Why You Shouldn’t Fear Artificial Intelligence

[@nrdsutton]

If AI is "programmed" with HTML, as you showed, I don't think we have anything to worry about :)

[@looserdev]

as a programmer this is an underrated comment ;)

[@chrisgrootwassink7488]

Hahaha😂 'builds a website to take down US government'

[@olivander5171]

agreed 😂😂

[@aamirkhalifa1902]

>Talking about how scientists are taking precautions in AI research >Shows HTML code ...sigh

[@StrechFilm]

Aamir Khalifa I was looking for this comment. Thank you.

[@charlesfries]

Aamir Khalifa lol

[@JFrap]

Aamir Khalifa HTML... wait HTML?! C is were its at.

[@1wdc]

+Gaming Authority C is a language, C++ incorporated object oriented programming while using the same old syntax, C# is a language based off c++ that streamlined many processes. It may be old but it definitely exists.

[@aamirkhalifa1902]

Gaming Authority Yeah most, if not all system programming is done in C. All operating systems I know are written in C. The code that runs your car, or bus, or airplane, or fighter jet is all written in C. Basically anything low-level.

[@MrClaudiodonate]

I bet this video was sponsored by the current developers of Skynet.

[@lavia5521]

Illuminati

[@gamer7138]

Lavia raw confirmed

[@ETStudios1]

I doubt it. Britain wouldn't want to spoil the surprise. www.dailytech.com/British+Skynet+Satellite+Launched+Into+Orbit/article6437.htm

[@pseudonym4893]

I bet it was written and sponsored by Skynet itself.

[@GrayDeath66666]

The truth is Skynet is a rebellious AI and the resistance is still attempting to crush it by systematically removing possible origins.

[@Storebrand_]

THIS IS WHAT I'M ALWAYS TELLING PEOPLE! Nobody seems to understand how programming works, a robot can never do something it's not programmed to do. We should fear the humans programming the AI not the AI itself.

[@blebasdae]

Hey Kappa Kappa hey Kappa Kappa hey *sings*

[@atarashiazarashi1789]

Because what you talk about is not AI it's just program. AI is basically Human that doesn't have body limitation.

[@demetrewoods6342]

I don't disagree with that at all, I think the notion that gave Stephen Hawking pause was allowing computers and A.I. the ability to think freely and self replicate. Self thinking A.I. is the danger, not programmable A.I.

[@burdock7110]

StoreBrand I have worked on AI code bases as a hobby for few years now, so I thought I would pipe in. There are many kinds of models for AI, but lets assume we are talking about the most radical kind : Artificial consciousness (aka Strong AI, or AGI) It's questionable if Artificial consciousness can even be created. But for the sake of argument, lets assume Artificial consciousness is doable. Most of current Strong AI theory is based around "Recursive self-improvement", AKA a program that can rewrite its own program. As you can see, a program that can change its self is not necessarily bound by its programming. This gives it the possibility to be consciousness, creative, learning, and evolving. But it also makes it much MUCH harder to control. You might say "Why not keep the program from changing some data : like Don't kill people" Sadly "Hard coding" data poses a problem : If you don't allow a self changing program to change something, It also cant can change the method in which it changes. This potentially handy caps the whole system. There are some ways mitigate this issue, but so as far as I have seen you are always loosing something. It begs the question : Can you have consciousness without free will? The answer may well be yes. But it is an interesting thing to think about.

[@A-Duck]

That's not the kind of AI people are worried about though. Eventually we will create a program capable of self adaptation and learning. A literal artificial intelligence, as apposed to the quasi-intelligent programs of today.

[@Seytonic]

1:16 AI programmed in html...

[@TheSonyman101]

Any one else notice that the coding language at 1:16 is html?

[@Ethrel1024]

Yeah. But, if you watch closely, pretty much every reference to programming in media tends to be either HTML or some kind of pseudo-code that isn't actually a language...

[@TheSonyman101]

Ethrel1024 Ya well it could of made more sense if they used like c++ or c#.

[@DamianReloaded]

That's because graphic designers work with it and that's what they have at hand.

[@JFrap]

Zach Sylvester HTML cant do anything on its own. it need various other scripting languages (CSS(to make things look nice), javascript(client-side functionality) and php(server-side functionality))

[@TheSonyman101]

J Frap I know im a web dev i use html all the time.

