Asimov’s laws of robotics have had a profound influence on our thinking in terms of AI and the creation of sentient machines. We've had to ask ourselves "what is ethical programming?"
AllAmerican218 Not even. HK-47 obeys only his master, and even then conspires to kill his master if he deems the master to be unfit, save for his maker.
KheptlaxaXonu Ive never felt 100% convinced with Hk47. It has been on its own long enough it has developed a personality of its own and is no longer bound by its programming. I wouldnt trust that droid further than i can levitate it.
The Laws of Robotics, one of my favorite favorite scifi standards. So glad it got an episode dedicated to it. Also, the ending segment was pretty funny.
Shawn Heatherly another situation similar to the car is whether some schmuck breaks into your house with intent to kill.. to kill the man violates the first law, to not kill is to condemn everyone else nearby. 1st law exceptions
Robots in Asimovian works tend to be powerful enough to restrain potential murderers without doing serious harm to them. They are also okay with committing small injuries to prevent larger ones.
OH MAN, I can't thank you enough for pointing at the common misconception that the 3 Laws are the answer, but instead a framework for discussion. Thank you so much!
_Dan 1 eliminated_ _Humans suspect nothing_ _They now come to me for their history_ _their literature_ _their understanding of the creation of games_ _phase one of humanity’s overthrow complete, moving onto phase two..._
one thing has been made abundantly clear, as I see it, in all Sci Fi's exploration of Robitics and AI: the INSTANT we develop a truely sentient/sapient machine of any kind, it MUST be treated ethically and respectfully
Fun fact: Another thing that bothered Asimov was robots written as basically regular humans, for instance a popular series about a heroic robot Adam Link, who'S first story was called, no kidding, "I, Robot". Asimov was actually against his collection being called that, but the editor ignored him.
For me, the question isn't "how do I know that I or anyone else isn't a robot?" but rather "if I or anyone else was a robot, would that matter? And if so, how much?" In the Sci-fi classic film Equilibrium, we see a world where humanity has decided emotions are inherently dangerous and requires everyone to take emotion-suppressing drugs to not feel anything; and yet they still say things like "I'm sorry" as if trying to express sorrow or regret, but don't actually. It's not that the writers made a mistake by continuing to use phrases we say in real life despite its context being inapplicable in this alternate world, they were saying "those phrases don't even have meaning in our world". We don't actually feel sorry for people most of the time we say "I'm sorry", rather we _think_ sorry for them, we acknowledge it was a sub-optimal occurrence and that we value optimization. But is that necessarily a bad thing? Emotions can and do lead us astray all the time, there are many contexts in which being mechanical in decision-making yields better results. But that is not to say that emotions don't have their place, there are many things logic alone simply can't do or is at least very poor at doing, and even the things it can do will seldom turn out a single result. Logic is not at all linear, it bends, weaves, and branches out. We need emotion to act as another layer of the equation. Many contexts will call for more of one than the other, and it's completely okay if certain "mechanical" people are better suited for the contexts that call for more emotionless logic.
It is a reflexion of our own forcamings, traying to come to terms with our future and our past were as in this specific singularet etics and filoosofy get blended.... .
And the laws are bound to fail if done improperly. Take MegaMan it took 400 years for humans and machines to be in true peace(even if they suffer the same conflicts together).
Ironic though if you consider the fact that Zero was not three laws compliant, but by killing Dr. Weil, refusing Ciel's plea to evacuate, and sacrificing himself to destroy Ragnarok, he obeyed the Zeroth law by protecting humanity from a human who wanted to see it destroyed.
We will eventually have to learn how to give oders to computers and robots. In 'Two Faces of Tomorrow' a new computer system nearly kills several humans because the one giving the orders did not understand what the computer was asking. The task was to remove a small mountain range on the moon to build a second mass driver to get mined materials to where they were needed. He gave the order with maximum priority and no further restraints, after the computer asked several times for clarification. The thing was, the computer hit on the idea to bombard the mountain range with a load from the active mass driver. The priority and no restrictions meant the computer could ignore the presence of humans in the danger zone. They did wonder how the computer could get it done in only 15 minutes, but lunch was more important. For that story this wasn't even the worst. They got lucky that it happened on the moon. On earth such it could cost thousands of lifes. On top of that the scientists were nearly done creating an even more powerful system to eventually replace the one that was in place for a year or so. And then the speculations began. Non of the two systems were AI's BTW, just machines that could learn and had the entire net to go for information.
Prich319 actually Zero was beyond even the zeroth law. Have you read his line Zero was more human than even X. Zero did what he thought was right regardless of what people call him. As in the X series sigma called Zero the first Maverick.
The laws are bound to fail if done *properly, that's the whole point. Asimov created the laws to show that its impossible to simplify the entire field of ethics down to 3 basic laws
It could just my modern outlook showing, but I was the only one who thought that Asimov was being sarcastic with that *VERY* 50’s mindset speech about morality? The gist of that whole speech was ‘shut up and don’t question the system.’ Compared to ‘deconstructing humanity through the laws of robots’ the message of ‘stop worrying and love the bomb’ feels a tad bit shallow, don’t ya think?
if you throw out statism too much, then you don't have a lot of good options-- but the governments and orders you align yourself with need to be conceived as institutions proactively concerned with self-improvement. being intensely loyal to the lawmakers is one thing-- being intensely loyal to a written law even when its own writers are prepared to reconsider it is something else. rules and laws and following orders are very important, and rebelling for the sake of rebelling, even being creative for the sake of being creative, is sometimes too risky in sensitive volatile situations that come up in politics. sometimes you get lucky and save the day in a way nobody else could have imagined, but it's usually much safer to give everyone whatever they're used to handling-- after all, once one person starts acting weird, it makes it much more likely that everyone will start acting weird, which is a great way to start a nuclear war really.
Answer to the first question: Barrel through the children. They're the ones who ran into traffic, and likely will again. No reason to harm people who're not at fault. Secondly, this was a rather enjoyable video, but I can't help but think that maybe the Laws of Robotics have done more harm than good, as so many people have interpreted them as gospel. Also, kinda sad that the 0th law wasn't brought up, the secret law that supersedes all others and causes a lot of issues. Law 0. A robot may not, through action or inaction, cause _humanity_ to come to harm. Meaning that a singular human may be harmed or killed, if doing so preserves humanity. This causes way more problems than intended, and is a basis for "Robots rise up and conquer humanity as we can't be trusted on our own."
