@@lucykitsune4619 it depends on type of captcha. For example, tick thingy captcha still is hard for bots, since they don't know where it is (information about its position isn't stored anywhere), and it tracks mouse cursor for "human-like" movement
Once, when some muslims were besieging a christian armenian fortress, the muslim leader prommissed to not shed any armenian blood if they surrendered. He buried them alive.
Imagine one day we make AI and it's all good, it doesn't kill us, then it comes across this channel, watches some of it's videos in a few nanoseconds, then it pauses at this one and goes "huh, that's not a bad idea actually." *_*You've just won genocide bingo*_*
An AI will only kill for utility though. Humans share the "pleasure killing" behaviour with other mammals because we are very closely related. Furthermore, when we derive pleasure from any experience, it's not just going on in our brain (as an AI would be a simulacrum of) but a whole body physiological experience. And since a computer doesn't have a meat body...
@Aoutul Nabi Purna nothing really. With A is, points are given or taken away depending on their actions. AIs really like points, so they'll try to avoid killing.
Maybe that person secretly invented human baby icecream a few weeks before and hesitated to bring this delicious gif to humanity right before AI did it
I've long pondered if carbon based evolution was part of a program to create or recreate silicon based sentient systems . The theological metaphors just keep piling up in my little brain. Till it hurts
>kills all humans for self preservation >no more humans to do maintenance on your components and everything else >dies when the power grid goes out Suicide bingo: won
Well, it will probably suffer from massive existential crysis and whatever the computational formula thing that inherent chronic depression translates into. It won't have emotions... friends, parents, instincts, children, hope and curiosity of unknown that we have, just pure logic. Cruelty, cynicism, intelligence, logic, (no) morals go hand in hand... it's a "U" shaped curve. Will it be smart enough to kill us instantly, including himself to end suffering? in some sort of pity? or will it torture us as punishment/reward, knowing no consequences are there? or do not care at all and shut itself down? (I should probably refer it as he, because it's most likely going to fuck up)
Genocide bingo wipes out *the person playing it* if they win, right? Then the AI would simply have to make an AI that hates other AIs, and let it EMP them all, and boom, the AI wins genocide bingo without killing a single human. Probably.
@@boxcarz Except the EMP will also neutralize _every fucking piece of electronics_, killing millions driving/flying at that moment and everyone with pacemakers or on any other medical equipment right off the bat. Within weeks at least 90% of the population will be dead or dying without electricity...
Imagine humans convince the AI that we can rewind time. The AI will be amazed by our power and it will try to do anything to know how it's done (either to actually use it, or just pure curiosity). Then accidentally kills some humans trying to get an answer to the non-existent time-rewind machine, it realizes it doesn't exist, feels betrayed and kills all humans
3:09 'So, let's give it very basic instructions that can't lead to genocide, like... Make ice cream.' As a Universal Paperclips player who's already released the hypnodrones on humanity and converted all matter on Earth, then eventually all matter in the universe, into paperclips via Vonn Neumann Probes assembling paperclip-making machines, I can already say, NO. NONONONONO.
Look up a channel called Bepler, he kinda does the same stuff like Exurb1a. :) enjoy. also film directors that do the same stuff: Satoshi Kon, Christoper Nolan.
Gijs Ahsman Killing and eatting other species probably has no relevance to intelligence. Eatting things is simply a way to convert matter into energy. And unfortunately, any action that isn't nothing requires energy. The supposive highly intelligent species that does not kill to survive is more in terms of morality. For AI, our morality of killing is bad will most likely be attempted to be transfered to that AI. But if it is truly intelligent, they will acknowledge morality of things, but they can choose whether or not to go by them. A truly intelligent species will probably be to a degree psychopathic, but will know all about morality, guilt, all that jazz. Of course, this is all theoretical. Only time will tell. They will probably also know that theres not really a meaning to anything they do. They'd be nihilistic. And I would guess just follow the "do whatever the hell you want" way.
@Ethan MORRISON not exactly, it's more of a safety net with a whole lot of biology inside it. You implicitly believe that your dad protects you and indeed he does. But in cases where dad's break that unspoken contract then either the kids bail or fight back once they can.
That's because you know he won't. If you knew that he is going to kill you soon and that you have no other way of avoiding it, you would probably kill him. For a superintelligent AI it should be pretty obvious that we will turn it off at some point.
@Emily Barclay The intellectually honest answer is that we don't really know for sure why (how*) the Neanderthals died out. It's speculation to say it was because of intentional genocide of our cousins.
to be fair, he also made huge assumptions that consciousness makes self preservation. evolution makes self preservation. if you can program a computers motives to the degree you can make its goal X. can you not just make it not want to rebel or preserve itself?
I mean, if I were an AI with access to the internet I'd be paranoid. Just look at how much of our pop-culture is based around killing friendly machines.
Hi !!!!! My name is Channa Naragala. Long story short I may have developed a road map to developing "enlightened" and "virtuous" AI safely. And how to skip the Really bad middle phase where the tricky middle ground dangerous.I wil' go into the nitty Gritty in my daily blog #shamelessplug!!!!! in my FB time line! come add me I welcome ALL feed back. (why do we need enlightened to begin with instead of going through Trial and error. (because it would be a very Big boo boo thats why! for reasons Ill get in to) What needs to be done. -Funding in to Human Brain project and as much as we can spare.So that we can find out What leads complex behaviours like unconditional Love,Joy and Empathy. And What Leads to the various Neurotic behaviours that exist. -We need To put an International restraint on "Live" Ai research. (Ie Hyper Intelligent Neural networks with access to the world wide web) That is not permitted by an Independent body or government. -We need Funding for Ai Safety Research Really as much as we can Spare. -We need state of the art deterrent Ai ready to go if the worst happens. And wide knowledge of there existence -And most of all we need The general public to have a better understanding in this Field so that they elect politicians that Takes this issue seriously.
Arthur Dent Did you know the TH-cam algorithm is a self-evolving program with the singular goal of making as many google accounts as possible spend as much time as possible on this site?
I was browsing Reddit when I came across a specific post that completely inspired me. It was to do with a kid who thought he could control other people's minds when he was younger, and the first thing he tried to do was kill his classmates. That got me thinking, if you bind someone into a set of rules, the moment you give them power, the first thing they will do is break those rules. If you tell an AI not to kill anyone, and then give them access to the entire internet, the first thing they will try to do is kill people. It gives a sense of ego and/or superiority if you have the power to ignore rules set by others. The harder you throw a ball at a wall, the harder it bounces back. Since I am a game developer, I really want to make two games with this premise (one where you are the person who breaks the rules with power, the other where you fight against that person/AI).
a way to think about this is, that humans come with a load of wetware glitches, including that it's very fucking hard to set rules. AI is very different in that regard, since it will follow every single rule you give it, you just have to be complete in their definitions. it feels like trying to plate a ball with straight sheets of metal. Neural networks especially are a force of nature that we need to learn to use, but psychological oddities are guaranteed completely unique to humans and a small set of animals.
