Music 17 Feb 2010 01:26 am

Pandora and Finding New Music

Pandora needs a “people who like what I like also like…” system, like Amazon’s. Actually Amazon’s is “people who bought this item also bought…”, but the concept is similar.

I find that when I create a new Pandora station seeded with a few specific songs/bands, the station is great for a bit but really doesn’t have much selection after a while — songs I’ve thumbed-up are repeated frequently, and if I listen for too long at a stretch, as the station “runs out” of songs matching the original criteria and starts varying the parameters a bit to introduce a bigger selection of songs it can play, the new songs added really are by and large not all that great. I end up continually using “thumbs-down” on new songs presented, which in turn means that new songs (as we go farther and farther out into the pool of available music) are even more random and dissimilar to the original seeds (with fewer and fewer things I like), and so on. At that point a given Pandora station has been “depleted”, and isn’t worth listening to for a while. (Since when I do listen to it, it’s mostly stuff I’ve heard already.)

On the other hand, I take as an axiom that there are many, many amazing songs out there in the space of all music which I have not heard; songs which may have dimensions and sounds widely or wildly different from those matching the songs I’ve already liked and which I might never get to, otherwise. I want to hear those ones! I want to know what “people whose tastes most match mine” like. The parameters of a specific song, which might look like this, for example, according to Pandora:

a Modern style
a symphony orchestra
harp playing
cello
violin
major key tonality
a singing, mellifluous aesthetic
a tranquil feeling
a well-known composer
a bittersweet sentiment

…are academically interesting and useful for finding similar-sounding pieces, but I don’t truly care about similarity of sound after a point. There’s no magic formula based on categorizable parameters (even if we were to come up with much more abstract ones than the above) for what I’ll like, I’m sure, but I know that my tastes do align very strongly with certain people’s. For example, I tend to like a lot of what I hear on KCRW.

Maybe last.fm or other online radio experiences are better in this respect?

And not to knock Pandora too much; it’s great and I use it frequently (to drown out the Nerf-dart fights and stuff going on around me in the office when I actually have to work), just with frequent use one starts to come up against some limitations.

Book Reviews & Face 26 Jan 2010 01:18 am

Chasing Medical Miracles: The Promise and Perils of Clinical Trials

A well-researched and personally-grounded exploration into the ethics and business of medical studies, and indeed: what a big business it has become. The author himself has diabetes, and decides to undergo an experimental procedure as a test subject, a human guinea pig as it were, where donor insulin-producing cells are transplanted into his liver as part of a potential cure. He decides to learn more about clinical trials and presents the results of his extensive research, while we’re kept in suspense until the end of the book as to whether his own experimental treatment was ultimately successful.

Here are a few things that come to mind when I think about medicine.

1. It’s amazing how far we’ve come, and how much we know.

2. I’m often reminded how basic/primitive, still, our understanding is of many of the body’s systems. Actually I wouldn’t use the word “primitive” with respect to our understanding, rather with respect to our current ability to manipulate and fix, compared to what should eventually be possible. We understand a lot but can’t do much about it.

3. Contrasting the above two ideas is probably a tired cliche. But I’m still continually struck by it. How far we’ve come, and how far we’ve yet to go. Yeah, yeah. Bear with me.

The author doesn’t take sides, and tries to present various concerns in a balanced way — and some of the ethical issues are pretty fascinating. For example, there was a trial of anti-HIV drugs done in Africa where participants in the study either received the drug being tested, or a placebo. Meanwhile, back “home”, there are standard, helpful HIV treatments available, and common practice is to compare a new drug with the current best-known treatment on the market in order to determine whether the new drug is any better than the current baseline. But because poor, HIV infected Africans didn’t ordinarily have access to the baseline drug in the first place, it wasn’t administered as part of the study, either. As a result, half of the study participants (who received the placebo and effectively went untreated) died or became sick. There was, not too surprisingly, an outcry about the ethics here. On the one hand, by doing a placebo-controlled trial, the new drug could potentially have been approved and brought to market much faster. On the other hand, placebo-based trials are almost never done in the U.S. or developed countries when baseline care (my terminology) is available. The drug company which sponsored this study argued that ultimately many more lives would be saved by faster approval of its drug, and futhermore that the people receiving the placebos wouldn’t have likely received any care at all, anyway. Others argued that people participated in the trials out of desperation and under the belief that they would receive actual treatment, and to have denied baseline treatment was unfair exploitation — that although we know that the one and only purpose of medical research is to gather data, not to provide treatment, in many circumstances it’s difficult for subjects, particularly those who don’t have any routine or even accessible medical care at all, to understand or believe this no matter how it’s explained to them (and researchers have an incentive to not explain it tirelessly but rather to allow or encourage the formation of hope).

