82
Comments
  • 6
    Left one is the interpretation of how the Alpha Go from Google works, right is just one box that uses image recognition based people's answers and improved upon it.
    I vote left.
  • 0
    I guess it depends on the use case
  • 13
    @sam9669 @codingghost

    No, because there is more to the right than is seems.

    The no-capcha works by analizing the threat level: i.e How was the page loaded..? How did the mouse move to the check box..? If it fails those, then you have to do an image test. I.e "Select all tiles containg street signs." even if you pass the test, if you mouse moved too quickly, and too perfectly, it blocks you anyways and makes you repeat the test.
  • 1
    This is bound to be won by the left side given an infinite amount of time and repeats.
  • 3
    I'm pretty sure you could train a deep neural network to mimic users behavior well enough to fool recaptcha...
  • 8
    @theCalcaholic captcha can always be fooled.. for any new captcha that comes, a crack can be devised.. that is why we have come from "what is 2+3" to "select the tiles with street signs".. the point here is how long does it take for you to devise a perfect crack for that captcha.. the more time it takes, the more successful that captcha is..
Add Comment