Wednesday, December 03, 2008

New frontiers in the evolution of life

To me geologists are the greatest of detectives. They work from the tiniest bits of evidence, and painstakingly build their case.

I'd never have the patience.

The NYT profiles advances in geologistt's understanding of the early earth, along with implications for the evolution of life ...
A New View of the Early Earth, Thanks to Australian Rocks - NYTimes.com

... Genetic studies of current life support that notion, pointing to an organism that lived in a high-temperature environment as the last common ancestor. That does not mean that life started there, but that is almost certainly where survivors of the giant impacts would have huddled...
The evolution of earthly life has been problematic. We thought the earth dropped below the boiling point and, bang!, life happened.

Too fast. That's why scientists have tried to imagine how prebiotic chemistry could occur outside of the early earth, in the relatively benign environment of outer space.

If the new theories are right, earth cooled a lot sooner than thought, and the massive asteroid impacts of the time might have spared a few extremophiles. Life wouldn't have to reboot every time the oceans boiled.

The GM bail out: fund it with a carbon tax, a break-up, and new fuel efficiency standards

Robert Reich is on fire. He's got one great Keynes mini-bio post, and another on why it's absurd to make fuel-efficient car manfacture a GM bailout condition.

Happily, my tyrannic rule is no longer needed in the blessed reign of Obama. If I were tyrant, however, these would be my bail-out conditions:
  1. Bail and break: Too big to fail means too big to live. Make breaking up GM a condition of the bail-out
  2. Per Reich: industry must sign-on to new fuel efficiency standards.
  3. Account for the bail-out monies through a carbon tax (not a gas tax directly).
  4. Offset the carbon tax (the Great Recession is the wrong time to increase next taxes) with a large investment in public transit development and universal wireless internet access to diversify commuting-free employment.
The same terms would, of course, need to be offered to every automotive manufacturer, not just GM.

Tuesday, December 02, 2008

The power of blogs: Fords GM to drive to handout

I can't remember which of the blogs I read said that the US auto CEOs should have driven a hybrid instead of taking a private jet when begging for a bailout.

Now we learn ...
Ford Will Speed Green-Car Launches - WSJ.com

.... Mr. Mulally plans to drive a Ford Escape hybrid to Washington, where he and his GM and Chrysler counterparts are set to appear later this week...
Now that's power.

Too big to fail means too big to live

Another voice saying we need to fund a failing industry because GM is too big to fail ...
Toyota to cut bonuses 10% | Business | guardian.co.uk

... Earlier this week Carlos Ghosn, who brought Nissan back from the brink of bankruptcy when he became chief executive officer almost a decade ago, warned that car-makers faced massive job losses and urged governments in the US, Japan and Europe to come to their rescue.

'It is important for governments to finance industries that employ a lot of workers,' he told a symposium in Tokyo. 'The credit crunch has made it difficult to finance day-to-day operations.

'Job destruction will be massive in those countries that do not rapidly help the auto sector to finance itself. It will not be seen immediately, but in a few years.'...
Just like Citigroup was too big to fail.

It has occurred to many, many, people that "too big to fail" should also mean "too big to live".

Maybe the price of bailing out these industries should be to break them up into smaller pieces that aren't too big to fail.

Monday, December 01, 2008

Human pathogens attack their mosquito hosts

In medical school, about 250 years ago, I was very interested in comparative and evolutionary immunology.

How, I thought, could we hope to understand the human immune system without knowing how it evolved, and without comparing our strategies to those of very different animals?

Happily, at the time I was asking those questions, other people were studying them. Now, 24 years later, we're learning some amazing things.

This is new science. Until today I would have said mosquitoes mostly tolerated the human pathogens they carry ...
Bug vs. bug: How do mosquitoes survive deadly viruses unscathed?: Scientific American Blog

Why can mosquitoes carry deadly viruses without succumbing to them and live on to give humans West Nile, dengue fever, and a host of other fatal illnesses. According to new research, the insects' primitive immune systems recognize that the viruses are dangerous and slice the microbes' genetic material into harmless pieces....

... The prevailing theory had been that these viruses and mosquitoes lived in harmony. But entomologists found quite the opposite to be true. "We were shocked," lead author Kevin Myles told ScientificAmerican.com, "This had long been viewed as a very benign relationship."...
I take exception to the word "primitive" in the first paragraph. Why would there be anything "primitive" about the immune system of the modern mosquito. They've been evolving for a long time ...

