| Blog: |
| In My Honest Opinon |
Topics: |
| science, data, research |
NASA and GM are working on robonauts..? Looks sorta like Honda’s bipedal robots, but these Robonauts seem to be focused on “waist-up” functions instead of walking.
The cost of school textbooks seems like a pretty heavy cost for school districts. But what if newspapers printed school lessons where the classifieds used to go? I think it’d be somewhat interesting if the state could subsidize failing newspapers by paying them to dedicate a section of their papers to textbook material. That would make these newspaper textbooks somewhat disposable and easy to obtain. Kids wouldn’t have to “share textbooks” or anything like that. But this idea would require “copyright free” textbook content, as well as more newspaper deliveries to schools. Newspapers would then get distributed by schools, instead of being thrown on your doorstep….
Machine learning will probably be one of the most important technological advances that affects the future of humanity and our ability to become more productive. It actually amazes me that I haven’t really heard of a machine learning project that attempts to augment human learning. (I assume such a project is fairly ambitious, but it seems like a worthy goal.) My understanding of machine learning is admittedly on the level of “Intro to Machine Learning” — so all of the potential sounds really exciting, but I’m sure the actual progress/development in machine learning algorithms is painfully slow and seemingly impossible.
That said, I think machine learning is a worthwhile field of study, and I think it’s good that there’s some gov’t stimulus money going towards teaching computers. There’s a $1 million NSF grant to create a machine learning system that improves the search results of limited databases. The program is called Osmot, and it looks like it’ll be open source. I’m not sure how they’ll measure improvements over other search methods, but any system that aims to improve the filtering digital data will probably have some value. There’s too much data for humans to sift through!
Here’s a Google Tech Talk cut down to about 10 minutes, discussing the benefits of using Thorium instead of Uranium or Plutonium in nuclear reactors. Will fusion beat out fission, though, for the future of energy? hmm..
Welcome to 2010, folks! Twenty-ten sounds so futuristic — but I guess not anymore… What year sounds more like the distant future now? 2100? 3001? Twenty-fifty doesn’t quite sound very distant to me. I think we’re supposed to run out of oil in 2050 or something… but I doubt it’ll happen. We’ll have biodiesel or nuclear fusion by then (hopefully).
I got my copy of this book… and I was avoiding the “unboxing” tweets and videos… but now that I’m halfway through reading it… I saw this and thought it was brilliant. I wonder what 7th graders think of Annabel Scheme….
This was a quick, not-quite-an-unboxing of Annabel Scheme that Anthony and the gang put together yesterday during Open Studio. For those not in the know, Annabel Scheme is Robin Sloan’s book. He’s the writer previously mentioned on this blog here, here, here, here, and here.
Note: Here is a site dedicated to unboxings. And here are some other unboxings of Annabel Scheme.
One of the differences between the US patent system and the patent processes in the rest of the world is that the US has “first to invent” rule — while the rest of the world goes by “first to file” — for granting patents. I say, “Why not make it BOTH?” Basically, instead of granting a monopoly to just 1 inventor, create a duopoly patent system — so that the public gets at least 2 competitors for every invention. Ok, so I’m pretty sure there would be legal chaos with this plan, but the system is already kinda messy… I wonder if there’s any way to model the possible effects of a duopoly patent system in SecondLife. hmm.
I’ve been sorta on the lookout for a cool DesktopAsAService to try out. Looks like Sun has one. What I’d like to be able to do.. is get a netbook setup that can log into a virtual OS X environment that’s hosted in a way that the lag isn’t noticeable. Maybe there’s a way to do it with Sun VirtualBox…
I randomly saw this article in the NYT about how Disney’s “Baby Einstein” DVDs made false claims that kids could be educated by videos. So Disney is offering a money back deal to parents who are unsatisfied with the DVDs… But what parent is going to admit their kid is dumb just to get back a few bucks? I think the refund rate (if it’s ever published) would make for some intriguing data on human psychology (if there were demographic data associated with the returns). Where are the Freakonomics guys when you need them?
The Callifornia “MegaMillions” lottery seemed to get some attention recently for its huge jackpot. Whenever this happens, I always hear people saying “try to pick numbers that no one else will pick, so that you don’t have to share the winnings (if you do, in fact, win).” So why doesn’t the Lottery sell unique sets of numbers for more? The odds of winning are no better, but then you also don’t have to share any of the jackpot if the numbers are picked. I’m pretty sure people would pay extra for that, but I’m not sure why the lottery administration doesn’t offer it. I suppose it might be difficult to implement a way to “guarantee” unique number sets? But it doesn’t sound that hard.
Another random lottery idea is to sell lottery data & analysis… I’m sure there are some crackpots out there who would love to pour through all the historical winning numbers to see if there are patterns. (And I’m sure the lottery administration must keep this data in order to verify that the system is fair and truly random.) Maybe the data is already free if you file with a Freedom of Information act request…?
