December 2014

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      
Blog powered by Typepad
Member since 08/2003

« Suck on This Again | Main | QOTD »

October 03, 2007


Feed You can follow this conversation by subscribing to the comment feed for this post.

Luis von Ahn is like the coolest dude ever, he has a Google Tech Talk you can watch on the internet that's really good. He does these awesome internet games that trick people into labeling images & stuff.

I vaguely recall that one of the search engines - Google, probably - was trying to get certain users to rank the results in terms of whether it was what they were looking for. Was that his idea?

that is really cool, but i don't get how it could work.

say, for example, facebook uses the scanned word "portion" from the graphic in this post as its CAPTCHA. some dude tries to comment on facebook and is flashed the image of the scanned word. he recognizes the word and types "portion" as a response.

how does the computer know that his response is correct? the key to CAPTCHAs is that it already knows that the "SKWCZ" graphic at the top of this post says SKWCZ. so when a user types it correctly, it knows the user is a real person but not a bot. it seems like the reCAPTCHA folks have it backwards. their computer doesn't know what the text days, so how can it verify that the user got it right?

err, i mean "what the text says" not "days"

Did you RTFA? It says it gives a known captcha along with an unknown recaptcha. If the user renders the captcha correctly, then it assumes provisionally that the recaptcha is correctly typed, too. Then it compares those provisional "correct" responses from different users to each other to see if they match.

err, i guess that does make sense.

and no, i didn't RTFA.

i guess reCAPTCHA is pretty clever after all.

The comments to this entry are closed.

Support This Blog

Philadelphia Bloggers