Sunday, June 04, 2006

People not like you and I ...

Three alleged humans have solved a five dimensional Rubik's "cube". A solution takes "thousands" of twists. I'd like to see these folks on TV somewhere ...

Relational and feature memory

Researchers of memory now think of it in terms of relational and feature memory.

Sounds good to me. I've long thought I was pretty good at remembering relationships between things, and average at remembering attributes or features of things. Researchers now believe these are handled by different neural subsystems.

Where do you hide a 300 mile crater?

A mile beneath the East Antarctic ice sheet. Theory is that a 30 mile meteor smacked the earth 250 million years ago, splitting Gondwana's plate and causing the Permian-Triassic extinction.

The surprising part is that 30% of land vertebrates survived! Land critters are hard to kill, 90% of sea critters died.

The crater sounds pretty real, its relation to the extinction is said to be more controversial.

Saturday, June 03, 2006

Da Vinci Code and the Catholic Digest: Oddly revealing

I find this oddly revealing, and a bit poignant.

I'd noted earlier that one good feature of the Da Vinci code is that it's causing some christians (probably just Catholics really) to examine the history of how their religion developed. In contrast to the often terrible and bloody history of the Catholic church, the history of Catholic thought is somewhat encouraging. So I was curious when I came across a version of that history at my parents house, in a pamphlet written by Catholic Digest. (I'm an agnostic/pantheist/atheist, but my mother is Vactican II Catholic - hence the pamphlet. I also grew up learning at my Quebec Catholic high school that the Children's Crusade was a noble endeavor, so I have a well-earned skepticism of church propaganda.)

The Catholic Digest represents one aspect of the modern Catholic church. I'd guess it's relatively mainstream. The pamphlet is a response to the Da Vinci code.

It's very well done, and it's quite fascinating, even erudite. Where else can one read, in about 3 brief pages, about early Jewish vegetarian Christians (the Ebionites), Marcion who felt that Yahweh was completely unrelated to the God of Jesus (isn't that obvious?), adoptionists who felt Jesus was born human and adopted by God, Docetists who claimed Jesus was faking suffering, the Shepherd of Hermas, the Letter of Barnabas, and Athansias of Alexander? And, of course, those Gnostics.

Pretty good stuff.

So the weird part? You have to order the pamphlet or pick it up at your local Catholic Church. In a few minutes of looking I couldn't find a web version on the Catholic Digest website. I'd naively thought it would be a big link on page one, or at least a link from where they sell the pamphlet.

My best guess is that they want the money for the pamphlet, so they didn't put it online. My next guess is that they really don't want it to be read without the intercession of a priest (or, since priests are rare these days, some other intercessor). Ironically, and I say this with sympathy, both attitudes are historically very Catholic.

It's a shame really. I hope they relent and put the text online, with links to additional educational material. On the other hand, they know their flock better than I. Perhaps they fear that could be more dangerous than a bestseller (which I'll probably never read, but which I now have great respect for).

Friday, June 02, 2006

Was the 2004 Election Stolen?

Robert F Kennedy Jr is not a rigorous thinker. His writings on autism are perfectly awful. So I'd prefer to have read about GOP election fraud from another source:
Rolling Stone : Was the 2004 Election Stolen?

... The mounting evidence that Republicans employed broad, methodical and illegal tactics in the 2004 election should raise serious alarms among news organizations. But instead of investigating allegations of wrongdoing, the press has simply accepted the result as valid. ''We're in a terrible fix,'' Rep. Conyers told me. ''We've got a media that uses its bullhorn in reverse -- to turn down the volume on this outrage rather than turning it up. That's why our citizens are not up in arms.''

The lone news anchor who seriously questioned the integrity of the 2004 election was Keith Olbermann of MSNBC. I asked him why he stood against the tide. ''I was a sports reporter, so I was used to dealing with numbers,'' he said. ''And the numbers made no sense. Kerry had an insurmountable lead in the exit polls on Election Night -- and then everything flipped.'' Olbermann believes that his journalistic colleagues fell down on the job. ''I was stunned by the lack of interest by investigative reporters,'' he said. ''The Republicans shut down Warren County, allegedly for national security purposes -- and no one covered it. Shouldn't someone have sent a camera and a few reporters out there?'
My guess is that the GOP had established a strong "ends justify the means" culture, and that their people took whatever measures were necessary. I'm not sure it required all that much coordination, all it took was a culture of victory at all costs. Was there fraud? It would be odd if there weren't - this is America after all. Was ther enough fraud to throw the election to Bush? Maybe, but I'll reserve judgment pending a more reliable source. Was it coordinated fraud? No clue.

Some of the most interesting evidence, by the way, comes from statistical analysis. The probability that all the exit polls would have erred to Kerry rather than Bush is said to be absurdly low (60,000 to 1)...

Thursday, June 01, 2006

Hyper-G: the road not taken

Charles Stross blew me away when he casually linked to BYTE's 1996 Hyper-G article. Hyper-G was Gopher's answer to the web ...
...Information Landscape offers an interactive, 3-D representation of the database structure. Users can 'fly' over the information hierarchy, represented as a virtual landscape.

... The color and height of specific landmarks, for example, represent document type and size. Two-dimensional maps are also standard. Any changes made to documents and databases are immediately reflected in both representations.

... Documents have attributes ... -- for example, author, keywords, and creation date -- that can be used in searches.

-- An underlying object-oriented database ensures data consistency and integrity.

You can appreciate some of Hyper-G's features only if you use a generic Hyper-G browser. Currently, two are available: Amadeus for Microsoft Windows and Harmony for the X Window system ... A client application for the Macintosh will be available soon. Generic clients are not really meant to compete with Web clients; besides the advanced navigation features, the main reason for using a generic client is authoring capability, so you can modify documents.
Ahh. Yes. The road not taken. No broken links (that's the database), real metadata (semantic web - 10 years ago), information visualization, built-in authoring ...

Oh well, it wouldn't have scaled. The web worked because it was crappy enough to scale (loosely coupled) and wonderful enough to be useful. Still, points to Stross for remembering the glories of the mid-90s. Now THAT was an era of innovation.

I was also surprised to learn that BYTE is still around in some form and that the 1994-1998 print archives are online. BYTE at its best was impossibly good, at its worst it was better than anything we have today. When we lost BYTE we lost 10 points off the world's collective tech IQ. Turns out Pournelle is still there, and still trying to get his computers to work. Don't confuse the web site, no matter how noble, with the print version. Do, however, read some of the 1995 to 1996 print archives. Those were the glory days ...

Charles Stross on the future of the identity card

Charles Stross writes science fiction. He's terribly bright. He's written a short essay on how identity cards, like our REAL ID, will work (not) in about 10 years. Some of it is more plausible than others (I think American's won't resist as much as his characters do), but it's persuasive.

I need to add him to my reading list and bloglines list.

Wednesday, May 31, 2006

DeLong on how to be a treasury secretary

Brad DeLong, a Clinton insider, has written a fascinating essay on how to be a Rubinesque treasury secretary. Whether or not the method would work, it's a unique introduction to high power politics.

New Orleans will not be entirely rebuilt

It's looked for a while as though there was reluctance to rebuild on the lower parts of New Orleans. This may end the debate:
New Orleans sinking faster than expected:

... The research, being published Thursday in the journal Nature, is based on new satellite radar data for the three years before Katrina struck in 2005. The data show that some areas are sinking - from overdevelopment, drainage and natural seismic shifts - four or five times faster than the rest of the city. And that, experts say, can be deadly.

'My concern is the very low-lying areas,' said lead author Tim Dixon, a University of Miami geophysicist. 'I think those areas are death traps. I don't think those areas should be rebuilt.'

For years, scientists figured New Orleans on average was sinking about one-fifth of an inch a year based on 100 measurements of the region, Dixon said. The new data from 150,000 measurements taken from space finds that about 10 percent to 20 percent of the region had yearly subsidence in the inch-a-year range, he said...

Molly Ivins: Enron's President Bush

Molly Ivins makes the populist case. It's a strong case these days. I liked the summary of the campaing contributions Enron and its execs made. Note the sum of the executive contributions (bolded by me):
WorkingForChange-The takeover is complete

...The extent to which not just state legislatures but the Congress of the United States are now run by large corporate special interests is beyond mere recognition as fact. The takeover is complete. Newt Gingrich and Tom DeLay put in place a system in which it's not a question of letting the head of the camel into the tent -- the camels run the place.

It has all happened quite quickly -- in less than 20 years. Laws were changed and regulations repealed until an Enron can set sail without responsibility, supervision or accountability. The business pages are fond of trumpeting the merits of 'transparency' and 'accountability,' but you will notice whenever there is a chance to roll back any of New Deal regs, the corporations go for broke trying to get rid of them entirely.

I'm not attempting to make this a partisan deal -- only 73 percent of Enron's political donations went to Republicans. But I'll be damned if Enron's No. 1 show pony politician, George W. Bush, should be allowed to walk away from this. Ken Lay gave $139,500 to Bush over the years. He chipped in $100,000 to the Bush Cheney Inaugural Fund in 2000 and $10K to the Bush-Cheney Recount Fund.

