Baumol's cost disease: medicine, education and post-AI disruption

William Baumol was born in 1922. In 2012, 90 years later, he's listed as first author on a new book: The Cost Disease: Why Computers Get Cheaper and Health Care Doesn't.

Damn. It's one thing to win the brain lottery, but winning the longevity lottery is really piling on. Even if all he did is read the page drafts he's doing pretty well.

That's not the most irritating thing about Baumol though. The most irritating thing is that I keep forgetting about his fundamental insight, one that I first blogged about 8 years ago...

... The disparity between rapid productivity growth in mechanized sectors and slow productivity growth in human-service jobs produces Baumol's disease—named after the economist William J. Baumol. According to Baumol, in a technological economy falling prices for manufactured goods and automated services eventually increase the relative cost of labor-intensive services such as nursing and teaching. Baumol has predicted that the share of gross domestic product spent on health care will rise from 11.6 percent in 1990 to 35 percent in 2040, while the share spent on education will rise from 6.7 percent to 29 percent.

The shifting of relative costs need not in itself be a problem. If Americans in 2050 or 2100 pay far more (as a percentage of their spending) for health care and education than they did in 1900, they may still be better off—if they pay correspondingly less for other goods and services. The problem is that as the relative cost of services like education and health care rises, more and more Americans will find themselves in service-sector jobs that, unlike the professions, have historically been low-wage...

Today Education and Health Care are famously afflicted by Baumol's disease. Law used to be, but then full-text search decimated legal employment (and yet, legal costs have not fallen ....).

Baumol argues that even if these professions remain labor intensive, and even if health care comes therefore to claim 50% of our GDP, that we'll be able to afford it nonetheless.

His argument is persuasive, but is that likely to happen? College education today is experiencing widespread disruption including iTunes U, coursera (Caltech, University of Toronto and many more), edX (MIT, Harvard, Berkeley), California open-source eTexts, Stanford Online, Khan Academy and numerous for-profit ventures. Education is deep in whitewater times.

Health care, particularly medical care, isn't changing as quickly. The fundamental tasks of sorting out what's going on with a particular patient, and how best to manage that problem in their personal context, and then how to manage the patient's psyche and health -- those haven't changed much [1] over the past century. 

We're accumulating more health care data though -- for better and for worse [3]. "Analytics" is the "hot" area in health care IT now, including running Google/Facebook style algorithms against large clinical and financial data sets [2].

That doesn't necessarily sound disruptive, unless you know that the techniques used in extracting meaning from large data sets are the same technologies that power our post-AI world. (Yeah, I used the forbidden acronym.) If you know that, then you know "Analytics" can be thought of as the current pseudonym for "Medical AI". Whether it's disruptive or not remains to be seen, but I suspect that we'll get to health care cost disruption well before health care hits 50% of a much larger future GDP.

 [1] It's interesting to read articles written in the 1970s during the early days of diagnostic lab testing. They imagined patients walking into a series of lab test queues staffed with low wage workers, then emerging with a set of diagnoses and plans. Similar plans arose during the last period of genomic enthusiasm. They will come again ... 
[2] The base stats is generally pretty simple stuff, if only because more complex algorithms don't scale well to terabyte data sets. The trick is that simple stats on large data sets enabled by cheap computation can produce surprisingly useful answers. This is best described in the terrific Halevy, Norvig and Pereira paper: The Unreasonable Effectiveness of Data.
[3] In 1996 I was part of a theater-style presentation called "Dark Visions: 1996-2010" that included a fanciful and intentionally dramatic timeline of dystopic data sharing. By 2005 India was the world center of clinical AI, and by 2006 elite health care providers had moved to more private paper records. Maybe we were a bit hasty :-).

See also:

• Free exchange: An incurable disease | The Economist 9/2012
• Baumol's cost disease - Wikipedia, the free encyclopedia
• Gordon's Notes: On redistribution 6/2004
• Gordon's Notes: Why college tuition continues to increase: Mankiw 12/2006 (before Mankiw sold out)
• Gordon's Notes: Life in the post-AI world. What's next? 9/2011

Why are OS X solutions easier to find than Windows solutions?

When something goes wrong on a computer, we all turn to Google. 

Almost all. I suppose Apple employees have to use Bing.

I can usually find OS X fixes fairly easily, even though when I need an OS X fix it's often a pretty obscure issue.

Windows 7 is another story. I rarely find solutions. Searches run into unanswered questions, paywalls, and wrong answers; it doesn't matter if I use Bing or Google.

Why is OS X search so much better than Windows 7 search? Windows 7 is still much widely used than OS X.

Has anyone else noticed this? Maybe it's just me.

Islamic rage and free speech.

A guy with a big chip on his shoulder isn't going to do well. He's easy to play with, but even without manipulation he's going to wreck his life.

A lot of the Islamic world is like that, probably because the population is young, social structures are intensely patriarchal, employment opportunities suck, education is distorted by religious doctrine and the world is a very scary place. Just be glad the "The End of Men" isn't a video. Threats to patriarchal structures are not entirely imaginary.

But it's not just Islam. The anti-Japanese riots in China look awfully to Cairo; and does anyone remember burning cities in 1960s America?

So national rage isn't as simple as a cultural trait unique to Islam.

What about free speech, how simple is that?

As any high school student should be able to describe, we don't practice completely free speech in the US. The NYT mentions a couple of examples but Wikipedia has more ...

Freedom of speech in the United States - Wikipedia

... the Miller test for obscenity, child pornography laws, speech that incites imminent lawless action, and regulation of commercial speech such as advertising. Within these limited areas, other limitations on free speech balance rights to free speech and other rights, such as rights for authors and inventors over their works and discoveries (copyright and patent), protection from imminent or potential violence against particular persons (restrictions on fighting words), or the use of untruths to harm others (slander)...

... Publishing, gathering or collecting national security information is not protected speech in the United States ... The unauthorized creation, publication, sale or transfer of photographs or sketches of vital defense installations or equipment as designated by the President is prohibited.[13] The knowing and willful disclosure of certain classified information is prohibited ... It is prohibited for a person who learns of the identity of a covert agent through a "pattern of activities intended to identify and expose covert agents" to disclose the identity to any individual not authorized access to classified information, with reason to believe that such activities would impair U.S. foreign intelligence efforts.. 

Julian Assange could say something about national security and restrictions on free speech -- though the material he published was widely redistributed in the US without much consequence. The US didn't exercise blocks on web servers. Other information is more restricted. Although the wikipedia article on nuclear weapon design is pretty detailed, more practical 'make your own terror weapon' recipes are hard to find.

Even videos of this type of restricted material would be unlikely to provoke popular outrage. Neither would videos attacking religious beliefs, politicians, science (the GOP makes those) or history -- denial of American slavery, Amerindian genocide or the Holocaust are all protected speech.

The only exception I can think of would be child pornography. Videos that use children under the age of 12 in sexual or harmful activities are strictly illegal in the US and would create anger and disgust. I can't see riots, though and, of course, the Islamic reaction would be at least as negative.

National rage is not uniquely Islamic, but protected speech is pretty distinctly American. Perhaps a consequence of that protected speech is that while sticks and stones will trigger invasions, "names will never hurt us".

Tom Cook's great test: the response to Apple maps.

Contrary to Joe Nocera's impression, Steve Jobs' Apple screwed up a lot. MobileMe wasn't the only disaster; even iCloud was a regression. OS X Lion was nothing to write home about. Apple did better under threat, worse when they felt confident. Ultimately, of course, Jobs cruel genius coupled with unprecedented power over a public company meant Apple escaped mediocrity.

Post-Jobs most of us expect Apple to behave more like a typically dysfunctional publicly traded company. We expect more products like iOS maps and Siri -- heavily marketed but only marginally useful. (Under Jobs Apple buried flops quickly.)

Of course there's still a chance Cook can find a way to escape this mediocrity. I'm watching for how Apple responds to user submitted map corrections. 

Historically Apple seems to ignore users. Even respected developers have a hard time getting their bug reports reviewed. Google, to their credit, has responded to every map correction I've made. Sometimes I get a response in a week or two, sometimes in a month or two, but I always find out which corrections were accepted and which were rejected. (So far 5/6 accepted, I think the other one had already been fixed.)

This is something Cook can change. If Apple provides feedback to users on the fate of their corrections, then I'll be hopeful that Cook can chart a new post-Jobs course. If they don't then they're on the fast track to Microsoft-land.

What has Apple done for me lately?

The new iOS connector cable. iOS 6 maps. iOS 6 Podcasts. RSS Deprecation. iTunes 11 (train coming). iCloud. Siri. Lion. Embedded webkit hole in iOS parental controls.

I'm sure there are more. Apple hasn't been doing much for me lately. Yeah, I'm in a grumpy mood. The good news for Apple is that Google's done even worse. The past two years have not been great for the 0.00001% like me.

It can't be all bad though. What, with it's zillions of dollars, has Apple done in the past 1-2 years that I like?

I'll grow the list, but here's the start. Send me additions if you can think of them...

• Mountain Lion looks promising.
• MacBook Air.
• iMessage.
• iOS 5 let me override exchange calendar color assignments.
• Six years late, Apple finally delivered iPhoto Pro with it's agonizingly slow and still incomplete Aperture/iPhoto integration project. Way, way late, but it counts.
• Find Friends (I think it will be useful for our family eventually).
• iPhone retina screen - good for old eyes.
• iPhone camera.

The bad and the good are close to balance, which is kind of sad for the world's richest tech company. Of course I'm not their market.

Tachytely and human evolution: implications for the Drake Equation and Fermi Paradox

I haven't done a Drake/Fermi Paradox post for ages. A lot has happened in the meantime; in particular estimates of the number of potentially-life-compatible planets in our galaxy has grown exponentially.

Of course not all life supporting planets will develop sentient tool using species. Unless there's something about sentience and tool use feedback loops that produce tachytelic development. That would boost the Drake estimate into the low thousands. We ought to be tripping over little green things.

But we don't. Of course if technological civilizations all self-destruct quickly this would all make sense.

See also:

• Refining the Drake equation: earth like planets more common? 10/2006
• Another stab at the Drake Equation – from the physics archives 10/2008
• SETI, the Fermi Paradox and The Singularity: Why our search for extraterrestial intelligence has failed 6/2000 originally
• Narrowing the Drake equation - if earth is common then ... 2/2009
• SETI: find one of nine galactic civilizations 11/2007
• How often will sentience evolve on earth? From anthropology to the Fermi Pardox - via the Drake Equation 2/2006

Why I'll be deferring iOS 6 and iPhone 5: podcasts gone

I listen to several Podcasts; using smartlists to identiy partly played casts and those I haven't started.

That won't be possible come iOS 6:

TidBITS Networking: Does Apple’s Podcasts App Suck Cellular Data?

...In iOS 6, the Music app removes podcasts altogether ...

Leaving us with the crummy Podcast.app.

We've also been told to expect an iTunes update. I wonder if that will drop podcasts too. I expect smart lists to go. We already know iOS 6 will lose Google Maps.

For me iOS 6 and iTunes 11 look to be big regressions. iPhone 5 and it's new cable hasn't impressed me either. I'll be going slow on both, which means I'll likely defer my iPhone 5 too.

Information archeology: a challenge for Google Reader (or its successor)

There's an immense amount of knowledge, insight and entertainment in the feeds I follow. Using Reeder.app on my iPhone, and Google Reader web or Reeder.app on my desktops, I learn something new every half-hour or so. I share some of the stream via Pinboard (archival), app.net, twitter and wordpress (archive and indexed into my custom search).

Works pretty well -- but it could be better.

I'm skimming the surface of the knowsphere. There's great material below, sometimes in once beloved blogs that have gone silent. We need a way to dig it up.

That's someting old-Google might have added to Reader. Using signals like links, non-spam comments and authorship Google Reader could define an "archeology" stream -- the best of the past, where 18 months might be the start of old.

