Author Archives: Michael

Robotic Piano Tuners

Robotic tuners actually do exist for guitars – but pianos have two to three strings resonating in unison per note, so tuning is not quite as simple. Nevertheless, it seems that a tuner could easily detect beats (probably using some sort of FFT/spectral analysis) and turn the tuning pegs automatically.

Human Pride

Humans are classified in the family Hominidae, along with the other Great Apes. Wikipedia has a history of the revision of the taxonomy of this particular family, and it’s an interesting modern example of the conflict between the human desire to set ourselves apart as special creatures and scientific objectivity dictating that we are not all that special from a biological standpoint. In particular, note the creation of tribe hominini after humans were merged into the same subfamily as chimpanzees and gorillas and the eventual reluctant merging of chimps into hominini as well.

Computer-Aided Cutaneous Testing?

I am wondering whether there are certain associations between the properties of the skin and, say, the presence of a noncutaneous infection in the bloodstream (or even a change in the normal levels of various hormones and other things that we currently need to take blood to test). At the least, one would expect quantifiable changes in the skin as a result of, say, leukocytosis, which can be used as a highly noninvasive biomarker for infection. It’s a potential area to explore using data mining.

Confirms what I said all along – the Medical Marijuana Debate on Digg

The disproportionate interest in legalization of medical marijuana on Digg is not due to any medical benefits. The development of a drug that decouples the medical effects from the high has shattered the pretense of righteousness that the adherents on Digg were previously clinging to. Take a look at the comments:

Digg – Cannabis like drug dims pain without the high.

Vs. a typical sort of article you’d see on Digg from last year:

Digg – Medical Marijuana User Commits Suicide After Long Being Denied Its Use.

I’m not here to engage in this debate (although it’s pretty predictable to see where someone who has lived a straight-edge lifestyle would fall), but I find hiding behind some flimsy pretense very hypocritical regardless of your position.

I can already see what the comments are going to look like, so I’m disabling them on this post.

A Counterexample

If it really takes 10,000 hours to master a skill, then my own mastery of programming is a counterexample. I literally coded for 8 hours a day (pretty much all of the time when I wasn’t asleep or at school, time that most people would allocate to homework included), which would translate into 4 years of coding to attain mastery. I wrote some amazing things at 12, but nothing I would consider a masterwork (Metasquarer included; the first version was riddled with bugs). I don’t think I truly mastered the skill until 17 or so – 9 years and approximately 26,000 hours of work.

And I was always way ahead of my peers in this skill, from the early stages onward, so it’s not as if I had some below-average progression or anything.

System administration, by contrast, is something I’ve spent very little time on. I started at 16ish and spent about 2 hours a day at most on it. By 19, I had attained mastery in the domains I had studied (which is to say that I’m not a master at every aspect of the skill; just the ones I’ve used. But this tends to be enough).

So I’m not sure I’m following the pattern here.

Dual-Locking Critical Sections

One potential way to avoid deadlocks would be to unlock critical sections if the main lock controlling the section was unlocked OR if another lock indicating the resource was unavailable was unlocked. The program could then deal with the case that the requested resource depends on another locked resource without locking up.

Mundane

I keep thinking that I should be doing something more exciting, more groundbreaking, more novel. There’s my university project, but I mean something that actually puts the considerable level of skill I’ve acquired in computer science to use. I didn’t go to grad. school to become an entrepreneur, after all; I went there to become a scientist.

But what grand challenges are there in my subfield? The only one I can identify is strong AI, and that is a problem I have no inclination to work on. There’s nothing like the development of MRI to be done here; the best I can hope for is development of some fancy image processing algorithm. Where are the deep, paradigm disrupting innovations? Mostly things that tend to be commercialized, rather than purely scientific innovations. Stuff like Google, which, had it not been commercialized, would have just been an interesting footnote in an algorithms journal somewhere. I’ll get to working on commercial things more at some point (I’m destined to, in a way), but not at the moment. I don’t have the time to split between what I’m currently doing and attempting to run another organization.

I am planning to pursue the biomedical aspect of my research more by enrolling in a medical program in 2010, with the hope of eventually convincing someone to finally train me enough to work on treating diseases, but I feel that there should be something more to show for my CS training than the work I’ve published.

The relationship between position in the classroom and performance appears to be causal.

Any teacher can tell you that there is definitely a correlation between distance from the front of the classroom and performance. But what I found interesting today was that my class, being confined to a smaller room and thus forced much closer to the professor, appeared bolder in answering questions and generally more engaged in the lecture. In short, they performed far better in class. This seems to suggest a causal relationship.