TPM: the polls in WW II

TPM, for the second time, deflates the theory that a lack of popular support for Bush somehow resembles a mythical lack of support for Roosevelt ...

Talking Points Memo | Polls

...The key point is that many polls were taken during the war. And approval of the president's conduct of the war, understanding and belief in the goals of the war and other similar measurements all remained constant at very high levels or in some cases actually went up. One key data point you can see on the chart is the number of Americans will to make peace with Hitler -- that is, an negotiated end to the war rather than the unconditional surrender which was a key allied war demand. The number was under 10% for most of 1942 and 1943. Then it briefly surged up to just over 20% in early 1944 (roughly the time of the invasion of Italy) before falling back down to about 15% for duration of the war in Europe...

I like the historical aspects of this most of all. BTW, TPM has moved to a proper blog format, and it's a great improvement.

Immunotherapy for sarcoma - in the 19th century

Damn Interesting is one of my favorite blogs, and this week's post is of particular interest to physicians (note the UK spelling of "tumor") ... 

Damn Interesting » Coley’s Cancer-Killing Concoction

... The story so convinced Coley that he– perhaps cavalierly– contrived to contaminate his next ten suitable sarcoma cases with Streptococcus. His initial approach was to inject a solution of live bacteria deep into the tumour mass on a repeated basis over several months. The first patient to undergo this treatment was a bedridden man with inoperable sarcoma in the abdominal wall, bladder, and pelvis. Using this experimental method, the patient was cured spectacularly. He staged a full recovery, and survived another twenty-six years before dying from a heart attack. But subsequent results were mixed; sometimes it was difficult to get the infection to take hold, and in two cases the cancer responded well to treatment but the patients died from the Streptococcus infection.

Coley’s discovery, as it turns out, was actually a re-discovery. The idea of a link between acute infection and the resolution of tumours was not new, and the phenomenon of infection-related "spontaneous regression" of cancer has been documented throughout history. A 13th century Italian saint was reputed to have his tumour-afflicted leg miraculously healed shortly after the malignant growth burst through the skin and became infected. Crude cancer immunotherapies working along similar lines to Coley’s early experiments were known in the 18th and 19th centuries, and may extend back to the time of the pharaohs. Ancient writings suggest that the renowned Egyptian physician Imhotep may have used a similar infect-and-incise method to treat tumours....

I've been fond of medical history ever since I enjoyed a thinly attended history of medicine course at McGill in the 80s. I don't recall ever hearing about Dr. Coley or his early use of immunotherapy for sarcoma -- a cancer that's often incurable even now. The article implies that the treatment has been long forgotten, which is not quite true, the work of Coley is periodically revisited.

Dr. Coley deserves a Wikipedia page, but as of 7/17/07 none exists.  Perhaps a scholar somewhere will insert one based on this article (A review of DI could be the source of several new topic pages really.)

Why the GOP dislikes McCain - it's not his support of the conquest of Mesopotamia

I've read several articles about McCain's campaign collapse; all of them cited his support for Cheney/Bush's war as the main reason for his poor prospects. I never questioned this oft-repeated theory, so I was impressed when Glenn Greenwald blew it away.

Of course Greenwald is right -- now that I've read his arguments I'm surprised I was ever so naive as to buy the party line. This is not Bush's war alone, it's the war of Bush's supporters, the hard-core dead-ender 30-percenter ... [update: decreased my"name calling" quotient] ... who are the heart of the modern GOP.

If I were a nobler soul I'd find some sympathy for dolts like Giuliani and Romney who have to kowtow to these people, but I'm not so noble.

Why I blog - Gordon's Notes and Gordon's Tech

I posted this as a Cosmic Variance comment, I've revised it a bit here ...

.. There's a meme about “commodity bloggers” and “echo chambers” that’s been simmering for a while but was fired up by a recent Jakob Nielsen post. CH has a good overview.