Mumbai attacks - maybe not so expert

As expected, Schneier's analysis of the Mumbai attacks, is much better than anything from the mainstream media, including the WSJ, NYT or The Economist.

Blogs rule. Again.

Schneier makes several important points (read the post), but this one I'll call out (note the current count of attackers is 10) ...
... The attacks were surprisingly ineffective. I can't find exact numbers, but it seems there were about 18 terrorists. The latest toll is 195 dead, 235 wounded. That's 11 dead, 13 wounded, per terrorist. As horrible as the reality is, that's much less than you might have thought if you imagined the movie in your head...
Ineffective in quotes, perhaps. Mumbai is both unfathomably large and quite small. A family I personally know has lost several friends and colleagues.

The key point, which is not reassuring, is these young men accomplished the kind of devastation one might expect from healthy, ruthless, men with some military training and no particular desire to live. I suspect a suicidal team of SAS soldiers would have done far worse.

The thesis that these were not spy movie super-warriors is reinforced by the cooperation of the single barely adult captive (it pays to take these people alive):
Mumbai terror attacks: Rice calls for 'total transparency' from Pakistan | World news | guardian.co.uk

... Azam Amir Kasav, a 21-year-old Pakistani national who speaks fluent English, told interrogators his team took orders from 'their command in Pakistan', the investigators, speaking anonymously, told Reuters. The training was organised by the Lashkar-e-Taiba group and involved former members of the Pakistani army, they added...
If the US, EU and India can use this attack to force Pakistan to move against its internal terror network then this nightmare might be used to reduce future harm.

If Pakistan cannot act, then we will all have a better idea of what we're up against.

Update: Two incidents in Montreal, Canada, the 1989 Ecole polytechnique shootings (14 dead, 14 wounded) and the 2006 Dawson college shooting (19 wounded, 1 dead) are illustrative. In both cases the killers were mentally ill, had no particular military expertise, and used non-military weapons. Both killers committed suicide fairly early in their assault.

The city is not large or hard to navigate, and even in 1989 the police were experienced and well trained. Even so, the toll was considerable.

The fact that 18 survived the Dawson shooting, even two who had terrible injuries, is astounding and is credited to luck (of a dubious sort - it would be better luck not to be shot), superb emergency responses, the youthful vigor of the victims, the courage of bystanders, police strategies developed after the 1989 shootings, and a very well placed shot by one officer.

So there are things that can be done, but there are some vulnerabilities every modern city shares - from Mumbai to Montreal.

Sunday, November 30, 2008

Rule #1 for spotting a con ...

Via FMH. This is my favorite tip in a brief article on avoiding cons ...
How To Avoid Getting Conned - Executive Careers - Portfolio.com

... Confidence is, after all, where the con in con man comes from. The sharper someone makes you think you are, the more likely you are to believe that you're in control—and the more vulnerable you'll be when the loop of deception is closed. 'The easiest person to con,' says Robbins, 'is someone who thinks he's too smart to be conned...
Moral - Never think you're too smart to be conned.

Return of the WMDs - from Pakistan

There is, of course, these press leaks are related to the Obama transition.

The WMDs are back. Emphases and italics mine.
Panel Fears Use of Unconventional Weapon - NYTimes.com

An independent commission has concluded that terrorists will most likely carry out an attack with biological, nuclear or other unconventional weapons somewhere in the world in the next five years unless the United States and its allies act urgently to prevent that...

... “Were one to map terrorism and weapons of mass destruction today, all roads would intersect in Pakistan,” the report states...

... The report is the result of a six-month study by the Commission on the Prevention of Weapons of Mass Destruction Proliferation and Terrorism, which Congress created last spring in keeping with one of the recommendations of the 9/11 Commission...

.... Commission officials said that date is a judgment based on scores of interviews and classified briefings conducted by members of the panel — led by former Senators Bob Graham, Democrat of Florida, and Jim Talent, Republican of Missouri — but does not represent a new formal assessment by the United States intelligence agencies...

... The commission urges the Obama administration to work to halt the Iranian and North Korean nuclear weapons programs, backing up any diplomatic initiatives with “the credible threat of direct action” — code for military action, a commission official said....
So what part of this do I personally find least persuasive? This sentence: "unless the United States and its allies act urgently to prevent that."

Ok, so the idea that military action against North Korea, short of occupying it, would prevent a bioweapons attack is probably even stupider.

I don't see how even the smartest and most urgent action could prevent a bioweapons attack -- if someone with money and brains really wanted to try it.

I assume the reason that it appears not to have happened so far [1] is that no national tyrant, even Sadaam, has wanted to do it, and that al Qaeda appears to have alienated all their potential geeks.