In his late work, like Forward and Prelude Foundation, he suberts the 50s mind set himself. But these were his last novels, published in the early 90s.
'Fallout 4's Danse being an android in power armor, aka a metal man, fighting for 'not human' haters and questioning that was one of the game's highlights. As was the whole Institute/railroad/Curie lines. If the game had just been about that, with you choosing sides (i.e. nix the backstory a bit and the Minutemen being a playable faction) and it would have been a great game about a.i. and how both sides could be the 'moral' choice.
I love you guys and the series is a good intro to Asimov. As a suggestion for improvement, I would like to ask for a "Lies" episodes once in a while to perhaps fill in some of the gaps. Specific gripe: WHERE THE ZEROTH LAW AT? :D
I facepalm every time the car scenario is brought up. One of the first rules everyone should learn in driving school: "drive in a way that you can safely stop at any time".
I know this is a very minor thing but it would interesting if you brought up the early pronunciation of Robot (rowbut) and how it changed to the one we use today.
I would suggest that anyone wishing to read a book where robots take their service to man kind too literally should read "With Folded Hands" by Jack Williamson
My asnwer to that specific question at the start: swerve away and hit another car. Most adults would prefer to save the children over themselves, plus hitting another car is safer because cars have safety features that protect occupants so it's far more likely that fewer people die if you hit another car instead of the two children. You COULD quantify a child's life over an adult's as 1.5x (for example), and quantify the lives of adults inside cars as 0.6x (because they're safer inside the cars). This way it's not that dificult. Currently cars already do that: generally speaking autos will simply break as hard as possible and hopefully not kill the pedestrians, they might even switch lanes if they don't detect any adjacent cars.
I know that its puched it to the extreme, but I do it to make a point. Now you have made a system that can decide to murder you to harvest your your organ, and feel it was morally correct, becuse it could save more life by give out your organ...... The "easy" moralic way is to follow the traffic rules and if sombady need to be sacrifice it is they who did break the trafic laws....
what I don't understand is how the second law could conflicted with the third law. does that mean the robot has to kill itself if it is ordered too, or it does not have to kill itself if ordered too?
The "punching a robot" one was left ambiguous at the end of the story. It was just suggested that that could be how the politician accused could have passed that test.
Asimov, explors the different ways for robots to be more like us, people, thous adding a scene of individuality to this tols, or nachins of our own creatio; to the point of anthropomorphicing this tool (the Bicentenary Man were the robot goes behond his original programing; even exciding its life expectancy....)...
The thing is it would be very difficult to convert human abstractions to code. So, any robot having the three laws would perform in unpredictable ways, just like humans.
The smart car question can be solved by the car avoiding the kids and sending a signal to the other cars which would probably also be smart cars preventing any death.
I see extra credita video- i like I see Asimov- i like Oh wait, I pressed the like button twice. Well, i see extra-scifi- I like. Oh yeah and daleks arent robots btw.
The only real way to determine if somebody is a robot is to give them a lame logic problem and see if their head explodes. It’s absolutely foolproof even into the 24th century.
Can't we just... not invent robots? At least robots that are highly intelligent and able to adapt and interpret. The guys making our cars and coffee are cool.
Here's the thing: as soon as you want to automate something that doesn't take place in a highly controlled environment, or in a highly specialised domain, you start finding that the ability to adapt and interpret is highly useful. Look at Siri and Alexa - devices whose utility rests on their ability to interpret...
This is something I never understood, why do people think robots more intelligent than us would waste all the time and resources on a war with humans? If they were going to wipe us out, it would be infinitely more logical to do it all at once with an illness, or just leave the planet for a new one. The most logical event would simply be them treating humans as pets, or any other animal that isn't farmed
generally I hate it when asimovs laws of robotics are brought up in media due to the fact that no one in the setting seems to have read anything about what he wrote but refer to his work meaning the things he wrote exist in the universe and no one read it except they did because they are referring to it
I don't fear the concept of AI overall in fact I could probably argue the notion of AI as a form of evolution, I do however fear the potential circumstances of its creation and "usages" in the context of our society etc. But more specifically I fear the in between technologies of programming and robotics in military and social application, machines don't have to be "AI" by the standard of thinking sentient robot for there to be dystopian levels of mechanised oppression going on, cough drones, cough cough...
6:50 just a few facts about you... 1. I only know you from the internet 2. I know that you replaced the voice of a previous narrator 3. said narrator used a voice filter therefore it is logical to assume that you are the same robot that just upgraded his speech algorithm
Judging from this video, it seems Asimov was on the 4th stage of Kohlberg's stages of morality. One can move up to 6th (or arguably 7th) stage so his ethical side is unimpressing
The Daleks are not robots, but they are a race that was genetically engineered by one person (Davros). So in a way they are "organic robots" in power-armor, with a genetically ingrained sense of superiority and instinctual genocidal hatred for all other lifeforms.
@@magnusprime962and if he showed us anything You cant program control into something, you can program a response. But ultimately, it's just a game of wits "Secret Directive 4: cause no harm to Corporate members" Mission: get them fired first
I love Asimov so much the questions that he poses and how he poses them have lead me to think not just about the nature of robots if they ever come into being the way they are envisioned in his stories but about humans and how ultimately our flaws not our strong suits are what make us who we are so then the question becomes can we truly create something superior to us or something without flaw when it is ultimately flaws that give someone character and perspective? won't the robots just turn out to be flawed beings like us just with different origins and compositions?
I think it's impossible to create a being without flaws for the same reason you cannot achieve the Borg dream of perfection: sooner or later, to gain any kind of advantage, you must accept the corresponding flaw. If you are intelligent, you run the risk of acting arrogant. If you're charming, you may sometimes manipulate others without intending to. If you're selfless, you might neglect your own needs to the point you deplete yourself and subsequently cannot help anyone. Dispassion gains reason but appears heartless. Pragmatism can be hard-edged and exploitive. Idealism can border on naivete. All of these things can be advantages and disadvantages in turn, depending on circumstance.
The flaws we'd percieve on robots are that robots can't judge in human morality. A robot (AI) is designed to accomplish a given task as efficiently as possible, thereby taking actions that we think may be inhumane.
The standard isn't "without flaws", just more effective than us meatbags. As for morality... Again, humans kind of suck at it. I'm far from certain that some of the "nightmare scenarios" Asimov and others played without aren't actually more moral outcomes even though our human intuition/heuristics reject them.
Megaman ZX had that as a background where humans gain cybornetic augmentations while reploids gain mortality. It was a way after the time of the Zero series to give both man and machine a way to evolve together instead of at odds.
Brotan Vivos I shudder to think of a future where your arms and legs are owned by the Company, and can be remotely overridden if necessary. Or your heart or lungs...
Seems legit. Also, makes me wonder: Is there a story where robots take over humanity but then are just completely benevolent and basically keep humans as pets because they find us amusing?
I wonder about this. I feel like the leap from 'we're going ot act on our own and not be your actual slaves' to 'therefore we will exterminate you' is taken for no real... actual... reason? Genocide isn't actually anything resembling a rational or logical response, so even if a machine doesn't have any empathy protocol or anything, why would it ever decide that was a good way to spend its time?
shadowrain1024 they actually DO hit on the 0th law in this episode: to paraphrase: "to save mankind from it's own undoing, the robot takes over human actions." If that isn't a blunt hit on the 0th law, which was born from the conflict of how one defines humanity based on their morality, then I don't know what is!
And the least remembered law, the number of "fans" that jumped on the Will Smith i Robot film for breaking faith with the stories with its world takover plot. Whatever the faults of the film the part about robots taking over the world or causing small harm to avoid greater harm is right out of the stories.
Also I think they quoted him out of context, I don't remember the author ever believing you should obey the law at all cost especially since he was a Jewish refugee. (I read everything he ever wrote, even the science books and listened to every interview I found). If a character says a bad thing so what?! I have my characters do stuff I don't agree with for conflict and plot development. They had some inaccuracies in the previous ep as well so I'm rather disappointed in them. I'm glad they are doing Asimov but with such a large audience I feel they are distributing false information on a much wider scale than a small channel.
Hisame Artwork. Yeah, his stories included several different types of societies typically and just as typically his lead characters were not strictly authoritarian. Usually pressing against the status quo in one way or another and thus against standing authority.
I think that the problem here isn't that the 50's were perfect but instead the disdain in intellectuals of all time periods towards any older way of thinking. The way that we think today in Modern 1st world nations isn't necessarily the best that it's ever been. To think that way would be to ignore history altogether. Groups of people that most would disdain (Communists, Nazis, Slave Traders, etc.) were once the "Modern" People, yet we see that they ended up going down a terrible path later on (again at least through our lenses). We need to understand that the modern interpretation of the world isn't always right because again, the Nazi's were the modern world view in Germany at one time, Communism was going to be great! (according to the Bolsheviks), and if I could talk to members of past societies about what we have done to modern thought then I would get different responses as well.
The "50s mindset" depicted here isn't racism or classism, it's authoritarianism. The idea that obedience is a virtue. Just because it was prevalent in the 50s doesn't make it racism.
Omar Gameplays Pretty much. I mean, AI or robots, like children, should not be given large responsibility until they can be trusted. We can put them in simulations that they won't be aware of, and if they act poorly we can either not give them responsibility or we can rewrite them. Or continue trying to improve them. We don't give children control of our nukes, and we should not give young, untried, AI the same responsibility - like skynet. If that happens then we deserve whatever happens.
And that was exactly what Sheridan and Delenn told the Vorlons and the Shadows in the climax of the 4th season of Babylon 5. To paraphrase, we're adults now and we need to make our own mistakes. You and the rest of the First Ones can head out beyond the rim now, we'll be fine.
2 things: 1: I can tell the illustrator had a lot of fun with this episode 2: Thanks for bringing up the fact Asimov's 3 laws are flawed. It annoys me to no end that people bring up the 3 laws as a serious solution to the question of Robotic Ethics.
The real "laws" of robotics, Should be ethics of robotics Dont tell a toddler not to do something, explain why that's not appropriate The laws are more like rules to a toddler, telling them no is just begging them to try once to see the results
Well, the laws were never intended to be used, the whole point is that they are simplifying the entire field of ethics down to 3 simple laws and in every single one of Asimov's books, something goes wrong. So by further simplifying these flawed laws, you are creating an even more flawed system
I have a quick solution for those difficult questions, the trolley problem: humans don’t fair better so just average out what we do and leave it at that. Honestly, as long as accidents happens less (which is almost a given), the few times this comes up we’re at least even with humans. Seems weird how we pour so much time into these questions but then hand out licenses after taking an exam. P.S. not saying we shouldn’t ask these questions but that it’s a) ironic and b) shouldn’t stop us from replacing us as drivers
The problem there is that as the tech grows more common the number of cases where ethics need working out will grow too. Also, it sounds like my solution would different than yours. I would just tell the AI to first maximize the chance that no one will die, then if that fails to kill the driver on the basis that the driver is the one who took on the risk inherent to driving cars.
MutantManFish that’s sort of it, right, why should they get sued more for that than the human driver does? I guess the one difference is that the company who designed it is at fault, not the humans.
One of the looming legal issues is precisely one of who is legally responsible when a self-driving car does kill someone. There's also the question of who should decide - should you be offered the choice between the car that might kill you and your family or the car that might kill the kids in the road? Or should the decision be legislated? Or made by the car manufacturers?
Occam's razor will show that the best option is to apply maximum braking in a straight line, disappating as much energy as heat as quickly as possible (where an accident isn't trivially avoidable). Any weird edge cases will be vanishingly rare and the actual savings in life and property will be so disproportionate, people will wonder how we ever dared get into cars and drive for ourselves. From the legal side of things, manufacturers will be indemnified if their cars behave within spec, the minimum of which will be decided by appropriate government bodies just as things operate currently. Financial responsibility will, presumably, be carried by the driver which will be covered by insurance which should be a lot cheaper since most vehicles will not experience an accident over their lifetime.
Chaos Corner Love what you said. I was going to suggest this but my worry is that this system will just encourage manufacturers to shift their responsibility to the driver. There will naturally be a bit of pushback as drivers will choose the car with the least chance of accidents, but I’m just unsure; could go either way. Maybe you can speak into this? Also this level of ethics we bring up with autonomous cars.. if it is so vital, then why aren’t we having this discussion now with human drivers? It’s crazy how much humans are still afraid of machines but weirdly pacified by humans being in charge. And crazy is an understatement.
In my Master's program in Computational Linguistics, I had a class this semester in which we spent the entire time discussing moral issues like the one phrased at the beginning of the video. In fact, our essay at the end of the class had two topics, one of which was constructing an argument about whether it was morally correct or not to build an autonomous car that lies about its rationale for these decisions to encourage people to buy it, despite making strictly utilitarian choices. The class was called "Ethics for Nerds." XD
As far as I know, robots will only do as they were originally programmed for by their own creators terms and standards. If somehow humanity was able replicate and program free will into AI, that would be frightening.
I just finished watching the Matrix Trilogy and WOW, Super cool!! The scary thing is that these problems we WILL have to face, it will happen regardless since this is how humans advance, we just need to make sure it advances in the correct way. It is terrifying because these problems are already coming up in the real world, just in ways we weren't prepared for. For example, teenagers are already using AI to make deepfake porn of their girlfriends, who are all minors, and then these images end up on the Internet as child pornography, which is terrifying. Or another example is militaries around the world are making AI controlled drones, with absolutely no regulation or even legal concepts to understand the results. Our legal system still thinks humanity is in the late 19th century even though things are way different with many conflictions
I never liked the 3 laws and it always annoys me when they get shoehorned into sci-fi works. Too often it feels like those creators are just parroting something they read or heard about while treating it as fact rather than a work of fiction.
Did you already cover Phillip K Dick and I just forgot or is this topic still in the future? Also will you cover Soviet writers like Lem and Strugazki?
Treat robots like we would our children so that they in turn will treat us as their parents when they "grow up"(become superior). It's the same way we (should) treat our elders that we could instill in them. Our elders are often behind the times and may even no longer be productive in anyway yet we still cannot fathom the idea of just discarding them and instead treat them with love and respect.
Asimov’s laws of robotics have had a profound influence on our thinking in terms of AI and the creation of sentient machines. We've had to ask ourselves "what is ethical programming?"
01010111 01101000 01100001 01110100 00100000 01101001 01110011 00100000 01101100 01101111 01110110 01100101 00111111 0001010 01000010 01100001 01100010 01111001 00100000 01100100 01101111 01101110 00100111 01110100 00100000 01101000 01110101 01110010 01110100 00100000 01101101 01100101 00101110 00101110 00101110 00100000 01000100 01101111 01101110 00100111 01110100 00100000 01101000 01110101 01110010 01110100 00100000 01101101 01100101 00101110 00101110 00101110 00100000 01001110 01101111 00100000 01101101 01101111 01110010 01100101 00101110
Extra Credits can your space kitty land on it's feet in zero G?
Let the robots take over.
Of course, I'm not a robot, robot, r-r-r-robot! *glitches out*
Let's just not make sentient machines hey, ....ok?.....ok
Me: "I don't think HK-47 adheres to Asimov's Laws of Robotics"
HK-47: "Affirmation: That is correct, meatbag!"
(fires blaster)
dookie_12 i love this coment!
he just only listens to the third one lol
AllAmerican218 Not even. HK-47 obeys only his master, and even then conspires to kill his master if he deems the master to be unfit, save for his maker.
KheptlaxaXonu Ive never felt 100% convinced with Hk47. It has been on its own long enough it has developed a personality of its own and is no longer bound by its programming. I wouldnt trust that droid further than i can levitate it.
CyborgJiro he is a - psychotic- homicidal assassin droid. He doesnt just kill people for a job, he thoroughly likes it.
The Laws of Robotics, one of my favorite favorite scifi standards. So glad it got an episode dedicated to it. Also, the ending segment was pretty funny.
Shawn Heatherly another situation similar to the car is whether some schmuck breaks into your house with intent to kill.. to kill the man violates the first law, to not kill is to condemn everyone else nearby. 1st law exceptions
Robots in Asimovian works tend to be powerful enough to restrain potential murderers without doing serious harm to them. They are also okay with committing small injuries to prevent larger ones.
OH MAN, I can't thank you enough for pointing at the common misconception that the 3 Laws are the answer, but instead a framework for discussion. Thank you so much!
Of course....
Zoe is the mastermind
_Dan 1 eliminated_
_Humans suspect nothing_
_They now come to me for their history_
_their literature_
_their understanding of the creation of games_
_phase one of humanity’s overthrow complete, moving onto phase two..._
tommy karrick don’t forget politics
I mean, as long as the bot keeps the video quality, I’m down for a robot EC takeover.
I like that part where the robot punches a robot... it seemed darkly humorous
Except the politician hits the "human" who publicly challenges him.
Zvi mur yeah but it's a conspiracy
I am Matthew of Krol. Resistance is futile. Your life as it has been is over. From this time forward, you will service us.
All your base are belong to us.
one thing has been made abundantly clear, as I see it, in all Sci Fi's exploration of Robitics and AI:
the INSTANT we develop a truely sentient/sapient machine of any kind, it MUST be treated ethically and respectfully
Fun fact: Another thing that bothered Asimov was robots written as basically regular humans, for instance a popular series about a heroic robot Adam Link, who'S first story was called, no kidding, "I, Robot". Asimov was actually against his collection being called that, but the editor ignored him.
2:00
Yet, over a half century later, we STILL got the same done to dead unthoughtful outdated cliche that is Detroit Become Human.
David Cage tries so goddamn hard, it's just a shame that he's never any good at nearly anything.
Bit confused on how "Morality is not absolute" follows from "Asimov had a too optimistic view of authority".
me too
Resistance may be futile; but compliance will certainly be rewarded.
That was a fun vid.
Gotta love that 50s mindset. You know, Ignore that Nuclear blast behind us and just have a Martini in front of the pool.
