By now, I've solved hundreds of captchas on Stack Exchange, maybe thousands. Often when I post (conservatively, at least 30% of the time), I'm presented with this:
We can't quite tell if you're a person or a script. Please don't take this personally. Bots and scripts can be remarkably lifelike these days!
I'm starting to take it a bit personally - a bot could not have posted thousands of well-received answers over the past few years, nor have I ever posted spam or with a bot - and it should be well within SE's power to recognize this. When I spend 30 minutes composing an answer, debugging, proof-reading, double-checking the logic and rendered Markdown, and then get hit with this when I try to post it, it feels like a minor slap in the face, especially when the captcha comes up frequently when I try to submit an answer.
(I'm sure I'm not hitting any of the listed rate limits - I never post more than one answer in under a minute, and I also always take more than 5 seconds to compose an answer, usually at least a minute)
The captcha usually requires me to analyze which of a bunch of squares (9 to 16) has a particular attribute. For some reason (maybe there's a bug in the captcha), it very frequently rejects my first attempt even when I'm sure I've clicked everything correctly, and I have to try again. So a thousand successfully solved captchas translates to maybe 20,0000 of these squares total that I've been forced to categorize before others have been allowed to view my solutions to their problems.
It's not uncommon for it to take more time to get through the captcha than for me to identify the problem in the question I'm trying to post an answer to.
Spam is certainly a problem here on Stack Exchange - it comes up multiple times every day, and we all do our part in flagging it. It's quite reasonable to require new users which trigger certain heuristics to solve a captcha or few the first time they post, or if they're posting too much in a short amount of time, or if they've posted spam before. But I think it is not reasonable to treat prolific, trusted contributors with the same suspicion when they post.
Would it be possible for Stack Exchange's algorithm for determining whether a user needs to solve a captcha be tweaked so that very likely honest contributors don't get hit by it? I don't know what heuristics are being triggered, but maybe add something like:
- If the user has made more than 25 posts on the network, and none of the user's posts have ever been removed as spam, and the user hasn't just made many posts in quick succession, do not require a captcha.
Or something along similar lines. I would really appreciate it, and I'm sure other trusted users who have to go through the captcha would too.