I think this is a sub-meme of the “blogging” is “destroying all that is good and pure and noble in human civilization” usually alternating with the “wikipedia is destroying education” meme. (More on the latter soon, maybe...)

Beyond these memes is an unspoken wariness about the increasingly subtle distinctions between an “echo chamber” blog and a splog — the more sophisticated splogs are eerily similar to low end commodity blogs.

I’ve nothing else to add to the good comments on the CV thread, save perhaps that my own very low readership blogs are written for these audiences in this order:

1. Myself. It’s how I learn and think.
2. The GoogleMind: building inferential links for search and reflection.
3. Tech blog: Future readers who find my posts useful to solve a problem they have that I've solved for myself.
4. Gordon's Notes: My grandchildren, so I can say I didn't remain silent -- and my tiny audience of regular readers, not least my wife (hey, we don't get that much time to talk!).

I, of course, agree with the obvious consensus that blogs are intended to be read by subscription tools (like bloglines) and that descriptive titles and label/categories should allow readers lots of tools to decide what to read. I do think the readers can, and will, make better use of metadata (themes, categories, labels, etc) in years to come.

Update: I revised my "why I write" list as I thought it over a bit more.

Update 7/17/07: Excellent comments by Rosenberg. For love, not money.

Reading to kids: don't do it for the test scores

I'd always assumed that the claim that reading to young children improved their academic performance was mostly wishful thinking, but I hadn't realized where the belief came from ...

Freakonomics Blog � The Benefits of Reading to Children, Tested With a Data Pool of One

... Children from low-income households average just 25 hours of shared reading time with their parents before starting school, compared with 1,000 to 1,700 hours for their counterparts from middle-income homes.

These oft-repeated numbers originate in a 1990 book by Marilyn Jager Adams titled, “Beginning to Read: Thinking And Learning About Print.” Ms. Adams got the 25-hours estimate from a study of 24 children in 22 low-income families. For the middle-income figures, she extrapolated from the experience of a single child: her then-4-year-old son, John …

This is a bit like the old saw about "using only 15% of your brain" and other urban myths. This one was a useful myth, it meant that poor test scores could be blamed on parents who, obviously, didn't read to their children enough (perhaps because they were struggling to keep the roof in place) [1].

It's fun reading to kids, though of our 3 we've had 2 that sat still for it and one who'd have needed four point restraints. Do it because it's fun, but don't get bent out of shape about it. There's no evidence, yet, that reading to children will make a significant contribution to their academic performance.

[1] Of course even if there was a causal relationship between reading to children and test scores that would be fallacious reasoning, but we're talking about hairless apes here!

Patrick Leahy is my hero. You must listen to this.

The video is quite short. It's your duty as a citizen to listen to it.



It is an undeserved miracle that America, in a dark time of failed citizenry, was given Patrick Leahy as head of the senate judiciary committee. He is an old man. This is his hour. Thank you Patrick, we don't deserve your service.

The WSJ's Laffler curve fitting: thank you Mr Murdoch

Rupert Murdoch cannot possible diminish the Wall Street Journal. It's already hit rock bottom.

Exhibit A: A physicist reviews the diagram the Wall Street Journal published. Yes, they really published the curve, just as CV describes. Editorial page, of course.

The WSJ is the laughing stock of anyone with any sense.

Do your worst Rupert.

The OS X iPods

Daring Fireball (48,800 bloglines subscribers*!) tells us Jobs confirmed that Apple is working on OS X based iPods. So the next generation of iPods ought to be feature competitive with the iPhone's music capabilities. That's great, but he asks the next, much more interesting, question:

Daring Fireball: Regarding OS X-Based iPods

...The biggest question, as I see it, is whether Apple plans to introduce iPods that are more or less just the iPod app from the iPhone (i.e. just music and video players), or iPods that are everything but phones, with Wi-Fi networking for email, web, and more...

By "more" I assume he means the thought that none dare speak - VOIP.

