Next fall, I'll be writing another grant application. The goal will be to secure a second round of sweet, sweet government research funding. Of course, there's an incentive to get that money: if I don't, my career is basically toast. My only remaining option would be to win the lottery and award myself a permanent fellowship. That's why it's on my mind many months in advance.

What I find interesting about applying for this government grant is that you're not held accountable to the objectives of your previous application. Sure, they gauge how successfully your spent their money last time. But it's simply not important whether you spent the money on the projects you described in your previous proposal, only that you achieved something. And thank goodness. It's important to have this as a safety net. If the project you apply for fizzles out after a month, you need the opportunity to shift your focus.

This fact makes it possible to hack the application process in interesting ways. My uncle, for example, always used to stay one project ahead of his grant applications. He would use the previous round's money to pursue some exciting new project, do all the research, and then request money in the next round to fund the research he just completed. It's very easy to write a convincing grant application when you already know the project will succeed! Sounds like a good technique, if you can work hard enough to get one project ahead of your funding.

But I think that this model can be exploited even further. Here's a strategy I would like to propose. You work really hard and craft the perfect research proposal. Full of suspense and redemption, awash with subtle wit, it would bring Nobel laureates to tears. Knuth would marvel at its flawless typography. You describe some hypothetical project that we'll call Project A. The proposal is so astonishing that it is fully funded. So you take the money, work on some other Project B, and simply re-submit the proposal for Project A. You're not responsible for having completed Project A in the first place, and you already know it's an outstanding proposal. Why shouldn't you get funded again at the same level? Sure enough, this buys you time to work on Project C. Repeat until retirement, picking whatever Projects D, E, F, and so on that strike your fancy. You only ever need to write one grant application, as long as you endeavour never to work on the associated project.

The only problem I can see with this approach is that sooner or later, you'll end up with reviewers who have read previous iterations of your proposal. Will they write negative reviews if they recognize your writing from five or ten years ago? Maybe. Or maybe they'll have a strong incentive to re-submit the review they wrote back then. Then everybody wins!

Our funding model is pretty flexible, I suppose. That suits me just fine. I'm not sure I could keep up with a rigid schedule of agency reviews or, um, site visits.

This is fantastic. There's a news story making the rounds about Michael Manga being named one of the Sexiest Men Alive by People Magazine. What's great is that Michael is not some vapid Hollywood heartthrob, but a professor in Earth and Planetary Sciences at Berkeley. Indeed, he is a recent recipient of the MacArthur genius grant. And there he is in the pages of People Magazine, right underneath Bono.

Naturally, this is deeply embarrassing for Michael, who's just trying to get research done. He'll no doubt be the butt of departmental jokes for decades. But still. Genius grant and sexiest man alive? Who wouldn't want to be recognized by those two disparate but equally important agencies? Will he put it on his CV? Will he be featured in the MacArthur Foundation's annual "Sexiest Geniuses" calendar? And of course, has he earned a chili pepper on ratemyprofessors.com? (Sadly, he doesn't seem to be listed there at all.) If there's a lesson to be learned from Michael's experience, surely it's that there's still hope for us all. I didn't forfeit my opportunity to be recognized as Sexiest Man Alive just because I chose to go into academia. That's a relief.

Naturally, I can't end this without including a note about my friend Erik Demaine. You think you're some kind of genius, Erik? You've got nothing! OK, well, you're halfway there. You've got the grant. But now I guess you've got something to really aspire to.

A few months ago, I mentioned that CBC Radio 2 was going to re-broadcast their excellent series The Wire. I knew that I very much wanted to listen to the whole series, but I couldn't know in advance that I'd be available every Sunday afternoon to sit and listen to it without certain three-year-olds yelling in my ear. So I resolved to capture the audio streams and listen to them later, when I knew I'd have some peace and quiet. If anyone from CBC is reading this, sorry! I promise not to share these recordings with anyone.

Of course, when trying to record audio streams from the internet, I run the risk of losing the connection to the server and recording silence while the connection is re-established. I didn't think I'd be able to capture the entirety of the eight episodes, but by golly I was going to try.

