If I Only Had a Brain

Via the always awesome Makezine blog*, this is an interesting article promoting "The Importance of Stupidity in Research." It's worth a read. The premise is that research, actual honest-to-goodness groundbreaking research, involves a lot of flailing about in the dark and feeling stupid, and people tend not to be accustomed to that feeling, because it's the opposite of how you're supposed to feel in your classes, where you're supposed to always know the answer.

This meshes well with a truism that I've been known to spout, which is, "If it was easy to do, somebody would have done it already." At least, for really novel and groundbreaking research, that's true. I suppose there's plenty of fill-in-the-gaps type of stuff that people have to do, the sort of stuff that requires labs full of technicians and grad students using time-honored techniques and a lot of elbow grease. But I think that's almost an entirely different category of research.

On the other hand, the converse of this sentiment was expressed once by Albert Young** when I was at Princeton, and he said, "90% of physics is just not being stupid."

*Speaking of which, if you're in the bay area, you should definitely attend the Maker Faire this weekend.

**Remembering Albert's name took a Herculean effort of googling. Basically, I had to remember that he was a classmate's thesis advisor (to whom he imparted the wisdom), and then I had to find a catalog of Princeton theses and look up that classmate's thesis. Albert is now apparently at NC State in the physics department, according to the intarwebs.

Dueling acronyms

Courtesy of Greg Snyder:
SHRiMP: Accurate Mapping of Short Color-space Reads

Not to be confused, of course, with:
SHRImP: Single-molecule high-resolution imaging with photobleaching

It's all in the placement of the capitalization (apparently.)

I also note that the first author of the SHRiMP paper is one "Stephen Rumble," who I can only assume is a third cousin once removed of Stephen Quake.

Nanorobotique

From Laboratoire De Nanorobotique (how freaking cool does that name sound?) at the École Polytechnique de Montréal:
Magnetic sensing bacteria are used as an external propulsion system to direct a solar-powered microsensing device. No published details yet as far as I can tell, but that is really awesome.

Non-Blinking Quantum Dots

Quantum dots are cool. They're bright, small, and they don't photobleach. The problem: they blink, and the "off" state can last a long time. Certain preparations can reduce the blinking, but personally my luck with this has been pretty middling. Now, apparently, a group of researchers has produced non-blinking quantum dots. All I can say is: wow! And: when can I buy them?

Arbitrary Friday: Sandwiches

Thinking lately about what it is I want to do with my life when I get out of here, and I have been thinking about things like professorship (hard to get in any place I want to live), biotech (I don't really want to spend the rest of my life at the bench), software (my first real job, and always a convenient fallback), VC (will six figures soothe a tortured soul?), and other possibilities. But, six or seventh down the list, I think I'd like to open a sandwich shop. Not really, of course. There's no better way to lose your shirt than to open a restaurant. But, I do love sandwiches. I mean, I lurrrrrve me a good sandwich. And I would love to spend my days perfecting the art of the sandwich, serving them to others, and making the world a better place by making it more sandwichy. Here are some of my favorite sandwiches:

• Primanti Bros in Pittsburgh. They put the french fries in the sandwich!!!
• Bat 17 in Evanston. Man do they make a good sandwich. Everything is toasted and smothered and cheesy and goddamn now I'm hungry.
• The Bread Company in Urbana, IL. Their swiss bread is amazing, and they make a really mean tuna salad.
• Jimmy John's. As fast food goes, these guys are the best. They use very high quality ingredients, and their selection is great. I would take a Jimmy John's over a Quiznos or a Subway anyday. Or over almost any other fast food, for that matter. And they're also headquartered in Champaign-Urbana.
• Lutticken's at Stanford. Lutticken's isn't amazing, but it's certainly the best you can get locally. They don't do sandwiches well here on the west coast, in my opinion, but Lutticken's gets points for their excellent Dutch crunch bread, their very tasty meatball subs (both meatful and meatless), and for being locally owned.
• Hoagie Haven in Princeton. Whenever I go back to Old Nass, I try to stop by The Haven for a peppersteak hoagie. They are legendary, for good reason. Fried on a greasy grill, and stuffed into your eager mitts at 1AM. Nothing satisfies like a Hoagie Haven hoagie.
• Mr. Beef in Chicago. Italian Beef is to Chicago what Cheesesteak is to Philly, and nobody does it better than Mr. Beef. Portillo's can stuff it, they don't even come close. One glance at the wall covered in the media coverage and the unsolicited celebrity endorsements is enough to convince you that the place has a well deserved reputation. (I've read two different interviews with Chicago native Joe Mantegna in which he extolls the virtues of Mr. Beef, apropos of nothing, he just feels the need to tell the world.)

I think you get the idea. I have to go make a sandwich now.

STEAMpunk Imaging

Holy Fast Fourier Transform, Batman! This is some seriously fast imaging: Serial time-encoded amplified imaging for real-time observation of fast dynamic phenomena. They claim imaging with 440 ps shutter time. Yowza! This is a pretty complicated engineering feat, but they dumbed it down into a really cute movie:



Basically, the way to understand this is that the two main time limiting steps with CCD imaging are CCD readout time and the inherent dark background, which requires you to have a minimum number of photons per image to not get swamped. The latter has been addressed in recent years by the advent of on-chip amplification, but it's still not perfect.

STEAM imaging (as they call it) is quite different. It takes a short laser pulse, which has an inherently broad spectrum due to the fourier transform of the pulse, and turns that spectrum into a 2D "spectral shower", where different colors map onto different parts of the image. The reflected light is then re-collected, and turned back into a pulse such that different parts of the spectrum arrive at different times. Then, each pulse carries the 2D image information in the time varying intensity of the pulse. This is amplified optically using a doped fiber (very high bandwidth, no electronics required), and it can be read out by a single pixel photodetector. The pulse intensity variation is mapped back onto a 2D image for human consumption.

Obviously, the system as designed can't be used for fluorescence. But for bright field imaging of metal particles (which they demonstrate in a microfluidic system) it's excellent. They image quite large particles (10 - 20 microns), and I'm not sure how well this would work with smaller particles. But, they suggest that this is intended specifically for biological applications several times in the paper, so I presume there's no intrinsic limit in that sense. I'm waiting to see if this will be available commercially any time soon.