Four short links: 30 September 2009

Smart Materials, Google OCR API, Teaching Webinar, HistEx

  1. Smart Materials in ArchitectureUsing thermal bimetals can allow architects to experiment with shape-changing buildings, Ritter said. Thermal bimetals include a combination of materials with different expansion coefficients that can cause a change in. Under changing temperatures this can lead one side of a compound to bend more than the other side, potentially creating an entirely different shape, he said. A little impractical at the moment, but think of it as hackers experimenting with what’s possible, iterating to find the fit between materials possibility and customer need. (via Liminal Existence)
  2. Google OCR APIThe server will attempt to extract the text from the images; creating a new Google Doc for each image. Experimental at this stage, and early users report periodic crashes. Still, it’s a useful service. I wonder whether they’re seeing how people correct the scan text and using that to train the OCR algorithms. (via Waxy)
  3. My O’Reilly Podcast: Dan Meyer — I’m not pimping this because it’s O’Reilly (O’R do heaps of stuff I don’t mention) but because it’s the astonishingly brilliant Dan Meyer. For everything it does well, the US model of math education conditions students to anticipate narrowly defined problems with narrowly prescribed solutions. This puts them in no place to anticipate the ambiguous, broadly defined, problems they’ll need to solve after graduation, as citizens. This webcast will define two contributing factors to this intellectual impatience and then suggest a solution.
  4. Inflation Conversion Factors for Dollars 1774 to Estimated 2019 — in PDF and Excel format. I’ve wanted such a table in the past for answering those inevitable “… in today’s dollars?” historical business questions. (via Schuyler on Delicious)
tags: , , , , ,
  • http://threepress.org Liza Daly

    Inflation Conversion Factors for Dollars 1774 to Estimated 2019

    This is the one thing I routinely use Wolfram Alpha for:

    $1 in 1980.

  • bowerbird

    > I wonder whether they’re seeing
    > how people correct the scan text and
    > using that to train the OCR algorithms.

    the google factory uses every part of the animal.

    so the answer to your question is likely “yes”…

    but compared to the extensive and plentiful data
    from their controlled-conditions book-scanning,
    where multiple copies give a “double-key” mode,
    the info they pull from this arena will be sketchy.

    -bowerbird