On the one hand, one might imagine that the AT&T contract forbids a "VOiPod". On the other hand, that is a delicious thought. OS X on the iPhone occupies 680MB and requires a serious CPU and battery, so it's hard to imaging putting it on an iPod just to plays music and movies ...

Of course there's more than one way to play video. Cringely thinks Apple will bundle a hardware H.264 decompressor chip into the next iPod. Jobs is on record as saying the current "Apple TV" is a "hobby".

Hmm.

H.264 HD decompressor chip. VOIP. Apple TV. Rumors that Apple has again bought up all Flash supplies over the next few months...

* That's just Bloglines. He probably has 2-3 times as many regular readers. That's a big enough lever to move even Apple ...

Broder and O'Reilly: comic relief from the right wing

Tom Tomorrow can't resist a Broder rant:

This Modern World » Blog Archive » Mr. McBobo strikes again!

... David Brooks takes wrongness to levels undreamt of by lesser mortals. David Brooks is the Olympic Gold Medal Champion of Being Wrong About Things. David Brooks could write a column about drinking a glass of water, and the end of it, you would sit at your kitchen table puzzled, thinking to yourself, “Has he ever actually had a glass of water?”

(… readers interested in a refresher course in the Unbearable Wrongness of David Brooks are directed to this classic article by Sasha Issenberg.)

Really, who can resist? Broder reminds me of John Dvorak, a computer industry writer who invents stories carefully calibrated to drive page views by easily outraged geeks. Over time we've mostly learned to ignore Dvorak, but Broder is so pompously vacuous we can't leave him entirely alone. (I never actually read his column though -- that would increase his page counts. I get the best stuff in excerpts.)

In the meantime, O'Reilly tells us to fear lesbian street gangs carrying guns, beating up heteros, and pushing Sappho. Of course he's making it all up.

In an era where malevolent incompetents lead the executive branch, Dvorak and O'Reilly are our best sources for accidental humor. Long may they writhe.

How will you know that the iPhone is ready?

Is there a test, a sign, that the iPhone is really ready for use by serious geeks? Something beyond the "we had to drop it at the last minute because we were going to miss our dates" omissions like "cut/paste, search, tasks, notes synchronization, working Outlook synchronization [1], etc". 

I think so. I think, assuming that aforementioned blindingly obvious last-minute panicky omissions are fixed, that a good test will be FileMaker Mobile 8 for the iPhone:

FileMaker Mobile 8 - FileMaker

FileMaker Mobile 8 is a companion to FileMaker Pro 9 and and FileMaker Pro 9 Advanced designed specifically for Palm OS and Pocket PC handhelds. FileMaker Mobile 8 lets you easily extend your FileMaker solutions beyond your desktop to mobile individuals and to workgroups by syncing with all FileMaker products....

Load tables, field names, and data from your existing FileMaker Pro databases with just a few clicks. Specify shorter field names for easier viewing on your handheld. Supports multiple field formats, such as checkboxes, date and time fields, and notes fields... View, modify, find, sort, add, and delete information on your hand- held in an intuitive interface. Sort data by tapping a column heading. FileMaker Mobile 8 works with select Palm and Pocket PC devices... FileMaker Mobile 8 now lets you take important information on the go and sync to solutions hosted by FileMaker Server 8 or FileMaker Server 9 for greater workgroup productivity.

If and when FileMaker Mobile works at least as well on the iPhone as it claims to work on Palm devices, then the iPhone will have arrived for the tiny, warped, twisted, group of people like me ...

[1] At least as well as a modern Palm device, which is a darned low standard.

Moments in history - the transformation of the print media

Bruce Bartlett, a respectable "second tier" (meaning not wealthy) columnist is  quitting the business. Extracting from DeLong's extraction:

... The Internet, in particular, has enormously changed the ability to get a message out.... Today, anyone with a computer and a modem can start a blog and, for all intents and purposes, be a columnist. ..

