• Advertisement


  • Content count

  • Joined

  • Last visited

Community Reputation

3010 Excellent

About cowsarenotevil

  • Rank

Personal Information

  • Industry Role
    3D Artist
    Technical Artist
  • Interests
  1. Why A.I is impossible

    I'm actually fairly optimistic that a "definitive" answer not only exists but might even be known in the next hundred years, if advances in AI and technology in general continue at a good pace, precisely because I think we can provide meaningful evidence for or against the possibility of a "philosophical zombie." If we can make a complete computer model of the human mind, then the only "difficult" problem is finding where, in that model, this notion of "subjective experience" actually comes from. A daunting task, for sure, but not necessarily an impossible one. It's already practical to use formal verification systems to prove things about models of certain biological systems, so I can imagine asking some modeled human mind whom I'll call Mr. Robot to attempt to describe "subjective experience," and then work backwards to study the processes and structures that lead to him discussing subjective experience. It's tough to say how satisfying such an answer will be, or exactly what other interesting conclusions we can derive from it, but this knowledge would at least be sufficient to rule out the idea that subjective experience is something external to the mind (or at very least to rule out that the subjective experience we can talk about, or that otherwise affects our behavior, is external to the mind). This would not, strictly speaking, rule out the possibility that Mr. Robot is a philosophical zombie, but the alternative leaves us stuck in a really weird place: we studied the brain, found out precisely why we can talk about philosophical zombies in the first place, only for this to actually be completely unrelated to real philosophical zombies. At this point, we'd absolutely never be able to study or analyze real philosophical zombies, because by assumption we literally cannot talk about them at all. On the other hand, if subjective experience actually turns out to really be something external to the mind (e.g. some kind of weird quantum behavior that's not computable and can't be modeled at all), I expect that that would become fairly clear as well -- as our ability to model physical systems and AI improves, I can only imagine it'll become increasingly clear where this "breakdown" occurs. That is, I imagine that the more "mind-like" things we can model computationally, the better hope we'll have of discovering exactly what portion of a real mind we can't model. Obviously this second case would be "nice" in that it suggests that there might be life after death, that humans really are "special," and that we don't have to feel guilty about experimenting on poor Mr. Robot because he doesn't really feel pain. I of course have no reason to think these things aren't true, they're just not scientifically useful until they can be used to make falsifiable predictions that differentiate them from the "brains are just computers" version -- which is something we get for free if we do discover a specific reason why we can't model a human mind computationally.
  2. Why A.I is impossible

    I have mixed feelings. On one hand, I agree that using the word "qualia" brings an implication of something more than simply "subjective experience," and I think that this adds yet another not-necessary-useful level of specificity to a concept that's already difficult to pin down. On the other hand, I still think that's pretty tough to talk about "subjective experience" on its own as well -- that is, in a way that actually distinguishes "subjective experience" as a concept from the physical and functional processes of the brain. There's a popular notion of a "philosophical zombie" that's defined as something that either recreates the physical processes of the brain, or emulates the functional processes of the brain, but without subjective experience. This seems like a good starting point for actually describing subjective experience, but I also think it's a potential trap: even though it's (at least sort of) possible to imagine such a zombie, it's not necessarily the case that such an entity could actually exist. For instance, it's totally possible to "imagine" some statement in formal logic, only to realize that the statement contains a contradiction. Basically, I think it's difficult to distinguish "qualia" from "subjective experience," but it's also surprisingly hard to distinguish "subjective experience" from "no experience at all," so I don't think dropping the notion of "qualia" really gets us much further.
  3. Why A.I is impossible

    Sure. All I'm saying is that there's a tremendous difference between assuming you won't fall into a black hole tomorrow and proving that it's impossible to fall into a black hole. Like I said before, I'm not even comfortable describing this kind of conjecture about consciousness as a theory, simply because it currently doesn't make any predictions that can be falsified. This is in contrast to other theories, like theories of gravity, which at least make predictions that can actually be measured. And even so, it's still impossible to actually prove that gravity works in a particular way. In fact, it's not even possible to say that gravity is likely to behave in some particular way, due to the problem of induction. I'm not aware of any claims that quantum processes aren't possible inside the mind. Isn't the argument more that there's no reason to believe that non-deterministic processes actually have a macroscopic effect on human behavior? To me, the problem with qualia as a concept is that it appears to be self-evident, rather than being something that can be derived logically or measured empirically. For this reason, it's unfortunately really hard to argue that qualia, which is essentially defined as any subjective experience that is distinct from the actual mechanical behavior of the brain, actually exists at all. For instance, I could just claim that all laptop computers experience "zualia," which, like qualia, cannot be explained in terms of a desktop computer, but that doesn't necessarily imply that there is, or needs to be, some other way of explaining "zualia." It also certainly doesn't imply that it's impossible to build a desktop computer that truly simulates a laptop computer. It would be different if we could actually identify and measure a specific set of behaviors that is unique to laptops, but so far, this hasn't been done. Likewise, if there is something that conscious humans can do that Turing machines can't, no one has found it yet. I also think that the notion that qualia is somehow based on quantum effects, or indeed anything that can't be described in terms of a Turing machine, doesn't really help to explain qualia, either. The idea that some behavior that can't be described computationally is able to affect our mind in such a way that we're actually able to refer to it is pretty weird. It would be one thing for these processes to affect our behavior in some subtle, difficult-to-describe ways, but having these processes actually affect our behavior in such a precise manner that the physical portion of our brain can actually reference those processes themselves and reason about them symbolically would seem to require some very complex machinery indeed.
  4. Why A.I is impossible

    Except not really, at all. It's tough to argue that his claims even meet the standard for being a theory (falsifiability), much less that they're something with any hope of being proven deductively. The conjecture that a human mind can be simulated by a Turing machine is just the Church-Turing thesis, and so far there haven't been any serious challenges to it. This isn't to say that Penrose's ideas about consciousness aren't interesting, or even that they're not (potentially) true, it's just that there are a lot of assumptions encoded in them, and many of those assumptions are actually pretty weird and not widely accepted. They're certainly not rigorous, either. In my opinion, the weirdest (but certainly not only) such assumption is actually in the text you reprinted about the incompleteness theorem. It's a personal pet peeve of mine to see people abuse this theorem (which is completely rigorous) to make non-rigorous claims about "consciousness" (or anything, really). In fact, I'd say that asserting that human minds have access to some kind of magic logic that cannot, by assumption be described formally, and then arguing from that to claim that Turing machines (which are, of course, describable by formal logic, since that's the whole point of them) must be "missing" something is roughly the mother of all circular arguments. How do you even argue that such a magic logic exists? You can't do it formally, by assumption. So what is even the point?
  5. Why A.I is impossible

    I feel like this argument is correct in broad strokes, but a bit imprecise. To clarify, decidability, computational complexity, and "one-way functions" are all distinct things. Factoring is treated like a one-way function, because checking whether a group of numbers are factors of a particular number is trivial (and consequently has low complexity), but there is no known way of factoring an arbitrary number with a comparable time complexity. Interestingly, it has not been proven that one-way functions even exist at all, to say nothing of whether factoring in particular is a one-way function. Factorization isn't undecidable, though. It's totally possible to program a universal Turing machine to factor a number in a way that it's always correct, even if it works slowly. For an undecidable problem, such as the halting problem, it is actually impossible to program a machine to solve it in a way that is always correct.
  6. Game design career interview questions

    I think Maya is the best bet in general. I'm pretty sure it's the most popular package for this kind of thing, and it seems to have the best support by major game engines (Unreal, Unity) as well. That said, your best bet is to look at teams/companies that you're specifically interested in, and see what they use. I also don't think learning a new package is quite as critical as some people would want you to believe. I started at my current job (which is in fact at a very large company) with essentially negligible Maya experience (I didn't have a personal copy of it and still don't) but many years of experience with Blender, and I picked up Maya without much conscious effort just by being around people who are comfortable with it, and now I use it every day. I'm more of a technical artist, so I'm more comfortable with the API/design of Maya than I am with the actual hotkeys, etc. needed to model efficiently, but I can model in it without completely embarrassing myself. For more advanced modeling/texturing I usually jump back into Blender. I don't think anyone really cares what package you use as long as a) you actually get your work done, b) your work is good, and c) you're willing to take on the responsibility of making sure your stuff plays nicely with whatever the rest of your team uses. That said, learning another package can only increase your usefulness. On the other hand, if you need to choose between either learning how to use Maya or having a good portfolio (even if it's Blender-only), focus on your portfolio first.
  7. Why A.I is impossible