Google's probably not going to do this, but a third party service could do this and create a feed. Yahoo!'s wonderful but forgotten Pipes! product could be modified along these lines.

Gerbil- the Achilles tooth

And now for something completely different.

This posts consists of two warnings about gerbils, the weird and probably inbred [1] semi-domesticated desert rat of Mongolia formally known as Meriones unguiculatus.

The first warning is that you shouldn't let your kids be exposed to a school gerbil that is going to be fed to the school snake unless someone adopts it. 

If this warning arrives too late, then you need to read the surprisingly detailed wikipedia entry and this UK article on Gerbil Health. From both you will learn about the Achilles tooth:

Gerbils can sometimes lose teeth ...  the first sign of this can be that the animal looses weight due to not being able to gnaw their food - they will stop chewing cardboard quite a while before this so if your pet is not destroying cardboard boxes with its normal enthusiasm have a look at its teeth.

If your gerbil looses one or more front teeth it is essential that any remaining teeth are trimmed by the vet every 2-3 weeks...

Trimmed ever 2-3 weeks at a cost per visit of $30-$50 or so (more if sedation is needed). And you thought Apple's #$@ iPhone 5 adapter cable was expensive.

Teeth, it turns out, are a rodent's curse. A pet gerbil can lose its tail and do well, but if anything disrupts chewing the teeth will grow until the gerbil dies [2].

Which brings me to the second warning. Don't buy your gerbil a house ...

These cute things are held together by tacks (and probably, since they're made in China, lead paint and radioactive waste). Gerbils, which live (and die) by chewing can chomp into the tacks and break a tooth. This is not a good thing. Don't buy these. Consider instead Totally Chewbular Tubes and Critter Caves.

You have been warned. No gerbils. If the kids get one, then no damned tack houses.

You're welcome. I don't think I'll have anything more to say about gerbils.

[1] Between 20 and 50% of all pet gerbils have the epilepsy. (according to Gerbil Health)
[2] Vets claim they don't tolerate suffering -- they dwindle and die instead. The lack of capacity for suffering implies that the great Programmer likes rodents, and that one escape from Samsara might be cyclic reincarnation as a gerbil.

Update: Another gerbil reference

Update 10/18/2012: Much to our surprise, a mere month later, he's fat and happy with both his upper teeth. He ended up having to regenerate two front teeth. During the long healing process he became addicted to Emily's deluxe ground gerbil food, devouring a dish in seconds. He's probably doubled his weight from his low famished state.

The Cosmo story, the facade of online security, and the US Postal Service.

Mat Honan, who is making a career out of being hacked, has a solid profile of a juvenile delinquent hacker [1] - "Derek", alias Cosmo (Cosmo, the Hacker 'God' Who Fell to Earth (via Schneier).

"Derek" is a troubled kid, but, in addition to hurting a lot of people, he's also done us a favor. He's become the latest in a series of people exposing the facade of online security.

Unsurprisingly AOL is the worst -- until recently you could reset someone's account just by knowing their address. Apple, Amazon, Netflix and just about everyone else isn't much better. Only Google makes a good try at it, and they just plugged a big hole.

This won't surprise anyone who knows the history of credit card hacks (example). The reasons are fairly easy to understand:

1. If your iCloud account is hacked, Apple loses approximately nothing.
2. Good processes and security are expensive. You have to train staff. To prevent one hack you probably have to irreversibly piss off somewhere between 10 and 1000 customers. Each of these customers will rage to at least five friends.
3. Less than 1 person in a zillion can manage password security, and that person's family will be completely screwed when they run off or die [2].

What we have here is a market failure. Market failures are one reason we have governments.

Governments, particularly post offices, have managed identities for a long time. Passports for example, are managed by Post and Passport Offices. There are laws and procedures in place.

Digital identity management in most nations will eventually be handled by some cooperative mixture of government and business within a regulatory framework. We'll use multi-factor authentication, and we will have "break the glass" functionality available through government when access is lost (for a fee).

