Wednesday, April 07, 2010

Minneapolis - best city in the US


It's true. Portland has more riders, but weaker facilities.

St. Paul doesn't do as well, but I bet Bicycling lumped us in with the bigger Twin (most mags do).

Since I've long argued that the best single measure of a city is its bicycling facilities, this means MSP is the best city in the US.

A sure thing bet on the LHC

I suspect overhead would eat away at the profit, but this seems like a sure thing ...
The Large Hadron Collider: Phew! | The Economist

...Paddy Power, an online bookmaker, is offering odds of 11 to 10 that dark matter will be found before black holes and 8 to 1 that black holes will be first. Dark energy, a mysterious force thought to drive the expansion of the universe, trails at 12 to 1. And for those who fancy a real outside bet, the firm is also offering 100 to 1 that the machine will discover God....
To make it worthwhile though you need to bet $100K to earn $1K. A CD would have comparable returns, which probably has something to do with the odds.

Google's machine learning system

Classifiers are a classic component of traditional AI systems ....
Official Google Research Blog: Lessons learned developing a practical large scale machine learning system

... Several years ago we began developing a large scale machine learning system, and have been refining it over time. We gave it the codename “Seti” because it searches for signals in a large space. It scales to massive data sets and has become one of the most broadly used classification systems at Google.

After building a few initial prototypes, we quickly settled on a system with the following properties:

Binary classification (produces a probability estimate of the class label)

Parallelized

Scales to process hundreds of billions of instances and beyond

Scales to billions of features and beyond

Automatically identifies useful combinations of features

Accuracy is competitive with state-of-the-art classifiers

Reacts to new data within minutes...
I can think of several reasons why they named it Seti. For one, HAL was taken.

One of these things is not like the other: Gravity

In this popular account from 1860 [1]) the observation that "electricity" traveled through a medium implied that "other" fundamental forces such as light and gravity must also travel through a medium (emphases mine) ...
April 1860 -Scientific American - Electrical Theory
... The results of the experiments instituted by Sir William Grove are exceedingly curious, and must be regarded as all but proving the truth of the modern theory, which assumes that electricity is not, in any sense, a material substance but only an affection (state) or motion of the particles of ordinary matter.
If electricity is unable to pass over or through a vacuum, it is probable that all the other so-called imponderable forces—light, heat, magnetism, and possibly attraction—obey the same law, and as these agencies freely travel the interplanetary spaces, the supposition of Newton that such spaces may be filled with an ethereal form of matter receives an indirect but powerful support....
There is no ethereal form of matter in the 19th century sense however. Electricity has a relationship to fundamental electromagnetic forces, but it is not the same sort of thing. Once electricity was divided from "light" it was possible to find common models for light and magnetism.

In 2010 some people are trying to deal with the unquantifiability of gravity through a similar approach ...
arXiv blog: Gravity Emerges from Quantum Information, Say Physicists..
One of the hottest new ideas in physics is that gravity is an emergent phenomena; that it somehow arises from the complex interaction of simpler things.

A few month's ago, Erik Verlinde at the the University of Amsterdam put forward one such idea which has taken the world of physics by storm. Verlinde suggested that gravity is merely a manifestation of entropy in the Universe. His idea is based on the second law of thermodynamics, that entropy always increases over time. It suggests that differences in entropy between parts of the Universe generates a force that redistributes matter in a way that maximises entropy. This is the force we call gravity.

What's exciting about the approach is that it dramatically simplifies the theoretical scaffolding that supports modern physics. And while it has its limitations--for example, it generates Newton's laws of gravity rather than Einstein's--it has some advantages too, such as the ability to account for the magnitude of dark energy which conventional theories of gravity struggle with.

But perhaps the most powerful idea to emerge from Verlinde's approach is that gravity is essentially a phenomenon of information.

Today, this idea gets a useful boost from Jae-Weon Lee at Jungwon University in South Korea and a couple of buddies. They use the idea of quantum information to derive a theory of gravity and they do it taking a slightly different tack to Verlinde.

At the heart of their idea is the tricky question of what happens to information when it enters a black hole. Physicists have puzzled over this for decades with little consensus. But one thing they agree on is Landauer's principle: that erasing a bit of quantum information always increases the entropy of the Universe by a certain small amount and requires a specific amount of energy.

Jae-Weon and co assume that this erasure process must occur at the black hole horizon. And if so, spacetime must organise itself in a way that maximises entropy at these horizons. In other words, it generates a gravity-like force.

That's intriguing for several reasons. First, Jae-Weon and co assume the existence of spacetime and its geometry and simply ask what form it must take if information is being erased at horizons in this way.

It also relates gravity to quantum information for the first time. Over recent years many results in quantum mechanics have pointed to the increasingly important role that information appears to play in the Universe.

Some physicists are convinced that the properties of information do not come from the behaviour of information carriers such as photons and electrons but the other way round. They think that information itself is the ghostly bedrock on which our universe is built.

Gravity has always been a fly in this ointment. But the growing realisation that information plays a fundamental role here too, could open the way to the kind of unification between the quantum mechanics and relativity that physicists have dreamed of..
In short, one way to deal with the gravity problem is to make gravity go away. It's merely a confusing epiphenomena.

Hey, it worked for electricity ...

[1] Just prior to the American civil war. In a few years the SciAm "150 year back" article excerpts will be all about military science relevant to the most bloody battles of the 19th century. That should be interesting.

Monday, April 05, 2010

Tech churn: Cameras yes, but laptops?

Until I saw this graphic I thought the compact, point and shoot, camera was doing just fine.

Right. Of course. My iPhone 3G isn't good enough to replace a point and shoot, but my June 2010 iPhone will. I haven't been paying attention.

That's tech churn. From nowhere to everywhere to nowhere in no time at all. Reminds me of books on CD.

So is Amit Gupta right that the iPad has made the laptop obsolete? I don't think so. A good laptop is much closer to a desktop than a Point & Shoot camera is to a DSLR; there's room for both in the "pro and enthusiast" market. I do agree that specialty OSs such as ChromeOS and iPhoneOS will dominate the "everyday use" end of the market. The form factor is, however, less critical than the usability, security, business model and low operating costs of these next generation environments.

Remember too that, when the market demands it, an iPhone will support a keyboard and external monitor. Then you have a phone and a laptop ...

Analyzing Financial Statements

I've had "learn to read financial statements" as a medium priority goal for a while. A VC has recently written a mini treatise on how to do this:
Analyzing Financial Statements

... This topic could be and is a full semester course at some business schools. It is a deep and rich topic that I can’t cover in one single blog post. But it is also a relatively narrow skill set at its most developed levels. If you are going to be a public equity analyst, you need to understand this stuff cold and this post will not get you there.

But if you are an entrepreneur being handed financial statements from your bookkeeper or accountant or controller, then you need to be able to understand them and I’d like this post to help you do that. I’d also like this post help those of you who want to be more confident buying, holding, and selling public stocks. So that’s the perspective I will bring to this topic.

In the past three weeks, we talked about the three main financial statements, the Income Statement, the Balance Sheet, and the Cash Flow Statement. This post is going to attempt to help you figure out how to analyze them, at least at a cursory level...
I've put time on my calendar to read and study these posts. (In my life only scheduled things happen. Your life may vary.)

Sunday, April 04, 2010

A conscience for robots - and for humans too

It costs a huge amount of resources to develop a functioning human adult. That's a price of being a social animal with a massively overclocked brain.

Not surprisingly, evolution has made us feeble killers. Our teeth, claws, and muscles are pathetic. Most adult humans are traumatized when they kill -- even when it's in self-defense and even after extensive training and conditioning. Only immature children kill easily, which is why child soldiers are valued by the world's most evil men. (See also: vertromedial injuries).

On the other hand, evolution hasn't had time to adjust to murder at a distance. Pilots do not seem to experience the trauma felt by marines, even though they may kill many more people. This has advantages for warfare, but there are problems with making killing too easy. These problems are showing up with drone use ...
Remote-control warfare: Droning on | The Economist

... If they have not been so commandeered, attacks on such sites may constitute war crimes. And drone attacks often kill civilians. On June 23rd 2009, for example, an attack on a funeral in South Waziristan killed 80 non-combatants.

