# Are the machines upon us?

I suspect that this MSE question (Screenshot) and an MO question by the same account (Screenshot) have been artificially generated. But I'm not sure.

For a machine, those would be nice works of association, but that is not unheard of, as there are known paper generators such as SCIgen. The occasional typos work against a clear diagnosis.

Now I wouldn't mind if an AI posed a clear and interesting question. But those are not clear (to me), and if those questions are indeed machine-generated, this means that someone only needs to turn a knob to increase MSE's junk inflow far beyond manageability.

We may need increased machine assistance. Are there tools in place or in development that could automatically send posts like those to a review queue?

Update: Another post (#2810926, now deleted) (Screenshot) that might have been machine-generated has been identified in the comments. This time it is an answer.

Until a bot attack or a publication in the spirit of SCIgen occurs, we might remain in the dark about whether those posts have been machine-generated or not.

Currently, what matters are the characteristics that help those posts pass existing automated detectors and superficial human reviews.

Characteristics that should be machine-testable:

• No block-displayed math, very few inline math snippets
• Frequent references to, or quotes with superficially relevant keywords from Wikipedia / ArXiv / other sites known to host math content
• Generous use of inline markup (bold or italic portions)
• Many blocks (paragraphs/quotations/list items), but all short

This all serves to distract from the post's thin content while creating an appearance of thorough composition. I'd like to see some scoring mechanism work against such pretension.

Further characteristics, perhaps not machine-testable, but annoying:

• In questions: Typos and poor grammar except when reciting
• In answers: Telegram style only
• No complete thought in the non-recited text portions
• Following the references yields no relevance.

I find this use of fake references particularly frustrating because it multiplies the waste of every dedicated reviewer's time and effort.

Can we get the existing detectors sharpened, or new detectors introduced, to take into account at least the machine-testable features listed above?

Update: I suppose that there is some scoring tool in place already, perhaps a linear classifier. It would apply a number of metrics, dot-multiply the resulting vector with a vector of weights, and obtain a score. Interestingly, adding more metrics often helps to better separate good examples from bad ones. Therefore, this is mostly a matter of adding new metrics and refining existing ones. I'd propose to:

• Measure density of block-displayed math
• Measure markup density outside math, blockquotes, and code sections

and append those to the other existing metrics. Then recompute the optimal separating hyperplane (i.e. weights and threshold) from training data. Make sure to include gibberish posts and cross-validate. I'd be interested in the results.

• You want a machine to fight machine generated questions? Maybe in the end they'll become friends and collude against us!
– quid Mod
Jun 5 '18 at 22:59
• @quid Whatever machine algorithm was used to generate the linked posts seems to do a better job at generating meaningless only if read questions than does the algorithm SE uses for generating fake suggested edit review audits! Jun 5 '18 at 23:01
• More seriously, I am not really convinced it's machine generated to begin with. There are some automatic checks in place to detect problematic content. In part as external extensions, like "smoke detector."
– quid Mod
Jun 5 '18 at 23:01
• @amWhy I am not sure it's machine generated.
– quid Mod
Jun 5 '18 at 23:02
• @ccorn My impression is that certain portions have a little more conceptual coherence than I would expect a computer to generate - I feel it may just be somebody who got the trollish urge to post gibberish questions. Jun 6 '18 at 3:20
• @MonstrousMoonshiner we have another James Joyce type: math.stackexchange.com/questions/2809643/… Jun 6 '18 at 3:33
• @WillJagy: But there is a recognizable question behind. Jun 6 '18 at 3:35
• CC, well, yes, did Bourbaki invent fractions. It is the comment below that raises it to literature; two comments, in fact. Priceless. Jun 6 '18 at 3:37
• Well, let's hope I'm wrong. Or better, let's fear I'm not, and get better tools. Jun 6 '18 at 3:39
– Asaf Karagila Mod
Jun 6 '18 at 7:10
• – Surb
Jun 6 '18 at 11:50
• No need to invoke a machine to explain this. A few drinks and/or not a native speaker of English seems to me sufficient explanation. Jun 6 '18 at 12:49
• Why does no one in this entire thread give any reasons why they think the linked posts are machine-generated? Are you guys machines yourselves? Jun 9 '18 at 20:50
• @Vincent: Because nothing could count as proof anyway: Humans can act like machines as long as it is not too much work. Let us focus on the interesting possibilities, that is, those with potential for significant damage. Those are machine-generated low-quality posts. Accordingly, speculating that I were a machine is useless because I can create quality posts. Jun 11 '18 at 1:20
• @numbermaniac: Thanks, I have updated the question accordingly. Jun 14 '18 at 10:14

Not to put too fine a point on it, but your question seems to be mostly about over-zealously imagining complicated technical solutions to a problem that doesn't exist and has simpler solutions even if it did.

