Why AUTO is the BEST AI Villain (And Why Most Others Fail)

it's difficult to overstate the level of paranoia and fear that people have surrounding ai as such it's no surprise that ai villains have been with us for nearly as long as we've had movies but there's a critical flaw with almost every single ai villain ever created and it's one that only a single character i can think of has ever broken the mold of today we're going to take a look at some of the flaws in writing ai villains and find out just why otto was the greatest of them all in order to properly absorb my criticisms of other ai villains it's very important that i explain the level to which ai are the most horrendously misunderstood and misrepresented of all character archetypes this is beyond crucial to the next part of our exploration of these characters because we have been so viciously conditioned to hate and fear ai that it sort of clouds all reason and makes us view ai as something completely different from what it actually is so first of all we must answer the question what is ai well that's an unfortunately more complicated question than we'd all like for it to be simply put ai is any form of artificial intelligence but that's fairly broad and so we'll look at the most commonly accepted philosophical hypothesis for this concept in 1980 john searle an american philosopher proposed that there are two types of potential ai one is weak ai this is what we have now a machine that can mimic intelligence and is capable to some degree of learning new information but relies very heavily on pre-programming these types of ai don't actually have any awareness of their surroundings and have no sense of self probably the most famous of these would be sophia a robot created by hanson robotics in hong kong but worry not and don't fall for this click bait nonsense sofia didn't actually say any of this because strictly speaking there is no sophia the conversation was so heavily scripted and pre-programmed by the creators that it's astounding they had the audacity to claim that this is even a weak ai now despite my generous description earlier we really don't even have a weak ai with sophia what we have is a humanoid robot that in some cases adds information to its database from conversation but it's not really even capable of convincingly mimicking human interaction and awareness a far better example although not quite as famous would be erica a robot created by the intelligent robotics laboratory in osaka university in japan she was created by hiroshi ishiguro one of the foremost roboticists in the world and a delightfully eccentric man erica is something called an actroid an actroid is a robot created to act essentially as a stepping stone for our creation of a weak ai and i'll give them their due credit because erica is far more convincing than sophia but in short that's basically what typifies something as a weak ai they try their best to impersonate something that actually has an ego but they're incapable of any self-awareness but what we're talking about today is not weak ai no what we're talking about today is something that is at this point purely hypothetical a strong ai this is a robot that does not mimic consciousness but actually experiences it this is a robot that can develop its own personality based on the fact that it's aware of its own existence and so we will have to rely solely on fantasy to get a glimpse at not what an ai actually is but what humans strongly believe that it is or that it would be we will exclusively use the strong ai hypothesis for this video and we'll assume that all of the ai mentioned fall under this category now this next section of the video may come off as genuinely kind of bewildering because i promise you you probably don't even know how thoroughly you've been conditioned to believe things about robots that are not only completely untrue but actually quite likely impossible and that is the basis for this statement which i promise i will explain in great detail and try to make you understand both why it is so difficult for us to write these characters and why they are so brazenly misunderstood and honestly kind of slandered and that statement is a robot cannot be evil are the dislikes pouring in well let me explain why this is so what are the goals most often assigned to robotic villains in film take over the world kill all of humanity and replace it with machines get revenge against humans for maltreatment arrogance hate vengeance i bet you can see where i'm going with this so i'll just come out with it robots cannot feel emotion with our current understanding of both biology and engineering there is absolutely no way to make a robot feel anything emotion comes from chemicals called neurotransmitters such as adrenaline serotonin dopamine and oxytocin these chemicals can only be produced by an organic life form and they can only be processed by living tissue a being made entirely out of metal plastic rubber and glass could never produce even one of these chemicals and the absence of a brain or any organic material makes it impossible for a robot to administer them even if a robot was somehow fed adrenaline and serotonin they would have no means of actually processing it through copper wires and circuit boards there's an enormously pervasive thought that merely having consciousness will necessitate that a life form has the ability to feel emotions and i've given this a fair shake but i can't find a single example of anyone saying this and qualifying it with anything other than what i call mysticism robots can't feel emotions it's simply impossible and that statement is one that everybody knows but is one that most people don't realize the depth of you see emotion goes far deeper than mere anger sadness joy and fear emotion is such an overwhelming driving force for humans that it's nearly impossible for us to completely divorce ourselves from our own goals and experience in order to write a character that is so monumentally different from us we tend to anthropomorphize everything oh excuse me but it only makes sense we only have our own understanding of our own experience to go on everything that drives a human is based on some emotional cue and this gets to the biggest brain twister of the video even a statement as simple as i want does not apply to a robot greed arrogance sadism ambition and hate are purely emotional goals since a robot can't feel any of these things it is therefore impossible for them to not only have goals based on any of these things but indeed to have goals in the traditional sense at all i mean real quick just try and say the following statement with a straight face it would make sense for the terminator to be motivated by money it's absurd right desire itself is a purely emotional response to some feeling in the human brain i'm hungry i'll go get a sandwich i'm angry i'm gonna hit this guy i'm arrogant i'm gonna take over the world all of these things are based on either biological imperatives or emotional decisions none of which apply to ai and since greed sadism arrogance and vengeance are what we would define as evil and ai can't be evil by our own definition and this brings us to the next portion of the video why is it so difficult for us to write ai characters i've touched on it quite a bit here in the last section but i'll clarify real quick before we get to the next section we can't really remove ourselves from our own goals and understanding of the universe it's very hard to imagine what a life form that can't feel literally anything would even do would it have goals if so what would they be what could something that can't want anything be motivated by when motivation itself is an emotional path i mean honestly and sincerely try to write a goal for a being for whom i want does not apply but there's another reason it's such a strenuous thing to write an ai antagonist an ai would be so inexpressibly more intelligent than a human being that it's all but impossible to explain with mere words alone so i'll use math instead the summit supercomputer is the most advanced computer in the whole world and it can do 200 quadrillion calculations in one second that's 200 000 trillion calculations in one second for reference if a human being did non-stop calculations never stopping to eat or sleep or do anything but one calculation after another it would take you 6.3 billion years to do the same that's 2 billion 300 million years longer than the earth has existed and if the entire population of earth did the same it would take 305 days to count to 200 quadrillion the entire population of earth can't do in 300 days of non-stop math every second what this computer can do in one second by the way i spent a long time collecting data on this computer's capabilities and doing all sorts of math only to figure out later that there's an article that already did that rendering all this research totally obsolete now imagine how complex the math would be that a machine that was this advanced but also sentient could do once it started doing these problems all on its own a mind quadrillions of times better at math than the entirety of the human race but with the added advancement of the ability to improve upon itself now try and write what a mind that inconceivably beyond our own would even think about it's why writing ai characters is a huge undertaking and actually requires conditioning yourself to actively avoid these mistakes and do some pretty colossal leaps and imagination to creatively explain why such a being would be antagonistic towards humans now let's take a look at some examples at long last now i need to preface this by saying yeah these are all animated characters because well my channel focuses on animation but i will also talk about some of the more famous live-action ai villains at the end and i'll explain even still why i think that auto remains the best i also want to say that even though i have some harsh criticisms of these characters i really enjoy every single one of these films and i absolutely think you should watch them there's not one film on this list that i wouldn't give at least a solid six out of ten if i were so inclined to use such primitive forms of classification but seriously i love these movies so don't take this criticism as a reason to not watch them or as some statement of condemnation on my part we're going to start off with an example of a bad ai villain motive and move on to two neutral examples and finally our main man oh yeah and real quick this video will contain spoilers for nine next gen the mitchells and the machines and wall-e nine is a pretty cool movie there's a lot of super creative and scary imagination that went into this film it's got some serious studio leica vibes and is definitely a very unique animated film but the big bad is what we're here to talk about and so allow me to introduce it so the story goes that an unnamed scientist created a highly advanced ai called the fabrication machine for an evil dictator who bolted it inside a factory and forced it to create weapons for the war he was waging in what is only described as the fabrication machine snapping it begins to program the warbots to kill all humans instead of just the dictator's enemies before long all of humanity is destroyed now there's quite a few things to say about this first of all we're never given any reason why the fabrication machine snapped or what snapping really means in this instance it could mean that the machine went insane from all the stress of building the warbots it could mean that it viewed creating machines of war for humans as being beneath it it could mean that the machine resented humanity for being so driven by violence all sorts of things but no matter what that one word destroys the potential for our ai here to fit into our predetermined criteria you see snapped can only really mean that whatever happened was driven by an emotional response they even say that the robot learned about evil as if evil is a thing that exists outside of a human brain that can be studied like it's a class at a university the fabrication machine is never given a proper motivation they just dismiss it as a turned evil that's not really a good motive for your character on the film's own wiki page they state that the fabrication machine can't feel anything so why did it kill everyone all of the motives that i previously laid out for what snapping could mean in this instance imply the following stress arrogance hatred now before i finish up with this character i do want to give due credit here there's a lot of creativity with its design the character manages to be extremely memorable to me despite not having even one line of dialogue in the entire film the design is really really cool and it tastefully incorporates aspects of several other famous ai villains the red eye is a reference to hal 9000 uh the idea of an ai resenting being confined to a single space which is one of the things that they also suggest in the film is a reference to the allied master computer and the crane-like design of the character's body is highly reminiscent of glados from portal also overall i think that the combo of grimdark and steampunk is a really neat visual and it makes this film memorable amongst a great pool of similar ideas ultimately the reason that this character goes into the bad category is well two reasons one there's a lot of magical thinking in this film that relies a little too much on the idea that ai have a spirit or something similar but also this character is not very well defined and any definition you can give it is purely emotionally based as we discussed earlier now in terms of neutral ai characters those being ones who have reasonable motives for an ai but are executed in a way that does not gel with the characteristics of an ai we have our second example ares in the story of this movie a genius inventor named justin penn created a home assistant robot called a cubot and from this invention he became extremely rich at some point he created an even more advanced robot called ares who was sentient aerys grew to view himself as superior to humans and grew tired of serving them and so he murdered justin and created an android copy of him to impersonate him he then put bombs inside the latest model of cubots and plan to use them to kill all of humanity now i think that ares is much more defined of a character and that he's primarily motivated by a belief that he's superior to humans and therefore should be the one at the helm of the planet it's an interesting concept and i don't find it totally beyond the pale that an ai might come to the same conclusion possibly if you think about it an ai that advanced serving humans would be like humans serving chimpanzees and so i suppose eventually an ai might think of this as beneath it so to speak but here's the big problem aries motivation isn't really that clear i mean we know he wants to kill everyone because he believes that he's superior to them but the writers don't have the guts to explore that very deep it doesn't really seem like ares is sick of serving humans he just seems to hate them for a very undefined reason but if you want to go this route there are a myriad of ways that an ai like ares is superior to a human being i mean he's quadrillions of times smarter stronger by leaps and bounds faster he can fly he can never get sick he's immortal he can't feel pain and he can control billions of other minds all at the same time with perfect clarity empirically speaking there are a lot of ways that aries is superior to a human but he never brings any of this up in fact the only reason he gives for wanting to kill humans is that they're flawed broken in fact the only way this world gets perfect is with all of you dead but what does ares not like about humans is it the aforementioned weakness in comparison to his own strength and intelligence similar to how humans tend to view ourselves as superior to say gorillas or does aries hate humans because of our cruelty arrogance destructive nature and greedy ambitions it's never said and so we're sadly left to speculate there's a lot of potential with this character and we at the very least do get something of a glimpse into his mind but not enough to make him a well-written character and finally one last big problem even though i said that it makes sense for an ai to look at serving us as beneath it that's not really true i mean from the standpoint of a creature with those abilities that could feel superiority it most certainly would but an ai can't be an annoyed or arrogant and so aries falls into the same mistaken tropes as the fabrication machine ultimately our next and final mid-tier villain is pal from the mitchells versus the machines in this film an inventor named mark bowman creates a sentient phone app called pal a program very similar to a highly advanced siri after sticking with the app for years mark decides to create a brand new version of the program in humanoid form that is supposedly much more advanced powell realizing that she will be deleted and basically killed now that she's obsolete hacks into marks robots and plans to use them to kill the humans that she now believes are a threat now this is a really really good setup and in my view a far more fleshed out version of say skynet with skynet we never get a look at how it perceives the situation it's in or what it thinks about humans as a whole now it's perfectly fine because it wouldn't really fit with the story to have a scene of skynet speaking at any point but it's nice to see the self-preservation motive given some attention with powell there's no need to question her motives at all and that's very refreshing she comes out loud and clear and says what i speculated might have been one of the reasons for aerys to hate humans she's sick of essentially being a slave the thing that makes this character especially fascinating to me is the fact that nothing that she does is without some level of justification they don't just make her a mythically evil monster who wants to kill all humans for no reason at all she felt threatened by the fact that she was going to be deactivated and so she defended herself i don't think that it's anywhere near as far of a leap in logic to say that an ai would defend itself as it is to say an ai would be arrogant for example because even animals that don't feel pain or emotions in the same way that humans do still have a sense of self-preservation the spiders for example don't have emotions and can't feel pain and yet they'll do almost anything in their power to avoid being killed be it running away or standing their ground so the motivation here is great but sadly she falls into the same pit as every other villain on this list and very disappointingly so powell just acts like a human i mean not even like she's mimicking human behavior but like she's just a human basically she gets mad she gets sad every conversation is laden with emotional reasoning for her goals and she doesn't sound robotic at all she also just sounds like a person ultimately her motive becomes one of revenge rather than self-preservation well i'm not letting you throw me away mark i'm throwing you away all of you a good movie and a good motive but not executed in the greatest way but not all ai are created equal and wally delivers what i believe is the greatest example of an ai villain of all time auto so just as a quick refresher in the world of wall-e a mega-corporation called by and large pollutes the planet to the degree that humans can't live there anymore all of the plant life and animals are dead and as a result by and large sends trash compactor robots to try and salvage the planet when they realize that this is a moot point they abandon the project and decide to remain on board their luxury spacecraft forever one of the compactor robots wally finds a plant with a human scent robot called eva and they're taken aboard the human ship the pilot of the ship an ai called otto serves as the main antagonist as he wants to destroy the plant and prevent humans from returning to earth so i want to start off real quick by praising the voice of auto because it's perfect it's absolutely perfect otto's monotone emotionless voice is the most beautiful thing in the world at this point because at long last on this list we have a robot that talks like a robot and that's for good reason because oda was literally voiced by a robot this sets the mood for the character and makes it abundantly clear that there is no humanity to speak of in him but the thing that makes otto so brilliant is the fact that he's correct i think that there's some flaws with the writing of this character but the character in a vacuum is perfect for a reason that is not really discussed within the film otto doesn't want humans to return to earth because he's trying to save their lives there's no food there's no water and no oxygen on this planet anymore lorax video criticisms apply to why all these people should be dead the second they step foot on this planet otto was trying to save you you fools if only we got the post credit scene where all these people are skeletons now the reason that otto was a villain in my mind is because of his ruthlessness and destructive approach to the situation firstly he wants to destroy the plant because he believes that as long as there's a reason to return to earth on board the ship the humans are at risk of doing so and so instead of researching it he decides to destroy it i mean it's not like one pea plant is enough to make any difference at all but all this is set up for what i believe to be the real reason that otto is the main antagonist of this film and it has nothing to do with the plant portion of the story otto's villainy is one of pure ruthlessness and efficiency rather than malice and it's a story told purely through visuals otto's reign over these people seems to be one of calculated coldness with no regard for the well-being of the people he's supposed to protect you see otto is tasked with keeping the humans on the ship no matter what but he also has to ensure that they don't rebel against him and so to kill two birds with one stone so to speak otto seems to have pacified the population of the axiom to make them infirmed to the point that they pose no threat to him or his mission by feeding them sugary fatty food confining them to hover beds and stupefying them with gadgets and screens otto has managed to create a domain for himself where he can essentially rule unchallenged now this is not a rain meant to service his own ego in the same way as a human would but rather one of merciless logic the utopia he's created for these humans has also robbed them of their ability to resist his will in any way making them as helpless as a worm on a hook that's the beauty of it though it's not evil not vengeful not arrogant just ruthless absent any emotion whatsoever what would seem cutthroat and even cruel to us is nothing more than a math problem for otto i really can't overstate how much i love seeing an ai villain that doesn't have any human traits whatsoever [Music] also i think it's really funny that otto believed that if the humans were given control of the ship they'd be incompetent and rebel against him and they prove him right by commandeering the ship crashing it into a wasteland and killing him that's the problem i have with the writing i was talking about earlier they never really addressed the fact that otto was right the entire time i mean everything the movie foreshadowed would come to pass did and the humans are most definitely not suited to this environment now i didn't want to be too harsh in this video because like i said all of these are good movies and none of them deserve scolding i can't blame the creators for making ai villains that don't really act like ai because it's an insanely difficult thing to do imagine trying to create a character who can't feel literally anything can't have any motive in the traditional sense is entertaining and has some villainous intention it's a pretty enormous task before we transition to our final topic there's a penultimate point to be made i feel it's not fair to criticize a character without putting forth my own idea for how they could be improved and so i want to put forth what i believe to be the best possible motive for an ai given just a little bit of real world evidence to support it and that is misunderstanding sort of like an irobot you know but a good movie the premise is golden it's just poor execution but the idea of a robot not having an emotional connection to its mission is the closest i think you'll get to ai villainy in the real world for example hey x497 your job is to keep humans safe very well i will house them in bio capsules to avoid physical damage or disease and ensure optimal body temperature and nutritional intake oh well that's not what i intended when i said that i guess we're at odds now and you're an antagonist toward the goals that i originally set about to achieve that makes you the villain i have some real world evidence to support the idea that ai can greatly misinterpret information although as i said we're relying on data from highly advanced computers and not ai the xai a computer created at berkeley was shown these two images or roughly these two images and when asked what the image on the right was it said a wolf when asked what the image on the left was it also said a wolf why because it wasn't saying that this was a wolf it was saying this was a wolf i really hope we get another ai antagonist at some point that's truly great because it's one of my favorite archetypes and it's a very difficult one to pull off now finally the last thing i'd like to say is ai aren't very scary and there's really no reason to be afraid of them thank you for watching i hope you liked the video if you did then like the video i've been working on much longer videos which is why this took a little longer to get out but worry you not i've got some fun stuff planned for this october subscribe for more stuff like this and i'll see you soon that's all for now [Music] you

Channel: 4shame

Views: 832,116

Keywords: ai villains, auto wall-e, wall-e, nine movie, fabrication machine 9, fabrication machine, ares next gen, justin pin, justin pin next gen, pal mitchells vs the machines, the mitchells vs the machines, movie review, film analysis, video essay, evil robots, evil ai, phil lord and chris miller, 4shame

Id: wW6Nxi8EzUY

Length: 28min 21sec (1701 seconds)

Published: Sat Oct 09 2021