The question Hacking the universe has shown up on several different "be aware" lists (particularly, answer flagging and number of answers in a short period of time). The question basically boils down to:

Imagine the whole universe is a simulation. ... What would be a reasonable attack vector for a simulated human to hack the very simulation that human lives in?

The question has attracted 12 answers in just over a day, over half (seven) of which are zero net voted as I am typing this. Several of the answers, while not necessarily bad, are also what I feel is below par for the site.

The main problem is that the question as it stands appears to me as very broad. It has been fairly highly voted, but a part of that appears to me to be that it offers the possibility of an in-universe explanation for magic. However, it also seems to be asking for a largely unbounded set of answers ("what would be a reasonable attack vector?" without establishing much in terms of constraints).

Again, the number of different answers attracted by that question, several of which by users new to the site as well, while not necessarily a problem in itself, shows how broad the question is and how differently it is being interpreted by different people.

I have put it on hold for the moment so that we can have a bit of discussion about it here and decide what to do with it.

  • Is the question too broad?
  • If it is too broad, then what can be done to it to make it less broad, such that it is a good fit for the Stack Exchange format and the Worldbuilding Stack Exchange?
  • If it isn't too broad, then what does that mean for our standards on what constitutes "too broad"?

In answering the above questions, you may also want to consider the meta questions listed below.

  • 1
    $\begingroup$ I've now added some clarifications which I hope resolve that problem. However, specifically Liath's answer, while not what I was originally after, is still so useful that I'd hate to lose it; unfortunately the clarification means it now no longer fits the question. I think Sheraff's also very useful answer still fits the refined question, although it doesn't completely answer it. $\endgroup$
    – celtschk
    Oct 27, 2014 at 22:00
  • $\begingroup$ At the moment, 8/13 answers do not have a positive score. Three are at -1, and have been deleted; an additional answer is at -1 but has not been deleted. Three out of these four are from new users, as is another one above. $\endgroup$
    – HDE 226868 Mod
    Nov 6, 2014 at 19:39
  • $\begingroup$ I would argue that the question is bounded by domain overlap. You could rephrase it "how does the domain of hacking overlap with the domain of physics assuming a simulated physics model. Whether or not that's an appropriately tight bound is up for discussion. I would argue that it is. $\endgroup$ Nov 7, 2014 at 11:33

5 Answers 5


I've been alternately trying to define to myself why this question is too broad and trying to answer the question (it is one of those questions that makes you go "I have an idea! Wait, ANOTHER idea!" ) and I think I've finally resolved both points to a single answer.

The answer to this question is effectively: Anything you want it to be.

That means there is no clear way to pick an answers, which in turn means that the question is too broadly scoped.

The top voted answer doesn't even answer the question specifically, it describes the process. Asking for what methodology you would use to make an attack seems reasonable, asking for a list that includes almost every set and subset of actions a person could take does not.


I read the question as "What properties should my world have so that this plot idea I have works?"

Reading back, this isn't quite what the person asked. It seems to be more of the format "Given an unspecified world, how can I develop this plot idea?"

The answers are again different, with the following interpretations showing up:

  • How can one discover bugs in a software system?
  • What aspects of reality appear to be bugs?
  • What kind of bugs can a simulated world have?
  • What might stop the characters from escaping a simulation?

Given this, it seems like the question is too open for interpretations, and should probably be respecified to be more focused. Many of the above interpretations could work, depending on what the OP actually wants to know. Different interpretations can be posted as different sub-questions.

If the original question is to be kept, I think the world should be specified. Many of the answers are of the form "Suppose your world is like this, then this can be done." If, instead, the properties of the world were fixed, these answers would have to be more specific (and probably more useful for the OP).


How broad is too broad?

Let's rephrase the question "How might I hack into a Linux web server". Well now we have a real world question we can answer. We could write an entire book on very specific vectors (heartbleed, shellshock, etc), or we could answer the question in broad strokes (find or inject a vulnerable component, hit that).

The broad stroke approach would arguably provide a sufficient answer for this question, but if someone wants to invest some time writing a canonical answer, well hey, why not. That would actually probably be rather useful.

So is a slightly handwavy answer sufficient? Is it OK to talk about general principles of hackery and how they might relate to the situation at hand, educating and enlightening the person, or do we want a single right solution (e.g. hit bash with a specially crafted request).

To use the hackneyed metaphor, do we want to give the questioner a fish, or do we want to teach them to fish? This is a question we will have to tackle as a community.

Well clearly questions have to be limited in some way, otherwise we have anarchy, but there are lots of ways to limit a question. We can limit to two domains and how they overlap (venn diagram style), or to a specific type of circumstance. The question needs to be clear, but it can still be a little open, because, by definition, the user doesn't know the answer before they ask it and may not be able to narrow it down without assistance.

