I once went into a few local restaurants to surreptitiously test the tuna they served. Some tuna was tuna, some was grouper, some was Nile perch, but all of it looked the same when cooked. In many cases this is not a case of restaurants misleading customers, or even being mislead themselves, but simply a problem with the length of the supply chain. The more intermediates that a piece of fish has to go through to get from the boat to your table, the more chances there are for it to be misidentified.
Southern Fried Scientist (commenting on a TED talk by Stephen Palumbi) on a challenge to supply-side conservation.
Like any new kid in class, RUBI took some time to find a niche. Children swarmed the robot when it first joined the classroom: instant popularity. But by the end of the day, a couple of boys had yanked off its arms.
The engineers went beyond stronger arms (or mounted weapons):
The RUBI team hit upon a solution one part mechanical and two parts psychological. The engineers programmed RUBI to cry when its arms were pulled. Its young playmates quickly backed off at the sound.
If the sobbing continued, the children usually shifted gears and came forward — to deliver a hug.
Students, Meet Your New Teacher, Mr. Robot
Last summer, the New York Times published an article about an IBM supercomputer being trained to solve Jeopardy clues. It’s not as ridiculous as it sounds. Jeopardy, for those of you who live under this rock with me, involves familiarity with astronomical amounts of trivia, something computers are great at, but it also trades on wordplay and allusions. A Jeopardy win would be a real score for computer understanding of human language. It’ll be a challenge:
Watson will not appear as a contestant on the regular show; instead, “Jeopardy!” will hold a special match pitting Watson against one or more famous winners from the past. If the contest includes Ken Jennings — the best player in “Jeopardy!” history, who won 74 games in a row in 2004 — Watson will lose if its performance doesn’t improve.
That episode is now scheduled to air—in February 2011. And indeed, Watson is facing Jennings (and Brad Rutter, another Jeopardy record-holder, in dollar winnings).
Update: Practice rounds against Rutter and Jennings
Update Feb 2011: Good thing you welcome your robot overlords, Mr Jennings!
I’ve been reviewing math lately. I have some specific, near-term goals but mostly a desire to demystify a subject that has caused me considerable anxiety in the past.
It’s slow going, though.
[Image source]
I’m tired of articles that oversell a perceived lack in a software-based product by assuming that the product is the be-all and end-all of what the maker envisioned. I am thinking in particular of iPad apps. I wish I had a dime for every person who has raged at the fall of Western Civilization (or destruction of journalism) because some iPad app they are using doesn’t have a bunch of linking and social features.
Building good interactive experiences—on the web, in apps, wherever—is hard. Everyone smart who is doing this, especially with a very young device like the iPad, is adopting a “build and then iterate” strategy. To do anything else would take too long, cost too much, and still get it wrong. Get it out there with the minimum feature set to be engaging, and then revise it to do more stuff, do more interesting stuff, do stuff better.
Wish you could email a friend an article, send a link to Twitter, or even, FSM forbid, “like” it on Facebook? Awesome, send the maker of the app a request, post to Twitter, write an article on your blog, shout it on the corner if that floats your boat—and here in San Francisco it might be surprisingly effective. Hey, hit all the channels you want. But do you honestly believe that anyone making an iPad app for subscription material is already completely done with the feature set? Really?
And when Murdoch’s iPad thingy finally comes out, and it omits all that stuff by design and has no plans to add it in, please don’t complain about that, either, because how could you not see that coming?
Basics of paleontology