263
‘Yes, I am a human’: bot detection is no longer working – and just wait until AI agents come along
(theconversation.com)
This is a most excellent place for technology news and articles.
Here's a crazy idea: make the CAPTCHAs so complicated humans can't complete them.
That way if someone does, you know they're a bot.
I should probably patent that or something. (Is joke, etc.)
This very circumstance is described in the book „Qualityland“ and probably many more literature. The book is about a man living in the near (and completely enshitified) future who is confronted by a defunct algorithm that automatically bought him a dolphin dildo. Seeing no way of returning it by normal means - the algorithm is always right and does not accept returns after all - he sets out to right this wrong and discovers the mechanisms and people behind the screens that dictate every aspect of your life.
Sounds straight out of Paranoia RPG. Friend Computer knows what is best for you. To imply that Friend Computer does not have your best intents in its algorithms is treason.
I think I have my next ready thanks!
https://en.wikipedia.org/wiki/Marc-Uwe_Kling ?
Choose at random a captcha that a human could complete or one a human couldn't complete (while the bot supposedly could complete all of them).
I give up on any captcha with mor than 1 image. The problem is you aren't letting people in, even if you are catching bots. You'll have to turn it off to let people in.
They usually only do it when you’re on a vpn. So you just have to give up your privacy and disable the vpn to be seen as human.
Generally true, but I had cases where I got multiple captcha without vpn, and not even a single one with VPN enabled...