Plus, Enron's PAC gave Bush $113,800 for his '94 and '98 political races and another $312,500 from its executives. Bush got 14 free rides on Enron's corporate jets during the 2000 campaign, including at least two during the recount. Until January 2004, Enron was Bush's top contributor.

And what did it get for its money? Ken Lay was on Bush's short list to be energy secretary. He not only almost certainly served on Cheney's energy task force, there is every indication that the task force's energy plan, the one we have been on for five years, is in fact the Enron plan. Lay used Bush as an errand boy, calling the governor of Texas and having him phone Tom Ridge of Pennsylvania to vouch for what swell energy deregulation bills Enron was sponsoring in states all over the country...
I like my campaign reform proposals better than Molly's generic appeal to public financing.

A diversity of minds: are humans less alike than we imagine?

Like everyone else who's worked with autistic children, I have a theory about autism. I have no data and no real expertise, so I'm free to speculate.

First of all, I think it's really not one thing. It's probably several disorders of brain development, largely arising from genetic and intrauterine effects, that manifest with a few common features and many atypical features.

That's not all that interesting, so I'll go further. I also think that what's common in autism is the brain's frantic efforts to adopt to being broken. So one or many things is/are going "wrong" (quotes are important) in brain development, but there are common compensatory mechanisms that still work. The developing brain is breaking and trying to repair itself -- all at the same time. Somethings get repaired well, some don't. Sometimes the result works -- but it works differently.

Temple Grandin's mind works differently. Grandin is a famous interlocutor for autistic persons. She has several books and essays out, including this one: My Mind is a Web Browser: How People with Autism Think. Grandin has since recognized that she's really describing how she thinks, and that not all autistic persons think the same way [1]. The point, however, is that she has an "alien" way to thinking. It doesn't work all that well most of the time, but it is well suited to certain types of problem solving. She is a visualization 'savant' [2], in a way that some autistic persons are numerical 'savants'. She has enough symbolic reasoning and language to be able to translate from her world of visualization to the more common world of language.

Variations in neurodevelopment resulting by "malfunction" and adaptation/healing, yielding diverse minds that function better in some environments, worse in others.

Remind you of anything?

Isn't that how natural selection works?

We think that humans haven't evolved much in the past few hundred thousand years because our bodies seem similar. We are creatures of the brain though. What if there's an adaptive advantage to a "flaky" neurodevelopmental process? Maybe there's a reason we have so many "malformed" brains. It's not only that brains are hard to make, but also that there's a species advantage to having diverse brains and minds being created - even if some individuals bear a heavy price for being maladapted to the common environment.

Perhaps if we truly understood the human mind, we would discover that we are much less alike than we think.


[1] Grandin is nothing if not definite, but she also changes her mind. It's an interesting trait, and very scientific in a way. Overgeneralizing from self to all is a classic autistic trait, and not uncommon in neurotypicals either.
[2] It would be interesting to see how well Grandin could talk to a dolphin, presumably they think in sonar.

Tuesday, May 30, 2006

War crimes at Hadithah: not unique, inevitable

This I too believed before I read Arkin. I suspect it's what wise military leaders also believe about the marine revenge killings that occurred at Haditha [emphases mine]...
Early Warning by William M. Arkin - washingtonpost.com

...The truth of the matter, I hear from military sources, and an explanation I suspect is completely true, is that what happened in Hadithah that day has happened more times than the Marines and the Pentagon would like to admit, and more times certainly than the American public would like to admit.

And that's the issue.

American volunteer soldiers are fighting a frightening and frustrating battle against a never depleting and highly motivated enemy. The enemy is not in uniform, chooses to fight on a civilian battlefield, intentionally using civilians as fodder and shields to manufacture enough blood and chaos to drive the conventional army from the country.

The fog of war in Iraq, and part of the inhumanity particularly of the current situation on the ground is that the enemy chooses to look like everyone else on the streets, thus eliminating the fundamental element of "distinction" between civilian and military that is so essential to fight any kind of a just and humane war.

We are left then with the U.S. military, the finest conventional military force on the planet, a force that does more to train and prepare and comply with the law of war than any other country, a force that is uniquely accountable not just to the American people but to the entire world.

I suspect in this impossible war there are hundreds of incidents of accidental and intentional killings of civilians that have gone by without an investigation of any sort.

This is not to say that I am excusing what happened in Hadithah in any way. Bu Hadithah happened as much because war never follows the predictable script, and war can never be fully controlled. No matter how grave the justification for war, no matter how grand the experiment, the unleashing of human beings to justifiably kill other human beings exposes an animal instinct so basic and horrifying that training and leadership and uniforms can only tentatively arbitrate.

