Bots are better than humans at cracking ‘Are you a robot?’ Captcha tests, study finds::New study suggests effort users put in cracking these puzzles every day may be more trouble than its worth

  • @cmrn
    link
    English
    521 year ago

    Those things have been kicking my ass whenever I have a VPN on lately.

    I’m out here clicking stairs, motorbikes, crosswalks, traffic lights, and it just keeps giving me “Try Again” …I’m starting to doubt I’m actually human.

    • @confusedwiseman
      link
      English
      251 year ago

      That’s the point. You can’t be identified. You will be stopped or harassed until you comply and be willing to be identified.

      Welcome to the brave new internet.

    • @Zeron
      link
      English
      171 year ago

      I’m convinced many of them fail you on purpose so they do multiples of them. Increasing the amount of training per user.

      My favorites are the ones i used to get on dread, jesus christ i’ve never sworn so much trying to do something so simple, it was ridiculous.

    • @[email protected]
      link
      fedilink
      English
      101 year ago

      I had the same problem. The image ones are ridiculous, so I started using the audio option. I’ve rarely had to try again on those and it’s usually only typing three words.

    • nostradiel
      link
      English
      81 year ago

      It’s actually machine-learning mechanism from google. They basically use all people on the internet to educate their systems about recognising stuff in the picture.

    • @Dempf
      link
      English
      31 year ago

      I’ve been having the same problem even on my home IP. It only happens when I’m using Firefox though. If I use Chrome, and oddly enough change my IP by routing through my rented VPS and pass a few captchas then Google will eventually start to trust Firefox from my home IP. It’s bizarre and Google is really starting to piss me off because I’m wasting so much time on this shit.

  • @[email protected]
    link
    fedilink
    English
    201 year ago

    I thought we were just basically training google’s image recognition AI by selecting the squares it asks etc. Nothing to do with security (in practise).

    • @residentmarchant
      link
      English
      201 year ago

      You’re probably correct now, but back when Captcha’s were introduced, machine vision was nowhere near as advanced as it is now

      It’s funny, but solving Captcha’s probably led to the downfall of Captcha’s

      • Jamie
        link
        fedilink
        English
        81 year ago

        We’re quickly approaching a world in which the smartest machines are more intelligent than the dumbest humans, which causes a problem for capcha creators.