Saturday, December 20, 2008

Oddities of popular posts

Based on comments, two of my most often appreciated posts:
The first is interesting from an anecdotal medical perspective.

The second is an aspect of one the most annoying problems with OS X -- device drivers are often very badly behaved. It comes up at the top of several Google searches on the topic, I get a comment every other day.

I've no way of predicting what's going to be popular.

WordPress's possibly related posts -- I want this from Blogger

Last July I wrote that I really wanted Blogger to support backlinks.

Well, Blogger still hasn't done anything. Meanwhile, I'm seeing more WordPress blogs using their "possibly related" feature.
Possibly an Announcement -- WordPress.com

... In a feature we’re calling possibly related posts we’ll now try to show posts related to yours a little section at the end. If we find any posts on your blog that are related, we’ll put those at the very top and in bold. Next we’ll show other posts from around WordPress.com, and finally we’ll check if there’s anything in the mainstream media.

The result is a handful of links that should provide you and your visitors something interesting to check out. On blogs that cover the same topics frequently related posts could cause a 5-10% increase in traffic overnight. You could also start to see traffic from lots of other blogs. It’s a bit of an experiment, and we’ll be tweaking it a lot based on your feedback and the data that we collect once everything is live.

Grrrrr. This feature is core to my memory extension strategy [1].

I want a Google Blogger "possibly related posts" feature that follows links and tags and, heck, textual analysis to create entries -- and that lets me choose whether to restrict to my own domains or open it up.

What do I need to do? I'm turning blue ...

[1] If Blogger label views had feeds I'd be able to create meta-tags that spanned Gordon's Notes and Tech. By the way.

Understanding the Collapse of '08 - the role of the deadbeats

I keep coming across the meme that the root cause of the Collapse of '08 was that the banks made loans to people who couldn't afford to repay them. To deadbeats, in other words.

Sometimes the implications is that this was because the banks were stupid, or evil but mistaken, and other times it's blamed on federal rules avowedly intended to end racist lending practices and to reduce poverty.

I thought stories like Michael Lewis's would slow that meme, but they haven't. So even though I wrote (emphases mine) ...
Gordon's Notes: Michael Lewis on the End of Wall Street. My God.
  1. We need to close the rating agencies down. Now.
  2. When you give 99% of the human race money and power, they cannot imagine that they don't deserve it. They are convinced they are brilliant. I know I would.
  3. The part where shorting the bad loans with credit-default swaps enabled the generation of 'virtual' bad loans involving real money is mind-boggling.
  4. I don't think this was caused by poor black people getting too many loans. Do you?...
The meme persists. Funny about that.

So let me try again.

Imagine that someone had built a massive amplifier. Huge. It takes anything it hears and it boosts it a thousand times.

Next to it, Yo-Yo Ma plays the Cello.

Everyone is happy.

Until the sound guy hits the wrong switch and the speakers connect to the microphone. The feedback loop blows the speakers, the audience, and the nationwide auditorium sky high.

So who's to blame? The sound guy? The speakers? Or the idiots who built a system that was built to self-destruct?

I guess you can tell what I think.

We'd built the mother of all firewall-free feedback loops. All it took was one thing to go wrong, and the entire setup was going to blow. If it hadn't been the sound guy, it would have been a short in the wires. The explosion was going to come -- one way or another.

Don't blame the sound guy. He's dead too.


See also:
  1. Complexity collapse
  2. Disintermediating Wall Street
  3. The future of the publicly traded company
  4. The role of the deadbeats
  5. Marked!

Bring down the Barons - the disintermediation of Wall Street

I'm doing my Roman Chair back rehab exercises, listening to the Teaching Company's History of the United States*, and learning about how the railroads built fortunes from the deluded Scandinavians of early Minnesota.

Those crazed farmers and woodsmen had only one way to get produce to market -- by rail. They couldn't leave the land, because they couldn't sell it until they owned it, and they only owned it by living on it for a time.

Chumps, in other words.

That feels ... familiar.

Eventually the balance shifted. Roads contested with rail, small farms disappeared, and now a few massive corporations dominate America's food chain. Now they have the power. And so it goes.

Today's Union and Southern Pacific are called Goldman Sachs and Merrill Lynch, though the comparison is kind to the latter. There's no doubt the railways delivered a vast amount of value to many people, the modern equivalents have been a bit more ... focused.

Still, there's no denying that our civilization depends on the flow of money and the funding of productive enterprise. The virtual railway of commerce is at least as important to us as the physical railway was to early 20th century America.

So we need the railway of finance, but do we really need to be paying the conductors tens and hundreds of millions of dollars? After all they've just driven ten thousand express trains off a cliff. Forget texting, they've been snorting coke and shooting 'ludes -- and they ain't giving their money back.

They're not worth what we've been paying, heck, they ain't worth nothin.

So, we need finance, but do we really need Wall Street?

Where's the brokering bottleneck that lets them siphon off so much value -- without competition? How can we break that de facto monopoly, and open them up to the bracing wind of competition?

Yes, we need a non-Bush SEC. Yes, Christopher Cox should join the Bush/Cheney tar and feather party. Those are good Democratic responses. I'm for 'em.

But we need good GOP responses too. True, there are few Republicans left who remember how those work, but we Dems can run a pre-Gingrich simulation of those forgotten voices.

How can we redesign finance so that a great money manager makes, say, $300,000 or so -- and a mediocre one becomes, say, an architect? Maybe we can start by figuring out where the chokepoints are, and how to open those up to some vicious competition.

It's time to break Wall Street, and make commoners of the Barons.

* Incidentally, it's a bloody and horrible story - at least in my 1st edition. Maybe they spruced it up for the current edition.

Friday, December 19, 2008

A world gone Madoff

Krugman is back, both barrels blazing. I really can't abbreviate much of what he wrote, every word counts. Emphases mine.

Op-Ed Columnist - The Madoff Economy - Paul Krugman

...How different, really, is Mr. Madoff’s tale from the story of the investment industry as a whole?

The financial services industry has claimed an ever-growing share of the nation’s income over the past generation, making the people who run the industry incredibly rich. Yet, at this point, it looks as if much of the industry has been destroying value, not creating it. And it’s not just a matter of money: the vast riches achieved by those who managed other people’s money have had a corrupting effect on our society as a whole.

Let’s start with those paychecks. Last year, the average salary of employees in “securities, commodity contracts, and investments” was more than four times the average salary in the rest of the economy. Earning a million dollars was nothing special, and even incomes of $20 million or more were fairly common. The incomes of the richest Americans have exploded over the past generation, even as wages of ordinary workers have stagnated; high pay on Wall Street was a major cause of that divergence.

But surely those financial superstars must have been earning their millions, right? No, not necessarily. The pay system on Wall Street lavishly rewards the appearance of profit, even if that appearance later turns out to have been an illusion.

Consider the hypothetical example of a money manager who leverages up his clients’ money with lots of debt, then invests the bulked-up total in high-yielding but risky assets, such as dubious mortgage-backed securities. For a while — say, as long as a housing bubble continues to inflate — he (it’s almost always a he) will make big profits and receive big bonuses. Then, when the bubble bursts and his investments turn into toxic waste, his investors will lose big — but he’ll keep those bonuses.

O.K., maybe my example wasn’t hypothetical after all.

So, how different is what Wall Street in general did from the Madoff affair? Well, Mr. Madoff allegedly skipped a few steps, simply stealing his clients’ money rather than collecting big fees while exposing investors to risks they didn’t understand. And while Mr. Madoff was apparently a self-conscious fraud, many people on Wall Street believed their own hype. Still, the end result was the same (except for the house arrest): the money managers got rich; the investors saw their money disappear.

We’re talking about a lot of money here. In recent years the finance sector accounted for 8 percent of America’s G.D.P., up from less than 5 percent a generation earlier. If that extra 3 percent was money for nothing — and it probably was — we’re talking about $400 billion a year in waste, fraud and abuse.

But the costs of America’s Ponzi era surely went beyond the direct waste of dollars and cents.

At the crudest level, Wall Street’s ill-gotten gains corrupted and continue to corrupt politics, in a nicely bipartisan way. From Bush administration officials like Christopher Cox, chairman of the Securities and Exchange Commission, who looked the other way as evidence of financial fraud mounted, to Democrats who still haven’t closed the outrageous tax loophole that benefits executives at hedge funds and private equity firms (hello, Senator Schumer), politicians have walked when money talked.

Meanwhile, how much has our nation’s future been damaged by the magnetic pull of quick personal wealth, which for years has drawn many of our best and brightest young people into investment banking, at the expense of science, public service and just about everything else?...

... Think of the way almost everyone important missed the warning signs of an impending crisis. How was that possible? How, for example, could Alan Greenspan have declared, just a few years ago, that “the financial system as a whole has become more resilient” — thanks to derivatives, no less? The answer, I believe, is that there’s an innate tendency on the part of even the elite to idolize men who are making a lot of money, and assume that they know what they’re doing...

... Now, as we survey the wreckage and try to understand how things can have gone so wrong, so fast, the answer is actually quite simple: What we’re looking at now are the consequences of a world gone Madoff.

Go back and read the the "corrupting ... society" lede. We need to recognize how corrupted and degraded our culture has become over the past eight years. That will not be easy to undo.

Elsewhere, Talking Points tells us more than a few people suspected Madoff was a con man ...

... a known reader passes on word from a hedge-fund analyst at one of the big (remaining) financial services powerhouses who says that Madoff was always on their short list of people he and his colleagues thought were crooks.

A bit more concretely, let's go back to Harry Markopolos, the money manager and financial fraud investigator who'd been blowing whistles on Madoff for almost a decade. In the report he submitted to the SEC in 2005, in addition to complex mathematical analyses showing that Madoff's system couldn't work, he said that "I have also spoken to the heads of various Wall Street equity derivative trading desks and every single one of the senior managers I spoke with told me that Bernie Madoff was a fraud." ...

I'd like to know who else is on this alleged "short list of ... crooks". Wouldn't you? Isn't that list worth a vast fortune -- since it predicts who will be next Madoff and thus enables various informed wagers?

While we wait for the rest of the "short list", can we tar and feather Christopher Cox?

Of course that's only going after a henchman, Cheney/Bush set out to destroy government, starting with the SEC. They get to own a good piece of this disaster, though I admit it's hard to find an open space for another pin. The SEC story strengthens Bush's case for being the worse president in history -- and that's a really tough contest.

The return of Bush 1998?

Years before Bush destroyed much of America and the world, he was a lucid speaker and even had a reputation as something of a "uniter" -- albeit in Texas, where he was "uniting" the moderate right with the extreme right.

Molly Ivins used to remind us of this, and how amazed she was by the incoherent loon of 2000 onwards.

I heard Bush this morning, and I had the same thought as Talking Points ...

Talking Points Memo | Who Is This Guy?

As one of my colleagues just remarked here, Bush is surprisingly lucid and cogent in explaining why bankruptcy is not a viable option for the automakers in the short run and why doing nothing is not viable in the short or long term.

So, did some brain sucking earwig fall out? Has he been abandoned by his KGB controllers and reverted to a prior mind? Does he have a split personality?

Weird.

Thursday, December 18, 2008

Annals of bizarre tripod ideas - the lampshade

When I think of digital photo tricks, I think of setting the ISO high then pushing the exposure to blow out the highlights. Seems to give less noise with pushed images.

Pogue thinks of weird tripod tricks ...
Pogue's Photo tricks

... It turns out that the threads at the top of just about any lamp--the place where the lampshade screws on--are precisely the same diameter as a tripod mount! In a pinch, you can whip off the lampshade, screw on the camera, and presto: You've got a rock-steady indoor tripod...
He's referring to a style of lamp where there's a fairly thick screw at the top of an elliptical loop that the lampshade hangs on.

Speedcubers and the 6.6 billion person world

Speebcubers compete to solve the Rubik's cube puzzle ...
Jessica Fridrich Specializes in Problems That Only Seem Impossible to Solve

... She has been far surpassed by speedcubers with records of 14, 13 and 10 seconds, some of whom can solve the cube blindfolded after studying it for less than a minute. “Today I would probably be in 20th or 30th place,” she said...
There are 6.6 billion people in the world. Even if only 1 in 10 million have the bizarre mix of talent and obsession needed to be a "speedcuber", that's 660 people.

Colonoscopy: As trustworthy as the markets?

A few months ago we learned that many cancers arise from flat lesions that can't be easily seen by colonoscopy. Two years ago we learned that gastroenterologists vary greatly in how many polyps they find, and that accuracy is inversely related to speed.

Now we find that even good gastroenterologists can't find right sided cancers ...
Colonoscopies Miss Many Cancers, Study Finds - NYTimes.com

.... In the new study, the test missed just about every cancer in the right side of the colon, where cancers are harder to detect but about 40 percent arise. And it also missed roughly a third of cancers in the left side of the colon.

Instead of preventing 90 percent of cancers, as some doctors have told patients, colonoscopies might actually prevent more like 60 percent to 70 percent ...
Of course it's possible right sided cancers are different from left sided cancers; maybe they don't have the same slow growth from polyps.

I also suspect that, like most routine procedures, gastroenterologists do best in the first 1-3 years of scoping, then do worse as the procedure becomes easy and routine.

In any event, it's been a bad few years for colonoscopy. The procedure now looks like a bit of a fraud.

We don't have know of anything better, but clearly we ought to review our options. Since it's not as good a test as we thought it was maybe we should start paying less for the procedure.

Greg Egan - strong drink for experienced readers

Greg Egan leads my list of mind expanding authors. He's strong stuff though. You don't start with Greg Egan, unless you like your reality straight, no water.

You start with Vinge, or Iain M. Banks. They're superficially benign. If an inattentive reader skips through the embedded essay on the relationship between the problem of evil and the Simulation Hypothesis in the midst of Bank's Matter the rest of the book is only mildly perturbing.

Of course Banks is a masterful deceiver. There are lots of perturbing ideas beneath the superficial space opera. I'm a veteran reader, and I'm sure even I've missed some of the many levels of meaning in Matter. There's always another layer of player in Banks.

Egan doesn't bother with mere accessibility. The first chapter of Diaspora outlines a persuasive model for the development of consciousness and self awareness, and a prescription for the creation of human-like minds within a digital world. In 27 pages. At the start of the book.

Jon Evan's essay on the mysterious non-stardom of Egan explains why Egan is not for the novice ...
... I love his short story “Wang’s Carpets,” (also part of his novel Diaspora) which is sort of about Fourier-transformed aliens4—but would I have understood what the hell was going on if I didn’t have an electrical engineering degree, the acquisition of which required the calculation of far too many Fourier transforms before breakfast? Do readers without any technical background have any hope of getting Egan at all?
Vinge, Stross and Banks do some deep dives, but their cover stories are accessible to the newcomer (Banks is especially good at this wonderful trick). Egan doesn't condescend -- he's strong drink, straight up.

I like my coffee black, my whisky rough, and I like Egan. His ideas haunt me like no others. Egan is the ultimate mind expanding author.

So don't start with Egan. But when you're ready, his books are waiting.

Update 6/16/09: See also my later post on Egan's Permutation City.

Bankruptcy for GM and Chrysler - opponents need a better case

Well written, sober, and understated - Roger Cohen of the NYT on why bankruptcy is the right path for GM and Chrysler.

His Pan Am and Alitalia comparisons are well chosen, but slightly misleading. When Pan Am went bankrupt it vanished. I've been flying on a bankrupt airline for months that's recently been acquired ("merged" technically) by another bankrupt airline. Bankruptcy has evolved.

Modern bankruptcy is not what it was. It feels like the right choice for GM, Chrysler and America. If it's not, someone has to make a much better case than we're hearing so far.

Madoff and how the victims con masters like best

Two weeks before my obligatory Madoff post I admired an article on spotting a con man:
Gordon's Notes: Rule #1 for spotting a con ...

[Portfolio.com] ... 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...
One of the many astounding aspects of the Madoff con was that people had to go to him to be conned. Only the savviest, the smartest, the most wealthy and clever, those blessed by fortune and fame could be accepted.

I wonder how long we'll all remember this lesson.

Six months?

No, we're not that smart.

Update 1/6/09: A wonderful and insightful essay by a writer who lost a third of her assets, a child of Auschwitz fleeced by Madoff.

Apple's quality problems continue ...

As usual, an essential [1] but "minor" OS X update with no significant new features causes widespread issues, though most can be resolved ...

Applers howl over Mac OS fix | Register Hardware

...Overall, the 10.5.6 rough spots are about par for the course for an Apple OS upgrade. However, we at The Reg still recommend that before you attempt the 10.5.6 update - or any update, for that matter - you should create a bootable backup of your entire system, including files and apps. Yes, doing so a time-consuming pain in the wazoo, but as your dear mum used to say, "Better safe than sorry."

Things are bad when reasonable advice includes a full backup bootable backup rather than checking the integrity of routine data backups. Of course since the update addresses issues with corrupt Time Machine backups that can't be restored the backup needs to be done with some other application. Incidentally, it's true that many Apple users have an unhealthy tolerance of Apple's failings. Cognitive dissonance, perhaps?

Apple's due for a leadership change, so it's possible there will be a different approach going forward.

PS. I don't do full system backups prior to these point releases. My procedure is:

  1. Wait two weeks from initial release.
  2. Check out any sync related products I use (Spanning Sync, etc) and any Preference Pane products. Those often break.
  3. Check Macintouch and Apple's discussion forums for any really bad news with products like mine. Especially check the PPC (legacy) discussions.
  4. Verify I can do a random file restore from my daily backup.
  5. Install first on our MacBook (MacTel machine).
  6. A week later install on the primary household machine (G5 iMac).
  7. When I install I first do a safe boot restart to clear out caches and test the machine. I then shutdown, and detach all peripherals. I then restart to an admin account and do the install. After install I do a regular restart, then a safe boot, then a regular restart. Tedious, but I attract nasty bugs.

The return of mad cow disease and ideas for writing about science

Fears raised over new vCJD wave tells us that the first wave of "mad cow" victims all shared a common (42% of the UK) gene variant called MM. New cases suggest a new wave in non-MM persons.

We don't know if this '2nd wave' will be more common as well as later onset, but the current guesses are roughly the same frequency -- meaning rare for the population as a whole.

I wonder why we don't see more retrospective reviews of the science of these historic disease outbreaks. We ought to have a tradition of five and ten year historic reviews of things like "Mac Cow" (variant CJD) and SARS-associated Coronavirus disesase.

How about it science writers? Just look back in the NYT's archives and find out what mysterious diseases were big 10 years ago. Do it every January 1st, and then put out a retrospective with the best current science. Could be a career.

Wednesday, December 17, 2008

The "No Rick Warren" facebook site

Well, that was a pretty short honeymoon.

Rahm Emanuel may yet end up in big Blago trouble, and now we have a colossal blunder in the choice of minister for Obama's inauguration.

From the Facebook activist group page:
Facebook | No Rick Warren at Obama Inauguration

... Anti-gay pastor Rick Warren has been selected to give benediction at Obama inauguration, this is an utter nightmare and a total insult to the LGBT community and their straight allies.

Send Obama a letter at

http://change.gov/page/content/contact/

And ask Rick Warren to be removed from the event!

Please email Parag Mehta, Obama's LGBT liaison on the transition team at parag.mehta@ptt.gov, to express your concern.
I wonder when Parag Mehta will resign.

Houston, we have a problem.