantir’s senior legal adviser, Bryan Cunningham, authored an amicus brief three years ago supporting the Bush Administration’s position in the infamous warrantless wiretapping case and defended its monitoring domestic communication without search warrants. Another event that got critics exercised: A Palantir engineer, exposed by the hacker collective Anonymous earlier this year for participating in a plot to break into the PCs of WikiLeaks supporters, was quietly rehired by the company after being placed on leave. . .

UPDATE: NPR has a two-part story on this topic. The first part discusses the uses of “Big Data.” The second part tells people who can help make sense of big data are located and groomed for the job.

Three things I noted: First, the articles make no mention of what is undoubtedly heavy government involvement (think DHS and NSA) in the entire enterprise: finding and grooming and recruiting talent, and developing applications to trawl big data looking for patterns.

Second, I would think it would be effective to use big data to find the people who can more effectively use big data: with school records, test scores, social networks, Amazon purchase history, movie-viewing history, and so on now available on-line (with a certain amount of hacking), finding people who can better find more people would be a logical direction of development.

Third, I find it interesting that the volume of digital data is driving development: new things and approaches being developed simply because the data are there.

Some people don’t seem to listen to themselves. Read this cringe-inducing story.

I blogged this well over two years ago, and yesterday I came across the recipe again: it still looks good. So here’s a repost:

Monica Shaw provides the background and also gives a very easy recipe for a healthful Bircher Muesli, following his original recipe, not the concoction available today. Take a look.

Also, here’s her earlier recipe, slightly different.

Both recipes look very good indeed, though now I would add a boiled egg for protein.