The series is over now, and I did alright. As far as I know, five of the eight episodes are recorded perfectly. Another has a ten or fifteen second gap in the middle.

The other two are mostly crap. One has five-second dropouts every few minutes. The other (the final episode) is particularly interesting. I simply couldn't connect to the CBC Radio 2 server that day. I wanted to have something recorded, so I plugged my radio into my computer, and recorded the analog signal picked up by my tuner. Unfortunately, I don't get very good reception of CBCR2, so the recording is full of static.

It's interesting to compare the dropout episode with the static episode. For the sake of argument, let's say that each contains about 75% of the total information content of the original audio source. In the dropout episode, the information content is all-or-nothing: 75% of the time you have the complete signal and 25% of the time you have silence. In the static episode, the 25% loss of signal is smeared continuously over the stuff you want. It's tempting to say that dropouts are the 21st century equivalent of static. No streamed audio signal will ever suffer from crackling noise, so this is the form of information loss we'll have to put up with.

Now here's the big question. Which is better? If you had to accept 25% signal loss, which would you prefer? And here's the answer: you'd always prefer analog static over dropouts. The reason is simple. Our brains are hard-wired to deal with that kind of noise. Every day, we need to filter lots of junk out of what we see and hear in order to make perception possible. After a while, we even adjust our gain and fail to notice the noise at all. If I'm listening to a speech where every sentence is missing two or three words, I'll have a hard time getting the speaker's message. But I can easily reconstruct the speaker's words if they're relayed to me through a walkie-talkie.

Unfortunately, streaming media just doesn't work that way. You either get 100% fidelity, or you get nothing. That raises an interesting question: is there an alternate encoding of digital audio that degrades (but doesn't drop out) if the signal goes missing for a few seconds? You'd need something that smears a low-res version of the signal out over a wide time window, and sends the high-res details right on time. This property of "degrading gracefully", recognizing the features of human perception, if highly desirable. Alas, while I'm interested in human perception, this isn't my department.

Note that these questions are particularly appropriate given they arose in the context of recording a show about the impact of electricity on music. Here, we are concerned with the transition from electricity to electronics, and analog static to digital static.

The hard part of the term has finally come to an end. I taught my last class two weeks ago, agonized over putting together a final exam, proctored the exam last Friday, and marked it immediately afterwards. I finished writing the nine paper reviews for two of the three program committees I'm on (the seven reviews for the third PC are due in February). Most of the internal committee work has wound down for the year. I now have some time to work at a more reasonable pace.

The main thing I need to think about between now and the end of the year is the new grad course I'll be teaching next term. No, it wouldn't do to teach the same grad course as last time, now would it? That would require less effort. On the other hand, I have a feeling that with these two grad courses to choose from in future terms, it's unlikely I'll need to come up with any new ones for a long time (ack -- who am I kidding?). Anyway, I need to put together an outline, prepare lectures, create assignments, and put the whole thing up on a web page. I know from last time that it's theoretically possible to do just-in-time course prep for a grad course, provided I want to stay up until 1:00am every weeknight. I'd just as soon avoid that.

The other important consideration is research. I'm told that many other professors occasionally find time to do it, and it's something that sounds fascinating. I hope to engage in it some time before the end of the year.

Seriously, I don't get nearly enough time to do research, especially at my preferred pace (slow). I'm seriously considering taking a six-month sabbatical in the second half of next year. My university makes this option so attractive that it's really hard to see why I wouldn't want to do it. I'm still trying to find a negative aspect of this sabbatical, just so that I can feel I thought about it hard enough. Well, I'll go talk to my chair about it soon enough.

For the record, I have no intention of leaving town. That's far too much effort for my taste, especially if it's only for half a year. It would be too much like being a CO-OP student again, except that I'd have to bring three humans and three cats along with me, not just a garbage bag of laundry and books in the trunk of my car. Besides, all I want to do is work quietly in semi-isolation for six months. All I need for that is an office-away-from-office. Finding such a thing will be the tricky part.