When killing in war become murder, we can delude ourselves into thinking that a few bad apples have stepped out of the "uniform" code and need to be punished. Right now though, we should be honest with ourselves and admit the hopelessness of our endeavor and the impossible situation we have created for our soldiers.
I am 100% certain that allied forces fighting in WW II committed many similar atrocities.

The difference between then and now is that our standards are higher than they were 70 years ago (progress happens), the battle conditions for US troops are consistently worse (occupation and insurgency) and the stakes are far lower.

Time to withdraw, and I believe the US military decided that some time ago.

Monday, May 29, 2006

The Deborah Howell affair: can bloggers improve a lazy (or corrupt) journalist?

DeLong makes it clear that the Washington Post's ombudsman, Deborah Howell, was either corrupt or lazy when she parotted the GOP party line.

She was pilloried by blog, and got some nastymail. She ranted about the nasty mail -- I suspect she's distorting how nasty it was but that's not the interesting topic.

The interesting question is whether this kind of feedback loop has any hope of improving someone who's a fairly weak journalist. If it does, that's something new in journalism.

Sunday, May 28, 2006

Establishing identity: Bloomberg's DNA database

NYC Mayor Bloomberg wants biometric ID methods for all:
BREITBART.COM - NYC Mayor Advocates U.S. Worker Database

... You don't have to work _ but if you want to work for a company you have to have a Social Security card,' he said. 'The difference is, in the day and age when everybody's got a PC on their desk with Photoshop that can replicate anything, it's become a joke.'

The mayor said DNA and fingerprint technology could be used to create a worker ID database that will 'uniquely identify the person' applying for a job, ensuring that cards are not illegally transferred or forged...
He's wrong about the SSN; even if one were as wealthy as Bloomberg, and didn't need to work, it's difficult to live in the US without having at least a fake SSN. The SSN is used very widely now; in fact one advantage of a true national ID number is that it would make visible abuses that are now obscure.

In any case, he's right that we will do biometric identifiers one day -- whether iris scans or thumbprints or DNA fingerprints. Our failure to deal now with identity theft (by, for example, making banks liable for financial losses) will eventually make Americans happy to accept an identity management solution they might otherwise have refused. Funny how that works.

In any case Americans would gladly trade their privacy for convenience -- really, Americans are not very private people. The convenience/identity theft issues are a double whammy for biometric identification.

Oh, yeah, and "security" and immigration too. Make that a quadruple whammy, even though Schneier makes a good case that the security benefits are marginal or negative. (When identity is trusted you can do far nastier things once you breach the system borders than when it's not protected.)

Eventually we will also do a full DNA registry, which will be handy for identifying children likely to sin; including identifying future secular humanists. and finding evildoers and rebels through their kin. Might as well sign up now. I'll be registering for the 'authenticated flyer program' myself...

PS. I was going to write something caustic about the NRA's belief that owning weapons was some protection against state tyranny, but then I realized they don't talk about that any more. They only talk about using guns against fellow citizens -- not the state. I guess they feel the state would be on their side ...

All alpha all the time - convicting the CEO

The clue, my sister-in-law says, is the salad bar at the Harvard Business School. A flock of CEO-wannabees contending for one optimal asparagus makes a rough scene. These are competitive people -- much tougher fighters than most of us. If they do make the alpha grade, riding on the dopamine infusion of power, they get much tougher.

Remember OJ Simpson? It's extremely likely that he visciously murdered two people (he lost the civil suit of course), and yet, despite strong evidence of his guilt, he never cracked. An alpha.

How many politicians, caught with both hands in the jar, stand their ground no matter what? Tough. How many senior mobsters fight for 30 years without tiring? Tough.

Victorious alphas fight to live and they live to fight. Bloody battle does not tire them, it engages them.

So it makes sense that if you're going to go after corporate criminals, you use the techniques developed for mobsters:
Tough Justice for Executives in Enron Era - New York Times

... When the former chief financial officer of Enron, Andrew S. Fastow, balked at cutting a deal with the government, prosecutors started putting pressure on his wife, Lea. She eventually pleaded guilty to income tax evasion for not reporting tens of thousands of dollars in kickback checks from one of Mr. Fastow's off-the-books schemes. Ms. Fastow went to prison for a year.
That's harsh. Very harsh. I'd have cracked long before my wife was sent to jail. But if you're going up against someone who makes the average gang-banger look like a kindergarden kid, what choices are there?

PS. Thanks to M for the "all alpha all the time" title!