# Should crazy bot-generated answers be flagged as spam?

I just stumbled upon this answer:

Here is a link to the answer, though it has already been deleted: https://electronics.stackexchange.com/a/378755/91862

Now, I scratched my head a little before finally flagging as spam. Whenever I see stuff like this I assume that it is posted by a bot, with the goal of becoming "trusted" enough on the network to eventually be able to spam more efficiently.

However, as it stands it is obviously not spam. There's no product links, no products to sell, etc.

Should this be flagged as spam?

• Obligatory XKCD: xkcd.com/810
– JAD
Jun 8 '18 at 9:03
• I'll wait for a few weeks before nuking that bot account, to maintain the context of this meta thread. Jun 11 '18 at 1:58

## 5 Answers

Since this is Meta, I suppose I could just add my reasoning as an answer for people to upvote and downvote at their leisure.

### Yes, flag as spam

This is most likely written by a bot trying to build up reputation and credibility on the network. Even though this will never work well on the Stack Exchange network, they are bound to try.

• I always flag these as spam - I feel they aren't rude or abusive, but they are used to build reputation for later spamming or worse. Jun 8 '18 at 13:25
• I've seen this bot "playbook" before. The next step -- which hadn't happened here yet, but probably would have given time -- is to edit spam links into the post.
– user39382
Jun 9 '18 at 7:18
• @Rory: They are abusive in that they are abusing the site. Jun 9 '18 at 22:24

It looks like quite some people disagree with my point of view, so I'll expand further.

### It looks like spam, but it shouldn't be treated as spam

I assimilate these posts as intentional gibberish, which, according to the offical rules, should not be flagged as spam, but rather as "rude or abusive" (which works almost the same as the spam flag).

Excerpt of the official rules regarding "spam" flagging (emphasis mine):

A post should be marked as spam only if it advertises a product, service, or similar and is unsolicited or lacks disclosure.

It should not be marked as spam when

• The post contains no useful information, such as an answer that says “I don't care about your problem”. Flag as not an answer instead.
• It contains only gibberish, such as “fsdguejgkfdlk”. Use the rude or abusive flag for these cases.

And regarding the "rude or abusive" flag:

A post should be marked as rude or abusive (formerly known as offensive) if it contains hate speech, obscenities, abuse against people, or abuse of the community or system, i.e., a clear violation of the be-nice policy.

Abuse of the system or community is everything that is created with the intention to harm them. This includes posts that contain no useful content at all – i.e. gibberish posts along the lines of:

asdasdasdasdasdasdasdasdasdasdasdasd

Both these flags work almost the same anyway, so it doesn't matter much. They will both lead to the post destruction, and a -100 rep penalty for the author (and potentially his permanent ban in the long run).

There are, however, two differences:

1. If a question gets two answers that are flagged as spam and deleted, it will be automatically protected. The same doesn't happen with rude or abusive flags.

2. If a post has helpful spam flags, then it can be used as an audit in review queues. If the post has only rude or abusive flags, then it won't be used.

This is where flagging as "rude or abusive" makes sense rather than "spam":

1. The gibberish posts are typically sent under random questions, so there is no reason to automatically protect such questions if they have been unlucky enough to get two gibberish posts as answer. On the other hand, real spam is typically sent to targeted questions, whose subject is more likley to attract spam (I have regularily seen spam posts from PCB fab houses on question about PCB manufacturing, for example). Automatically protecting these questions makes sense.

2. The gibberish posts are too easy to identify, so it does not make sense to reuse them for audits.

• Functionally the spam and rude/abusive flags are exactly the same, they're just two different labels on the same thing. Six of either in any combination automatically removes the post, for example, and will hide the post content and penalise the poster by 100 rep. Jun 8 '18 at 13:28
• @doppelgreener I'm just applying the rules as stated by the SE community/team. I have no reason to not apply these rules just because these flags would be the same under the hood. But in any case, have a look at the "How does the spam flag differ from the rude or abusive flag?" chapter in the link I provided. There are a few differences between them.
– dim
Jun 8 '18 at 13:36
• OK, I see -- there's a couple of differences. You may benefit from questions that get hit by multiple nonsense posts like this getting protected though. What I mean to suggest though is that whether it's spam or rude/abusive doesn't make a huge difference and sort of splitting hairs -- what's important is that the content gets purged from the site, and either flag will do that job. It's sort of spam (it may be a spambot probe) but it's also gibberish nonsense getting posted on multiple other sites. Jun 8 '18 at 13:41
• To my knowledge (not having an official resource link handy right now), the main difference between spam and R/A flags is that spam posts may end up in review audits, whereas abusive posts will not. It doesn't really matter here probably, as having that post as audit is fine, but e.g. NSFW content and cat-on-keyboard should rather be flagged R/A if possible because it's not pleasant to look at or too easy as an audit, respectively. Jun 8 '18 at 13:50
• @Byte You're right (and the link to the official resource is the one I provided). The other difference is the fact two spam posts under a question will lead the question to be automatically protected. Which shouldn't apply to gibberish posts: the gibberish posts are most likely sent under random questions, whereas spam posts are typically sent under questions that attract... well... spam (e.g. PCB fab houses-related questions).
– dim
Jun 8 '18 at 13:59
• Regardless of what it is, a person who posts such gibberish is not welcome here and should be dealt with in an identical way as a spammer. Delete, user ban, account nuke etc. So use the spam flag. It is not "rude or abusive", it is copy-pasta of some generic wiki article. Jun 11 '18 at 6:58
• @Lundin again, I'm stating the official rules, which are certainly there for a reason. "Rude or abusive" also translates to delete, user ban, account nuke etc... in the same way a "spam" flag is.
– dim
Jun 11 '18 at 7:42
• @dim This says "Do not use this flag because:" (rude/abusive) "Somebody appears to have posted nonsense due to an innocent mistake such as a copy-and-paste error". But I don't think it matters much how the post is flagged, as long as it is removed from the site in a swift manner. Most people think "spam" when they see a post like this, so the natural flag to use is spam. Jun 11 '18 at 8:17
• @Lundin You really think this nonsense post is due to "an innocent mistake such as a copy and paste error"? No, it is intentional gibberish, which is explicitly asked to be flagged as "rude/abusive" a few line above above the reference you gave (which is the same I provided, by the way).
– dim
Jun 11 '18 at 8:24
• @dim Yeah I suppose the part you quote is more accurate for this case. It is a weird case though, it doesn't really fit either category. Jun 11 '18 at 8:27
• @Lundin And, by the way, this same meta post explicitly says "It should not be marked as spam when [...] it contains only gibberish, such as “fsdguejgkfdlk”. Use the rude or abusive flag for these cases."
– dim
Jun 11 '18 at 8:28
• There is a difference on the back end. Post flagged as spam are used to train the spam filters. This is spam. Gibberish per rude/offensive is humans posting gibberish to annoy, not a bot posting machine learned language to get around the spam filters. Flag it properly! Jun 12 '18 at 2:16
• @Passerby. I have evidence of the contrary. SpamRam, the SE bot for spam filtering, treats both flags the same. SmokeDetector, the third-party spam filter, uses manually-crafted regexps. And if at some point it had to automate things, it would probably take R&A flags in consideration, just like the spam flags (why wouldn't it?).
– dim
Jun 12 '18 at 7:25

