I’m at the 2015 AAAS meeting in San Jose, California. This is definitely not my typical meeting: too big, too broad, and I hardly know anyone here. But here’s a quick (ha ha; yah sure) summary of the meeting so far.
Gerald Fink gave the President’s Address last night. He’s the AAAS President, so I guess that’s appropriate. But after five minutes of really lame simplistic crap (for example, he said something like, “A single picture can destroy our known understanding of the universe,” like innovation and improving our understanding is a bad thing), I left.
Oh, and before that: the emcee of the evening, who introduced Janet Napolitano, totally couldn’t pronounce her last name. (Her remarks, particularly her comments in support of public universities, were quite powerful.) Old dude: practice such things! Your ineptness reveals that you haven’t paid proper attention to her.
A huge meeting, but I know next to no one here. But I ran into Sanjay Shete in the exhibit hall, where I attempted to get two of every tchotchke. (My kids will pitch a fit if one gets something, no matter how lame the thing, and the other doesn’t.) Sanjay was named AAAS Fellow, that’s why he’s here.
I went to a dozen talks. A half-dozen I really liked.
Alan Aspuru-Guzik talked about how to find (and visualize) useful organic molecules among the 1060 (or 10180?) possible. Cool high-throughput computing and interactive graphics to produce better solar panels (particularly for developing countries) and huge batteries to store wind- and solar-based power.
Russ Altman talked about how to search databases, web-search histories, and social media, to identify pairs of drugs that, together, give bad (or good) side effects that wouldn’t be predicted from their on-their-own side effects.
David Altshuler had a hilarious outline slide for his talk, but the rest was really awesome. A key point: to develop precision medicine will require hard work and there’s no magic bullet. And basic (not just translational) research is critical: we can’t make a medicine that gets to the precise cause (and that’s what precision medicine is about) if we don’t understand that basic biology.
I gave a talk myself, in a session on visualization of biomedical data, but it was definitely not the best talk in the session, nor the second best. Mine might have been the worst of the five talks in the session. But that’s okay; I think I did fine. It’s just that Sean Hanlon (brother of my UW-Madison colleague, Bret Hanlon) put together a superb, but thinly-attended, session.
Miriah Meyer’s was my favorite talk of the day. She develops visualization tools to help scientists make sense of their data. And her approach is much like mine: specific solutions to specific data and questions. She talked about MulteeSum, PathLine, and MizBee. Favorite quote: “It’s amazing how much people like circles these days.”
Frederick Streitz from Lawrence Livermore National Lab talked about simulating and visualizing the electrophysiology of the human heart at super-high resolution using a frigging huge cluster, with 1.5 million cores. I loved his analogies: if you are painting your house, having a friend or two over to help will reduce the time by the expected factor, but having 1000 friends or 100k friends to help? In parallel computing, you need to rethink what you’ll use the computers for.
His second analogy: The DOE cluster at Livermore is 100k times a desktop computer. That’s like the difference between PacMan (1980, 2.1 megaFLOPS) to Assassin’s Creed (2011, 260 GigaFLOPS). And their cluster is 100k times that.
At the end of the day, Daphne Koller talked about Coursera. She’s awesome; Coursera’s awesome; I’m a crappy teacher. That’s my thinking at the moment, anyway. (A video of her talk is online. Have I mentioned how much I hate it when people screw up the aspect ratio? It seems like they screwed up the aspect ratio.) University faculty exist to help people, and with Coursera and other MOOCs, we can help a lot of people. Key lessons: the value of peer grading (for learning), not being constrained by the classroom or the 60-min format, ability to explore possible teaching innovations, and just having a hugely broad reach.
College is a place where a professor's lecture notes go straight to the students' lecture notes, without passing through the brains of either.
Boy, am I old
I seem to be staying at the same hotel as the American Junior Academy of Sciences (AJAS). Are these high school or college students? Man, do I feel old.
My contribution to education, today: if all of the elevators going down are too packed to accept passengers, press the up button and ride it up and then down. (Later I learned, from one of the AJAS youth, that the “alarm will sound” sign at the bottom of the stairs is a lie. You can take the stairs.)