Such errors are not only tragic, but also counterproductive. Sympathetic local politicians will be embarrassed and previously neutral non-combatants may take the enemy’s side. Moreover, the operators of drones, often on the other side of the world, are far removed from the sight, sound and smell of the battlefield. They may make decisions to attack that a commander on the ground might not, treating warfare as a video game.

Ronald Arkin of the Georgia Institute of Technology’s School of Interactive Computing has a suggestion that might ease some of these concerns. He proposes involving the drone itself—or, rather, the software that is used to operate it—in the decision to attack. In effect, he plans to give the machine a conscience.

The software conscience that Dr Arkin and his colleagues have developed is called the Ethical Architecture. Its judgment may be better than a human’s because it operates so fast and knows so much. And—like a human but unlike most machines—it can learn.

The drone would initially be programmed to understand the effects of the blast of the weapon it is armed with. It would also be linked to both the Global Positioning System (which tells it where on the Earth’s surface the target is) and the Pentagon’s Global Information Grid, a vast database that contains, among many other things, the locations of buildings in military theatres and what is known about their current use.

After each strike the drone would be updated with information about the actual destruction caused. It would note any damage to nearby buildings and would subsequently receive information from other sources, such as soldiers in the area, fixed cameras on the ground and other aircraft. Using this information, it could compare the level of destruction it expected with what actually happened. If it did more damage than expected—for example, if a nearby cemetery or mosque was harmed by an attack on a suspected terrorist safe house—then it could use this information to restrict its choice of weapon in future engagements. It could also pass the information to other drones.

No commander is going to give a machine a veto, of course, so the Ethical Architecture’s decisions could be overridden. That, however, would take two humans—both the drone’s operator and his commanding officer...
Even if this particular implementation doesn't succeed, it makes a great deal of sense to build in this kind of automated oversight.

Obviously it's not simply of interest to weapons, though that's where the initial funding will come from. Even if we don't get sentient machines in the next fifty years (if you're the praying type, pray we don't), we will be deploying systems that make many risk/benefit trade-offs in many contexts. We will benefit if they evolve a conscience.

Some humans too would benefit from a prosthetic conscience. It might allow persons with disorders of conscience to function more effectively in the modern world. Our prisons are full of low IQ individuals with a limited capacity to model the impacts of their actions on other persons. A prosthetic conscience might allow them to avoid prison, or to have great success after prison life.

Of course if we do develop non-human sentience, it might be very much to our advantage if they felt qualms about hurting us ...

Thursday, April 01, 2010

South Dakota - state of denial

Just west of the Minnesota Saint Paul border is the Catfish Bay water ski park.

It's a puddle with a grandstand.

Some obsessed maniac dug a hole in the ground and turned it into a water skiing attraction - in South Dakota. A few hours west of the land of 10,000 lakes.

It's insane. It's pathetic. It's glorious.

The Corn Palace. Wall Drug. Mount Rushmore. Crazy Horse Memorial. They're all in South Dakota, at the heart of the Buffalo Commons. They're all crazy. Doomed.

Just like humanity, not to mention our ever dissolving universe.

Catfish Bay is an honored member of the trans-galactic museum of the Deniers of Doom.

Monday, March 29, 2010

Apple's next surprise: A Facebook alternative

It's easy to make fun of Google's "Don't be evil" philosophy. Just like it was easy to make fun of Al Gore.

Remember how "fun with Al Gore" turned out? We got Dick Cheney.

Maybe we should give Google a bit more credit, because the main rival to Google is the Dick Cheney of the online world - Facebook. Their motto isn't so much "be evil" as "whatever it takes".

Facebook is bad. I no longer suggest friends join up. Fortunately most of my friends who did join are drifting away, tired of account hacks, app scams, and the hopeless task of working with a demonic vendor (shades of pre-reform health insurance). Facebook's numbers keep rising, but there are fewer people I know participating. That's a bit funny. Does anyone audit that stuff?

It's a bit of a shame though. There are a lot of non-App things that Facebook did that I like. I hoped Google would give us a real alternative, but instead we got Buzz. There's more wrongness at Google than we know about.

So if Google can't do it, who has the brand and the money and the reach and the business savvy and just-enough evilness to deliver Facebook type services and a first-class user experience without Facebook-class evil?

Cough.

Ok, so I admit Apple has struggled with the online world for a while (emphases mine) ...
AOL - Wikipedia, the free encyclopedia 
... In May 1988, Quantum and Apple launched AppleLink Personal Edition for Apple II and Macintosh computers. In August 1988, Quantum launched PC Link, a service for IBM-compatible PCs developed in a joint venture with the Tandy Corporation. After the company parted ways with Apple in October 1989, Quantum changed the service's name to America Online ...
I used AppleLink, it was pretty good. So too was 1990 AOL for that matter. Since then, not so good. Online hasn't been Jobs thing.

Still. We know Apple has built data centers they're not yet using. We know Apple has an increasing consumer platform reach with the iPhone and the iPad. We know Apple isn't adverse to entering a market and turning it upside down.

I believe Apple is going to go up against Facebook. Soon.
--
My Google Reader Shared items (feed)

The Lonely Planet blog

I had my time to wander, long ago. I used it well. I kept my Lonely Planet books at hand.

I travel different roads these days, but I've not forgotten the road. I've only now discovered that two years ago LP launched the Lonely Planet blog. The web site is slow to load, but the posts do fine in Google Reader.
--
My Google Reader Shared items (feed)

Sunday, March 28, 2010

Health care reform – the road ahead

Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning.
Winston Churchill, 1942.

One way or another I'm gonna find ya
I'm gonna getcha getcha getcha getcha
.
Blondie.

In the original Senate version of the Health Care reform bill we find this clause …

Subtitle D: Improvements to Medicaid Services - (Sec. 2301) Requires Medicaid coverage of: (1) freestanding birth center services …

Yes, there are a lot of funny bits in this sausage. We don’t know the half of ‘em, but I do know my family will pay for them. Now that the bill done (for now) I can say that the wingnuts were right about some things; my team’s “$250,000 and above” slogan was nonsense. “One way or another” upper 30th percentile “tax-equivalents” [1] will go up.

I’m good with paying for this; I expected to pay for it when I campaigned for Barack Obama. Altruism aside, there are tangible personal benefits for my family:

What’s next? We’ve been stuck in the headlights for a generation, but now we’re moving. We might be charging towards the oncoming truck, but even that may be an improvement on standing still. At least with motion comes opportunity.

The question isn’t where we need to go. We’ve known that for at least thirty years. We’re going to the same place as everyone else on earth – good enough care for everyone and luxurious care for those with money. The question is how we get there.

Until now the American people have been completely unwilling to think about health care cost. It would have been nice if we could have cut costs before increasing access, it would have been nice if we’d come up with a reform plan that made it easier to cut costs, it would be nice if we weren’t charging towards the right fender of that oncoming truck. Nice – but not going to happen in a world where the GOP has gone mad.

Now, however, we’ll all be, at last, considering costs and value. It won’t be pretty, but if it were pretty it wouldn’t be real.

That’s progress.

See also:

Post-passage commentary

Older general discussions that are still relevant

The future: “pretty good care”, aka “good enough care” – where we’re going

[1] Federal “Taxes” rarely include anything so obvious as a rate increase. Tax equivalents include

  • actions that shift services burdens to the states – which either reduce state services or increase state or local taxes
  • unfunded federal state or local mandates of all varieties including regulatory or reporting burdens
  • means testing that remove tax breaks (the AMT is the mother of means testing)
  • elimination of tax dodging programs such as the Flex plan we enroll in (I’ll be glad to see that evil scam die)
  • user fees
  • service taxes (such as the 10% tanning tax – which, amusingly, is aimed squarely at the Tea Party demographic of less educated paleskins).

Rwandan genocide and the Zani score

I made up the "Zani score" as "... a measure of how concerned one should be about nascent entities or organizations. It tries to measure social structures that, 999/1000 times go nowhere, but 1/1000 times lead, given chance and circumstance, to very bad things..."

I imagined it in the context of an industrial society, but I asked a friend who's an expert on the Rwandan genocide to try to apply the metric to that setting:
  1. A belief that the ends justify the means, or, in other words, "extremism in the defense of liberty is no vice". NOT SURE.
  2. A sense of grievance and injustice. YES.
  3. A charismatic leader. NOT REALLY. NO ONE WAS CHARISMATIC, BUT AUTHORITY WAS TOTAL.
  4. Celebration and admiration of violence. WELL, SINCE 1990. IT WAS SPECTACLE FOR SOME.
  5. Tribal or ethnic boundaries; a division into the "chosen" and the "other". YES.
  6. Anti-intellectual, in particular anti-geek. YES, INTELLECTUALS WERE AMONG THE FIRST KILLED.
  7. Denial of skepticism. Skeptics are outcast, dissent is forbidden. NOT SURE.
  8. Welcoming and affirmation of the convert. NONCONFORMITY WAS NEVER ENCOURAGED GENERALLY.
  9. Membership alone is proof of virtue. IN TERMS OF ETHNICITY.
  10. Scorn for the weak; denial of pity or sympathy for the other. NOT SURE.
The only real conflict to the model was that the leader of the genocide was not particularly charismatic -- but since he had total authority that wasn't much of an obstacle. There are a number of unknowns, but not bad for something that was designed with a very different society in mind.

My Google Reader Shared items (feed)

Saturday, March 27, 2010

Three NYT OpEds on the latest American extremism

Shortly after writing my post on the Zani score, the NYT has published 3 OpEds over 3 days on the latest Zani-score bumps in American culture:


The GOP's Zani score is probably about five. We should really worry if it hits 8.

Incidentally, Palin's Facebook page still uses gun sight cross hairs to mark our her enemies.
--
My Google Reader Shared items (feed)

Miep Gies and the Zani score

At age 50, at a local theater, I attended a performance of The Diary of Ann Frank.

I knew the story of course, but, until now I'd missed the book, the movie and the play. Seeing it at this point in my life I am awed by the endurance and compassion of Otto and Edith Franck, sympathetic to the less favorably portrayed refugees, and curious about the heroes Victor Kugler, Johannes Kleiman, Jan and Miep Gies, and Johannes and Bep Voskuijl. Curious too about what kind of man Otto Franck was to create such love and loyalty in his employees.

Of the heroes we know the most about Miep Gies, in part because of her astounding longevity. She passed for an ordinary person before and after World War II. She claimed, somewhat convincingly, that she was motivated not by courage but by a fear of unbearable guilt should she fail to perform her duty. It may be relevant that she was, by necessity, given up for adoption by her birth mother.

I wondered then, and wonder now, how extraordinary Gies was. In coverage of her death this past January I recall that of 81 people asked by the Dutch resistance to shelter Jews, 7 accepted. Clearly they did not ask just anyone; if we guess that only 1/10 were considered candidates, and 7/81 of those accepted, then Gies-class heroes were, and are, perhaps 1/100. Unusual certainly, but more common than world class athletes.

That feels right. I can believe that somewhere between 1/30 and 1/100 of humans are heroes born, and another 1/10 to 1/20 heroically inclined. Likewise it feels like 1/5 of us are Nazi-capable and 1/50 Nazi born.  The rest of us, in most circumstances, favor the good. Which is why civilization is possible.

I expect the epidemiology of heroism has been studied by scholars of later genocides in Cambodia and Rwanda. I know one such, so maybe I'll update my post with some real data.

Do the demographics of hero and villain vary by society? Obviously some societies are far more evil than others; Germany of 2010 is not Germany of 1940. I would not be surprised to learn, however, that the frequency of fundamental human heroism and villainy is fairly constant. It might instead be chance and circumstance that leads to the rare, but cataclysmic, ascendance of the villainous.

Could villains win in modern America? Obviously yes. Even if there had been any past doubts, the recent widespread public support for governmental torture has put them to rest. We, like most nations, are quite capable of industrial evil.

Given that we Americans, like most nations, have a low but real risk of repeating the worst of modern human history, shouldn't we put some measure in place so we can estimate and track our risk?

We can't call this the "Nazi score" because the word Nazi has too much baggage. It cannot, for example, be applied to readily applied to Israel and it is historically bound to a peculiar form of industrial organization. In any event  a Nazimeter score would be a Godwin's Law violation.

Still, the lessons of Nazism are so powerful, and so often studied, that it would be insane to ignore them. So I'll permute some characters and name this metric the Zani score.

It only remains then, to assemble the metric. Tradition dictates a 10 point scale, so we need to come up with 10 distinct indicators of roughly equal weight. As a rough guide we can assume that the National Socialists get 9-10 points and the American Tea Party movement must score less than 5.

Given that rough outline here's my start on the 10 indicators that sum to a Zani score for any social movement or organization. Suggestions are most welcome and I hope to refine the scale over time.

  1. A belief that the ends justify the means, or, in other words, "extremism in the defense of liberty is no vice".
  2. A sense of grievance and injustice.
  3. A charismatic leader.
  4. Celebration and admiration of violence.
  5. Tribal or ethnic boundaries; a division into the "chosen" and the "other".
  6. Anti-intellectual, in particular anti-geek.
  7. Denial of skepticism. Skeptics are outcast, dissent is forbidden.
  8. Welcoming and affirmation of the convert.
  9. Membership alone is proof of virtue.
  10. Scorn for the weak; denial of pity or sympathy for the other.
Any suggestions on additions or deletions? Does anyone know of a genuine, empirically tested, Zani metric?

Thursday, March 25, 2010

Dinosaurs born of CO2

The new bit is the Triassic transition, but an Economist article is a handy summary of current thoughts on the Permian and later extinctions (emphases mine):

Economist: Rise of the dinosaurs

… This cycle has happened five times in the history of modern life. The most famous occasion was 65m years ago, when the dinosaurs were wiped out and the mammals emerged victorious from the wreckage. A bigger mass extinction, at the end of the Permian period 251m years ago, killed 70% of the world’s land vertebrates (and 96% of all marine animals) and paved the way for the age of reptiles.

Exactly which sort of reptile would come out on top, however, was not something that was decided until later—201.4m years ago, to be precise. This was towards the end of the Triassic period. Then, the ranks of aetosaurs, phytosaurs, shuvosaurs and many other uncrocodile-like relatives of the crocodiles were suddenly thinned, and a previously obscure group came to the fore. The result, once natural selection had done its work over the course of millions of years, was the now familiar cast of Allosaurus, Diplodocus,Triceratops and Tyrannosaurus rex

… The dinosaurs were done for, as everybody knows, by a collision with an asteroid. The Permian was curtailed by massive volcanism. But what exactly happened towards the end of the Triassic has been much debated. A study just published in the Proceedings of the National Academy of Sciences, by Jessica Whiteside of Brown University in Rhode Island and her colleagues, pretty well nails it down. It was the geological chaos that created the North Atlantic Ocean.

… The initial volcanism as North America split from Europe released carbon dioxide from deep inside the Earth. That produced a greenhouse effect which, in turn, melted seabed structures known as methane clathrates, which trap that gas in ice. This caused a massive release of 12C-rich methane into the atmosphere, explaining the initial drop in 13C concentrations. The methane, being a much more potent greenhouse gas than carbon dioxide, exacerbated things, while the carbon dioxide acidified the oceans, killing most of the animal shellmakers and fertilising the photosynthesis of planktonic plants. The subsequent plankton bloom sucked up the12C and the isotope ratio veered off in the opposite direction.

The greenhouse warming and the acid rain also did for the forests and many of the reptiles. Only once things had settled down could the survivors regroup. New species of trees took over. The forests grew back. And a bunch of hitherto not-so-terrible lizards began their long march.

So a spike in CO2 from deep sources led to a methane spike. Together the two baked and acid burned the planet. A plankton bloom sucked down the CO2 and things settled down again.

We, of course, are on track to repeat history.

CO2 or not, we are in the midst of a mass “holocene” extinction anyway. What comes from that remains to be seen, but if humans last a bit longer it might be retrospectively labeled the transition to the age of the machines.