How Smart?

Are you old enough to remember the commercial “Is it live or is it Memorex?”  If so, you can probably appreciate how we work to get technology to do its best.  If not, you probably believe that it’s always been this way.  Either way, we know that it’s only going to get better with time.

Your reading today in this vein should include this article from the New York Times.  “If an Algorithm Wrote This, How Would You Even Know?”  Of course, articles written for the media are created by humans.  Right?  Read the article and you may never feel the same way about it again.

Computers try.  They try hard, and they’re getting better.

Now, on Twitter, I always encourage Ontario Educators to follow other Ontario Educators.  There’s huge power in learning from others who have the same conditions.  But, for today, I’m going to ask you to consider following another Ontario entity.  It’s the user Interesting_JPG.  From the University of Toronto Deep Learning on the Computer Science website, it’s part of a demonstration of a project to turn images to text.  I spent a great deal of time poking around their website to get a sense of how this might work.  It’s absolutely interesting and I’m guessing that the point is to make image recognition smarter by learning.

Attempts are posted to Twitter with image and descriptor.


In the beginning, it will be tough going, I suspect.  The images I’ve seen so far are not easy to interpret, but what happens as the machine learns and gets better?

By the way, I keep enough spelling and grammatical errors in this blog just to confirm to y’all that this is written by a person…


OTR Links 03/10/2015

Posted from Diigo. The rest of my favorite links are here.