I'm one of the lucky ones who gets to respond to messages that you guys mark as potential spam.
Most of the time, it's fairly obvious that something is spam. Especially if they post a link. Other times, it's not so clear. But one good indicator is if the user joined, read a message or two, then posted right away.
And when I'm pretty sure it's spam, I chose the option to delete the message & user, and block the IP address. That's the best I can do.
Today, I saw a new reply to an old thread, and this is what it said:
Ohh great! i too needed this, the steps really worked for me!!!
Judging by the fact the user just joined, I would guess this is spam, but I can't be sure.
So here's my question: What's the point of these types of spam messages? What does anyone, human or bot, gain from this?
If I were going to create a spambot nowadays I'd incorporate AI somehow and train it to "outwit" anti-spam measures.
In the case of the KM forum, the AI will by now have realised that new users with names like "abc123xyz" who post obvious spam get detected pretty quickly and banned. Therefore maybe this AI is now going to try posting obvious spam after one or more of its new-user posts have been accepted by us as bona fide.
Now that I've woken up, I realise this may not just be a bad dream...
Possibly it's a method to build up the fake user's trust level in a forum so they can then more easily (freely? under less scrutiny?) make a full spam/scam post? It's so difficult to fathom ill-intent and the reasons that drive it.
On the negative side, I was thinking like @hemicyon described, trying to get a spam user accepted so it can later make a play for something.
On the more positive side I was thinking of AI training, just in terms of determining where is the line between a reply being flagged versus passing muster. Sure,
Ohh great! i too needed this, the steps really worked for me!!!
does not look legitimate. But what about
I needed this. Thank you.
This is the solution I needed.
Perfect! Been looking for how to do this!
I have been trying to figure this out and never could. Finally!!! So glad these steps worked.
Maybe there's value in AI systems learning the first two will not get flagged but the last two will.
Iâm responding to the question âwhat should we do with this one?â I have sometimes wondered whether forum leaders ever DM such an account to request more information. I have further thoughts along those lines, but it occurs to me that a publicly available part of the forum may not be the place to discuss tactics which could be scraped by the very spammers that cause the problem.
I would actually have expected a post like this to have appeared under âForum Adminâ but anyway, that has the same issue.
I would think it is similar to those "silent" telephone calls that phone scammers use to test if the number their bots are dialling will be answered. It saves the human operators in those scam companies wasting their time dialling numbers that won't be answered.
By seeing if any of their fake posts survive a day or two on a Forum it must test to see if it worth their trouble to post actual scam/spam posts. I would guess it is many different scammers sending these test messages, not just one.
I agree, it is sometimes hard to tell, but this one looks like Spam so I have deleted it.
I get the same with some email spam, where there is no link and no ability to do anything and it's really unclear why the spam email was sent - but sending email is so cheap, that is probably doesn't mean anything. Maybe it was just a mistake, the link was missing, or as some have suggested, it's an attempt to build up trust level. Probably a lot of forums are configured to disallow linking in initial posts, so maybe this is part of that. Who knows.
As for DMing users, I have very very rarely done that in cases where it was really unclear, but in most cases I either accept that the post might be legitimate and allow it and wait for further evidence, or alternatively just delete them and in the exceptionally unlikely event it is a real person they can figure out contacting me and asking what happened.
Maybe add something to the FAQ and/or the pinned thread "Getting Started with Keyboard Maestro and the Forum" notifying users to reach out by email if they have a post inadvertantly marked as spam and deleted.
or more specifically, measuring the mean time to take-down of spam posts or user accounts,
would both have some commercial value to operations selling a capacity to place links for a certain amount of time, or harvest a certain number of clicks.
The big red spam flag for me is "these steps worked". I find it hard to imagine anyone getting into any KBM macro enough to implement it who would then refer to the process as "these steps". Its broad generality betrays a fundamental lack of app-specific vocabulary. I've done lot of what I call "forensic editing", teasing apart confusing constructions from non-native-English-speaking engineers whose first language was Russian, German, Polish, French, Chinese, as well those peculiar convoluted dialects of English that are used in India, in order to determine what they were trying to say while they were having difficulty with the English, because I knew the software subject they were attempting to talk about. And I have trouble imagining even any of those speakers using that particular construction when talking about Keyboard Maestro macros.
Doesn't look like spam to me. The user may have gone a bit overboard with exclamation marks but maybe just joined to post this out of enthusiasm. I'd just flag the user for observation, especially in light that the message does not solicit an action, so worst case it would be a warm-up post to legitimize the account.
Anyway, I'd say 80% this is just a new enthusiastic user, potentially non-native speaker, so the tone is off.