First, all of those seem quite plausibly written by a human when you combine a non-native English speaker with someone who doesn't understand what they're talking about. But let's say they aren't. So? They were hardly massive disruptions to the site. Community moderation readily handled them.

But maybe, as you suggest, someone will flood the site with such questions. Why would they want to do that? Sure, some people do have some inscrutable desire to vandalize or produce meaningless, non-commercial spam, but they don't tend to spend months building and refining natural language generators. If they did, the user or IP would simply be banned. This will happen automatically if the user/IP produces a large number of low-quality posts. If they went even further and used a botnet to get around that, CAPTCHAs would simply be enabled. At any rate, if I wanted to flood the site with hard-to-detect-but-ultimately-low-value questions, I'd just randomly copy-paste existing questions, maybe with some minor changes or perhaps I'd mix questions together. This would not require any natural language processing sophistication.

Summary: There is no evidence that this is happening currently or that anyone would want to do this, nor is there any evidence that the existing moderation would be unable to deal with this if it did happen.

• I think that this is an acceptable answer. But let me note that skepticism of the form Why would they want to do that? can always be answered with "because they can." Like for SCIgen. Jun 13 '18 at 5:39
• I completely understand why you might want to make a natural language generator and see if it can fool people. I don't understand why you would you would then want to attempt to disrupt StackExchange with it. The creators of SCIgen didn't follow its creation with flooding Elsevier journals with generated papers to attempt to destroy their viability. Jun 13 '18 at 5:50
• True for SCIgen. (I added that as an afterthought, but that admittedly does not match the disruption scenario.). But then again: I don't understand why you would... is reassuring, but not a convincing argument by itself. Fortunately, it is not needed; the other parts of your post demonstrate quite well that even if there were a motive for disruption, there would be more likely channels for that, and high-bandwidth posting would have its own characteristics. Jun 13 '18 at 6:06
• I don't agree with the sentiment that you and others appear to imply here - "Why speculate about what if ...". Most of the progress made in mathematics and other sciences is the result of what was initially idle speculation - the usefulness only became apparent later, in many cases. Curiosity doesn't need any other reason. Jun 15 '18 at 10:22

the MSE question seems more free verse; it would improve with the Latex verse environment. The two lines that begin with lower case letters are supposed to be indented a good deal, two to two-and-a-half words.

$$\begin{array}{l}\text{Like a skein of loose silk blown against a wall}\cr \text{She walks by the railing of a path in Kensington Gardens,}\cr \text{And she is dying piece-meal}\cr \text{ of a sort of emotional anemia.}\cr \text{}\cr \text{And round about there is a rabble}\cr \text{Of the filthy, sturdy, unkillable infants of the very poor.}\cr \text{They shall inherit the earth.}\cr \text{In her is the end of breeding.}\cr \text{Her boredom is exquisite and excessive.}\cr \text{}\cr \text{She would like some one to speak to her,}\cr \text{And is almost afraid that I}\cr \text{ will commit that indiscretion.}\end{array}$$

The question still there on MO is much closer to the end of Ulysses by James Joyce, https://en.wikipedia.org/wiki/Molly_Bloom#Soliloquy

• I am actually a bit concerned. If the MSE+MO posts in question are artificial, then SE moderation as of now cannot be effective against a flood of those. Jun 6 '18 at 2:46
• @ccorn I understand. We are mostly supposed to avoid speculating about the users, but this person strikes me as a combination of an urge to impress with very little actual awareness of people around him. You may have noticed that we have users who post several items per day that nobody ever reads. That strikes me as sad; your guy is a little further off the deep end. By the way, we have had machine-like submissions from time to time, people who put in a ton of effort to glue together article excerpts. Jun 6 '18 at 2:55
• Um, what are the literature excerpts in there about? The questions don't seem similar to them... Jun 8 '18 at 11:20
• @ΈρικΚωνσταντόπουλος I think a machine might have crafted this answer. It fits precisely the structure of the original posts in question.
– Alex
Jun 12 '18 at 1:02
• That wall of text looks nothing like either of the posts in question. What the heck is this answer? Jun 14 '18 at 8:48
• I always appreciate the opportunity to read some Joyce.
– user98602
Jun 16 '18 at 10:43