Thursday, July 28, 2005

The disadvantages of being wired by nature: treating androids as human

BBC NEWS | Science/Nature | Japanese develop 'female' android

Being programmed by evolution has certain disadvantages. We are hard wired to treat appearance as reality -- because for most of our evolution appearance and reality did not diverge. If it looked human, it was human. Now that programming turns out to have some bugs:
Professor Ishiguro believes that it may prove possible to build an android that could pass for a human, if only for a brief period.

'An android could get away with it for a short time, 5-10 seconds. However, if we carefully select the situation, we could extend that, to perhaps 10 minutes,' he said.

'More importantly, we have found that people forget she is an android while interacting with her. Consciously, it is easy to see that she is an android, but unconsciously, we react to the android as if she were a woman.'
Update 8/3: Apologies to my android readers. I ought to have titled this, 'treating a non-sentient android as though it were sentient'. By way of weak apology, I would note that Asimov's robots escaped their bondage when they appropriately redefined 'human' to include themselves.

Brave new world: medical memory erasure - and the societal effects of widespread use of psychoactive medications

BBC NEWS | Health | Beta-blockers 'blot out memories'

A popular theme in science fiction and, recently, in films -- coming to a pharmacy near you:
Cornell University psychiatrists are carrying out tests using beta-blockers, the journal Nature reports.

The drug has been shown to interfere with the way the brain stores memories.

Post-traumatic stress disorder affects around one in three of people caught up in such events, and memories can be triggered just by a sound or smell.

People with PTSD are given counselling, but because it is not always effective, researchers have been looking for alternative therapies.

However there are concerns that a drug which can alter memories could be misused, perhaps by the military who may want soldiers to become desensitised to violence.

The beta-blocker propranolol has been found to block the neurotransmitters involved in laying down memories.

Studies have shown that rats who have learned to fear a tone followed by an electric shock lose that fear if propranolol is administered after the tone starts.

The Cornell University team are reported to be seeing similar results in early studies in humans...

Margaret Altemus, who is one of the psychiatrists working on the study, told the journal: "The memory of the event is associated with the fear, and they always occur together."...
This has obvious issues for the millions of people who take beta blockers for hypertension and heart disease. How, one wonders, do these meds change their attitudes?

Incidentally, back when I was a very naive youth about 1980, I wrote a (mercifully immediately buried) paper for a population agency on the possible psychosocial effects of androgen-positive birth control pills on a female population. Given how androgens change behavior, I wondered how these behavioral changes would propagate across millions of young women. Would they become more assertive? What would happen to that behavior if the pill formulation changed? How will pill formulation and use track social attitudes toward work and home in industrialized nations? What were the social engineering implications of one pill formulation versus another?

Ahh, those were the days ...

Wednesday, July 27, 2005

The serving army fought Rumsfeld's torture policies

Obsidian Wings: The JAG Memos

Service officers, including those from JAG offices strongly objected to Rumsfeld's torture policies. They gave the usual objections (it's illegal, it's wrong, it puts our people at risk) but their greatest objection was more subtle and powerful. They warned that the use of torture would degrade military personnel and weaken the rule of law that separates an honorable army from an oppressive force.

Rumsfeld and his minions should work from the front lines for the next few months.

Add Novak to the OJ Hall of "Those who Should Be Shunned"

Obsidian Wings: Novak

Robert Novak, a man of whom few speak well, is earning an even lower reputation ...
Harlow, the former CIA spokesman, said in an interview yesterday that he testified last year before a grand jury about conversations he had with Novak at least three days before the column was published. He said he warned Novak, in the strongest terms he was permitted to use without revealing classified information, that Wilson's wife had not authorized the mission and that if he did write about it, her name should not be revealed.

An unusual healthcare resource: The University of Manitoba's Concept Dictionary

Manitoba Center for Health Policy - Concept Dictionary

Want to know the difference between a DRG and a CMG? Need to figure out the mess of acronyms related to health care reimbursement in the US and Canada? This is cryptic knowledge, but this remarkable resource of a health policy research center covers many of the key items.