Preposterous? No. Six years ago these kinds of proposals generated snort-milk-out-the-nose laughter. I don't hear the laughter any more. It will take a decade, just because these things always stagger on for longer than I can imagine, but it will eventually happen.

See also:

• Gordon's Notes: Net security is completely broken 8/2012
• United States Postal Inspection Service US Mail identity theft report
• Digital Identity - Opportunities for the Postal Service: www.uspsoig.gov/foia_files/RARC-WP-12-011.pdf: 5/29/2012

[1] Steve Jobs was the most famous member of this cohort.
[2] Number of people who have both a highly secure password system and a method to pass information to spouse in event of death or disability? Does your spouse have your list of ten Google two-factor bypass codes? What if s/he dies in the car crash with you? Does your estate have them?

iPhone 5: meh.

What I wanted from the new iPhone and Apple

• Ability to specify calendar colors for Exchange Server/ActiveSync calendars
• Water resistance
• Lower cost data plans
• Parental controls that work (ability to disable embedded browsers).
• Fix the Apple ID debacle
• Bicycle directions on the Map app

What I got

• A new connector with a $30 adapter.
• No more parental controls for YouTube (since it's a separate app).
• A map app without Google bicycle routes
• A bunch of features I don't care much about
• A dumbed down version of iTunes that will probably omit much of many of the query (smart list) abilities I rely on
• More iCloud fail

Meh. I'll probably buy another AT&T iPhone and extend my contract, but I'll wait until I see how the different data plans shake out. The $30 adapter is particularly arrogant.

Home maintenance: Our toilet leak, why Google searches about home problems suck, and my fix

Emily claimed the toilet was running. It seemed fine to me. Flush, fill, stop.

She didn't believe me, despite my stellar home repair history (I know how to stand aside while she calls a contractor). She bought a new flap valve, but still she could hear a leak. Obsessed, she took to spending hours staring at the toilet, listening ...

Or maybe minutes. She showed me that the tank water level was dropping until it triggered the float valve, and the tank refilled. There was no sign of water entering the bowl -- so where was it going?

Google was no help, but eventually we figured out that when I last replaced the float/valve mechanism I'd failed to trim the refill tube. It was so far down the overflow pipe it was below the fill line, and water was being siphoned into the overflow pipe. I assume this must run into the bowl, but we couldn't detect the flow (ok, that puzzles me - but I can't believe there's another flow option.)

That solved our home problem, but not the Google problem. It's not the first time I've noticed that searches for home issues don't work well. For some reasons, Google seems to have lost the SEO battle for the home maintenance market.

The answer to that problem is a free (ad-supported) Google custom search engine.

First, now that I knew the problem, I searched for a site that would have solved it:

Toiletology ... : "Another occasion when siphoning is a problem in a toilet tank occurs when the refill tube drops too far down into the overflow pipe. Then the water is siphoned from the tank into the overflow pipe and down the drain. While this scenario won't harm you, it will wreak havoc on your water bill, because you have water constantly recycling through your toilet. This problem often arises when a new refill valve is installed. They usually come with extra long refill tubes that are meant to be cut to size, but instead a do-it-yourselfer just drops the long tube down inside the pipe. The refill tube should be cut to just reach the top of the overflow and then be clipped to the top edge of the pipe. "

Then I started my home repair and maintenance custom search engine (see all of mine) with toiletology. Next I added the site for a magazine Emily subscribes to - "The Family Handyman". I'll gradually add sites to the CSE, but I don't want to add too many. That will produce Google's behavior -- lots of duplicative references on common problems, and the real result buried on page 114. 

See also:

• Google Custom Search - Special needs services in Minneapolis and St Paul
• Gordon's Tech: Making your Google custom search the IE and Firefox defaults 9/2008
• Gordon's Tech: Google Custom Search with Bloglines Search OS X mashup 11/2007 - custom search is old.
• Review: Custom Search Engine (Matt Cuts - Custom Search Engine 10/2006 - early introduction)
• Gordon's Tech: My Google Custom Search just died. Did I offend the GoogleNet? (fixed) 5/2012 - avoid the linked pages option
• Gordon's Notes: Strange loops: Google custom and customized search - and a memory blog 8/2008

Learning French: using free speech recognition services to improve pronunciation and accent

I can learn most things pretty easily; but not music or language. Tone deaf is an understatement.

This was a problem for a lower class anglais caught up in Quebec's Quiet Revolution. Around the age of 12 I had to learn French; at a time when nobody knew how to teach it. This did not go well. Much later a delightful summer immersion in Chicoutimi was more successful, but I still struggled to be understood. My French diction is almost as bad as my Thai.

This is a hard problem. I needed a demanding listener who could tolerate repetitive errors without annoyance. It would help if they were available at any time, and were willing to work for free.

Happily, that teacher has long been available. It occurred to me long ago that I just needed to practice with a French speaker-independent non-adaptive continuous speech recognition engine until the output matched my input. I waited for that to be built into language education software.

And I waited. Sometimes I idly thought about putting the software together myself.

Today I realized I could simply use Mac speech recognition and language support to learn French speech. It works rather well. I expect many iOS, Android and OS X language education apps will be able to integrate this  capability into language education for French, Chinese and English. Android will have an advantage here, since iOS is more restrictive about what developers can do.

I'm looking forward to buying those apps when they become available. In the meantime my son will be using this during for his High School French class.

Education - how much has math teaching changed?

I have thought of myself as somewhat mathic. I'm certainly no mathematician, but I did it well back when raptors roamed the earth. When I did another grad degree in the 90s I enjoyed my grad stats.

I've changed my mind though, now that I'm reading math blogs like Gödel’s Lost Letter and P=NP, and now that I'm seeing the 13yo cover math topics years ahead of when I did them (MN public school). I was mathic in my day, but the curve is wider now.

Math education has changed, perhaps more than most of us realize. It's likely to change a lot more. Math classes still require medieval calculators, and math exams often mandate particularly primitive calculators. Meanwhile the cost of a used iPod Touch is falling to $120, and Wolfram Algebra Course assistant sells for $1.99.

Sometime in the next few years, despite the drag of obsolete standardized testing [1], math classes will switch from primitive calculators to symbolic math software.

Times are changing.

[1] Old-fart anecdote. I grew up with Quebec provincial standardized exams. I don't know when they started, as a teen they seemed eternal. In the 1970s they still included exercises that involved logarithm table lookups. Paper logarithm tables -- what people used before slide rules. So we learned how to use paper logarithm tables (not hard) at the expense of slide rules (that was dumb). This didn't turn into much of a handicap because calculators came along, so the exams just dropped the log tables and never had to address slide rules. So these transitions are not without precedent.

Do evolutionary strategies evolve?

Biologists study evolutionary "strategies", such as r and K selection.

These are the strategies deployed the Great Programmer as she fiddles with the game states of the multiver... erkkk. Just kidding. These are, of course, human terms for the emergent phenomena of natural selection.

At a more granular level, a predator's niche might be contested on the basis of bigger teeth, stronger claws, faster moves, greater endurance, or bigger brains.

Likewise, microbes, who rule the earth, have a range of "strategies". Symbiosis, parasitism, fast reproduction, encysting and so on.

Presumably the catalog of strategies changes over time. Before there were teeth, big teeth strategies were not available.

Before there were neurons, big brain strategies didn't work.

So that leads to the obvious question, do evolutionary strategies evolve? That is, do new strategies emerge from variations of strategies such that the strategies themselves are subject to selection pressure (a sort of meta-selection I suppose)?

Seems an obvious question, but as of Sept 2012 Google has 9 hits on that precise phrase, none by biologists.

So I guess it's an obvious question, but maybe obviously dumb. I'm surprised though that I didn't find a blog post explaining why it's dumb.

(A bit of context, this came up in a discussion with my 13yo about what species would fill our ecological niche (global multicellular apex predator). Having hit upon the strategy of investing in brains, would natural selection keep returning to the theme?)