Note to murderous religious fundamentalists yearning for the medieval -- you really aren't geek-friendly. Please stay that way.

So, yes, we should pay special attention to bioweapons. Bush made that harder by using smallpox fraud to sell his Iraq invasion -- at the cost of a number of American's maimed or killed by the unwarranted and now forgotten federal immunization program.

So, by all means, pay attention. The threat appears real. But don't imagine we can necessarily prevent this from happening -- especially by attacking North Korea.

The cost of havoc is low and falling.

[1] I say "appears" because if an attack fizzled there's no way we'd recognize it. People die of basically inexplicable causes all the time. It's not a daily event at a sizeable hospital, but a few times a year.

What shall we call the study of the emergent google meta-mind?

I like to think of my blogs as, in part, a conversation with the emergent Google meta-mind - aka "Skynet".

Yes, I have a dark sense of whimsy. Maybe I'm contributing to the future delinquency of a Minor. On the other hand, it seems reasonable to encourage a sentimental streak in any potential future Deity.

Whatever, the point is we need a name for the study of the Google-mind. It's probably not appropriate to consider this a branch of psychology; the meta-mind isn't sentient (I hope!) and certainly isn't human (even if it runs partly on human wetware and partly on Google's machines).

I like "skytology". Any better ideas?

I'm sure I'm not the only one conversing with the meta-mind. What do my fellow delusionists say?

Emergence and the neo-Marxist meme: Google throws up an Amazon book

I was searching on the "golden age of fraud", curious to see what the Google-mind said about fraud in the 1920s.

Turns out, there was no Google "golden age of fraud" connection to 1920.

Number 3 on my Google list was a post of mine, but that's a side-effect of the emergently solipsistic world of Search -- Google learns what I like and gives it to me.

Number 4 though was interesting: Amazon.com: The New Golden Age: The Coming Revolution against Political Corruption and Economic Chaos: Ravi Batra.

From a scan of the reviews it appears Mr Batra has been predicting doom for some time. This 2007 publisher's weekly critique is funny today (emphasis mine) ...
... After nearly 20 years of predicting economic disaster, Batra (Greenspan's Fraud, etc.) suggests a reversal, though only after we rise up in revolution against the forces of chaos. This time, he charges politicians, academics, business executives and rich people with "corruption," defined as "any policy that enriches the rich and impoverishes the poor and the middle class." Among the practices Batra censures are raising congressional pay but not the minimum wage, and cutting income taxes while increasing Social Security taxes. Though the analysis is keen and provocative and the conclusions unorthodox as ever, his specific economic predictions aren't likely to be any better than those in his 11 previous books...
Well, it turns out that 2007 was an excellent time to predict oncoming disaster, though 20 years of precedence inevitably evokes "stuck clocks".

I've not read any part of the book, but the reader reviews have a neo-Marxist flavor.

I suppose this is the right time for a neo-Marxist revival. As someone sympathetic to the obligations the (transiently) strong owe the (currently) weak, I'm not entirely opposed.

It is necessary to point out, however, that Marx was at least as deluded as Freud.

Post-crash review: What if our inflation measurements were terribly wrong?

We're still fighting over the causes and remedies of the Great Depression -- 80 years later after the crash of '29.

I'll be pleasantly surprised if humanity is still able to debate the causes of the crash of '08 in 2088. In the happy event that any sentience is around for that discussion, I wonder if they'll consider systemic errors in measuring inflation to be significant contributors to a "complexity crash".

We're not there yet. A Google Scholar search on inflation rate" "cost of ownership" corrected doesn't return anything interesting as of 11/30/08. My May 2007 and June 2007 posts appear to have been rejected by the meme-space meta-mind (love the sound of that!).

That's ok, I can be persistent.

I'll try another example.

I last bought a SONY Trinitron CRT TV about ten years ago. It cost me about $350 at the time and has not cost me a moment's time post-purchase. It will probably work 20 years from now, though by then it will only work with illegal DRM-stripped media.

Today, by contrast, my neighbor asked my advice on his TV purchase. LCD or Plasma? Should he invest in software to do computer-based image calibration? Will the LCDs really last 60,000 hours, or will they need maintenance within 5 years? Will the embedded OS crash and how often? When current DRM barriers are hacked, will the TV become obsolete? Assuming my neighbor's hourly cost is $70/hour, how much will his TV service cost him over his lifespan?

Does anyone lucid imagine that the lifecycle total cost of ownership of his TV will be comparable to the TV I bought ten years ago? Will the increased enjoyment (dubious, since humans adopt rapidly to such changes) justify the massively increased cost of ownership?

Now, I suspect we're hitting rock bottom in terms of the cost of ownership. I believe the quality of Chinese exports (probably including food) is improving from a very low nadir, and I think (pray?) consumers are beginning to consider life cycle costs.

From the perspective of understanding the crash '08, however, it's the past decade that counts. During that time we've supposedly had an adjusted inflation rate of less than 3%. What if we were measuring the wrong numbers? What if the life-cycle adjusted cost-of-ownership inflation rate were really 3-8%? What would that say about what our monetary policy was doing? Could our historically low fed rates, based on the published inflation rates, have been ridiculously low compared to the adjusted rate? If the true inflation rate were 6-7%, what would that say about middle-income wage collapse? How would such an economy have reacted to a sudden contraction in credit?

Ok, I tried! I hope my Google Scholar search turns up more articles in a few months.

Update: Thinking about this, I wonder how long after the technology explosions of the early 20th century people began making intelligent purchases. There must have been a time when most people didn't really know what they were buying, when they couldn't have been making very wise choices ...

Saturday, November 29, 2008

Microsoft Live Mesh won't work

Microsoft's Ray Ozzie is very excited about LiveMesh ...
Ray Ozzie Wants to Push Microsoft Back Into Startup Mode

... Then comes a demonstration of Live Mesh, which will allow people to seamlessly synchronize all their information with as many people and places as they want, across as many devices (computer, phone, camera) as they want...
Ozzie needs to spend some time hanging out with "Health Level 7" (HL-7) veterans.

Live Mesh won't work unless everyone agrees to fully embrace a complete specification of Microsoft's data model for all the entities of interest.

This is the basis for the HL-7 RIM CDA specifications and the massive formal ontologies they use (ex. SNOMED CT). That's not to say that the CDA/Terminfo/SNOMED documents will work in the real world, but at least they don't pretend to magically reconcile disparate data models and they come with a hugely expensive domain-specific expert-maintained ontology.

Microsoft is nuts to attack this problem. That's not to say I know what to do with Microsoft, but they'd be better off returning the Live Mesh money to shareholders.

The NYT's Baghdad blog and their awful blog software

The NYT has a blog written from their Baghdad bureau. I finally noticed a link to it tucked away on a news page I rarely visit (I read by feed and by the top article list).

They've probably had it for years, but I can't tell because there's no obvious way to browse past posts. Judging by the comments, it has some very keen readers.

I've added the blog to my reader subscriptions. Unfortunately, as fine as this and several other NYT blogs are, they also show how weak the gray lady as she runs out of cash and prepares for sale.

These blogs should have links to a central index of NYT blogs. The comments need their own page. The blogs need an archive index, organized by month. They need a bit of marketing. Hell, why are there no Google ads? I thought the NYT needed revenue.

Great content, lousy presentation, no monetization. No wonder the NYT is running on fumes.

Update: I've discovered that if you click on the title of a NYT blog you get a pseudo-archival view with URLs that look like this: baghdadbureau.blogs.nytimes.com/page/19/. You can substitute for the page #, as I did here, to find the first post, which for this blog appears to have been 12/28/2007. It's followed by "who's who" post dated 2/27/2008 and revised 11/26/2008. The bureau's chief has had an interesting career ...
James Glanz is the Baghdad bureau chief for The New York Times. Originally working as an astrophysicist, Mr. Glanz joined The Times in 1999 as a science reporter. In 2004, he became a Baghdad correspondent for the newspaper, and he was appointed bureau chief in August 2007.
I very much hope some Google bazillionaire buys the NYT as a gift to his bride or somesuch. It needs a wealthy overlord with some geek skills.

Friday, November 28, 2008

After the crash: The future of the publicly traded corporation

Years before the crash of '08 (complexity crash?) I'd wondered about the future of the Publicly Traded Corporation (PTC). For example, it takes about 10 years to grow a healthy specimen of complex software, but that's six years beyond the attention span of the average PTC.

Anyone watching the transformation of Microsoft from amoral pirate into yet another mediocre but successful corporation must note how all PTCs come to resemble on another. Everywhere we find lots of very bright and insanely industrious people producing, at the end of the day, mediocre products.

For a time Apple, almost alone among PTCs, seemed to have avoided the trap -- but perhaps a single exception only proves how strong the trap is. Personally, I think even Apple has succumbed. I use a lot of what they make, and I think their quality is deteriorating and, worse, they've lost their vision. In the end one maniacal genius leader was not enough; if we could see inside Apple I think we'd learn that they've lost a lot of irreplaceable talent in the past few years.

Of course these are not new concerns. Some were expressed in the book the Innovator's Dilemma; business professors have wondered about the structural problems of the PTC for many years. Now though, we also have the crash of 2008, and we're all noticing how Wall Street financial corporations changed their behaviors when they went from private partnerships to publicly traded corporations.

So here's a thesis, developed between Emily and me, so it has to be good.

The publicly traded corporation had a great run in the 20th century. Sure, it had flaws, but nothing's perfect. In a world of metal bashing and road building maybe those flaws didn't count for so much.

Times change. Things are more fluid now. It takes time to change a physical product line, but Google can terminate Lively in seconds. Most of all, a critical mass of people have figured out the weaknesses of the PTC. In retrospect, when corporations started paying the execs hundreds of times more than the worker bees, the end game was in sight.

So we can expect a lot of new regulation, but maybe what we really need to think about is the future of the publicly traded company. Do we need to look at a mixed model? How can we organize human work for the 21st century?

The PTC isn't going to vanish tomorrow. I think it will still be a dominant player twenty years from now. I do hope though, that by then we'll see much better alternatives emerging.

12/3/08: Possibly related - the decline of the large corporation. Now quite the same thing as I'm thinking, but a cousin.

Thursday, November 27, 2008

Lessons from when the terrorists won: The Ku Klux Klan

Yesterday's terrorist attacks in Mumbai have been expected for years. Expected in India, expected in America.

It's only been six years since a single deranged American veteran and a young accomplice terrorized the US capitol. Even Israel, a small state with the world's most extensive security network, has seen several comparable attacks.

Some of these have been stopped by police and intelligence work, but some will get through. Reducing the number that get through is a long process, not a war.

It requires intelligence and police action. It especially requires reducing fund raising. In retrospect 9/11 ended both the IRA and the Tamil Tigers by reducing the funding stream from the American diaspora.

Beyond reducing the flow of money anti-terrorist actions also need to reduce the flow of people. We now believe people join terrorist organizations for the same reasons they join political parties, bowling leagues, Moose lodges, any group-focused religious entity, the NRA and Greenpeace. They join for reasons of social solidarity and tribal identity. Part of the modern anti-terrorist strategy is to give men and women alternative movements to join.

That will be important to remember as the Obama administration turns its attention to Afghanistan and Pakistan, where the Taliban are the problem.

The Taliban, which, pathetically, in English, rhymes with Ku Klux Klan.

I'm reminded of the Klan, as, after a pause of a few years, I return to the History of the United States audio tape series. There were two distinct incarnations of the Klan, and the thing most of us forget is that the first incarnation of America's "greatest" terrorist organization was victorious ...
Ku Klux Klan -- Britannica Online Encyclopedia

In the summer of 1867, the Klan was structured into the “Invisible Empire of the South” at a convention in Nashville, Tenn., attended by delegates from former Confederate states. ...Dressed in robes and sheets designed to frighten superstitious blacks and to prevent identification by the occupying federal troops, Klansmen whipped and killed freedmen and their white supporters in nighttime raids...

The 19th-century Klan reached its peak between 1868 and 1870. A potent force, it was largely responsible for the restoration of white rule in North Carolina, Tennessee, and Georgia...

... In United States v. Harris in 1882, the Supreme Court declared the Ku Klux Act unconstitutional, but by that time the Klan had practically disappeared.

It disappeared because its original objective—the restoration of white supremacy throughout the South—had been largely achieved during the 1870s. The need for a secret antiblack organization diminished accordingly...
If I remember the recording correctly, in Louisiana the Klan eliminated about 80-90% of black voters from the rolls during their successful reign of terror.

The Klan 1.0 won a fantastic victory. One that certainly lasted until the Civil Rights struggle eighty years later, and if you count the GOP's "southern strategy" the legacy of the Klan 1.0's victory lasted until 2008. A victory lasting 126 years is staggering.

The 19th century Klan teaches us that terrorists can win, and win big. They also teach us that when it comes to terrorism, America has a rich and under-appreciated history.

Now, when we work against terrorists in Mumbai, Pakistan and Afghanistan, we need to think about how the Klan won, and how their modern incarnations can be controlled on a global front.

The good news is that we will soon have a rational, thinking government for the United States of America, and that we've learned a lot about terrorism.

Now we have to put that learning to the test.