3

I'm writing two academic papers currently, and have discovered the point where things get absurd and I need something that is lighter and more enjoyable - the kind of feeling that spawned The Journal Of Irreproducible Results.

As a fan of the show Westworld, I'd enjoy producing a paper which uses it as a medium to examine how AI can be implemented to fit grand contexts like storylines, a large park with unpredictable interference, and the social interactions that humans consider normal. There are more than a dozen Ph.D thesis that use Star Trek extensively - but they are all in the humanities, and they tend to examine things like the show's allusions to ancient mythology and history (which were rife in the show), or cultural perceptions of the show. I do not know of any papers in the hard science that used Star Trek in any significant way to explain or illustrate concepts.

Westworld, however, is different. There are several scenes that display real issues in computer science, the problems that can be encountered, and potential solutions. The show also included a well-architected software framework that all the hosts work within, and an excellent division of responsibilities within the code. So, while Star Trek used fictional physics, the computer science in Westworld is realistic - the AI may be exaggerated, but the fundamental approaches and issues were real. This paper would be weekend work and I'd write it to distract myself from the toils of the more traditional papers. Spoilers follow.

So my question is: can a fictional world be the basis and foundation for a serious academic analysis? I would diversify the subject to include other examples, but I don't think there are any that are elucidated well enough.

A perfect example is in Season 1 episode 3, when a woodcutter wanders off because he was tampered with. The park was only alerted to the problem because the woodcutter's group hadn't moved for several days. Upon investigation, the humans found the woodcutter's camp and the rest of the crew bickering over who should cut some wood for the campfire so they could cook dinner. The lumberjack was the only host allowed to use the ax - which is registered as a deadly weapon and the other characters were not authorized to use it. The circular bickering among the group is very similar to deadlock - they are all waiting for a resource and requesting it from each other, but the entity that can provide the resource is unresponsive. The group had no handling for a condition where the woodcutter wandered off, even though it presumably had some form of handling for a situation where the woodcutter was killed (that problem would be solved by the role-based plotlines, which I would cover in another part of the paper, and these are an extremely consistent and well-illustrated part of the show.) I would argue that the bickering is an optimal way for AI to handle that situation. The only solution the AI could create on its own - to let someone else do the work - would violate the safety protocol which has god-level precedence. The solution for park management is to kill the woodcutter host (or at least terminate him in that storyline), so the role can be re-designated to another host, and that can be programmed into the group's storyline as an exception case.

The show has many aspects like this; I think it has enough to reverse-engineer the highest levels of the software framework and treat it as a case study. That architecture would be applicable to current or near-future video games that are built around a storyline such as Fallout, Witcher, and similar. Other sources have already noted things like markov chains on the control tablets in the show.

I would write the paper to the highest standard possible, especially regarding diagrams and with a focus on making it more understandable to a broader computer science audience than just AI or systems architecture researchers. The references section may even be larger than usable, as I would compare a number of existing frameworks to show how they could be applied in the show. And I would try to do it without killing the enjoyment of the show.

  • 1
    What is your question exactly? You can write any paper you like. If you are asking if it’s a good idea, I doubt anyone can make predictions about that until they’ve seen the final product. (As a general rule, I’ve seen a fair number of published papers in pure math and computer science make references to pop culture as a way to motivate and inspire the reader, and even included such a reference myself in one of my recent papers.) – Dan Romik Oct 31 '17 at 18:42
  • 1
    That's pretty much what I'm asking - is it acceptable to use a pop culture element as the basis of the paper? Basically I'd treat the Westworld park as though it were real, and write a case study of it. My concern is that it's not really "valid" to pretend that a fictional item is realistic, even if the fictional nature is irrelevant. Maybe I could compare it to a climate analysis of Game Of Thrones, or a study on the administrative effectiveness of Hogwarts. The analysis can be done, even if the subject is fictional. So does a fictional subject make the analysis invalid? – user3685427 Nov 1 '17 at 14:10
  • 1
    It is “acceptable” in the sense that there is no rule saying what you can write in an academic paper. If you want to write it, go ahead - no one will stop you. However, as for the separate questions (which I suspect are what you really want to know) of whether it’s a good idea, whether it is likely to be accepted for publication in a good journal or conference, whether it will impress your peers and help your career, etc - as I said, I don’t know the answers and suspect no one can meaningfully answer them without seeing the finished paper. Good luck anyway. – Dan Romik Nov 1 '17 at 14:16
  • 4
    All of what @DanRomik says is of course correct, but as a first-order approximation I would say that a STEM paper that "uses Westworld as a case study" would be better appreciated as a nice blog post than as a formal peer-reviewed paper. – xLeitix Nov 1 '17 at 14:21
  • It sounds like I wouldn't be laughed out of the room just from the title, so that's helpful! I'm happy to get some chuckles and grins, but I hope those would be more appreciative than derisive. The subject may be a tightrope in some sense, but trapeze acts are impressive and easier to watch than a chess game. – user3685427 Nov 1 '17 at 14:23
3

You can write whatever you want. However, if you want to call it computer science, it had better teach us something about computation. Yes, a fictional world can be "the basis and foundation for a serious academic analysis" but literary analysis is literary analysis, not computer science.

You talk about using Westworld as a case study. How can it be that? It isn't a "case" because it doesn't exist. It's a work of fiction. The systems within it don't exist and aren't fully worked out. Using a work of fiction as a source of illustrative examples can be a useful device, as your example with the deadlocked work party almost1 shows but there's a big difference between using a fictional computer system as an example and making a detailed study of that fictional system.

Ultimately, anything that happens in a fictional world happens because the author of the fiction wanted it to happen. For example, you could use the woodcutter wandering off as evidence of a bug in his AI. Perhaps the author never intended that bug to be there but, at some point, realised that the bug was a consequence of some other thing. At that point, they had two choices: they could either rewrite the fiction to remove the cause of the bug, or they could have the thing wander off and write about the consequences. They chose to keep the bug because it made the plot more interesting. Likewise, they chose how severe the consequences would be. That's not how real-world systems work, and a real-life example of the problem you wish to illustrate would be much more convincing.


1 The problem is, it's not deadlock at all: it's resource starvation. Deadlock is a situation in which two or more processes cannot make progress because each needs resources held by one of the others. For example, consider two mechanics who each need a hammer and a wrench to carry out their task. The first mechanic picks up the hammer, the second mechanic picks up the wrench. The first mechanic says, "Please give me the wrench." The second says, "Sorry, I'm using it. I'll give it to you as soon as I've finished my task. Please give me the hammer." The first mechanic says, "Funny, I'm using the hammer. But I'll give it to you as soon as I'm done with it." That is a deadlock: neither mechanic can finish his task because each one is holding a resource that the other needs, and refusing to give it up. However, in your example, all of the processes are waiting on a single resource (firewood) that is unavailable.

  • 1
    Thanks! The magnitude of the author's choice in the show hadn't quite hit me as strongly as you describe it; I think I got lured in and it flows so well that every action inside the park feels so natural (and they are supposed to be scripted, thus that feels 'real') - it's the people running the park who seem more arbitrary and unnatural. And you are entirely correct that the term "resource starvation" is the correct one. I avoided using the term because I didn't want to imply that the robots were literally starving, considering that they couldn't cook for three days. – user3685427 Nov 3 '17 at 19:18
2

Boiling down your question based on a comment you made:

is it acceptable to use a pop culture element as the basis of the paper?

The answer to this is yes. And not just in an obscure journal - if it's a useful worked example, tells us something new about the field, suggests new directions, etc. there's no reason why it couldn't be. Indeed, I've published several papers using pop culture and an event within a fictional setting as a basis to think critically about my field.

1

Yes. This could be published. It would be worthwhile to purse the Journal of Science Fiction or G|A|M|E Journal. Computer Science journals are less likely to take something of this nature.

You should write an abstract and "test the waters" with it.

  • While I believe accurate, had I taken this advice when I was considering doing this, I would have regretted it. – Fomite Nov 3 '17 at 7:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.