The spam bots are pretty good. Anyone with an app with a comment system can confirm. So, since they know to solve them, they do it pretty quick, as apposed to humans, who take on average 28.4 seconds ( http://www.stanford.edu/~jurafsky/burszstein_2010_captcha.pdf ).
So, can we say that bots are statistically better at captcha than users?
While not obvious, I do think there is a UX angle to this question;
If bots can read a Captcha better than a human user, why include them at all?
After all, it more often then not, is damaging to the user experience. 28.4 seconds is a long time for a user to be stuck at such a hurdle, and we've all been there - to say it's annoying would be an understatement.
So if a Captcha isn't able to do the job it set out to do (by that I mean preventing bots from submitting data), then it serves no purpose and thus should be removed.
Of course, I recognize that there is a need for Captcha's as not all bots are able to read them, and without a Captcha spam may just build up all too rapidly.
My preference would be to refrain from including a Captcha, and instead moderate submissions via filters etc. Perhaps with final approval coming from a webmaster.
At the company that I'm working I have suggested the same problem with Captcha. After days of trying to persuade the team not to use it we came up with an alternative.
Captcha isn't visible to the user. It only appears if the form submit button is pressed within 2 seconds of the page load. We didn't experience any problems yet.