I’ve encountered a few friends or colleagues recently who have tremendous mistrust for the medical establishment. (That goes together with stories in the news about parents who somehow remain convinced that there’s a link between childhood vaccines and autism, I think.) One person believed that chemotherapy is essentially a scam perpetrated by doctors and hospitals in order to make money, an expensive and ultimately unnecessary treatment because there are other, cheaper, more natural, less harmful treatments available. (That view disgusted me, since I and most of us have friends who’ve dealt with cancer.) Or that in the field of psychiatry, drugs are “pushed” on patients by doctors who are paid by drug companies to write prescriptions. (Incentivized? An area of concern, but not something with which I’d condemn a whole field of medicine.) In reality, the world of medicine is extremely complex, and simplistic conspiracy theories like that just don’t ring true. The key is, everything in medicine depends on data. If there’s some compelling evidence that, for example, eating precisely ten navel oranges a day cures cancer better than chemo, somebody will probably fund a clinical trial to test that. In general drug companies do fund the majority of studies, but there are layers of independent review boards, politicians, research hospitals, and lawyers ready to swoop in, all of which creates an atmosphere of oversight and regulation. Private donors, too (as in the case of the diabetes-cure trial the author underwent). It’s not perfect, and there are murky ethical boundaries sometimes crossed, policies which need to be tweaked, incentives and regulations to be adjusted and so on, but as a whole, the machine works.

Overall, Chasing Medical Miracles conveys a message of hope. More clinical trials than ever before are being done, more stuff is being tested, and although “miracles” are unlikely, we’re chipping away at the rock face, gathering data a little at a time, and I think it’s okay to put some trust in the system.

Books & Face 26 Jan 2010 12:31 am

Ender’s Shadow

A small comment about writing style. (I wrote this a long time ago. Found it in a drafty place.)

I read the original Ender’s Game by Orson Scott Card years ago, and at the time it seemed “simplistic”. The hero (Ender) and the path he took were just too perfect. He faced his challenges almost robotically and we never had to struggle; we were never really in any doubt as to his ultimate success. The story was direct and told without flair, almost as if it itself were written by a child. What set the book apart, though, was the depth of analysis of the characters’ absolutely every action and thought and intention; the lack of subtlety and, truth be told, the refreshing departure from “show, don’t tell”, that old writers’ maxim which was drilled into me in elementary and high school English classes. Okay, I confess at this point that I don’t remember the prose of Ender’s Game all that well, but with respect to Ender’s Shadow, this commentary seems very applicable.

The boy ate the banana and was no longer hungry.

That would be “tell”.

The banana peel lay on the table, and the boy’s stomach stopped rumbling.

That would be “show”.

The second seems far more literary. Want to get even more literary? In the “show” version, I still “told” you some facts (about the banana peel and the boy’s stomach) which implied the idea I really wanted to convey. But why not show the above facts, themselves, instead? This’ll sound a bit absurd, but that’s the nature of the beast, as we go one level of abstraction further up the ladder:

If someone had walked on the table, he would likely have slipped in a classically comical way, while meanwhile in the boy’s stomach elephants were no longer on parade.

That’s a pretty absurd way of describing lack of hunger due to having consumed a banana, yet we see prose which aspires to that level of obliqueness all the time, and it’s often venerated for that. What does it add, really?

My point is that Card’s novels work because he just flat out “tells” and doesn’t pretend to be doing anything more than that. Although the reader isn’t left anything to figure out, I maintain that for the most part, making readers figure things out is mostly tiring, as is the telling of irrelevant facts (e.g., the fact that the banana peel is on the table; the fact that the boy’s stomach was rumbling) in order to be “showing” the important facts which are illuminated only by implication. In truth: take the flowery language and non-linear narrative blocks out of many books, and the story itself would compress quite a bit. Perhaps the same mood wouldn’t be conveyed, true, but sometimes “he felt sad” is just as good as “he hung his head in despair”, unless he really did hang his head in despair, but that’s just a literary cliche at this point and we don’t even know what he really did with his head when you say that, if anything (other than feel sad with it) unless more specific non-cliche actions are described.

Card just says “he felt sad” and uses the extra space to explain exactly why he feels sad, what his sadness means, what he is doing differently now because of his sadness, what he feels about the fact that he feels sad, how other characters react to his analysis of his own sadness and, in turn, their analyses of their reactions to his analysis of his sadness and their reactions to their analyses of such, and so on. And all of this is somehow relevant to the story.

That makes it atypical and fun.

Database & Uncategorized 25 Jan 2010 11:46 pm

Categories

Why does WordPress have a selectable “Uncategorized” category? Isn’t that contradictory?

Further, although it automatically sets “Uncategorized” upon save if no other categories were selected, it doesn’t unset that category when selecting other categories, leaving open the possibility of a post being both definitively categorized and categorized as Uncategorized all at the same time. How silly is that?

Face 27 Oct 2009 12:12 pm

Virgin Math

From Los Angeles to Seattle.

Standard ticket price: $64
Refundable ticket price: $222
Cancellation/change penalty fee for standard tickets: $50

Hmm…

Uncategorized 25 Oct 2009 01:12 am

Passing By

I was walking home from work, heading along Wilshire, passing 4th, when an elderly gentleman driving a car towards the ocean called out through his passenger-side window: “Sir, is the ocean that way?” I called back “yes it is!” and we continued in our respective directions.

Uncategorized 14 Oct 2009 04:30 pm

Database Artistry

I wanted to register the domains “databaseartistry.com” and “dbartistry.com”, but these were already taken. Figures; anyway I can be more creative than that. :)

Uncategorized 20 Aug 2009 03:02 am

Relationship Generator

Text of ad:

Are you single? My name is Markus and I created Plentyoffish.com; my site is completely free and we generate about 800,000 relationships a year…

Hmm, generate relationships? Interesting language. I guess you could put it that way from a statistical perspective, as in, salesmen, sorry… sales_people_ generate leads. Perhaps that’s a poor example, though; salespeople don’t just discover folks who happen to be already interested in the product being sold (but just don’t yet know it), although that’s one way of putting it, rather they also (and maybe this is the bulk of the marketing effort) create that interest, and in the case of leads-generation we might imagine they do it one folk at a time. [Side note: "folks" is one of those words without a singular case, it seems. Like cattle; see the "Singular terminology dilemma" section.]

Are you single? My name is Markus and I created Plentyoffish.com; my site is completely free and we foment about 800,000 relationships a year…

Nah, that’s not it…

Are you single? My name is Markus and I created Plentyoffish.com; my site is completely free and we hook up about 1,600,000 people a year…

Hmm, that facilitates a larger number, and plays to the reputation I hear the site’s earned, ha ha, but nah, I’m sure that’s not the angle they’re going for, either.

There’s gotta be something better than generate

Internet 20 Aug 2009 01:45 am

Evite: A Blast from the Past

I dislike the expression “blast from the past”, as if the fact that it rhymes with itself makes it somehow worthy despite the fact that “blast” hardly makes sense here, semantically. But, wow, the Evite photo viewer //stinks//. Is this still 1995?

  • Maximum photo size appears to be 400 × 300.
  • Uploaded mages are re-compressed at very low JPEG quality.
  • Normal photo viewer does not use AJAX / JavaScript when going from one photo to the next.
  • Website is slow.
  • …So it takes several seconds to skip from each photo to the next.
  • “Slide show” feature is faster at skipping from one photo to the next, but does not pre-load the next photo, so it still takes a moment.
  • Ads are shown all over the place.
  • Uploading requires Java applet.
  • Photo viewer often renders random JavaScript code as visible on the body of the page.

On the positive side, it doesn’t matter if I have my eyes closed, because the resolution is so low that you can’t tell.

Evite has been sitting pretty for too long; they really should upgrade their technology. The problem is that the //name// is so evocative that it has virtually become part of everyday language, and there’s a lot in names. (“Add me to the evite!”, you say, using a word that’ll be added to the O.E.D. any day now, I’ll betcha.) Speaking of names, I don’t remember if I posted about this, but I firmly believe that Dukakis lost against Bush because… who would want a president named Dukakis? Say it to yourself. Dukakis, Dukakis, Dukakis. Now //Bush//, on the other hand, that sounds positively presidential. I’m so glad people follow their conscience(s) when they vote.

Uncategorized 20 Aug 2009 12:39 am

Unlinking Myself

I’ll keep this short, because it’s a meta post. And “meta means murder” (I read somewhere). And if I read it somewhere, it must be true! Anyhow, let’s continue.

After much figurative navel-gazing, I realize I’ve been reluctant to post here because I’d subscribed my Facebook profile to this blog, such that everything I wrote was being cross-posted over there. And as cool an idea as that was, it meant that in the back of my head I was writing for a lower-common-denominator audience. Oh, not to denigrate my Facebook pals in relation to all of you dear readers and dear leaders (even though they won’t see this, ha ha); rather, I know what those folks are interested in, and there are pretty large swaths who don’t care about random musings on JavaScript replete with code examples, which is perfectly fine; I still love them all. Facebook being a “social” network, just as I’d probably not walk into a party and subject a relatively random group of friends to a discussion of Objects versus Arrays (complete with code examples), not even at a Google party, the same goes for the large virtual party which is my Facebook profile. Yes, that’s where the party is happening; don’t forget: you read it here first!

A better solution by which burnination can be achieved: create a separate RSS feed which only incorporates items I’ve tagged with “facebook” and link that up; only those party-worthy pearls will get auto-posted. Like the proverbial hole in the fence around the nudist colony, I’m looking into that.

Uncategorized 02 Aug 2009 04:34 pm

Single Junior Efficiency

Terms for apartment units evolve. What was once a “bachelor” (and perhaps before that an “efficiency”) is now (probably for political correctness) a “studio”. What was once a “single” is now, apparently, a “junior one-bedroom”. Why? Because they added a little three-foot-high divider around the area where one would most probably want to put a bed. Bingo, instant bedroom. Or “junior bedroom”, whatever that means. Smaller than a…? Subordinate to but still a…? Really, it’s just a single. Those clever real estate salespeople! Living in a one-bedroom apartment at present, I wouldn’t immediately consider a much smaller-sounding single; that’s presumably a downgrade. But “junior one-bedroom” sounded intriguing, so I checked it out. They got me in the door, at least…

Uncategorized 24 Jun 2009 02:27 pm

How to Get a Million Comments on your Blog Post

…Post [http://www.codinghorror.com/blog/archives/001278.html something] about the Monty Hall problem.

Poetry & Wordplay & Randomness 22 Jun 2009 08:15 pm

Quoth Willard R. Espy

I have a selective memory (don’t we all). One of the things it selectively remembers is silly poetry from my youth [father William, the young man said]…

: //I have a little philtrum//
: //Wherein my spilltrum flows//
: //When I am feeling illtrum//
: //And runny at the nose//
:
: –W.R.E.

Technology 22 Jun 2009 08:07 pm

Click Hijacking

…When I’m about to click on something, when the “press down now” signal from my brain is already en route to my finger, but an instant before the actual click takes place a dialog box from some other program pops up on top of the window and steals my click. I then have no idea what I just clicked on, because the dialog has immediately gone away and was only onscreen for part of a second. Perhaps it was a box that said:

: “Would you like to format your hard drive now? [ Yes ] [ No ]“

All I know is that that just happened, and that now my hard drive is making lots of noise. Hmm…

Uncategorized 19 Jun 2009 10:28 am

In Soviet Russia…

…Nikolai Ivanovich Lobachevsky plagiarizes //you//.

Life 16 Jun 2009 03:46 pm

Mechanical Difficulties

There’s a weird dance you have to do with auto mechanics. I strongly believe in second opinions, but that culture doesn’t really exist, as far as I can tell, with respect to mechanical repairs. I feel like if I’m told “your O-ring distributor valve spark-plug pan gasket is broken” and I say “I want to get a second opinion on that”, that I’m directly calling into question the mechanic’s honesty, a clear affront. We all know the human body is extremely complicated and that when a doctor makes a diagnosis, it’s a sort of very educated guess, basically a deduction based on symptoms, and doctors themselves are (or should be) happy to have a diagnosis checked — (strongly encouraging towards, even). It’s not about honesty in medicine, but about the realities of the trade. Some conditions might have a very obvious diagnosis (in which second and third opinions will always agree), and some might have subtle symptoms where other doctors would not agree — in which case the other opinions give you (and each doctor) more information to work with. But a car is just a mechanical thing, and we assume a good mechanic should be able to tell what’s wrong with 100% accuracy — if there’s doubt, just take it apart some more until he //knows// what’s wrong for sure, although it should generally be obvious. You never hear “well, we //think// it’s the master slave cylinder, but we still have to run some more tests…”

Anyway, the fundamental issue is that I don’t implicitly trust auto mechanics, even when they’ve been specifically recommended to me. Well, why should anyone, given everything you hear? (…All those “hidden camera” operations where 4 out of 6 garages will claim they’ve fixed something but really haven’t done anything, and so on. E.g., the [http://www.snopes.com/inboxer/scams/jiffylube.asp Jiffy Lube scam]. If it happens in big chains with lots to lose for getting caught, you can bet it happens in independent garages.) Actually, the recommender’s word carries some weight, but then (in a specific case of my own yesterday, moving from the hypothetical to the real) I got a call telling me I needed a couple thousand dollars’ worth of additional repairs. Power steering rack and pinion system needs to be repaired because of a leak ($700)? Huh? Master clutch cylinder has a leak ($300)? Why have I never heard of that part before? Anyway, my point is that because I don’t have a detailed understanding of everything under the hood, I don’t have a way of fact-checking what a mechanic tells me besides my intuition, and that’s (as mentioned) limited — all the knowledge I have is from past repairs and from what I’ve tried to pick up by looking under the hood myself now and again when smoke starts pouring out, or something.

But I do have one trick up my sleeve, and that’s the ability to play one mechanic against another — to get that second opinion. What I like to do sometimes is bring my car to a dealership, pay $100 for a full inspection, and see what they find. I feel like dealers are pretty (or at least //more//) reputable in telling you what’s wrong with a car, because they charge enough for actual repairs to not have to bother with faking problems and/or work. Anyway, I take that list of carefully itemized stuff, call up independent garages, and ask how much said repairs will cost: There’s your second opinion on parts & labor, as opposed to the problem itself. (I’m not even bothering to cross-check the dealer’s price, here; that’s going to be several times other mechanics’ prices just as a given. But people actually pay those prices simply because they don’t trust the average independent garage — and understandably so. And I’m making an assumption that dealers are trustworthy, too, but it also seems more likely they would be simply because they have more on the line if caught than independents.)

So (back to reality), I brought my car in yesterday to a garage recommended highly by a coworker (and highly rated on Google Maps, too), because the engine was overheating, according to the guage. No problem, they diagnosed that as a faulty themostat. Okay, I’ll pay $160 to fix that. Maybe it’s on the high end of the price scale, but it’s not worth the time to shop inexpensive things around. But their suggested $2000 of additional repairs? I’m usually told about those various leaks (if they seem serious) whenever I get an oil change, so the sudden appearance of all these expensive mechanical issues puts me on guard, and I have to tell the nice mechanic or “garage salesman” (since I’m sure he wasn’t the one actually doing any of the work) not to fix anything other than the thermostat (and change my front tires, all right). After having described and detailed all those other problems in such a grave tone of voice it was almost strange that he didn’t further attempt to sell me on the other repairs, as in “…but your car’s gonna explode if you don’t fix this!”, like there was some sort of miscalibration in the story — it was just a light “okay” now, as if he tacitly recognized that I wasn’t a sucker and immediately backed off. I threw in a “…but I’ll watch out for the other issues” to validate his mechanicsmanship, and asked him to itemize all of the findings. So now I’m tempted to bring my car in for a dealer inspection and compare the findings from that, side by side, with these — the proof should be in the pudding, as it were. Or maybe my car did have these problems, but “seriousness” and “price-to-fix” were both exaggerated. But this had me wondering, how often do people do this kind of thing (seek second opinions under some other guise)? I’m sure there are trusting and mistrusting folks. Some people (“suckers”, assuming widespread dishonesty by mechanics, which there’s evidence is the case) who do exactly what’s suggested and pay full price for it every time. But of the ones who compare, who get second opinions… What if mechanics colluded, and had a secret online database where they shared fake repair findings tagged by license plate or car description? Okay, that seems too far fetched, now that I write it. Especially because other fields with even more nebulous findings don’t need to stoop that low. (I was reading about chiropractors the other day, and how second and third opinions very frequently report completely unrelated conditions, as if diagnoses had been pulled out of a hat.) So I don’t worry about a hypothetical level of collusion among independent mechanics (even the ones next to dealerships) but still, my intuition pinged, and now I want to get that second opinion just so I can personally validate what //seems// to be a general trend in the car repair community. And show up the guy who recommended this particular shop.

Life 16 Jun 2009 02:46 am

Red Light District

I hope I didn’t just get a ticket. 2 AM, no cars in sight, stopped at a light at the intersection of Wilshire and Sepulveda. I’d come to a full stop, and moments later the red-light camera at that intersection flashed a bunch of times. I guess because my front wheel was half a foot over the limit line (of the crosswalk — not even the intersection proper). I was about to get out and take a picture of my own from the side to show where exactly the wheel was, but a couple cars were coming from behind and by that point the light had changed, so it was safer to just proceed, and anyway the evidence had already been collected. Hopefully a false alarm and the camera was just doing it’s thing, starting off an attempt to construct incontrovertable proof of antisocial driving behavior from the moment I first poked my nose across the line, but since the remainder of the series just shows me sitting there and not proceeding across the intersection, it’d be ridiculous to say I “ran a red light”. I don’t have high expectations for //not// getting a ticket in the mail and having to invest a few man-hours fighting it (and ultimately failing) so that I don’t have to burn my “get out of jail free” traffic school card (or perhaps I’ll pay now and save that for my next ticket for going 80 mph on the freeway just like everybody else). I’m not bitter! There we go, that was cathartic. On to better things.

Randomness 15 Jun 2009 05:09 pm

Arbitrarity

There’s a store on Santa Monica Blvd. called Tuesday Morning. I wonder if there’s anything special about shopping there on Tuesday morning, like some kind of discount given. I don’t shop there, but it would just satisfy my sense of world order and make me happy if there were.

Books 13 Jun 2009 02:31 am

Book Piracy and Publishing Shenanigans

Why is piracy of books in electronic format not an issue? Why don’t we find large volumes of PDFs of scanned books or cracked e-books on peer-to-peer file sharing networks? Why aren’t book publishers up in arms, suing book pirates left and right?

Book sharing isn’t rampant, but imagine if it were. Now… would we (as a society) really mind? Because people are //stealing//, fine, but also people are stealing //books//. There’s a little bit of cognitive dissonance in this.

Different but related topic: Why do book publishers get away with releasing a new edition of, say, //math// textbooks every year? My little brother at UCLA had to buy a new $100 textbook because the used edition from last year didn’t have the right homework problems or the right problem sets in the right order, or something like that. You’d think that for a subject like math, the 1950’s edition of the book would be just as good, as long as it didn’t require the use of a slide rule.

Artwork 09 Jun 2009 01:17 am

Twisties

A couple random designs, done at some sort of gathering at [http://www.bikerowave.org/ BikeRoWave].

[img[/design/random/twist-a-t.jpg]] [img[/design/random/twist-b-t.jpg]]

View them in their [http://mh-z.com/design/ natural habitat].

Next Page »