Google Image Labeler

Very interesting, I must say!  When I first launched the game, my initial thoughts were:  "I wonder what you get for the points?"  Google answered that very quickly on the starter page, there are no prizes or rewards for points.  I can admit, that took the excitement for me out of playing the game. 

So I watched the video of Dr. Luis von Ahn, and I was quite impressed with his talent, and advanced thinking on how we can "make our computers smarter".  In the beginning when he said he encourages people to work for free, I was wondering what he meant by that.  Now the puzzle is put together, and I think it's brilliant. 

So I began to play the game for a second time after the video to reach my goal of 900 points.  I thought, that if this is anything like the first go around, I may be sitting here for a while.  By the way, my first two partners on the initial game hung up on me, I guess I was submitting to slow.  I noticed that on their counts, they were entering probably 5 or 6 words for my every 1, and they knew in order to gain more points, they needed a faster partner.  So, this second time, I didn't even look at the "off limits" words, I just typed whatever came to mind, and found that in doing it this way, I was much faster and my partner actually stayed online with me.  By the time I had had enough, I had 1040 points, plus whatever points I had gotten the day before.  So, competition is definitely a major motivation for the players of this game, I noticed the high scores were in the millions. 

How fascinating is this game??? Also the other two games Dr von Ahn talked about.  Getting people to work for free (although they don't look at it like it's work), and gathering valuable data about images on the web to be a more accurate web.  So, is Google Image Labeler (game) Dr von Ahn's program with their name on it?  I'm sure it is, every concept is the same, but not for sure.  Dr von Ahn answered someones question and said that the intent and purpose of the game was disclosed before the player begins, but I didn't see this with Google Image Labeler.  Did anyone else?  Maybe I just missed it. 

On this type of input, I think it's a great idea for people to do this work.  But as someone said in the forum, are the answers bias?  Absolutely!  There is a certain type that plays games online, that has the time to play games online, etc.  So, the input data will be bias, and you really don't have a global perspective on the image.  But, as said by Dr von Ahn, the images put in the bank are basic like dog or car, they aren't as specific to really change the meaning.  Not sure if Google adopted that portion of his game, but I had some challenging images, many I had to pass on.  In my field of sales training, this mights come in handy when we have objections.  Objection faces, objection words and phrases, and objection body language.  We could put an image of an objection, and have the people agree on what the objection would really mean.  Just a guess....I think this would be a great tool for education, specifically in foreign language.  The game would be fun practice in helping with vocabulary.

Well, these are my thoughts...until next time.