... Those who wanted more biting opinion gravitated to the Internet, where vast numbers of people offer commentary along every single point on the political spectrum. It became very easy to find writers expressing exactly one's own personal opinion about everything. Bloggers also have the advantages of no space constraints, an ability to post comments in real time, and to offer links to supporting documents and sources. Now they even have audio and video.

As a result, the demand for traditional column writing has pretty much dried up.... I don't mourn the old system. I am a great fan of bloggers and learn far more from them than I do from the Broders and Friedmans of the world, who have largely become irrelevant to serious political discussion....

I think a future historian will want to take note of this column. It is a milestone, of sorts, in the long delayed transformation of print media. Murdoch's WSJ acquisition is another milestone.

Where it will go we cannot know, but I agree with Bartlett that almost any blog is an improvement over Friedman, and radio silence is an improvement over Broder.

PS. I suppose we call our DSL/cable thingies "modems", but it does sound a bit anachronistic in this context. It reads like he was thinking of a device that beeped and hissed and sang off key ... Just thinking of that makes me remember Hayes command tweaking, and that makes me think of heavy drink ...

Friedman actually makes a suggestion worth considering - briefly

I'm not saying it's a good suggestion, but given Friedman's record of years of no advice, or nonsensical advice, this recommendation is remarkable because it seems almost reasonable ... 

In or Out - Tom Friedman - New York Times

...  I prefer setting a withdrawal date, but accompanying it with a last-ditch U.N.-led — not U.S. — diplomatic effort to get the Iraqi parties to resolve their political differences. If they can, then any withdrawal can be postponed. If they can’t agree — even with a gun to their heads about to go off — then staying is truly pointless and leaving by a set date is the only option...

If Bush/Cheney were gone it would be possible to consider a suggestion like this. They're not gone. We have to accept the ongoing reality of supremely incompetent [1] leadership. This means the definition of what "resolution" is would be redefined by Bush/Cheney to support whatever incompetent direction they wish to pursue.

Given that problem, this suggest too must be reluctantly rejected. Now if Cheney and Bush were to both resign ...

[1] Of course if their agenda is the destruction of America and the collapse of human civilization they're actually quite competent. Maybe I'm looking at this the wrong way ...

The case of the disappearing Amazon reviews

In at least one case, a good number of Amazon reviews have gone missing: 

Scott Rosenberg’s Wordyard » Blog Archive » The case of the disappearing Amazon reviews

...As of about three weeks ago, we had 33 reviews posted. Most were positive, a handful were negative; either way, each one meant that some reader cared enough to take the time to post their reactions, and that meant a lot to me.

Then something weird happened about ten days ago. Suddenly, Amazon showed only 10 reviews. Two dozen reviews posted between mid-February and the end of June had simply disappeared. In the time since then, a couple of new reviews have joined the total, but the missing reviews have not reappeared...

Disappeared without explanation or even recognition. User reviews constitute a large portion of Amazon's value, if Amazon is not treating them wisely their investors need to wake up.

In addition to an example of a worrisome behavior by Amazon (reminds me of a Sprint cell phone review I had to rewrite six times to sneak by the cell phone company's Amazonian censors), this is an example of the general principle that one should own one's own data. When I write reviews I first publish them here, then submit them to Amazon as well. At the least, the data will stay on my blog.

Speaking of nefarious tricks and reviews, has anyone noted that some hardware vendors are now revising "model numbers" (Hint: HP) every few weeks? It's almost as though they're trying to escape from Amazon's reviews ...

Google maps: more exciting developments

I was thrilled by Google's integration of 'My Maps' and Google/Picasa's web albums. Commentators who think Google's only success is search are ignoring Google Maps and Gmail, both of which are steadily consuming market share (ex: Google Maps):

Google Inc. to unite mapping mashups - Yahoo! News

... Google has been steadily gaining ground in recent months. In June, Google's maps attracted 28.9 million U.S. visitors, a 28 percent increase from the same time last year, Media Metrix said. Meanwhile, Yahoo's mapping traffic fell 12 percent to 29.6 million visitors. Mapquest continued to hold a comfortable lead with 53.9 million visitors, a 3 percent increase from last year.
[jg: See update re: Mapquest ..]