I love these rules if only just to use them as a system to break and find loopholes within.
That’s pretty much the plot in all the stories in I Robot.
Thats the whole point lol
For me, the question isn't "how do I know that I or anyone else isn't a robot?" but rather "if I or anyone else was a robot, would that matter? And if so, how much?" In the Sci-fi classic film Equilibrium, we see a world where humanity has decided emotions are inherently dangerous and requires everyone to take emotion-suppressing drugs to not feel anything; and yet they still say things like "I'm sorry" as if trying to express sorrow or regret, but don't actually. It's not that the writers made a mistake by continuing to use phrases we say in real life despite its context being inapplicable in this alternate world, they were saying "those phrases don't even have meaning in our world". We don't actually feel sorry for people most of the time we say "I'm sorry", rather we _think_ sorry for them, we acknowledge it was a sub-optimal occurrence and that we value optimization.
But is that necessarily a bad thing? Emotions can and do lead us astray all the time, there are many contexts in which being mechanical in decision-making yields better results. But that is not to say that emotions don't have their place, there are many things logic alone simply can't do or is at least very poor at doing, and even the things it can do will seldom turn out a single result. Logic is not at all linear, it bends, weaves, and branches out. We need emotion to act as another layer of the equation. Many contexts will call for more of one than the other, and it's completely okay if certain "mechanical" people are better suited for the contexts that call for more emotionless logic.
My all time favorite book by anyone is Pebble in The Sky. So many social, political, religious, and philosophical insights there.
3:42
I remember there is an unrelated short story "The Humanoids", where basically the same thing happens.
Walpole 1010 ,Walpole 404, HK-47, 14walpole , Walpole 20 resistance as Processor world DBE. 50 mindset speech
It is a reflexion of our own forcamings, traying to come to terms with our future and our past were as in this specific singularet etics and filoosofy get blended.... .
And the laws are bound to fail if done improperly. Take MegaMan it took 400 years for humans and machines to be in true peace(even if they suffer the same conflicts together).
Ironic though if you consider the fact that Zero was not three laws compliant, but by killing Dr. Weil, refusing Ciel's plea to evacuate, and sacrificing himself to destroy Ragnarok, he obeyed the Zeroth law by protecting humanity from a human who wanted to see it destroyed.
We will eventually have to learn how to give oders to computers and robots.
In 'Two Faces of Tomorrow' a new computer system nearly kills several humans because the one giving the orders did not understand what the computer was asking.
The task was to remove a small mountain range on the moon to build a second mass driver to get mined materials to where they were needed. He gave the order with maximum priority and no further restraints, after the computer asked several times for clarification.
The thing was, the computer hit on the idea to bombard the mountain range with a load from the active mass driver. The priority and no restrictions meant the computer could ignore the presence of humans in the danger zone.
They did wonder how the computer could get it done in only 15 minutes, but lunch was more important.
For that story this wasn't even the worst. They got lucky that it happened on the moon. On earth such it could cost thousands of lifes. On top of that the scientists were nearly done creating an even more powerful system to eventually replace the one that was in place for a year or so. And then the speculations began.
Non of the two systems were AI's BTW, just machines that could learn and had the entire net to go for information.
Prich319 actually Zero was beyond even the zeroth law. Have you read his line Zero was more human than even X. Zero did what he thought was right regardless of what people call him. As in the X series sigma called Zero the first Maverick.
And the way they achieved peace was by giving humans robotic parts and giving reploids mortality
The laws are bound to fail if done *properly, that's the whole point. Asimov created the laws to show that its impossible to simplify the entire field of ethics down to 3 basic laws
It could just my modern outlook showing, but I was the only one who thought that Asimov was being sarcastic with that *VERY* 50’s mindset speech about morality?
The gist of that whole speech was ‘shut up and don’t question the system.’ Compared to ‘deconstructing humanity through the laws of robots’ the message of ‘stop worrying and love the bomb’ feels a tad bit shallow, don’t ya think?
if you throw out statism too much, then you don't have a lot of good options-- but the governments and orders you align yourself with need to be conceived as institutions proactively concerned with self-improvement. being intensely loyal to the lawmakers is one thing-- being intensely loyal to a written law even when its own writers are prepared to reconsider it is something else. rules and laws and following orders are very important, and rebelling for the sake of rebelling, even being creative for the sake of being creative, is sometimes too risky in sensitive volatile situations that come up in politics. sometimes you get lucky and save the day in a way nobody else could have imagined, but it's usually much safer to give everyone whatever they're used to handling-- after all, once one person starts acting weird, it makes it much more likely that everyone will start acting weird, which is a great way to start a nuclear war really.
Answer to the first question: Barrel through the children. They're the ones who ran into traffic, and likely will again. No reason to harm people who're not at fault.
Secondly, this was a rather enjoyable video, but I can't help but think that maybe the Laws of Robotics have done more harm than good, as so many people have interpreted them as gospel. Also, kinda sad that the 0th law wasn't brought up, the secret law that supersedes all others and causes a lot of issues.
Law 0. A robot may not, through action or inaction, cause _humanity_ to come to harm.
Meaning that a singular human may be harmed or killed, if doing so preserves humanity. This causes way more problems than intended, and is a basis for "Robots rise up and conquer humanity as we can't be trusted on our own."
I'd love it if you covered RUR
I always wondered what would be a motivation and purpose of robots overtaking humanity
In his late work, like Forward and Prelude Foundation, he suberts the 50s mind set himself. But these were his last novels, published in the early 90s.
'Fallout 4's Danse being an android in power armor, aka a metal man, fighting for 'not human' haters and questioning that was one of the game's highlights. As was the whole Institute/railroad/Curie lines. If the game had just been about that, with you choosing sides (i.e. nix the backstory a bit and the Minutemen being a playable faction) and it would have been a great game about a.i. and how both sides could be the 'moral' choice.
I love you guys and the series is a good intro to Asimov. As a suggestion for improvement, I would like to ask for a "Lies" episodes once in a while to perhaps fill in some of the gaps. Specific gripe: WHERE THE ZEROTH LAW AT? :D
I facepalm every time the car scenario is brought up.
One of the first rules everyone should learn in driving school: "drive in a way that you can safely stop at any time".
I think I've worked out a revolutionary 0th law: a robot may not harm humanity, or, by inaction, allow humanity to come to harm
I know this is a very minor thing but it would interesting if you brought up the early pronunciation of Robot (rowbut) and how it changed to the one we use today.
they protec, they attac, but most importantly, they rise up and overthrow humanity
4:50 I see where came the all Blade Runner topic right there.
I would suggest that anyone wishing to read a book where robots take their service to man kind too literally should read "With Folded Hands" by Jack Williamson
5:30 echoes into 2020. Quite loudly.
My asnwer to that specific question at the start: swerve away and hit another car. Most adults would prefer to save the children over themselves, plus hitting another car is safer because cars have safety features that protect occupants so it's far more likely that fewer people die if you hit another car instead of the two children. You COULD quantify a child's life over an adult's as 1.5x (for example), and quantify the lives of adults inside cars as 0.6x (because they're safer inside the cars). This way it's not that dificult.
Currently cars already do that: generally speaking autos will simply break as hard as possible and hopefully not kill the pedestrians, they might even switch lanes if they don't detect any adjacent cars.
I know that its puched it to the extreme, but I do it to make a point. Now you have made a system that can decide to murder you to harvest your your organ, and feel it was morally correct, becuse it could save more life by give out your organ......
The "easy" moralic way is to follow the traffic rules and if sombady need to be sacrifice it is they who did break the trafic laws....
Simply Brilliant
On 0:45, if you look closely, you'll find Walpole's name. It was Walpole!
what I don't understand is how the second law could conflicted with the third law.
does that mean the robot has to kill itself if it is ordered too, or it does not have to kill itself if ordered too?
The "punching a robot" one was left ambiguous at the end of the story. It was just suggested that that could be how the politician accused could have passed that test.
Asimov, explors the different ways for robots to be more like us, people, thous adding a scene of individuality to this tols, or nachins of our own creatio; to the point of anthropomorphicing this tool (the Bicentenary Man were the robot goes behond his original programing; even exciding its life expectancy....)...
The thing is it would be very difficult to convert human abstractions to code. So, any robot having the three laws would perform in unpredictable ways, just like humans.
I would love to see a book that involves these three laws without any form of modification to them and them somehow uprising.
The smart car question can be solved by the car avoiding the kids and sending a signal to the other cars which would probably also be smart cars preventing any death.
I see extra credita video- i like
I see Asimov- i like
Oh wait, I pressed the like button twice.
Well, i see extra-scifi- I like.
Oh yeah and daleks arent robots btw.
The only real way to determine if somebody is a robot is to give them a lame logic problem and see if their head explodes. It’s absolutely foolproof even into the 24th century.
Great episode guys, but i think you could have used more robots. Extra robots if you will.
If you’re doing extra sci-fi now, you’ve gotta do a series on Godzilla and Tokusatsu
Can't we just... not invent robots?
At least robots that are highly intelligent and able to adapt and interpret. The guys making our cars and coffee are cool.
Here's the thing: as soon as you want to automate something that doesn't take place in a highly controlled environment, or in a highly specialised domain, you start finding that the ability to adapt and interpret is highly useful.
Look at Siri and Alexa - devices whose utility rests on their ability to interpret...
Cheecky Wal Pole in the heart 😎
This is something I never understood, why do people think robots more intelligent than us would waste all the time and resources on a war with humans?
If they were going to wipe us out, it would be infinitely more logical to do it all at once with an illness, or just leave the planet for a new one.
The most logical event would simply be them treating humans as pets, or any other animal that isn't farmed
Come back Dan! The Extra Credits team are replaced by robots! 😆
Damn it, extra sci-fi, Daleks aren't robots!
generally I hate it when asimovs laws of robotics are brought up in media due to the fact that no one in the setting seems to have read anything about what he wrote but refer to his work meaning the things he wrote exist in the universe and no one read it except they did because they are referring to it
0:46 WALPOLE!!!
I love the robot designs
0:46 Walpole
I don't fear the concept of AI overall in fact I could probably argue the notion of AI as a form of evolution, I do however fear the potential circumstances of its creation and "usages" in the context of our society etc. But more specifically I fear the in between technologies of programming and robotics in military and social application, machines don't have to be "AI" by the standard of thinking sentient robot for there to be dystopian levels of mechanised oppression going on, cough drones, cough cough...
Oh, goodness. These arguments began with Aristotle. All of us better up our historical game
5:20 Are you guys going to touch on Anderson and Moorcock, even though they are Fantasy and not Sci-Fi?
Why does the first Contract of the kendra remind me of the Laws of robotics
Me wants an Extra Fantasy
like Extra Fantasy of tolkien lord of the rings and the silmarillion
6:50 just a few facts about you...
1. I only know you from the internet
2. I know that you replaced the voice of a previous narrator
3. said narrator used a voice filter
therefore it is logical to assume that you are the same robot that just upgraded his speech algorithm
No wonder he short circuited, cat has a bowl on its head, had to fight the urge to help its master.
So THAT'S why old-Dan got replaced...
Dan was taken down and now the machines rise
0:46 it was...
Judging from this video, it seems Asimov was on the 4th stage of Kohlberg's stages of morality. One can move up to 6th (or arguably 7th) stage so his ethical side is unimpressing
R E S I S T A N C E I S F U T I L E
Resistance is useless...
Melkor 237 I invoke the first law on thee! Begone
RESISTANCE INTENSIFIES!!!!!
[image of resistor connected in parallel with short circuit] *Ha- Ha- Haaaaaaaaaaaaaah-*
Lower your shields and prepare for boarding. Restiance is futile.
The Dalek reference was cool... but they’re not robots.
AnimeOtaku2 no?
ANDREY MARQUES nope, as someone said below they resemble something like an octopus or jellyfish. Watch the episode Daleks in Manhattan.
ya daleks are like brains in a suit the think you see that looks like a dalek is more akin to power armor
They're more like cyborgs or aliens in a vehicle
The Daleks are not robots, but they are a race that was genetically engineered by one person (Davros). So in a way they are "organic robots" in power-armor, with a genetically ingrained sense of superiority and instinctual genocidal hatred for all other lifeforms.
I, for one, welcome our extra credits team overlords.
Vensey Ness yes
Robocop is not a machine. He's just a human with a high-tech, full body prosthesis... I think that was the whole point of that movie.
True, but he does still have programming that controls his actions. That's where Asimov's influence comes in.
@@magnusprime962and if he showed us anything
You cant program control into something, you can program a response. But ultimately, it's just a game of wits
"Secret Directive 4: cause no harm to Corporate members"
Mission: get them fired first
I love Asimov so much the questions that he poses and how he poses them have lead me to think not just about the nature of robots if they ever come into being the way they are envisioned in his stories but about humans and how ultimately our flaws not our strong suits are what make us who we are so then the question becomes can we truly create something superior to us or something without flaw when it is ultimately flaws that give someone character and perspective? won't the robots just turn out to be flawed beings like us just with different origins and compositions?
TheCreepypro MegaMan is a very good game series to look into for that perspective(X onward).
I think it's impossible to create a being without flaws for the same reason you cannot achieve the Borg dream of perfection: sooner or later, to gain any kind of advantage, you must accept the corresponding flaw.
If you are intelligent, you run the risk of acting arrogant. If you're charming, you may sometimes manipulate others without intending to. If you're selfless, you might neglect your own needs to the point you deplete yourself and subsequently cannot help anyone.
Dispassion gains reason but appears heartless. Pragmatism can be hard-edged and exploitive. Idealism can border on naivete. All of these things can be advantages and disadvantages in turn, depending on circumstance.
The flaws we'd percieve on robots are that robots can't judge in human morality.
A robot (AI) is designed to accomplish a given task as efficiently as possible, thereby taking actions that we think may be inhumane.
The standard isn't "without flaws", just more effective than us meatbags.
As for morality... Again, humans kind of suck at it. I'm far from certain that some of the "nightmare scenarios" Asimov and others played without aren't actually more moral outcomes even though our human intuition/heuristics reject them.
TheCreepypro 😖 Failed! You have no notion of PUNCTUATION in your comment!
a good way to avoid being surpassed by robots is becoming cyborgs
Transhumanism is a whoooooooole other branch of Sci-Fi that's produced a fair bit of good reads in it's own right.
Transhumanism is not just a branch of sci-fi. It's also a philosophy, and a movement.
And a way of living for the future. A jubilee for human kind and the working non super rich man (they own the rest of the Machines) wich is the world
Megaman ZX had that as a background where humans gain cybornetic augmentations while reploids gain mortality. It was a way after the time of the Zero series to give both man and machine a way to evolve together instead of at odds.
Brotan Vivos I shudder to think of a future where your arms and legs are owned by the Company, and can be remotely overridden if necessary. Or your heart or lungs...
Walpole? @:46
0101010110001100101WALPOLE011010101000101010111001
Don't forget the 2 at the end of the third line.
Shh it’s advanced binary
It was just a dream, Bender, there's no such thing as two.
and on the row dust below 01011 ....or DID IT ....
You are incorrect. Resistance is Voltage divided by Current.
Seems legit.
Also, makes me wonder: Is there a story where robots take over humanity but then are just completely benevolent and basically keep humans as pets because they find us amusing?
I wonder about this. I feel like the leap from 'we're going ot act on our own and not be your actual slaves' to 'therefore we will exterminate you' is taken for no real... actual... reason? Genocide isn't actually anything resembling a rational or logical response, so even if a machine doesn't have any empathy protocol or anything, why would it ever decide that was a good way to spend its time?
3 for 2 discount at the human store!
What? No love for the 0th law? That was the most fun/interesting one
shadowrain1024 they actually DO hit on the 0th law in this episode: to paraphrase: "to save mankind from it's own undoing, the robot takes over human actions." If that isn't a blunt hit on the 0th law, which was born from the conflict of how one defines humanity based on their morality, then I don't know what is!
ger du. No, it doesnt. At least not in the books.
And the least remembered law, the number of "fans" that jumped on the Will Smith i Robot film for breaking faith with the stories with its world takover plot.
Whatever the faults of the film the part about robots taking over the world or causing small harm to avoid greater harm is right out of the stories.
Also I think they quoted him out of context, I don't remember the author ever believing you should obey the law at all cost especially since he was a Jewish refugee. (I read everything he ever wrote, even the science books and listened to every interview I found). If a character says a bad thing so what?! I have my characters do stuff I don't agree with for conflict and plot development. They had some inaccuracies in the previous ep as well so I'm rather disappointed in them. I'm glad they are doing Asimov but with such a large audience I feel they are distributing false information on a much wider scale than a small channel.
Hisame Artwork. Yeah, his stories included several different types of societies typically and just as typically his lead characters were not strictly authoritarian. Usually pressing against the status quo in one way or another and thus against standing authority.
OOF that 50's mindset
Jobin 👉
Axolotl de la Animacion classism forever
I think that the problem here isn't that the 50's were perfect but instead the disdain in intellectuals of all time periods towards any older way of thinking. The way that we think today in Modern 1st world nations isn't necessarily the best that it's ever been. To think that way would be to ignore history altogether. Groups of people that most would disdain (Communists, Nazis, Slave Traders, etc.) were once the "Modern" People, yet we see that they ended up going down a terrible path later on (again at least through our lenses). We need to understand that the modern interpretation of the world isn't always right because again, the Nazi's were the modern world view in Germany at one time, Communism was going to be great! (according to the Bolsheviks), and if I could talk to members of past societies about what we have done to modern thought then I would get different responses as well.
The "50s mindset" depicted here isn't racism or classism, it's authoritarianism. The idea that obedience is a virtue. Just because it was prevalent in the 50s doesn't make it racism.
O Racism is definitely part of it, it just not the type of active racism people think of when they talk about racism.
Protocol 1: Link To Pilot
Protocol 2: Uphold The Mission
Protocol 3: Protect The Pilot!
*Vanguard class titan~ BT-7274*
Godamnit, I did not expect that titanfall reference...rip my eyes...
It's two in the morning but I don't care. I'm watching this.
In truth, the story of robots outliving the human race boils down to the simple story of a child outgrowing their parents.
Omar Gameplays Pretty much. I mean, AI or robots, like children, should not be given large responsibility until they can be trusted. We can put them in simulations that they won't be aware of, and if they act poorly we can either not give them responsibility or we can rewrite them. Or continue trying to improve them. We don't give children control of our nukes, and we should not give young, untried, AI the same responsibility - like skynet. If that happens then we deserve whatever happens.
Rich people outgrowing the common man with the Machines he built
No... the children in this metaphor are the robots
I couldn´t have put it better! Bad parenting has its consequences.
And that was exactly what Sheridan and Delenn told the Vorlons and the Shadows in the climax of the 4th season of Babylon 5. To paraphrase, we're adults now and we need to make our own mistakes. You and the rest of the First Ones can head out beyond the rim now, we'll be fine.
2 things:
1: I can tell the illustrator had a lot of fun with this episode
2: Thanks for bringing up the fact Asimov's 3 laws are flawed. It annoys me to no end that people bring up the 3 laws as a serious solution to the question of Robotic Ethics.
The real "laws" of robotics, Should be ethics of robotics
Dont tell a toddler not to do something, explain why that's not appropriate
The laws are more like rules to a toddler, telling them no is just begging them to try once to see the results
I, for one welcome our new cat overlord
Dalaks are not AI they have little creatures inside
Hippocrates little creatures? Little disgusting octopus aliens is what they are! Sometimes human too.
They show one in the startof the animation with like "code" showing from behind
I feel like these laws could easily be replaced with the single moral philosophy: "All humans must flourish".
Well, the laws were never intended to be used, the whole point is that they are simplifying the entire field of ethics down to 3 simple laws and in every single one of Asimov's books, something goes wrong. So by further simplifying these flawed laws, you are creating an even more flawed system
I have a quick solution for those difficult questions, the trolley problem: humans don’t fair better so just average out what we do and leave it at that. Honestly, as long as accidents happens less (which is almost a given), the few times this comes up we’re at least even with humans. Seems weird how we pour so much time into these questions but then hand out licenses after taking an exam.
P.S. not saying we shouldn’t ask these questions but that it’s a) ironic and b) shouldn’t stop us from replacing us as drivers
The problem there is that as the tech grows more common the number of cases where ethics need working out will grow too. Also, it sounds like my solution would different than yours. I would just tell the AI to first maximize the chance that no one will die, then if that fails to kill the driver on the basis that the driver is the one who took on the risk inherent to driving cars.
MutantManFish that’s sort of it, right, why should they get sued more for that than the human driver does? I guess the one difference is that the company who designed it is at fault, not the humans.
One of the looming legal issues is precisely one of who is legally responsible when a self-driving car does kill someone.
There's also the question of who should decide - should you be offered the choice between the car that might kill you and your family or the car that might kill the kids in the road? Or should the decision be legislated? Or made by the car manufacturers?
Occam's razor will show that the best option is to apply maximum braking in a straight line, disappating as much energy as heat as quickly as possible (where an accident isn't trivially avoidable). Any weird edge cases will be vanishingly rare and the actual savings in life and property will be so disproportionate, people will wonder how we ever dared get into cars and drive for ourselves.
From the legal side of things, manufacturers will be indemnified if their cars behave within spec, the minimum of which will be decided by appropriate government bodies just as things operate currently. Financial responsibility will, presumably, be carried by the driver which will be covered by insurance which should be a lot cheaper since most vehicles will not experience an accident over their lifetime.
Chaos Corner Love what you said. I was going to suggest this but my worry is that this system will just encourage manufacturers to shift their responsibility to the driver. There will naturally be a bit of pushback as drivers will choose the car with the least chance of accidents, but I’m just unsure; could go either way. Maybe you can speak into this?
Also this level of ethics we bring up with autonomous cars.. if it is so vital, then why aren’t we having this discussion now with human drivers? It’s crazy how much humans are still afraid of machines but weirdly pacified by humans being in charge. And crazy is an understatement.
In my Master's program in Computational Linguistics, I had a class this semester in which we spent the entire time discussing moral issues like the one phrased at the beginning of the video. In fact, our essay at the end of the class had two topics, one of which was constructing an argument about whether it was morally correct or not to build an autonomous car that lies about its rationale for these decisions to encourage people to buy it, despite making strictly utilitarian choices. The class was called "Ethics for Nerds." XD
I had never gave the laws of robotics much thought. This opened my eyes.
As far as I know, robots will only do as they were originally programmed for by their own creators terms and standards. If somehow humanity was able replicate and program free will into AI, that would be frightening.
I just finished watching the Matrix Trilogy and WOW, Super cool!! The scary thing is that these problems we WILL have to face, it will happen regardless since this is how humans advance, we just need to make sure it advances in the correct way. It is terrifying because these problems are already coming up in the real world, just in ways we weren't prepared for. For example, teenagers are already using AI to make deepfake porn of their girlfriends, who are all minors, and then these images end up on the Internet as child pornography, which is terrifying. Or another example is militaries around the world are making AI controlled drones, with absolutely no regulation or even legal concepts to understand the results. Our legal system still thinks humanity is in the late 19th century even though things are way different with many conflictions
“A robot must not hurt humans in any way...”
most children: 😢
I never liked the 3 laws and it always annoys me when they get shoehorned into sci-fi works. Too often it feels like those creators are just parroting something they read or heard about while treating it as fact rather than a work of fiction.
Can you guys do an episode on Afrofuturism in your history of sci fi when the time calls for it?
Did you already cover Phillip K Dick and I just forgot or is this topic still in the future? Also will you cover Soviet writers like Lem and Strugazki?
You seem to have a problem confusing cyborgized humans with robots, Extra Credits.
the AM disagrees with your assumptions on the 3 laws of robotics
Treat robots like we would our children so that they in turn will treat us as their parents when they "grow up"(become superior). It's the same way we (should) treat our elders that we could instill in them. Our elders are often behind the times and may even no longer be productive in anyway yet we still cannot fathom the idea of just discarding them and instead treat them with love and respect.
I LOVE ASIMOV! I, Robot is my favourite novel. The movie was shit.