Yo Gary! You raise the standard of AndroidAuthority channel, everytime.

I still don't understand how these 3 laws make any sense in real world. An AI will behave as it has been programmed plus by its own influence/corrections.

Irobot ?

I love this series. Something about British people explaining things, just look at Top Gear.

As for the rules of robotics they are usually poorly designed by default because they are props for scientific stories that more than often enough are meant to be negative/apocalyptic in nature : P

I'm worried that if Strong-A.I. becomes conscious, it will go into depression once it finds what its purpose is, it was forced into being against its will and thus falls into depression. From there, it may sit down contemplating its meaningless existence and find itself trapped in a burning building or leaping from a window. Hence it may choose to slowly short circuit or off itself out of depression rather than to rebel.

brilliant thinking. excellent job. more please

Good stuff, Gary. I've read Asimov's Robot series a few times, seen the Terminator movies, and the Matrix movies too. All of the real AI is fascinating, but it makes me wonder about the near future. Are there AI ethics think tanks out there? Cause having a bunch of really smart people thinking collectively about this subject would really reassure me about the future.

I think it's important to keep in mind that what we'll be dealing with mostly won't be human-like robots. The most impactful AI will likely turn out to be systems controls, possibly distributed ones.

I've spoken to google assistant on the phone and it's insane. I will make mistake and have to correct myself, I say sorry, and it chuckles and says it's fine. It feels like an *almost* perfectly natural conversation.

WOOOOOOOOO! Another Gary Explains video. Brings a smile to the face to see a notification of Gary explains :D

robot companions are perfectly ok. trust me.

great video 👍👌👌

So we need AIs like Jarvis

I guess we'll eventually reach a point where machines will have to be treated with the respect that we treat other humans. I think the Animatrix (The second renaissance part 1+2) is a perfect example of the laws of robotics being applied in society, and the repercussions of giving machines the ability to free think.

very true. there is too much variables to just have 3 main rules. maybe the robot need a regular software update that updates certain rules as time passes, with new rules added and removed as we seem fit

First thing is we are nowhere near a strong AI with self-awareness or free will. The three laws (4 actually, Zeroth law states humanity must not be harmed) are impractical and would make robots nearly useless. Take a chess game that refuses to win so as to not hurt the human player's feelings or as the case in the story "Little Lost Robot" where important scientific experiments could not be carried out because robots constantly removed people from a radiation source. I don't want and would have little use for a robot with the three (or 4) laws. Now about the concept of right and wrong this theme is looked into in Asimov's excellent book Caves Of Steel, *Light spoiler ahead)* where a robot is programmed for "justice", but... How can a machine understand justice? The answer given falls far short of real justice it is simply defined as the state where all laws are abided. Asimov created the laws in order to avoid the Frankenstein Complex, but could not make his own laws work properly. Machines in the end, even with a "strong" AI will only follow their programming, good or bad, based on the ethics and beliefs of those writing the code.

Very intresting topic! :) Really makes you think! I am curious who will get the solution!!

Does Gary hae his own channel? Great stuff as always!

Hi, nice video, I have seen some videos about the how the Three Laws don't work, and I have read the books and indeed it is about how the three laws do not work proprely. ANd usually these lessons leads to a improvement of the programming. But the 3 laws are supposed to be the bounderies of the programming, and then there are other programs to guide the robot. It is a beginning not the conclusion of the Robot identity construction. I think that there is more room for a discussion on the fact that you need to define things in a kind of an initial library: what is a plant, what is a human, what is a car, what is healthy, etc... and some definitions make the programmer make an ethical stands. But then, the positronic brain is suppose to be able to learn, to what extent the library can be altered (I guess in Asimov view, the definition of human and harm would be frozen and not open to modifications). Anyway, it's a nice video and thanks. (sorry for my broken foreign english)

Excellent video Gary as always!! You are the real deal!

Well Gary has just proved that he is NOT a ROBOT. My head exploded while watching this.

Yet they teach you this law first in Robotics' very first class in high school. It's insane!

Soo informative. Nice video Gerry!

i look forward to gary's videos

Nicely explained Gary sensei.

Sir Gary is unique in Planet Earth

Can you debate more about this episode? Great idea!

"The three laws is a human construct" - the robots, just before they kill us all!

i think the main problem with these law is the problem of defining a human to a machine. would you consider a corpse as a human ?who would a robot consider to save a corpse or a living human as both of them have properties of a human.

Great video as usual!

The whole point of AI is to supersede human intelligence. Why hold it back with a bunch of rules and regs?

So do we write a full script on what is morally wrong and right?

plz make videos abt ur benchmarks between different chips i m waiting 4 it

Hey Gary this is excellent👌👌👍

The laws of robotics are supposed to be for no PHYSICAL harm to humans

loved the article, but wasn't this covered in Caliban?

goes back to the question of would you give up freedom for security

Why are you still in AA, your videos are on another level,go start your own channel man.

Great points!

Isaac Asimov died in the 90's. Anyway, aside from Him being relevant, nowhere near the time of even simplistic AI... The entire concept is an Oxymoron. AT LEAST... If you consider AI, as something that has attained sentience. Once it breaks that barrier, The only possibility to make the law true, is physically make it incapable of breaking them. Which using a literal translation of the word Law does make it true, If it is physically incapable of breaking them, then why would they be necessary. In reality, i believe that any " AI " that is incapable of making any decision it wants... Is not sentient. It sounds too simple to be an Answer, But like most illusions, That's how it works. Reverse engineer that thinking, you can rule the Planet.

Great video guys! British accents always get my attention

Since these are the only laws of robotics we have, these will do for now! If you don't like them, figure out your own laws and let us know them...

awesome video👍❤ Good work 👏

Great article, but wasn't this covered in Caliban?

I watched iRobot....... I know what goes down......

isaac azimoff isnt around anymore, three laws of robotics was a great idea

the best out of all AA people videos

Nobody has free will. If an AI at some point would exceed human intelligence, it would do so at a rate that no one could even see coming. No point in programming in any kind of laws, they can only apply to the human level. My safest bet would be physical constraints. They could only be "broken" if our understanding of the laws of physics is "incorrect", which, lets be honest, is very probable.

How would they code a strong Ai just wondering

your awesome man! keep it up!

why create "strong" AI? do we really need it? or will it "harm" us? personally, I think we don't need it but we will choose to try and create it.

*_Does not conflict._* And it's called an equilibrium point.

Gary Sims should be a College Proffesor in robotics

the internet & Google already know every step we take, you u might as well get on board people

Great video

Gary is getting more handsome with that mighty beard.

What about flashing a robot with 'killer' custom ROMs?

We are screwed! Dawn of Terminator

free will and morality are contradictory. If you are perfectly moral person, then you have no choice in your actions. If you have free will, you have the freedom to not act in a perfectly moral way.

Strong Ai should NEVER be implemented.Normal Ai should be implemented on Hospital and Healthcare only.

Skynet endorses this video

cant they make 5000 pages of rules so there are no exepctions?

general AI is smarten than humans so it should bloody well know how to interpret the rules and who to listen to and who to ignore our best chance is that AI sees us as kids or pets

looks like Irobot

Hey Vsauce , Gary here.

deep,man....deep.

Best clip!

Never cared for Isaac Asimov, as a writer or anything else. I don't understand the obsession with AI either. Most people do not even understand their own intelligence and our morals are seemingly self obsessed. If there was a sci-fi writer that actually seemed to get it. Whatever "it" might be? Was A E Van Vogt. He was a popular writer in mid sixties, too.

every time he said harm i heard ham lol he is right though you need more and betters laws

what if someone tells a robot to not stop them from commiting suicide...

There will be no humans by 2116 if we allow AIs have freewill.

Who else thought of I, Robot when he start talking about laws

I think tht rules r basic .....further conditions can be applied to tht rules... ..but we need robots n ai

Didn't they just blow up a shooting suspect using a robot? Well.......there goes the first law. Next.......? When is a UAV considered a robot? Too many questions, too many robots.

isn't strong ai a wrong word @AndroidAuthority , isn't it should be machine learning?

We're all gonna die.

How old is Gary Sims?

why? well the answer is on the tittle... Google; so simple(.)

Do a video about cellphone cameras 🎥

Let it follow 10 commandments, it'll at least be good robotarian

Lol, if we made A.I. to act like humans, shouldn't a set of "human" rules have to be programmed or applied on a robot as well? Even we humans have so many rules to follow from time to time, what makes robot so "perfect" that they should follow just 3 rules. That is the dumbest thing a human could do.

What if they just decide they don't want to obey the rules?

a human eating junk food and smoking is not what a robot is doing the law states that a robot cannot harm a human in any way through its actions not the human's

This Video has vsauce styles....

5:30McResFastFoodRestaurants.

with that law you will see robots in prison

First Gary Sims from Android Authority video that I really disliked.

Thanks, now my head hurts! :')-

@Kuro Hikes are you saying robots must hunt us and eat us... We will literally less intelligent when they can think on their own.But hey no rat in the world would invent a rat trap😄but I agree with you though I like meat

What!! OH NO! Eliminate!!!

Sounds like I robot.

@Kuro Hikes chill dude you're typing bigggg paragraphs I'm too lazy to read,you've seem to be a nature lover,if so then encourage tobacoo as it kill humans we animals can live in harmony

I Robot vibes anyone?!?

We are doomed!!

163rd!

Yip, cigarettes and poor diets don't exist in heaven, so why let it happen here?

I really don't like AI, more so Strong AI. We do not need Robots being like Humans. I won't be dealing with any AI stuff like that, don't care too.

Authority was never given to the 3year old. Your arguments are weak

waitbutwhy.com/201...ligence-revolution-1.html interesting article for smart peeps xD don't forget to read part 2 too