    I don't think I was actually disagreeing with you, in that I'm definitely not saying that being able to refer to consciousness is necessary for having it. I'm also not claiming that consciousness as a metaphysical thing does (or does not) actually exist. If I do disagree, it's only partially with the statement that "it's only an assumption of me that other people than me have consciousnesses." Of course it is impossible by definition to (directly) verify someone else's subjective experience, so you can't be sure that other people experience consciousness the same way that you do. But if consciousness is some metaphysical thing, it'd at least be weird (or coincidental) for lots of other people to talk about "consciousness" when you're the only one who actually has it. That is, if other people than you do have consciousness, then we can plausibly all talk about it, and everything is good. If other people than you don't have consciousness, then you can potentially talk about it, but all of the other people who talk about their own "consciousness" must actually be talking about something else, even though we all seem to be talking about roughly the same thing. You're absolutely right that a program that just prints "I have consciousness" also appears to be talking about consciousness, but since other people were presumably not programmed (at least not by you) to do this, it seems at least difficult to explain why they'd make such a statement in the absence of a conscious entity external to you. This again assumes that consciousness isn't an illusion -- if it's an illusion, then of course we all could be (falsely) claiming to have it for the same reason.
  8. Why A.I is impossible

    Why not because they can talk about it? As you point out, your own consciousness (insofar as that's a thing that exists at all) is self-evident to you, but when you talk about it, are you actually referring to it? If so, then it would, at least, be pretty implausible that other people would appear to talk about their own consciousness if it weren't something that they themselves also actually have. If not, then there's some even weirder coincidence afoot: you experience consciousness, but when you talk about your own consciousness, you're actually talking about something different than the consciousness you actually experience. Basically, either consciousness manifests itself physically to the extent that people are at least able to refer to it in speech and writing, or it doesn't, meaning we can't actually refer to it at all despite the fact that we appear to be discussing it. In the former case, the fact that people outside of your own perception of consciousness claim to refer to consciousness would suggest that they too actually can refer to it, and thus experience it in some way. In the latter case, either it's pure coincidence that we merely appear to be discussing a phenomenon that actually exists (but cannot actually be discussed), or consciousness doesn't exist at all.
  9. Why A.I is impossible

    I agree, and I think that perhaps the more relevant question would be why would we want to channel this magical spirit consciousness? What would that actually accomplish? There are two possibilities worth thinking about. Either the magical spirit consciousness does something that can also be fully described by an algorithm running on a computer or some other formal machine (which of course means that there's no need for merely channeling it -- it can be replicated instead), or that it can't be described computationally. The latter case sounds interesting: magical spirit consciousness generates output in a manner that can't be described by an algorithm. Great! What does that even mean? Superficially it seems like it could be something like "free will," but unfortunately that doesn't really work: it might be "free," but it's pretty hard to think of an action as willful if it doesn't actually follow in any describable way from the initial conditions. In fact, behavior of this sort isn't even meaningfully distinguishable from randomness (if it was, we'd be able to describe the non-random aspect of it computationally, which by assumption we cannot). Now, I certainly can't argue that channeling this magical spirit consciousness won't magically result in something desirable (it is magic, after all), but I think it's also safe to say that arguing about something that by definition cannot be described formally probably won't be very productive.
  10. I'm so confused

    I don't know, to be honest it seems to me like just deleting the post and sending the member a message would probably have been more than sufficient, given that this is pretty clearly an honest mistake (and especially now that this person has gone to the trouble of trying to learn from their mistake). Besides, a 20 day warning doesn't really seem like it'd accomplish anything unless the member was already planning to continue participating in the community for at least another 20 days. If anything, it seems like it'd discourage someone from being an active member of the community, whereas someone who made one post and then left forever wouldn't even know they'd gotten a warning. If there's really such a commitment to enforcing this rule, why not just have the forum software automatically disable posting until some minimum threshold of participation is met? At least it'd ensure consistency in enforcing this kind of thing, because right now I feel like it's a bit haphazard -- those weird "Kenneth Eng" posts are still around, for example, and that guy seemed to be deliberately failing to be a productive member of the community, even after having the situation explained to him repeatedly. I guess I just prefer to err on the side of having a welcoming community that can be flexible about this kind of thing, rather than one that says "nope, not going to remove your 20 day warning even though you apologized and tried to understand what you did wrong. Rules are rules, no flexibility, etc." Anyway, that's what I think.
  11. What game was this? FPS, low poly

    It's probably way to well-known to be the right answer to this question, but Deus Ex 1 seems to fit this bill fairly well. On a side note, someone needs to make a bot that can find the names of games and movies based only on half-remembered descriptions. Trivially, you could train a classifier based on descriptions after which the author successfully discovers the identity of what they were describing. Human memory is very strange, but I can't imagine that there aren't also a lot of patterns in which sort of details players are likely to remember/describe, etc. or even which elements are likely to be remembered incorrectly.
  12. WoA V - Afterparty/Judging thread

    Well I enjoyed this immensely, although I still haven't managed to review a single one of the other games yet. I've liked what I've seen so far, though! I guess it's a good thing I wasn't a judge. I was really pleasantly surprised by the response to my game. It wasn't a great fit for this contest, since it was much more about style than gameplay in the traditional sense, but I feel like it was judged based on its own merits. Even with some major pacing issues, I feel like people still manged to actually engage with it. I wish I'd had a chance to playtest a bit more, since I probably would have realized that going back upstairs after finding the floating orbs was not at all an obvious choice. I also noticed that it's a bit too easy to step into the final trigger zone while pointed away from the window, thus missing the ending. There are some pretty obvious things I could have done to avoid this had I discovered it a bit earlier. The idea for the aliens actually came from modeling the scroll ornaments on the columns. I thought "hmm, this looks sort of like some weird alien face," remembered that aliens were in fact part of the contest theme, and just rolled with it. I'd have liked to incorporate the scroll pattern into the actual alien design, but I didn't quite get to that in time. I've already added a lot of detail and polish to the floating spheres and secret staircase area, but I think this whole baroque + alien aesthetic might actually work if I can get it to gel together with a bit more nuance. It's been a few years since I've actually made a serious attempt at a game jam, and even longer since I've attempted a solo entry, so I'm definitely trying to polish and expand my game into something longer and less broken. The two biggest things I'd like to add are some more involved gameplay with the alien, including possibly some kind of chase sequence, plus actually getting to meet the king. If this works out I might try to have it tested more thoroughly and release it as some kind of donationware. Also, my castle looks a lot like the Cuban embassy for some reason? Especially now that I've gotten around to polishing it up a bit more.
  13. WoA V - Afterparty/Judging thread

    That was so great. I love being able to see people's live reactions, and, despite the fact that I know how playtesting is supposed to work, I have the hardest time not interfering when I see someone playing something I made, so a "naive" playthrough like this is extremely valuable. There are some obvious problems with my game, but I consider myself extremely lucky that people are consistently getting stuck in the same couple of areas. I'm totally looking forward to fixing up the ending as right now a lot of people miss all/part of it. I want to talk about it in more detail but I'm going to wait until all of the judging is done. Finally, and this is purely my own vanity speaking, but I did make 100% of the geometry myself specifically for this game. In fact, aside from some primitive shapes (and a couple of spoilery things that appear at the very end), I managed to make everything out of the same 8-ish models that I made available earlier in the contest thread. I definitely took some of the base texture images from free sources, but all of the actual materials were mine as well (although for these I re-purposed some of the effects, like the dust, from my own other projects).
  14. Hobby: How do you finish your projects?

    One fairly straightforward thing that I've found helps is to tell a bunch of people what you're working on and letting them follow along with what you're doing. If people find what you're doing interesting enough, they'll sort of automatically put some pressure on you to finish. Another thing I've found that works well is to make a short and simple "finished" version of your game, even if you've only been working on it for a week or so. It's a lot less daunting to add features to a game that you know is already good (or at least something) than it is to try to get through a massive list of features before your game be playable at all.
  15. WoA V - Afterparty/Judging thread

    This has actually been a semi-known problem for some people with Unreal for a while, although I've never actually seen it firsthand. It only pops up on some machines, and the symptom is a pretty generic low-level error which made it harder for Epic to pin down. The bad news is that thee "Real-Time Thumbnails" and most other workarounds only help with the editor, not already-packaged games, and apparently none of the workarounds work for everyone. The good news is that most people report that it doesn't happen in Unreal 4.17, and, if you didn't run into this problem with my game, I think it's safe to say that that was the difference. I bet if Newtopia is able to re-build with 4.17 it'll run properly for you. Anyway, I'm hoping to get to reviewing all of the other games some time this week. I'm also polishing up my game a bit. I'm sure I'll get more feedback regarding what's actually wrong with it, but I've already fixed a few things other people have caught, and there's a bunch of cosmetic issues that have been driving me crazy. Mainly some of the geometry that's conspicuously unfinished/un-detailed, particularly when seen in comparison to the more-polished stuff I made earlier in the process.
  • Advertisement