I would say there's a place for all these approaches on this site. For exact answers, for broad principle answers limited to a very specific domain, and even for long form answers if the user has the time, provided of course they are limited to a very specific domain.

How broad is too broad?

I would argue that a question asking for a limited set of principles from a well specified domain and how they might in principle apply to a situation is acceptable. Perhaps that's just me.

I also think denying questions like this one sucks some of the joy out of the site. Perhaps that's just me too.

  • $\begingroup$ The question has been edited since this question was raised to the satisfaction of those concerned about its broadness. " How might I hack into a Linux box is indeed answerable. There are resources one can point to, evidence one can use to form an answer. The problem is that this question previous was perceived to have asked for what hacks could be done (rather than what methods could be used) without defining a clear system ("universal simulation" is not a definition of a system with rules), nor is "doing thing you wouldn't normally be able to do" a point to aim for. $\endgroup$
    – Mourdos
    Nov 7, 2014 at 10:46

See I'm really fairly new here but having just had a careful look at a couple of things I should have checked out day one but couldn't find (because I'm really bad at navigating websites) I don't think the question under discussion meets the basic criteria of answerability, there seems, to me, to be too much missing context for a clear cogent answer.

Edit: Actually it's not only too broad it's also completely and blatantly off-topic, this question is the antithesis of worldbuilding, it's literally asking "please help me cut the world to pieces".

  • $\begingroup$ Ash, I'm typically willing to include "world breaking" under the general umbrella of world-building, because usually the author has a purpose for the broken world. Whether it be as prosaic as "having a shattered moon is good visuals", or as plot-centric as "I need frequent moon-rock meteor showers as a mana source", a request for help in destroying the world usually betokens a desire to do something with the smoking wreckage. ;D $\endgroup$
    – akaioi
    Nov 28, 2017 at 17:04

I'm going to play devil's advocate here.

Firstly I agree that the large number of posts from new users is an indicator that it could be too broad however it could equally be an indicator that this is a really good and interesting question, it's made the Hot Topics list - these questions always attract new users. In addition I know I'm as guilty as anyone of not reading every other answer before adding my own, I don't think the fact their are lots of answers means that each one is a new idea (I'm afraid I've not read each one in detail yet).

I think the question is actually fairly well limited:

  • It presents the world succinctly (in an artificial reality)
  • It describes the people/tech... us and now
  • It's a concept we're familiar with (The Matrix)
  • It gives a clear goal to hack (or cause mischief with the system)

It could be scoped further by giving a more defined goal

  • Gain root access
  • Bend physics to fly/spontaneously combust/teleport

However by doing so you're effectively asking how could you do these things given what we know about the world. That's not what the OP wanted, they're after breaking an imaginary (we hope) system which we live in. You need to give scope here otherwise you're just asking "How could I hack the universe to fly?" it's unanswerable.

I feel that the fact this question has received so many upvotes is an indication that it's an interesting question. I think the large number of answers only shows that people WANT to add their own opinions. I believe it's our responsibility as SE users to:

  • Go through these answers and vote up/down appropriately
  • Mark the question as protected if there are a lot of low quality responses

I think the closing the question as "Too Broad" reaction hurts us because it's shutting down a question which people clearly like and want to have answers to. I think we need to go through and critique the answers instead.

  • 3
    $\begingroup$ Note though that for now, the question isn't closed. It's been put on hold until we figure out what we want to do with it. $\endgroup$
    – user
    Oct 27, 2014 at 9:35
  • 2
    $\begingroup$ Also, on what basis do we decide whether to vote answers up or down? "How we feel about the answer" isn't a good basis for voting; we should be able to judge how well the answer answers the question. That's pretty hard, with the question as it currently stands. $\endgroup$
    – user
    Oct 27, 2014 at 9:36
  • 2
    $\begingroup$ That said, I put the question on hold fully with the intention of starting a discussion about it here on Meta, so that we can decide how to best handle it. $\endgroup$
    – user
    Oct 27, 2014 at 9:36
  • 4
    $\begingroup$ Unfortunately, a popular question is not the same as a good question. Especially since a lot of too broad questions make you want to answer because you just have this really awesome idea that nobody else spotted, which is mostly due to the question being to broad! $\endgroup$
    – Mourdos
    Oct 27, 2014 at 13:43
  • $\begingroup$ "On hold" just means "has been closed for less than 120 hours". Or from another perspective, "closed" means "on hold for too long". $\endgroup$ Oct 28, 2014 at 23:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .