Ive heard many times that google gathers data for their self driving cars through captchas.
But how does it train the AI if the AI already has the solution to the captcha im solving.
Its like a school test. The student can give right or wrong answers. No matter what answer the students gives, the teacher doesnt learn anything because he already has the answers
In the same way how does the „AI“ know my answer for the captcha is wrong
And how does my answer train anything
In: 8
for what I know they’re mainly used to process a lot of information about images using same object in different context, color, light and shape, then they give those images and informations to an AI that, with a lot of data, can learn to recognise already analysed object.
This is useful to self driving cars bc they need to recognise rapidly static and moving object on the road and then act based on what they recognise.
Plus I read that actually Captcha arn’t used to detect bot (there are other and more useful method) it should be just a way to make a lot of people work without needing to pay them (like wage-free microworks)
Latest Answers