During one of our email exchanges regarding OoL, James, Julia and I were discussing the sudden increase in spam we’ve been getting. James wondered why Spamblock did not appear to be cutting the mustard any more. My own gut feeling was that these spammers were not computers but people. The comments appeared as if they were written to the post, rather than the jumbled garbage we are used to getting – or the simple “wonderful post” stuff.
Take a look at one that landed on the post about my MP signing me up to his newsletter without permission:
He gave you response – just via newsletter 😉
Or this one:
click report spam button on gmail, gmail will let u to unscribe automaticly
Okay, so they do not address the subject matter in a sensible manner – no, the MP did not give me a response (he harvested my email without responding to my request for help) and the reference to Gmail is irrelevant to the context. However, they are both clearly pertinent to the content of the post; i.e. being subscribed to a newsletter. In order to do that, someone must have read it, surely. Or have they developed algorithms to read the subject matter and write a comment?
Nah…
I am a moderator of a forum, and in recent months we’ve received some posts from people who, from the links in their profiles, are obvious spammers, but which seem almost relevant to the topic in question.
Weird.
I suspect that paying real people in the third word is pretty cheap.
Yeah, it’s usually someone in the Philippines or India.
Hence the flaky syntax and comprehension.
You’ll drive yourself mad worrying about this stuff. Not allowing comments is the best policy.
Actually, my defence is very simple. Those not picked up by Akismet will automatically go into moderation as any new commenter has to be approved before they can comment without moderation.
These twats never get past first base.
I think the recent change to Blogger’s Turing word generator might be in response to the realisation that real people are behind the recent spam activity, because it seems designed to frustrate real people rather than automated functions! 😆
Yeh, it usually takes me three or four attempts at trying to enter the garbled words before I get one which is half way readable.
As usual XKCD got there first.
Oh dear