It's precursor spam. It is obviously nothing to do with electronics, and it isn't the sort of thing a newbie electronics person would post. It falls into the category of:

This user was created to post spam or nonsense and has no other positive participation

The user in this case has posted "spam or nonsense" and has no other positive participation.

Just flag it as spam, rather than making a flag that the moderators have to actually look at. Posts flagged as spam are automatically deleted once enough people flag them.

Googles definition of spam is pretty straightforward:

irrelevant or unsolicited messages sent over the Internet, typically to a large number of users, for the purposes of advertising, phishing, spreading malware, etc.

Any irrelevant message sent over the internet qualifies as spam. If the bots start posting coherent answers though.. I think there was an xkcd for that

• Your own clarification misses the qualifier though: for the purposes of advertising, phishing, spreading malware, etc. The first statement only applies if the second does. I don't know the purpose of the message here, I can only guess - but it is not phishing or spreading malware, and it doesn't seem to advertise anything, so according to google it is then not spam.
– pipe
Jun 12 '18 at 14:42
• @pipe the ", etc." part is vague enough to cover any purpose Jun 12 '18 at 18:48
• @user81993 But then, if the "etc." part is vague enough to cover "any purpose", should it really be considered a definition?
– dim
Jun 12 '18 at 20:28
• @user81993 Then your post also qualifies as spam. Shall we flag? :) Jun 18 '18 at 7:45
• @DmitryGrigoryev if you find the post irrelevant or unsolicited then go ahead Jun 18 '18 at 22:07
• @dim the reasoning part is vague but the characterizing part is not and I think that's enough. We can theorize about the purposes of spam and conclude that its for some roundabout kind of personal gain but then again some of it could be for the purposes of transporting all the dolphins to the moon for all we know. Jun 18 '18 at 22:12
• @user81993 I don't feel like I'm theorizing anything. The thing is, this definition isn't really one, and in any case, I don't understand why you would use Google's definition of spam rather than the one from StackExchange, when it is about what rules to apply on the StackExchange network. Refer to the "What makes something spam and when should I flag it?" chapter in the reference I provided in my answer.
– dim
Jun 19 '18 at 10:40

Yes, it is spam. It is a wall of gibberish, not parsed into paragraphs at all. Spelling is spot-on, which a bot can do with ease. But bots are not yet smart enough to understand proper English or American grammar, as in paragraphs, and where to put them.

We mildly chide those who post a question as a wall of text, but they usually have spelling or mild grammar mistakes as well.

This bot used perfect spelling, even with Unicode symbols, but did not understand parsing by context at all. Also all words begin with a capital letter, as this is a fine detail that takes foreigners years to learn. So by default it used capitals in many places we would not.

• Why is that so strange? You write "Yes, it is spam.", someone disagrees, and downvotes. Three people agree, and upvote.
– pipe
Jun 12 '18 at 8:22
• I downvoted because your answer is wrong. As dim showed in his answer (which I upvoted), the official definition of the rude or abusive flag includes gibberish. Abusive doesn't just mean abusive language, but also abuse of the system. Your answer therefore goes directly against what SE wants us to do. Jun 12 '18 at 10:58
• @dim, Olin. Thanks for the input. I now understand meta a little bit better.
– user105652
Jun 12 '18 at 17:53
• Should crazy bot-generated answers be flagged as spam? I took the question for its literal meaning and based my decision on that. I did not split hairs over 'abusive' or 'rude' context. It was my opinion that (obviously) robot created text is spam.
– user105652
Jun 19 '18 at 4:36
• I hate to stir this back up, but it's splitting hairs at this point. It's like debating if the post should be killed with red flames or blue flames... it's still going to be killed with fire.
– W5VO
Jun 19 '18 at 19:49