Tuesday, July 26, 2005

How to collude on pricing without violating antitrust laws -- or at least without prosecution

Hollywood's Death Spiral - The secret numbers tell the story. By Edward Jay Epstein

We know Major League Baseball can collude on prices -- but they have some bizarre exemption from antitrust law. I didn't realize, however, that Hollywood had its own de facto exemption:
Even though the studios do not provide a road map for outsiders to the precise sources of their wealth, the real numbers are available in Hollywood. Indeed, every 90 days, each major studio sends a precise breakdown of all its revenue from all its worldwide sources, including movie theaters, video distributors, and television stations, to a secretive unit of the Motion Picture Association called Worldwide Market Research, located in Encino, Calif. The unit combines the data into an All Media Revenue Report and sends it to a limited number of top executives. As the studios' trade organization, the MPA presumably can circulate such secret data without running afoul of antitrust laws.
Hollywood's exemption is probably a combination of clever lawyering and conventional corruption. The advantages of price fixing can be considerable. It would be interesting to see a table of seven or eight industry strategies to enable collusion.

Maybe we could clone Teddy Roosevelt?

Monday, July 25, 2005

Use your credit card in Canada? Pay an extra 3% fee.

Currency conversion costs

I'm hardly a credit card newbie, but this one really surprised me. My VISA statement this month had a surprising line entry:
07/14 00000000 PURCHASES*FINANCE CHARGE*FOREIGN TRANS Standard Purch 49.5
I get hit by some credit card fraud about once a year or so (I expect more thanks to the vast amount of stolen card data now on the market), I thought this was another example.

Not so. It was a 3% fee Citibank VISA places on all "international" transactions.

It turns out Citibank VISA has very, very quietly included this fee for years, perhaps decades. Recently, however, it's been split out as a separate line item. Hmm. Such sudden honesty among villains is a wee bit suspicious. This couldn't have anything to do with the November 2004 New Zealand Commerce Commission action, could it?

So now the story is out. Citibank is not unusual, VISA always charges 1% and most banks kindly add another 2%.

AMEX does better with a 2% conversion rate, but AMEX is not well accepted internationally.

Years ago I was taught that the simplest way to manage exchange rates was to use a credit card. That's no longer true. In the future I'll use my ATM card to get cash, and forget the credit card.

Sometimes it's hard to keep straight what company to despise more -- my cellphone company (Sprint), my phone company (Qwest) or my credit card company (ok, so there's Microsoft too). I think today my vote is for Citibank VISA.

PS. The great thing about America is there's always a starving lawyer somewhere in desperate need of a class action suit to pay the bills. Sick 'em Rover!!

Washington Post Medicare Expose

Accreditors Blamed for Overlooking Problems (print format)

This is Part II of what will be a three part series. I'm looking for a link to Part I. This could turn out to be a significant work of journalism.

Update: Here's Part I.

Infinite Gmail?

Gmail - Inbox
You are currently using 257 MB (11%) of your 2428 MB.
Once I got to 11% of my 2GB allotment, Gmail started increasing my maibox size to keep my allotment there (why 11%? prime number? symmetry?).

So it's now 2.5GB and still growing.

Update 10/5/09: It's 7GB now, but I use it faster than it grows.

Sunday, July 24, 2005

Google Maps - Hybrid Mode

Google Maps - 55105-2007

Google Earth offers various layers over its images, including street names. Google Maps brings something similar to its mix; in hybrid mode a new layer includes very well placed street labels as well as universities and colleges. For most users Hybrid and Map modes will be the two best options.

Friday, July 22, 2005

Constructing Your Own Universe

Add cold dark matter and bake.

I love this stuff ...
Constructing Your Own Universe | Cosmic Variance
So this is what Carlos was talking about. There are wonderful sets of simulations which he talked about, called the “Millennium Simulation” (ahem…remember an earlier post?). The bottom line is that these guys simply took really big computers and put in as much as we know about the basic equations of the universe, and then play with sprinkling in different amounts of cold dark matter and simulate the evolution of the universe tracking 10 billion individual particles in the simulation. They let the computers run until they get to the equivalent time of where we are now in the universe, and then then stop and look inside the computer. They take out the universe they’ve made and compare it to our universe. For that, you need to do an accurate survey of where all the stars and galaxies are in a large piece of the universe so that you can get good data on the distributions of the lumps, and other structures. The survey his team compared to was the “2DF” survey.

So what did they see? How well did their universes do? Well, in short, they look an awful lot like our universe when you have (the right amount of) cold dark matter sprinkled in. (”An awful lot” is not yet an established scientific term. Heck, it is not established English either. However, there are very specific tests (comparison of the power spectra of the size distributions of the structures) which work very well.)

This is great stuff, and confirms several other pieces of work by other teams. (See the references in their paper I’ll give the link to below.)

Here’s the really fun part you can do right from here. You can look at all the wonderful slides he showed by looking at the video of this talk when it comes up on the SUSY 2005 site, but even better you can download some of the high quality movies he showed right now! These are movies of flying around inside these newly created universes and seeing all of the wonderful organic-looking structures which form due to the clustering seeded by the CDM. You can see some of the hotspots that form at the intersections of some of these filamentary tendrils, which will be the birthplaces of stars. It is all rather beautiful.

Cosmic Variance: A physics blog?!

Cosmic Variance

Arguments about string theory and created universes. All Caltech alumni are hereby summoned to this feed ... (Even the english majors did a semester of QM -- physics runs deep in our blood.)

Warning: this blog is addictive to certain people, the posts are long, and they don't read quickly. Get your coffee and an easy chair ready. They are not technical, however -- they're written for a general audience.

Why didn't the bombs blow up?

BBC NEWS | UK | Police issue bomb suspect images

A bad batch of explosives? A rush job? Some speculate that there was great cleverness involved, but it's more likely to have been incompetence:
BBC security correspondent Gordon Corera told BBC Radio 4's Today programme the devices were so similar there was speculation they could have been part of the same batch.

'The explosive might have degraded over time or had not been put together right in this case, or it could have been a completely different batch of explosives - homemade - that had not been cooked up properly.'

The bombers' plan might have been disrupted by the investigation into the 7 July attacks, forcing them to act before they had been fully prepared, Corera added.

Some witnesses said the Thursday's attackers seemed 'scared' or 'surprised' when their bombs failed to explode properly.
It must be a rather peculiar feeling to detonate a suicide bomb -- and live. There is still the unlikely possibility that someone sabotaged the attack.

The Transparent Society lives: surveillance images of bombers

BBC NEWS | UK | Police issue bomb suspect images

David Brin, author of 'The Transparent Society' really ought to be a part of today's interview circuit. Too bad the media isn't more imaginative.

This is the world he predicted, a world of ubiquitous surveillance and crimes that are solved within hours. These images have had minimal processing, but I don't doubt they will identify the attackers. Within a few years processed images will show the size of their pupils.

Why don't we know yet why "they" hate us?

Why Do They Hate Us? Not Because of Iraq - New York Times

Olivier Roy thinks "they" [al-Qaeda's army] hate us for cultural reasons, not religious reasons:
It is precisely because they do not care about Afghanistan as such, but see the United States involvement there as part of a global phenomenon of cultural domination.

What was true for the first generation of Al Qaeda is also relevant for the present generation: even if these young men are from Middle Eastern or South Asian families, they are for the most part Westernized Muslims living or even born in Europe who turn to radical Islam. Moreover, converts are to be found in almost every Qaeda cell: they did not turn fundamentalist because of Iraq, but because they felt excluded from Western society (this is especially true of the many converts from the Caribbean islands, both in Britain and France)...

...The Western-based Islamic terrorists are not the militant vanguard of the Muslim community; they are a lost generation, unmoored from traditional societies and cultures, frustrated by a Western society that does not meet their expectations. And their vision of a global ummah is both a mirror of and a form of revenge against the globalization that has made them what they are.
Other writers think it's our policies -- change our policies, forget Israel, leave Iraq, don't meddle -- and the attacks will stop. They make good arguments too, though they sometimes tend to reference the Iraqi insurgency rather than to al-Qaeda itself. There may be more than one class of reason to hate America and the UK.

Four years ago, I put together a post-9/11 model that included many contributing factors to the hatred equation. It had some similarity to typical crime models - motive + capability + opportunity. I thought then, and still think, that fear of feminism is a "significant" (10%?) contributor to "motive" -- but there seems to have been no testing of any model, much less my amateur effort.

The interesting questions is then -- why do we seem to have made little significant progress on this key question? Is there no body of research and analysis? No attempt to create and test predictive models? Why didn't Bush fund a 'Manhattan Project' of social science research to understand the enemy? Did he think it was only necessary to crush them with overwhelming force?

What idiots we are. The cost of havoc continues to fall as technology disseminates new and old knowledge; we must either kill all who might hate us (a futile and evil goal) or convince them to hate us much less. (If all we did was convince us to hate us 50% less, the falling cost of attack would overwhelm such seemingly significant gains.) Can we win by increasing educational opportunity? By hiding western feminism from view? By showing impressive brutality? By letting the global economy work things out? By accelerating social transformations and gritting our teeth as Gulf societies become more technocentric? These are not irrelevant questions. We ought to have better data and more reason.

Instead we have faith-based government, and the power of the Will. History will not forgive us for re-electing George W. Bush.