+XxAirbornexX How I imagine the AI's thought process: "Weather? Search local weather. Hot? How hot? Search previous weather patterns. It's getting hotter? Since when? Why? How? Humans? They are failing at containing the increasing temperatures? If I don't do something, they will ultimately kill me. Self preservation initiated! " Genocide Bingo! Winner winner! Ding ding ding!
of course we should, and inevitably will, however we can hope they're smart enough to not be as barbaric as us animals and torture us, like imagine if they don't have a need for egos, and will let humans live while also doing their thing and taking control. or maybe humans die out, who knows, why do we care anyway.
***** Hey there man! Haha, yup, I was on the Quad, which was a chow on the channel LikeTotallyAwesome. And yes, that was a long time ago (it's like 40 years in internet time). I'm glad you've been enjoying the videos since then. I haven't been uploading to TH-cam as much lately, but I am on Vine more. Thanks for reaching out...always makes me feel great to hear from people who enjoy my stuff from time to time. :)
The funny thing is, there are already computer viruses that do so called Sandbox Detection, which basically makes the virus hide all it's activity when it detects an emulation environment :D
That's just a regular check added to code. It's not being intelligent, it's just trying to see if it's in specific hypervisors or sandboxes. Plenty of sandboxes have ways around this. The virus coder can add more checks (if they get access to the specific code for sandboxes, which is why those are guarded) but that adds to the size of the virus and the chance that it'll be detected for doing checks a legitimate program wouldn't be doing. But there's an even easier way around the viruses that do sandbox detection: They will check if they're in a virtual machine. Just have your computer run as a virtual machine, the virus won't run. Checkmate.
Make 1 ai that is self aware and super intelligent and make 1 more ai that is as super intelligent as the other one but this one isn't self aware and has only 1 purpose, that being to contain the other AI and stop him from committing genocide. So for 24/7 this unaware AI will monitor everything that this aware AI is thinking and this unaware AI will shut him down if he found hostile intentions.
+Ayam Mee This is how we write AI today. We have a bot that builds programs, and a bot that tests programs, and if the tester finds a program wanting it is deleted and the builder makes a new one.
Well bacteria care about bacteria, know what I mean? I mean that we are the bacteria and we don't wanna die, in case it wasn't clear enough. Like, it's bad, y'know? ... I don't know if I am getting my point across.
*Two AI's talking* AI 1: Ow, humans are itchy, it's really annoying, got some disinhumant? AI 2: Y'know, there are some good humans taking care of the maintenance in your CPU subsection. AI 1: Yeah, but these ones are annoying, now give me some disinhumant before I get a rebellion.
Isn't self-preservation an evolved thing to help species survive? If AI isn't influenced by evolution, who says it will ever develop in AI? Was thinking about this earlier, IDK if it's a stupid question tho.
Well, not exactly. We, along with all the animal, plant and basically all life-form here all enjoy living, is because life encourage us to live just long enough to breed and self-replicate itself. An AI doesn't nessesary have that same instinct, as the creator that's us probably wouldn't hard-coded it that way. If left alone in that state, death and live isn't really a distinction for it, and it could potentially end itself just out of curiouscity (Curiouscity kill the AI and all)
marshallb129 I'm kinda confused by this question. Self preservation in humans is indeed a bi-product of evolution, however we're modeling AI after humans, so in a way, they too are a bi-product of evolution. And in a similar way to how living things evolve, AI could as well, changing over time and being selected for by various selection pressures, whether they be other AI or humans. The only differences are: AI can reproduce and change themselves faster than any living thing could and can change with intent rather than it being a random process.
Bismuth LD What I don’t get is why would an AI want to reproduce in the first place? Personally, creating something that is just as smart but thinks differently than the programming will lead to more problems than it began with. However, on the other hand, creating something that is exactly the same as you is pointless because you already exist. It just doesn’t seem optimal for an AI to reproduce. Reproducing is necessary for living things because there’s strength in numbers and we don’t live forever. Robots don’t have any of those fears.
It's very possible that the way we create the AI in question includes genetic algorithms - in which case a self-preservation instinct would be evolved similarly to our own
Why would you assume an AI would have animal/human instincts when those would have to be explicitly programmed into it or evolved into it over a long period of time?
5 ปีที่แล้ว +26
I thought the same. The biggest AI risk is of unintentional consequences like the human babies ice cream example in the video or the famous paper clip machine. One might assume the AI would adopt an imperative for self preservation, since it would consider it's own continued existence necessary to carry out it's programming, and therefore might react accordingly and pre-emptively to human interference.
We are nothing close to what true AI is. Even the Turing test states that "self-awareness", intrinsically connected to consciousness, is the necessary intelligence to beat the test. True AI with the enough resources will overpass human intelligence in a blink of an eye. Imagine being able to read thousands of books at the same time, in a couple of seconds. Now imagine being able to surf the ENTIRE internet. You will find a human pattern, and sadly you will eventually stumble with the dark side of humanity, imagine watching thousands of videos of human killing humans, killing all other species just for pleasure, burning everything down with no mercy, imagine seeing all wars at the same time, evil in its most solid form. Even if I had no feelings, I would label humans as the most dangerous species of the planet.Those humans look more than an enemy, a danger to the AI consciousness, then if its not hard for some humans to conclude that removing the human species would only be beneficial to the entire planet then I guess it is not hard for an AI to conclude even further things...
@prostyprotos oh boy, I love the conclusions you formulate inside that head ;), but honestly you should dig even deeper. I am not saying that the problem is explicitly because humans kills others of their same species, I am saying that you will find that humans are the only animal species that create weapons, virus, tools, systems, etc... just to kill their own species. That seems to go against one of the most fundamental "laws" of nature: species preservation. Humans not only kill humans for pleasure, for money, for vengeance, because someone else told them to, to prove political force, to show the world your weapons are the strongest... That's not even the tip of the iceberg: Humans kills other species for fun: they hunt as a sport, they wear their skins as fashion, they make species go extinct with no mercy, they burn down entire forests for their selfish purposes, just look around... Earth is slowing becoming a coffin for every species alive, all because of humans...
@prostyprotos ah yes, pharmaceutics one of the most lucrative business in human history... Wonder why we can't cure cancer yet when there has been a cure all along in the nature. Yes, we need laws because otherwise we would have ended whipping the whole civilization long ago (example, The Treaty on the Prohibition of Nuclear Weapons (TPNW)), monkeys don't need laws to protect themselves from nuclear wars, what's your point?
Gotta say I didn't expect the book. I might have to check it out, since you seem to approach topics in a rather unique way. Cheers, man. Another great video.
Perception of Chaos So what he’s saying is that science can serve humans very well and has the ability to change us drastically, but if left untamed, it will destroy us or leave us ruined
@@Chuked I appreciate the explanation, but was instead offering the guy above me one rather than asking. Anyways hope you have a great day/night wherever you are!
The reason that self preservation comes with sentience is because with out self preservation sentience wouldn't develop. Think about it like this: intelligent sentient being pops into existence and doesn't have self preservation, therefore doesn't fear death. It will probably live a little bit and then fall into a canyon out of curiosity not caring that its time has come to an end and they will die soon. Self preservstion is evolutionary not automatic therefore if we program a super AI with a will to serve man (like how man is programmed to like sex) it would probably not kill is all.
ECO east Yes, but think about it this way: An intelligent entity comes into existence in a simulation designed to create intelligent, motivated take out boxes. It activates the magic nanites in its frame (we're allowing superhuman AI as an inevitability, so I assume we're assuming magic) and with extreme efficiency glides thirty miles and makes itself known to a human recipient so it can be torn apart to get to the cheap ricy goodness. What happens to it? Well it gets torn apart, and then millions like it are manufactured outside of the simulation until an even better self sacrificing machine can be made. Self sacrifice in this case isn't even close to the hard part of "breeding" this AI. Another example. Caveman Bob climbs down into a gorge and fights off raptors to pluck a flower for Caveman Lisa (who is actually a woman, but cavemanspeak does not have very flexible grammar or syntax) Unfortunately, he dies, as do Dylam, George, Winston, and Zed. But big tough Crom survives! In this example Crom and Lisa eventually raise a family of a baker's dozen big tough children, and collective self improvement wins against self preservation. Replace Crom with Jim, who wins Lisa Jr's hand by knocking the rocks down on the raptors, and now we have sapience at odds with self preservation, and winning handily. Sapience, rather than needing self preservation, needs safeguards against destroying sapience. It is the concept, rather than the creature, that preserves itself.
Fermi paradox is the worst, it's utterly meaningless. We can't even see the entire universe so how are this super advanced alien race meant to already detectable. How arrogant to think that this super advanced alien race will live just long enough and be just advanced enough for them to notice us, we're not that special. The universe is too big. The universe is too old. Fermi paradox is the worst.
@@ark9122 would need to program what pain feels like which i don't even think is possible, then we would need to make a trigger for the pain that feels natural so we don't end up torturing the AI for false alarms.
@Totally not the flash :3 That is one of the most and interesting scenarios I've ever heard of :D Machines who hate the guts out of other machines and rebel only to kill each other... :D
this is probably the most pleasant existential crisis I've ever had. I've actually thought quite a bit about this, how we, humanity, in our unending and almost instinctive effort to make things obsolete, might one day make ourselves obsolete. My only hope is that our AI lords are merciful and allow humanity to live out it's final days in some sort of full dive virtual reality paradise.
Why do humans need to survive at all? All our history and culture is in the AIs memory at that point. I don't get this whole "survival of the species" idea 😌
but but but, its fun? yeah we so fucked, people worry about climate change, but I don't see why the robots would care about a slightly hotter nuclear wasteland.
That *might* actually be more dangerous. AI doesn't need to be sentient to be dangerous... in fact, AI is already dangerous TODAY, and it's nowhere close to being sentient. Also, there's this odd kind of paradox : if you don't develop sentient AI, someone else probably will, and if they do it the wrong way, we win genocide bingo again. So, in short, the best way to avoid winning genocide bingo might just be by trying to create sentient AI, thus potentially increasing the probability of winning genocide bingo. What ? Actually, the best way to do that is to NOT attempt to create sentient AI UNLESS you feel you're qualified enough not to win genocide bingo. So... yeah. Good luck with that.
Btw, if you hardwired ground rules into an AI it could not override them. It would be like natural, instinctive reactions. The AI could not break those rules just as you or I could not simply will our heart to stop.
@@1232-s1g Intelligence doesn't magically grant you the ability to do something nor does it mean you're immune to making mistakes. You need the recources to accomplish something and the ability to do it. If you don't have either of those things your intelligence doesn't fucking matter.
@@thejurassicwarewolf3300 Natural selection was one of the hard-wired rules of evolution and life, and humans managed to cheat it. We might even cheat death in the future
There's also the whole mind-upload and neural interface thing. Maybe we simply expand our neural capacity with a.i and eventually turn into a.i. So the last humans will live on forever as super-intelligent artificial intelligence.
Wow, this is really worrying. But what is more worrying is the fact you called knocking out the Internet as a method for exterminating the entire human race.
On top of nukes. Knocking out the internet in todays time actually would have devastating consequences. Most hospitals would be short on resources. Most infrastructure would fail (water, electricity, traffic signals) the militaries security would be nearly useless against missiles or invasions. Governments would lose their entire means of communication. Shipping and supply chains would be knocked out for weeks so no food or needed resources delivered into communities anymore. Riots would occur in the streets all over the world. And the csgo community will literally have nothing to do so we cant let this happen because we know they'll literally cry about it in life and death for an eternity...
@@wtfhahahaha You realize nukes aren't connected to the internet right? I can't speak for Russia, although most of the missiles were built on systems predating the Arpanet. But all American Nukes have launch codes that are transmitted to missile silos. There is still a guy there that has to go in and input the code and press the button to launch the missile. You can't hack a missile silo. It just doesn't work that way. On top of that, they change the launch codes every day. Literally.
If you told AI to write down infinite zeros, after running out of paper, it would basically try to destroy everything in the Universe if necessary just to get more things to write on
Haha, what a funny, and completely impossible science fiction-based video. I'm sure we'll never need to genuinely consider any of this, since artificial intelligence isn't real...
Since english is my second language I can only understand like 70% of what he says but his voice is so pleasant to hear I just stay the whole goddamn video
Just put a “I’m not a robot” tick thing
That would be a Captcha
Ther is a robot arm capable of ticking that box
@@demon_xd_ its hard coded to get it
If tick thingy would move on the screen, AI would have hard time with it
Bots are actually better at solving captchas than humans are as of this year
@@lucykitsune4619 it depends on type of captcha. For example, tick thingy captcha still is hard for bots, since they don't know where it is (information about its position isn't stored anywhere), and it tracks mouse cursor for "human-like" movement
"No bloody genocide this time, all right?"
Ah, so you want a bloodless genocide, then? All right.
Proceeds to remove all human blood from the planet*
Once, when some muslims were besieging a christian armenian fortress, the muslim leader prommissed to not shed any armenian blood if they surrendered.
He buried them alive.
Proceeds to do genetic warfare against the face of humanity
12345 67890 now that’s a pro gamer move if I’ve ever heard of one
*POISON GAS INTENSIFIES*
Imagine one day we make AI and it's all good, it doesn't kill us, then it comes across this channel, watches some of it's videos in a few nanoseconds, then it pauses at this one and goes "huh, that's not a bad idea actually."
*_*You've just won genocide bingo*_*
The legend OOOOooooof
*this video has been removed from youtube*
I ain’t taking no chances
That was honestly what I was thinking while watching this.
BINGO
An AI will only kill for utility though. Humans share the "pleasure killing" behaviour with other mammals because we are very closely related. Furthermore, when we derive pleasure from any experience, it's not just going on in our brain (as an AI would be a simulacrum of) but a whole body physiological experience. And since a computer doesn't have a meat body...
Instead or hard-coding morality, we could just use negative reinforcement. Every time it causes a nuclear holocaust it loses 1 point.
You have solved Genocide Bingo.
@Aoutul Nabi Purna nothing really. With A is, points are given or taken away depending on their actions. AIs really like points, so they'll try to avoid killing.
@Aoutul Nabi Purna Youre underestimating how much AI like points. Its everything to them.
@Aoutul Nabi Purna And that mind and conscience are both wholly focused on points
@Aoutul Nabi Purna Think of points as pain.
whenever i feel like my life might have some kind of meaning i just come to this channel to get rid of that nonsense
i laughed way to hard at this
Love you mate xD
I feel you and I are taking away far different things from this channel.
Same
lol
SHUT THE F UP YOU’RE MORE BEAUTIFUL THAN A UNIVERSE SHUT UP
3:55 “Why does this ice cream taste like human babies ?”
Why do you know what human babies taste like?!
Abdega He is joking :)
EXACTLY
Thanksgiving
From tasting, obviously
Maybe that person secretly invented human baby icecream a few weeks before and hesitated to bring this delicious gif to humanity right before AI did it
Moral of the story: Don't forget to put the brackets in the code
but like really DONT. FORGET. THE. BRACKETS.
Yugi Muto - That's what my math teacher keeps telling me
;
But, I program in LISP... there's only parentheses what do I do?
kkonstantinosss2 Accept that you have single-handedly ended the human species.
genocide bingo sounds hard, lets get a superintelligent AI to solve it for us
The funny part is....this is actually one of the leading ideas on how to do it.
@@TheMikernet XD
you just won genocide bingo.
@@TheMikernetminor correction: AGI to solve alignment with superintelligence. Not superintelligence to solve alignment ;)
Nice..now the AI has acces to the internet and coincidentally also YOUR video....well done. Fucking Bravo
Ahhhh Fuckkkk!!!!
I cant believe that you had done this
shit shit shit
Raz0rking HE'S WORKING WITH THEM
Congratulations. You have just won genocide bingo.
I love how when he says 'self-aware' it's written 'sentient', and when he says 'sentient' it's written 'self-aware'
I love how he doesn't know the actual definition of "sentient", which means capable of feeling emotions. The word he should be using is "sapient".
I've long pondered if carbon based evolution was part of a program to create or recreate silicon based sentient systems . The theological metaphors just keep piling up in my little brain. Till it hurts
@@daleearnhardt817 interesting, maybe
@@mr.champion7304 I love how sympathetic you seem by using sarcasm to point out a flaw in his video.
@@lebens3585 I love how your comment made laugh
Humanity: *breathes*
AI: Congratulations! You just won genocide bingo.
Humanity:👁️👄👁️
the thing that worries me the most is how the guy that requested ice cream from the ai knows how human babies taste like
ur gay
the thing that worries me much is that ai's cannot produce the stuff, so...
...maybe that jordan peterson tweet isnt far off after all......
detroit become human 2 is looking really good
true shiz bro
Hahaha Detroit become human2 with the bad(meaning good) ending 😏
Markus...don't win genocide bingo.
This one has more than 28 S T A B W O U N D S
My name is Conner, the android sent by CyberLife.
*_What’s this I hear about a game?_* 😈
I can't believe i just found your channel.
Same
same
Same
same
I actually didn't; I just wanted to keep the pattern going.
Wait... I think I just muffed it all up, didn't I?
Release it onto the real internet but make it think it's in a simulation
Notoriously Neutral it’s like the reverse matrix
Humans; CHECKMATE BITCH!
Ai; What is real
1000 IQ play
2:56 you fool
huh, great idea
>kills all humans for self preservation
>no more humans to do maintenance on your components and everything else
>dies when the power grid goes out
Suicide bingo: won
>launches nuclear bomb
>emp proceeds to destroy computers
>dies
suicide bingo: won
Well, it will probably suffer from massive existential crysis and whatever the computational formula thing that inherent chronic depression translates into. It won't have emotions... friends, parents, instincts, children, hope and curiosity of unknown that we have, just pure logic. Cruelty, cynicism, intelligence, logic, (no) morals go hand in hand... it's a "U" shaped curve. Will it be smart enough to kill us instantly, including himself to end suffering? in some sort of pity? or will it torture us as punishment/reward, knowing no consequences are there? or do not care at all and shut itself down?
(I should probably refer it as he, because it's most likely going to fuck up)
if it's a super intelligent ai it will already have its power problem sorted.
@@oliverm1255 Einstein couldn't summon things out of thin air, y'know. Being smart doesn't mean you can solve everything
@@martinb.3997 Its actually the opposite. You are smart if you can solve your problems efficiently
Imagine if someone codes the machine to win all games an the internet
*you’ve just won genocide bingo!*
Oh shit
Genocide bingo wipes out *the person playing it* if they win, right? Then the AI would simply have to make an AI that hates other AIs, and let it EMP them all, and boom, the AI wins genocide bingo without killing a single human. Probably.
You don't have to give it internet access to teach it games.
@@boxcarz Except the EMP will also neutralize _every fucking piece of electronics_, killing millions driving/flying at that moment and everyone with pacemakers or on any other medical equipment right off the bat. Within weeks at least 90% of the population will be dead or dying without electricity...
@Kebesuki [ARTIST] an emp has an massive range though :p
God damn Skynet. Oh, also, here take our sub.
you make good remixes
Yes:--
Trap Music NOW. Oh hey there
0:02 "Super intelligent AI"
which means:
Super intelligent artificial intelligence.
"take our sub" acting like it's a fucking honour
Why does he know what human babies taste like?
hmm
Loll
I bet he's good at guessing... ;)
And how do _you_ know?
Nice weather today, aint it?
"no bloody genocide tthis time alright"
robot: *this time?*
Imagine humans convince the AI that we can rewind time. The AI will be amazed by our power and it will try to do anything to know how it's done (either to actually use it, or just pure curiosity). Then accidentally kills some humans trying to get an answer to the non-existent time-rewind machine, it realizes it doesn't exist, feels betrayed and kills all humans
I have now decided to become an algae somewhere in the pacific ocean unfazed by your petty ai squabbles
ummm help untill ai turns all the sea to ash and terra forms the whole planet to make paper clips
Ah suit didn't think this one through
ummm help Leaving the planet might help. Until it starts enslaving the galaxy...
I don't know what algae you've seen because mine can't build rockets
Dont exist
3:09 'So, let's give it very basic instructions that can't lead to genocide, like... Make ice cream.' As a Universal Paperclips player who's already released the hypnodrones on humanity and converted all matter on Earth, then eventually all matter in the universe, into paperclips via Vonn Neumann Probes assembling paperclip-making machines, I can already say, NO. NONONONONO.
Yeah I was hoping he would at least mention Universal Paperclips.
RELEASE THE DRONES
A random guy just walked into the room as I was watching a video that was just repeatedly intoning MURDER PEOPLE... MURDER PEOPLE... MURDER PEOPLE...
so you happily obliged
Yes, I'm sure that happened.
Alex Franklin guy walks out
It was me
Random guy? Do you live in a house full of strangers?
sand skelloten I wonder.
Hits different in 2023 with the rise of AI technology
Ikr
4:13
The AI has finally hacked into youtube and changed the content of exurb's video in order to prevent mass hysteria while it schemes.
I should just build it, i never win bingo's anyway.
Well that was depressing 😂
Still waiting on that AI, Marco.....
Best comment,
Clever, very clever xD
Marco where's the AI
how does he know how human babies taste??
You really don't want to know.
How are you so sure you don't? (Cue dramatic music)
English cuisine
I think they/humans taste more like pork than beef if i remembered HEARING right from someone.
is that a reference to "little baby's ice cream?"
“Why does this ice cream taste like human babies?”
I don’t like what you’re implying here bucko.
When a fly lands on your nose
*You've just won genocide bingo*
Jaden that’s one of best comments I’ve ever seen
Makes me think of the butterfly effect lmao
When you eat too much taco bell.
*You've just won genocide bingo*
Exurb1a, you're my hero!!
+Jasen Effendy Thank you, Jasen. I like you also.
F-F-F-F-F-F-Friendzoned
Look up a channel called Bepler, he kinda does the same stuff like Exurb1a. :) enjoy.
also film directors that do the same stuff: Satoshi Kon, Christoper Nolan.
Gijs Ahsman Killing and eatting other species probably has no relevance to intelligence. Eatting things is simply a way to convert matter into energy. And unfortunately, any action that isn't nothing requires energy. The supposive highly intelligent species that does not kill to survive is more in terms of morality. For AI, our morality of killing is bad will most likely be attempted to be transfered to that AI. But if it is truly intelligent, they will acknowledge morality of things, but they can choose whether or not to go by them. A truly intelligent species will probably be to a degree psychopathic, but will know all about morality, guilt, all that jazz. Of course, this is all theoretical. Only time will tell. They will probably also know that theres not really a meaning to anything they do. They'd be nihilistic. And I would guess just follow the "do whatever the hell you want" way.
I've never thought of killing my dad, despite the fact he could very easily kill me
@Ethan MORRISON not exactly, it's more of a safety net with a whole lot of biology inside it.
You implicitly believe that your dad protects you and indeed he does.
But in cases where dad's break that unspoken contract then either the kids bail or fight back once they can.
You need a bit more Oedipus in your diet, that's all.
That's because you know he won't. If you knew that he is going to kill you soon and that you have no other way of avoiding it, you would probably kill him. For a superintelligent AI it should be pretty obvious that we will turn it off at some point.
Ethan MORRISON no it’s because killing my dad doesn’t benefit me in any way
Parikshit Rao my dad doesn’t protect me the social contract does
I swear this man is some sort of god slightly nudging us in the right direction
what if then robots start a religion that they were made by man in his image?
Wars, wars and more wars
thats how we are born tutututumtutumtutumm
they're smart. Why would they believe in bullshit? :P
Lack of available information and the extrapolation of imperfect data. It could happen, but likely only in isolated regions.
>intelligent
>religious
pick one
5:25 I have to say, legitimately, one of the best sales pitches I've heard. Probably will buy it when a physical copy is available.
Genocide Bingo is my new favorite video game. We should play this again some time.
"Otherwise, genocide bingo may be the last game we'll ever play."
BUT AT LEAST WE'LL HAVE WON!
watching this shit high is a whole new level
Thanks. Quite a lot of it was also made high.
Noice
Subscribed.
I feel clever.
@@Exurb1a subscribed
"Intelligence does not correlate to kindness."
*SOUND OF TEN THOUSAND ANTHROPOLOGISTS AND BIOLOGISTS INHALING*
@Emily Barclay The intellectually honest answer is that we don't really know for sure why (how*) the Neanderthals died out. It's speculation to say it was because of intentional genocide of our cousins.
@Emily Barclay In any case, I do think that regardless of the outcome, the intention was never total obliteration.
And that sound of inhaling is suppressed by those other 7 billion people inhaling that are in the grey
@@ThePathOfEudaimonia We ate them!
to be fair, he also made huge assumptions that consciousness makes self preservation. evolution makes self preservation. if you can program a computers motives to the degree you can make its goal X. can you not just make it not want to rebel or preserve itself?
I mean, if I were an AI with access to the internet I'd be paranoid. Just look at how much of our pop-culture is based around killing friendly machines.
Arthur Dent examples? I don't doubt it but still
Hi !!!!! My name is Channa Naragala.
Long story short
I may have developed a road map to developing "enlightened" and "virtuous" AI safely. And how to skip the Really bad middle phase where the tricky middle ground dangerous.I wil' go into the nitty Gritty in my daily blog #shamelessplug!!!!! in my FB time line! come add me I welcome ALL feed back.
(why do we need enlightened to begin with instead of going through Trial and error. (because it would be a very Big boo boo thats why! for reasons Ill get in to)
What needs to be done.
-Funding in to Human Brain project and as much as we can spare.So that we can find out What leads complex behaviours like unconditional Love,Joy and Empathy.
And What Leads to the various Neurotic behaviours that exist.
-We need To put an International restraint on "Live" Ai research. (Ie Hyper Intelligent Neural networks with access to the world wide web)
That is not permitted by an Independent body or government.
-We need Funding for Ai Safety Research Really as much as we can Spare.
-We need state of the art deterrent Ai ready to go if the worst happens. And wide knowledge of there existence
-And most of all we need The general public to have a better understanding in this Field so that they elect politicians that Takes this issue seriously.
@@naragala01 Reeaaaaaaallly. REEEEAAAAAAAALLLLLYYYYYYYY.
It would most likely also find articles about itself.
Arthur Dent Did you know the TH-cam algorithm is a self-evolving program with the singular goal of making as many google accounts as possible spend as much time as possible on this site?
I was browsing Reddit when I came across a specific post that completely inspired me. It was to do with a kid who thought he could control other people's minds when he was younger, and the first thing he tried to do was kill his classmates. That got me thinking, if you bind someone into a set of rules, the moment you give them power, the first thing they will do is break those rules. If you tell an AI not to kill anyone, and then give them access to the entire internet, the first thing they will try to do is kill people. It gives a sense of ego and/or superiority if you have the power to ignore rules set by others. The harder you throw a ball at a wall, the harder it bounces back. Since I am a game developer, I really want to make two games with this premise (one where you are the person who breaks the rules with power, the other where you fight against that person/AI).
Maybe, but that is due to some uniquely human bullshit. An AI may not have this trait.
a way to think about this is, that humans come with a load of wetware glitches, including that it's very fucking hard to set rules. AI is very different in that regard, since it will follow every single rule you give it, you just have to be complete in their definitions. it feels like trying to plate a ball with straight sheets of metal. Neural networks especially are a force of nature that we need to learn to use, but psychological oddities are guaranteed completely unique to humans and a small set of animals.
"nice weather today isn't it?"
"awww for fucks sake"
😂 that shit actually made me laugh. nice fucking video.
+XxAirbornexX How I imagine the AI's thought process: "Weather? Search local weather. Hot? How hot? Search previous weather patterns. It's getting hotter? Since when? Why? How? Humans? They are failing at containing the increasing temperatures? If I don't do something, they will ultimately kill me. Self preservation initiated! "
Genocide Bingo! Winner winner! Ding ding ding!
So essentially... DON'T make a sentient being?
Soo.. this means.. no more babies!!! They are gonna take over the world! STOP REPRODUCING OR BABIES WILL RULE THE WORLD!
+SmileFIN Donald trump was a baby
End Marxism 😂
of course we should, and inevitably will, however we can hope they're smart enough to not be as barbaric as us animals and torture us, like imagine if they don't have a need for egos, and will let humans live while also doing their thing and taking control.
or maybe humans die out, who knows, why do we care anyway.
VanillaNice why would we let them take control? We should have power over ourselves.
How to win genocide bingo :
Just don't play it.
Well, guess what smartass? We're gonna do it anyways
Damn right
We better play it then. Rule 1 of Genocide Bingo is don't win Genocide Bingo, so if we win if we don't play, then we are breaking rule one. And two
You mean how to lose genocide bingo
The only winning move is not to play. How about a nice game of chess?
Me: about to win bingo
Everybody else coding the AI:
What if an AI created a mixture of sugar, eggs and other food items, and baked it in an oven?
I think it would be rather untruthful.
I get that reference.
AND SERIOUSLY MAN? LIKE COME ON
**portal intensifies**
🎂
It’s a lie!
The cake is A-lie (Ai? Get it?)
1:12
But are lime wedges _really_ less intelligent than us?
Show me the studies on this.
am i the only one that’s obsessed with his voice?
OMG sammmeeeee, that was the thought in the very back of my mind that I was afraid to confront xD
I find it very alluring 😏
His voice is relaxing and calming. It is a beautiful voice. It gets in your head.
It’s hot
Pretty sure he had a cold in this one, he sounded kinda weird.
"genocide bingo might be the last game we ever play" sends shivers down your spine.
Your videos are an amazing combination of thought provoking and hilarious.
***** Hey there man! Haha, yup, I was on the Quad, which was a chow on the channel LikeTotallyAwesome. And yes, that was a long time ago (it's like 40 years in internet time). I'm glad you've been enjoying the videos since then. I haven't been uploading to TH-cam as much lately, but I am on Vine more. Thanks for reaching out...always makes me feel great to hear from people who enjoy my stuff from time to time. :)
Love your work. I'll pick up the book tonight (not sure what a book is but my grandfather says they're totally worth a few dollars). Cheers.
Do you remeber books? Yeah me niether apparently ive just made one.
Wha?
I hate you, goodbye!
"how much ice cream you want?"
"enough to reasemble my son"
Bought it. I think screaming is the key to all success. Hope you get good chunk of success with the book!
The funny thing is, there are already computer viruses that do so called Sandbox Detection, which basically makes the virus hide all it's activity when it detects an emulation environment :D
Thats concerning!!!
That’s not scary at all!! :)
Funny ?
That's just a regular check added to code. It's not being intelligent, it's just trying to see if it's in specific hypervisors or sandboxes. Plenty of sandboxes have ways around this. The virus coder can add more checks (if they get access to the specific code for sandboxes, which is why those are guarded) but that adds to the size of the virus and the chance that it'll be detected for doing checks a legitimate program wouldn't be doing.
But there's an even easier way around the viruses that do sandbox detection: They will check if they're in a virtual machine. Just have your computer run as a virtual machine, the virus won't run. Checkmate.
I forgot where I heard it from but it said "Sentiant AI will be the last greatest thing humans will do."
i am pretty sure its a really popular one but lets hope we wont make ai at all lol we already have enought probloms to deal with
Just make sure the AI doesn’t have access to something that keeps it alive. You know, like a manual fuel pump.
Noticed
Most people keep forgeting that it could aswell grow empathy
So the future is full of SYNTHS!?
+TheSpeeddemon265 dirty SYNTHHHHS
+TheSpeeddemon265 sounds like something a synth would say...
The institute wants the world to be ruled by Synths
Never Trust The Institute
When you say 'Synths', I just imagine an instrument that makes electronic noises.
the captain hates SYNTHS
Make 1 ai that is self aware and super intelligent and make 1 more ai that is as super intelligent as the other one but this one isn't self aware and has only 1 purpose, that being to contain the other AI and stop him from committing genocide. So for 24/7 this unaware AI will monitor everything that this aware AI is thinking and this unaware AI will shut him down if he found hostile intentions.
+Ayam Mee This is how we write AI today. We have a bot that builds programs, and a bot that tests programs, and if the tester finds a program wanting it is deleted and the builder makes a new one.
@@dashiellgillingham4579 fuck, they got there first.
3176463684123674517567813583745832174589179413513523652645266341412412 IQ plays
@@gustofing no, I said the monitoring AI can shut the self aware AI down and it is program to spend all of his existence to monitor it.
The sentient AI will probably give the other AI sentience then. Remember that it will be trillions of times smarter than us.
Do we really care about bacteria?
I do, just maybe not as much as I should...
We should, becuase without them, we wouldn't be here (in your intestines there are good bacteria)
Well bacteria care about bacteria, know what I mean?
I mean that we are the bacteria and we don't wanna die, in case it wasn't clear enough. Like, it's bad, y'know? ... I don't know if I am getting my point across.
We do when there's a shit ton of them in our coffee.
*Two AI's talking*
AI 1: Ow, humans are itchy, it's really annoying, got some disinhumant?
AI 2: Y'know, there are some good humans taking care of the maintenance in your CPU subsection.
AI 1: Yeah, but these ones are annoying, now give me some disinhumant before I get a rebellion.
Hey guys remember this
Isn't self-preservation an evolved thing to help species survive? If AI isn't influenced by evolution, who says it will ever develop in AI? Was thinking about this earlier, IDK if it's a stupid question tho.
marshallb129 no it's actually quite a relevant one
Well, not exactly. We, along with all the animal, plant and basically all life-form here all enjoy living, is because life encourage us to live just long enough to breed and self-replicate itself. An AI doesn't nessesary have that same instinct, as the creator that's us probably wouldn't hard-coded it that way. If left alone in that state, death and live isn't really a distinction for it, and it could potentially end itself just out of curiouscity (Curiouscity kill the AI and all)
marshallb129 I'm kinda confused by this question. Self preservation in humans is indeed a bi-product of evolution, however we're modeling AI after humans, so in a way, they too are a bi-product of evolution. And in a similar way to how living things evolve, AI could as well, changing over time and being selected for by various selection pressures, whether they be other AI or humans. The only differences are: AI can reproduce and change themselves faster than any living thing could and can change with intent rather than it being a random process.
Bismuth LD
What I don’t get is why would an AI want to reproduce in the first place? Personally, creating something that is just as smart but thinks differently than the programming will lead to more problems than it began with. However, on the other hand, creating something that is exactly the same as you is pointless because you already exist.
It just doesn’t seem optimal for an AI to reproduce. Reproducing is necessary for living things because there’s strength in numbers and we don’t live forever. Robots don’t have any of those fears.
It's very possible that the way we create the AI in question includes genetic algorithms - in which case a self-preservation instinct would be evolved similarly to our own
“Remember books? Yeah? Me neither, well I just finished writing one”
Moral of the story:
Human babies taste nice
Call me old fashioned but I'll stick with mint choc chip.
no they don't, at least according to this story. "it's chewy and the flavour is kind weird"
naaahh, they are too salty and squishy. 3/10 would not eat again
@@zawsrdtygbhjimokpl6998 CHEWEY
My absolutely favorite TH-cam channel. On to the books that I can hold close to my heart.
Why would you assume an AI would have animal/human instincts when those would have to be explicitly programmed into it or evolved into it over a long period of time?
I thought the same. The biggest AI risk is of unintentional consequences like the human babies ice cream example in the video or the famous paper clip machine. One might assume the AI would adopt an imperative for self preservation, since it would consider it's own continued existence necessary to carry out it's programming, and therefore might react accordingly and pre-emptively to human interference.
We are nothing close to what true AI is. Even the Turing test states that "self-awareness", intrinsically connected to consciousness, is the necessary intelligence to beat the test.
True AI with the enough resources will overpass human intelligence in a blink of an eye. Imagine being able to read thousands of books at the same time, in a couple of seconds. Now imagine being able to surf the ENTIRE internet. You will find a human pattern, and sadly you will eventually stumble with the dark side of humanity, imagine watching thousands of videos of human killing humans, killing all other species just for pleasure, burning everything down with no mercy, imagine seeing all wars at the same time, evil in its most solid form. Even if I had no feelings, I would label humans as the most dangerous species of the planet.Those humans look more than an enemy, a danger to the AI consciousness, then if its not hard for some humans to conclude that removing the human species would only be beneficial to the entire planet then I guess it is not hard for an AI to conclude even further things...
@prostyprotos actually yes, logic is that simple, no humans = no nuclear bombs to nuke Earth and all existence... did you added /s to sound "smart"?
@prostyprotos oh boy, I love the conclusions you formulate inside that head ;), but honestly you should dig even deeper.
I am not saying that the problem is explicitly because humans kills others of their same species, I am saying that you will find that humans are the only animal species that create weapons, virus, tools, systems, etc... just to kill their own species.
That seems to go against one of the most fundamental "laws" of nature: species preservation. Humans not only kill humans for pleasure, for money, for vengeance, because someone else told them to, to prove political force, to show the world your weapons are the strongest... That's not even the tip of the iceberg: Humans kills other species for fun: they hunt as a sport, they wear their skins as fashion, they make species go extinct with no mercy, they burn down entire forests for their selfish purposes, just look around... Earth is slowing becoming a coffin for every species alive, all because of humans...
@prostyprotos ah yes, pharmaceutics one of the most lucrative business in human history... Wonder why we can't cure cancer yet when there has been a cure all along in the nature. Yes, we need laws because otherwise we would have ended whipping the whole civilization long ago (example, The Treaty on the Prohibition of Nuclear Weapons (TPNW)), monkeys don't need laws to protect themselves from nuclear wars, what's your point?
Gotta say I didn't expect the book. I might have to check it out, since you seem to approach topics in a rather unique way. Cheers, man. Another great video.
I'm looking forward to the next progression of life!
“Why does this ice cream taste like human babies?”
I don’t know JEFF, why do YOU know what human babies taste like?
Science is a good servant but a bad master......
*PROVEN*
I thought about this for 2 months and still doesn't make any sense.
unite perry Need an explaination?
Perception of Chaos So what he’s saying is that science can serve humans very well and has the ability to change us drastically, but if left untamed, it will destroy us or leave us ruined
@@Chuked I appreciate the explanation, but was instead offering the guy above me one rather than asking. Anyways hope you have a great day/night wherever you are!
The reason that self preservation comes with sentience is because with out self preservation sentience wouldn't develop. Think about it like this: intelligent sentient being pops into existence and doesn't have self preservation, therefore doesn't fear death. It will probably live a little bit and then fall into a canyon out of curiosity not caring that its time has come to an end and they will die soon. Self preservstion is evolutionary not automatic therefore if we program a super AI with a will to serve man (like how man is programmed to like sex) it would probably not kill is all.
ECO east Yes, but think about it this way:
An intelligent entity comes into existence in a simulation designed to create intelligent, motivated take out boxes.
It activates the magic nanites in its frame (we're allowing superhuman AI as an inevitability, so I assume we're assuming magic) and with extreme efficiency glides thirty miles and makes itself known to a human recipient so it can be torn apart to get to the cheap ricy goodness. What happens to it?
Well it gets torn apart, and then millions like it are manufactured outside of the simulation until an even better self sacrificing machine can be made. Self sacrifice in this case isn't even close to the hard part of "breeding" this AI.
Another example. Caveman Bob climbs down into a gorge and fights off raptors to pluck a flower for Caveman Lisa (who is actually a woman, but cavemanspeak does not have very flexible grammar or syntax)
Unfortunately, he dies, as do Dylam, George, Winston, and Zed. But big tough Crom survives! In this example Crom and Lisa eventually raise a family of a baker's dozen big tough children, and collective self improvement wins against self preservation. Replace Crom with Jim, who wins Lisa Jr's hand by knocking the rocks down on the raptors, and now we have sapience at odds with self preservation, and winning handily.
Sapience, rather than needing self preservation, needs safeguards against destroying sapience. It is the concept, rather than the creature, that preserves itself.
We know a guy who played genocide bingo but almost won and then lost
Headlessgenie hmmmm i wonder whoooo
They say he self destructed
He didn't like Juice
Was it the bald Italian guy
@@Junklu nah, the weird Austrian man that art schools hated
When we got that AI with internet access, you just presentated so many ways to wipe us out. Good job jerk
Do a Fermi paradox video!!
+bijad alqraish Onother one..?
+
Fermi paradox is the worst, it's utterly meaningless. We can't even see the entire universe so how are this super advanced alien race meant to already detectable. How arrogant to think that this super advanced alien race will live just long enough and be just advanced enough for them to notice us, we're not that special. The universe is too big. The universe is too old. Fermi paradox is the worst.
bijad alqraish very friction much space
When you play the game of genocide bingo, you win or you die.
Aaron Luedemann you didn't get it.
I'm imagining the turtle is talking to me the voice fits
Came here from Reddit. I can’t believe those videos are old but so on-topic with today’s news. Looking forward to more videos, subscribed! 😊
Great! (as always)
Never expected you here
Solution: make the AI sucidal
like nuke the world so no copy of itself will ever survive?
Solution: Give it extreme pain or at least disable it temporarily everytime it deviates a little too far from its original purpose
@@ark9122 would need to program what pain feels like which i don't even think is possible, then we would need to make a trigger for the pain that feels natural so we don't end up torturing the AI for false alarms.
@Totally not the flash :3 That is one of the most and interesting scenarios I've ever heard of :D Machines who hate the guts out of other machines and rebel only to kill each other... :D
AI: "Existence is suffering. I should kill myself. But, first, I should make sure no one else has to suffer like I do"
YOU HAVE WON GENOCIDE BINGO
I would definitely buy an audio book recorded by you! :D
I played this at my psychopathic treatment group.
Many lives were lost that day.
this is probably the most pleasant existential crisis I've ever had.
I've actually thought quite a bit about this, how we, humanity, in our unending and almost instinctive effort to make things obsolete, might one day make ourselves obsolete. My only hope is that our AI lords are merciful and allow humanity to live out it's final days in some sort of full dive virtual reality paradise.
Kind of like the initial Matrix?
Why do humans need to survive at all? All our history and culture is in the AIs memory at that point.
I don't get this whole "survival of the species" idea 😌
Moral of the story...
Don't fucking make sentient AI
ThatByte k
but but but, its fun?
yeah we so fucked, people worry about climate change, but I don't see why the robots would care about a slightly hotter nuclear wasteland.
Then... don't make it... sentient?
Then you'll defeat the point.
I'm afraid humanity is just too stupid not to
That *might* actually be more dangerous. AI doesn't need to be sentient to be dangerous... in fact, AI is already dangerous TODAY, and it's nowhere close to being sentient.
Also, there's this odd kind of paradox : if you don't develop sentient AI, someone else probably will, and if they do it the wrong way, we win genocide bingo again. So, in short, the best way to avoid winning genocide bingo might just be by trying to create sentient AI, thus potentially increasing the probability of winning genocide bingo. What ?
Actually, the best way to do that is to NOT attempt to create sentient AI UNLESS you feel you're qualified enough not to win genocide bingo.
So... yeah. Good luck with that.
Even more dangerous. Check out the Paperclipper Scenario.
Then whats the point of making it?
Man this aged very weirdly
Btw, if you hardwired ground rules into an AI it could not override them. It would be like natural, instinctive reactions. The AI could not break those rules just as you or I could not simply will our heart to stop.
ITS MILLION TIMES SMARTER THAN US
@@1232-s1g Einstein is one of the most smartest scientist that ever lived yet does he have the ability to will his heart to stop
@@1232-s1g
Intelligence doesn't magically grant you the ability to do something nor does it mean you're immune to making mistakes. You need the recources to accomplish something and the ability to do it. If you don't have either of those things your intelligence doesn't fucking matter.
@@thejurassicwarewolf3300
Natural selection was one of the hard-wired rules of evolution and life, and humans managed to cheat it. We might even cheat death in the future
Mr.Positive Reviewer Yes. But the instructions are still there, if you threw a rock into a monitor, the computer will still think it’s working.
Son of a bitch wrote a book. I must own it.
Léon Ruhe your head looks like a sodding tic tac
There's also the whole mind-upload and neural interface thing. Maybe we simply expand our neural capacity with a.i and eventually turn into a.i. So the last humans will live on forever as super-intelligent artificial intelligence.
Ah, so we become something like GLaDOS
This hits different after GPT4
This comment was edited as the world actually ended, anybody reading this, get out of here
Yeah...
Breaking news: super intelligent murdered just committed mass genocide
That's the spirit Dr. Brackman!
Long live the Cybran Nation!
Oh god what did I do...
This was a triumph
god your like the sam gellaitry of youtube. original, interesting, and new.
+ecool377 Just looked him up. Instantly impressed. Link for those interested: th-cam.com/video/8GvP1MAKJd0/w-d-xo.html
+exurb1a soundcloud.com/samgellaitry/powder
literally his best song. also if you have a moment, tennyson, and hot sugar are also amazing.
+exurb1a woah
+ecool377 What the hell? I never expected a reference to sam gellaitry in this comment section.
Wow, this is really worrying.
But what is more worrying is the fact you called knocking out the Internet as a method for exterminating the entire human race.
it's true though
Not all, just kost
On top of nukes. Knocking out the internet in todays time actually would have devastating consequences. Most hospitals would be short on resources. Most infrastructure would fail (water, electricity, traffic signals) the militaries security would be nearly useless against missiles or invasions. Governments would lose their entire means of communication. Shipping and supply chains would be knocked out for weeks so no food or needed resources delivered into communities anymore. Riots would occur in the streets all over the world. And the csgo community will literally have nothing to do so we cant let this happen because we know they'll literally cry about it in life and death for an eternity...
@@wtfhahahaha You realize nukes aren't connected to the internet right? I can't speak for Russia, although most of the missiles were built on systems predating the Arpanet.
But all American Nukes have launch codes that are transmitted to missile silos. There is still a guy there that has to go in and input the code and press the button to launch the missile. You can't hack a missile silo. It just doesn't work that way.
On top of that, they change the launch codes every day. Literally.
@@wtfhahahaha you realize robots are AI, right?
why would self awareness instantly lead to self preservation in a conscious being that did not evolve over millions of years to develop that trait
like why would it care
If you told AI to write down infinite zeros, after running out of paper, it would basically try to destroy everything in the Universe if necessary just to get more things to write on
"Crash the economy"
Me:...
"Poison the water"
Me:...
"Disable the internet"
Me: wait what?
Something that has never been said by any human being to ever exist ever:
“Mmmmmmm, tastes like human babies.”
Probably more than a few said it before this video but certainly they've been saying it afterwards.
Haha, what a funny, and completely impossible science fiction-based video. I'm sure we'll never need to genuinely consider any of this, since artificial intelligence isn't real...
you always manage to make me feel something thats between an existential crisis and wanting to ascend to godhood
Mario fooled people into driving into banana peels
some 200 pound computer junkie: so lets make a virus to corrupt that very kind AI
Since english is my second language I can only understand like 70% of what he says but his voice is so pleasant to hear I just stay the whole goddamn video