writeyourdarlings

I originally posted this on reddit, but a user recommended I posted this here as well.

I see a lot of talk about bots on fanfiction sites, but hardly any about why they do what they do, which is why I decided to summarize and analyze the patterns between their comments.

Praise Bots

If I were to define praise bots, there would be two tell-tale signs, and three possible reasons behind their comments;

(The Signs)

a) The exaggeration of praise is a notable piece, because it includes strange metaphors that have little-to-no correlation to the work, such as “This work is so amazing that it might be the reason electricity flows through the world”, or ”Your characters are so lifelike and multidimensional that they could run governments, launch space missions, or negotiate peace deals.”

b) The usage of the third person perspective in the comment, rather than something that feels personalized, such as “readers would love this” or ”this will appeal to readers of (listed genre)”

(The Reasons)

c) The first reason that a bot would leave a praise comment is to gauge the audience, because the goal is to find out which users are vulnerable to those kinds of comments, and who they can snare in their plan.

d) The second reason that a bot would leave a praise comment is to surreptitiously solicit the author via flattery, in which they lure them to another site upon response. I’ll touch up on this in the commission bots summary.

e) The third reason, which is fairly unlikely, is that this is the work of a reader that wants to support authors, but doesn’t understand how to do so.

Hate Bots

If I were to define hate bots, there would be two tell-tale signs, and three possible reasons behind their comments;

(The Signs)

a) The bait and switch is a common tactic, in which it begins with a sweet, loving comment that quickly changes into a hateful speech about the work, despite including little-to-no references as evidence to their claims. It usually takes on the format of ”This was the greatest thing I’ve ever read! Is what you want me to say, but in reality it’s awful.” or another similar variant.

b) The ‘author should stop writing’ phrases, or other variants that convey the same meaning, with terms such as “slop or “garbage”. This method of recognizing a hate bot is (generally) unreliable, because it follows the same formula that a real person would use.

(The Reasons)

c) The first reason that a bot would leave a hate comment is to scrape an author’s work without repercussions, in which they pressure the writer into removing their works, hence losing the ability to prove that they wrote it.

d) The second reason is related to solicitation, in which the bot would prey on the insecurities of the writer in order to promote their brand, generally a form of artificial intelligence, which I will elaborate on in the business bots’ section.

e) The third reason is unlikely, but it could be related to users that are against the content of the archive, hence spamming authors with hatred in order to get their works removed. I say this is unlikely because (most) bots are not free, and require time and effort to maintain.

Misinformation Bots

If I were to define misinformation bots, there would be one tell-tale sign, and two possible reasons behind their comments;

(The Signs)

a) The ‘concerned commenter’ is the tactic that is commonly used in this scam, in which they come at the author with a polite, ”hey, I say this out of worry for you,” or ”the archive is removing works from scarce fandoms to make room—“ esque attitudes in order to scare the author. In particular, this type of comment is hit-or-miss, but is relatively easy to snipe if the author knows the content rules of the archive.

(The Reasons)

b) The first reason is similar to hate bots, in which the goal is to remove the author’s proven association to the work. The motivation behind this is unknown, but it could vary from scraping to blatant plagiarism.

c) The second reason is gauging user vulnerability, alongside the kind of comments they can utilize to fear monger. It’s a research study on an audience, which is likely related to a third-party audience, which I will cover on business bots.

Commission Bots

If I were to define commission bots, there would be two tell-tale signs, and three possible reasons behind their comments;

(The Signs)

a) The ”I would love to bring this—specific or non-descript—scene to life!” followed by the comment requesting a form of socials outside of the archive. A piece that separates these comments from genuine artists wanting to draw is the mention of ‘discussing’ the terms of the art on another site.

b) The flexibility in pricing is another giveaway. If you consider that the artist had some form of business or portfolio, they wouldn’t have the time to spam advertisements in the comment section of an author that doesn’t even earn profit from their works, nor would they be able to cut down their pricing to something that fit the author’s budget. It’s impractical.

(The Reasons)

c) The first reason that the commission bots would work is to gain profit. It’s a simple reason, but there isn’t much to it; the ‘artist’ will spend a couple of minutes producing an AI image, or they’ll repurpose a pre-existing piece of art. It takes effort on the author’s end, but not the scammer.

d) The second reason that a commission bot would target a writer is to collect their data, whether it be to sell to another site, or to use their data for another purpose. It would likely be enacted through clicking a link to wherever they want to ‘discuss terms’.

e) The third reason is to gauge user interests, and find out what kind of advertisements appeal to their audience: If they find that enough users are susceptible to their market, they have reason to believe that they can amass profit.

Business Bots

If I were to define business bots, there would be two tell-tale signs, and one possible reason behind their comments;

(The Signs)

a) The first format is related to what I previously mentioned in the hate bots summary; the comment preys on the insecurities of a writer, saying things such as ”This is the worst piece I’ve ever read.”, which encourages writers to pursue whatever options they suggest to improve.

b) The second format is to accuse the writer of using AI to write their works, which sets off another chain of events. It tends to look like this; ”you used this very specific AI program to write this.” or ”could you prove this isn’t (specific program) AI?”

(The Reasons)

c) The reason for talking about artificial intelligence in comments is to advertise their business. If an author is accused of having written with a certain form of AI program,—”you generated this with (AI name), didn’t you?—the author may feel inclined to check out the program in order to compare their work, which is the goal of the bots’ purposes.

The Patterns

In acknowledging the varying comment types, the commonality that they all share is that they are all looking to make a profit by preying on vulnerable users, whether it be through a third-party commission or direct contact with the victim.

The Result

As this analysis comes to an end, I have concluded that the common goal of bots on the archive is to profit, whether it be through the theft of fic ownership, gauging public interaction through a third-party transaction, or through the means of AI-produced imagery.