Now Google is taking this to the next level:

Google Inc. to unite mapping mashups - Yahoo! News

... Google introduced My Maps in April to give users a way to save and share their own mashups.

Now, users with Google log-ins will be able to pick from more than 100 mapplets to customize and save their own maps. Google expects the number of mapplets to increase as word about the service spreads. To encourage the phenomenon, Google's own engineers also contributed mapplets...

All of which, we can expect, will work on the iPhone. By the way [1].

[1] There is supposedly no GPS in iPhone 1.0 Apple has come up with some semi-reasonable workarounds. I wonder why they couldn't do it ...

Update: I had the impression Mapquest sent money back to companies that included their services and I commented on that, but "NoTime4Foolz" corrected me (see comments):

Actually, companies pay Mapquest to use their maps. They have a business solutions group that sells enterprise software. Outside of those clients, they have market share because they're still excellent. Not as pretty as Google's maps in some cases, but dependable, easy to use, and accurate.

The Hutter prize: AI pattern recognition developed through compression algorithm tests

To parse out the Slashdot | Text Compressor 1% Away From AI Threshold story, first read the Wikipedia article on the Hutter prize. That's the real story.

In essence, pattern recognition (search) has always been thought to be a key skill for any mind, biologic or abiologic. Pattern recognition is also a key component of compression algorithms (most people are familiar with .zip, .tiff, .jpg and other kinds of compressed files). The Hutter prize funders want to advance AI development (Google preserve us from well intended fools!), so they fund efforts to improve pattern recognition technology by awarding prizes for compression algorithms. (Incidentally, "prizes" as incentives were big in the 19th century and have made a come back in the past 10 years.)

The latest attempt to win the Hutter prize was the subject of the slashdot article. It suggests that this particular skill is close to the theoretic optimum - for text:

Slashdot | Text Compressor 1% Away From AI Threshold

Alexander Ratushnyak compressed the first 100,000,000 bytes of Wikipedia to a record-small 16,481,655 bytes (including decompression program), thereby not only winning the second payout of The Hutter Prize for Compression of Human Knowledge, but also bringing text compression within 1% of the threshold for artificial intelligence. Achieving 1.319 bits per character, this makes the next winner of the Hutter Prize likely to reach the threshold of human performance (between 0.6 and 1.3 bits per character) estimated by the founder of information theory, Claude Shannon and confirmed by Cover and King in 1978 using text prediction gambling. When the Hutter Prize started, less than a year ago, the best performance was 1.466 bits per character. Alexander Ratushnyak's open-sourced GPL program is called paq8hp12 [rar file]...

and from a later comment suggesting the text optimum may be misleading:

... Compression is about recognizing patterns. Once you have a pattern, you can substitute that pattern with a smaller pattern and a lookup table. Pattern recognition is a primary branch of AI, and is something that actual intelligences are currently much better at.

We can generally show this is true by applying the "grad student algorithm" to compression - i.e., lock a grad student in a room for a week and tell him he can't come out until he gets optimum compression on some data (with breaks for pizza and bathroom), and present the resulting compressed data at the end. So far this beats out compression produced by a compression program because people are exceedingly clever at finding patterns.

Of course, while this is somewhat interesting in text, it's a lot more interesting in images, and more interesting still in video. You can do a lot better with those by actually having some concept of objects - with a model of the world, essentially, than you can without. With text you can cheat - exploiting patterns that come up because of the nature of the language rather than because of the semantics of the situation. In other words, your text compressor can be quite "stupid" in the way it finds patterns and still get a result rivaling a human.