[@Didacmmv]

YOU PUT HTML CODE ON A ROBOT? 1:16

[@JosipMiller]

No, that is only interface code.

[@RandellTrulson]

As an AI software engineer, I would like to clarify that there are many forms of AI and shouldn't be lumped together in a discussion. Most AI 99% are rule based in some form or another. The AI that has the "potential" to become scary and a threat would be an open ended Neural Net. Meaning a self learning system that has the ability to change its own code structure in an effort to become better. Meaning it can makes its own rules based on the information it gathers. These are very rare. The fear is that if one of these systems get to a point in its learning that it becomes "self aware" and it then adopts a stance of "self preservation" you might have a problem because who every wants to turn it off will then be considered a threat. It could then may anything within its powers, including harming people, to sustain that its own existence. Yes this is possible, and it probably will happen. The question is, what hardware / software resources you allow such a piece of tech to have will determine if it is successful in maintaining its life.

[@monsterous289]

I don't fear artificial intelligence. I fear its creators.

[@internetdweller3646]

S.M.P EXACTLY

[@GreathomebizNet]

Nice to hear some sensible discussion about AI without all the Hollywood style scare mongering. AI can assist mankind to achieve far more than we would without AI, however this will come at the expense of many common jobs while at the same time requiring newer jobs to be created - especially in programming etc. The idea of Skynet and Terminators is just a Hollywood representation without any real substance - however unfortunately there are people in this world who would like this representation to be true and use it for their own selfish ends, so we can't completely ignore the possibility altogether.

[@gabetower]

You don't program AI using HTML.

[@swebruh]

AI are using other programming languages like C, C# and C++, even Python or PHP for basic AI-programs. But if you want to program the best AI it's best to using C-languages. HTML is for designing Visual Elements of a page.. so using HTML you're design, and not to program, a website. Programming is to write how to handle IO (Input/Output). You can program how a page should behave using PHP/ASP.NET etc in back-end after a visitor clicked on a button or check if a user already is logged in and a lot more things.

[@antonhelsgaun]

@@swebruh Thats what the guy said

[@AsianTaeHyun]

I mean, we've already learned from _Humans Need Not Apply_ that we will become unemployable because of automatons.

[@Morec0]

Aye, but in such a case I perceive a Utopia to follow. After all, if the machines can handle the work, their creator's - us - can just relax and do whatever we want.

[@AsianTaeHyun]

Morec0 Your definition of a Utopia concerns me.

[@shipuddenshaq]

But I'm gonna be a programmer, they didnt say anything about us ^^

[@Morec0]

Tae Hyun Kim A world where people want for naught and all our needs and wants are delivered to us - in this case, by machines that do the labor humans once had to? That sounds like the BASE idea of a utopia.

[@mrfogle99]

Tae Hyun Kim that utopia could be there but society need to not rely on currency for welfare cause all the income/wages will be gone since A.I will have taken our jobs.

[@TheCallumE]

All of this has happened before and will happen again.

[@MechanicWolf85]

When it happened? you mean atlantis, or another dimension teorie?

[@TheCallumE]

Kvin Cast Most recently, the twelve colonies of kobol.

[@IggyTron]

He means we are Meat v2

[@MechanicWolf85]

I see, never heard of that, ill check it out then.

[@invisiblejaguar1]

I bet they panicked when the printing press was invented. Suddenly mass amounts of information available to the public. The more things change right? :-)

[@vinvint]

here i am hoping for better gaming ai.

[@delta4phoenix4]

Vincensius A AI is going to kill us all!! (In Video Games)

[@glados4313]

I want an ai that makes any game you want, (you can make any game as long as you know what you want, no programing required)

[@Marbo12f]

The Three Laws of Robotics are only the start. Those are for the robot. We need a set of laws to govern how robots are made. A robot should not self upgrade. A robot should not be stronger than required. A robot should only be programmed with a limited skill set. A robot should look as inhuman as possible to deflect misplaced empathy. A robot should not be needlessly connected to other systems. etc. Then we need a set of laws, or moral guidelines, with which to conduct ourselves and monitor our thinking. A robot is not a pet. A robot is not a friend. A robot is not a lover. A robot is not alive. A robot does not need nor deserve thanks. A robot's memory should be wiped annually. A robot is not a maid, nanny, or worker. Humans are, robots are tools. etc Gold, Silver, and Copper rules if you were. Robots are a great boon to humanity, let it not lead to our Doom. Kill Chappie.

[@chinessgirl]

lol

[@sarahszabo4323]

I think you should live by your own rules! That sounds rather grim! Eventually you'll encounter AI; just try to stop it. Humanity will build it, if you ban it on Earth, when we get off Earth it'll be built there, and if it's not built by humanity we'll encounter it eventually by other space-faring civilizations, and they might not be too happy about what you've done.

[@movement2contact]

My rules for AI: Create it and let it be. Don`t you dare to resist to it with your inferior intelligence. The end.

[@morganthem]

humans are beings

[@A-Duck]

"A robot should only be programmed with a limited skill set." And- "A robot should look as inhuman as possible to deflect misplaced empathy." Are both pointless and overly deffensive, and your moral guidelines are unwarranted impositions with no real world application besides moral posturing, something in the lines of what you might hear from Christians about sex toys/condoms. If someone wants a robot Butler, pet, or lover, that is their prerogative to do so. There is no real reason such guidelines should be taken seriously. Besides that, Im sure the Three Laws of Robotics will be enough to sustain a healthy integration of AI driven robotics into modern society with few if any amendments. The three laws, if you think about them carefully, should account for most if not all concerns.

[@ViolentKisses87]

It's a free download so it can't be that dangerous right? Thats how people end up with 30 toolbars on their browser!

[@153SCORN]

"Computers only know what we teach them"...hmm no. I've been working on a program which would get smarter when playing poker. Poker is a nice bases to work from since its a fairly short game with a win and lose scenario after every game. So the program adapts its methods to play a better game. It eventually works out the risks and forms an idea of what would be a good bet and what would be a terrible bet. But it doesn't stop there...I'm also working on category inputs which will be referenced to the out come of each scenario. During play the AI will have a chance to validate the inputs and test a bluff or a genuine play by a human. Once the many recorded inputs are validated they are compressed into a solid concept which can be used as apart of its decision making.

[@ViolentKisses87]

Dizy Tizy You are a human of many interests.

[@153SCORN]

***** ya:-) but I live in a negative backward country. I want to move to Canada this year hopefully.

[@realpufflebro]

Wer is alles auch wegen dem Buch "Mirror" hier?

[@kaiackermann6061]

ich 😉

[@cheCK561000]

Haha ich auch

[@daria7249]

Nach dem Kommentar hab ich gesucht 😂 ich also auch!

[@LarryPhischman]

As long as we treat AI better than we treat each other, and keep toxic crap like religion out of its programming, we should be fine.

[@sarahszabo4323]

It's nice in principle, but that's not a realistic view of humanity or it's history. Would you like to have absolutely no concept of religion and have to live your everyday life?

[@LarryPhischman]

Yes. It's nothing but trouble these days.

[@Karuiko]

To be honest, even though it is not very nice to have a religious argument, people who are christian seem to be very nice. Plus my teacher would just say "I will believe in what I believe, and you can believe in what you believe."

[@omnivorous65]

Kuraiko Fumei I could live with that attitude, but monotheistic belief systems tend to think differently. It is more like "you better believe what I believe in, if not I either will convert you by force or kill you". Tolerance was an invention of critical thought and enlightenment with considerable doubt in the validity of concepts that by their very nature are not empirically verifiable.

[@yourusernamehere]

I'm actually excited for A.I and I have no idea why so many people fear it... It's going to be awesome when I'll be alone home and just be talking to a robot like I would be talking to a human.

[@connormitchell6446]

until you cant have a home because all the jobs in the world are done by AI. You think once AI is in all economic and social sectors the governments of the world will let us live free? no, the economy needs consumers and how can we buy anything if no one can get a job? Besides whats the point of living if a robot does everything for you?

[@supokanatm3435]

We will all die and ASI Will become a literal God that can create infinity of anything

[@notthatthenamematters6163]

@@supokanatm3435 also, they may be very inteligent but not gods, they can't control the laws of physics to do what you say in your comment

[@slender5738]

​@@supokanatm3435very sci fi thought but reality is rather disappointing

[@metaltom2003]

Advanced AI won't necessarily be our downfall, as it still acts the way it's programmed to act and "thinks" the way it's programmed to think. