Hacker News new | past | comments | ask | show | jobs | submit login

> ie, humans disagree with other humans more often than their systems disagree with the average human rating[0].

I wonder why google has that new image categorization captcha when it's a task where computers fare better than humans.




How do you suppose they obtain the data to know how humans fare?


Sure, it's great for training/testing their AIs. But if the result is AIs better than humans at the task then it doesn't make a great captcha anymore.


Maybe they actually don't care about the captcha aspect - they just found a good mechanism to easily acquire data from humans that no one will complain about because it's so entrenched.


It still protects against spammers that cannot deploy such AI. Which is probably very close to all of them.


Hence the fancy "click if not a robot" captcha. :)




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: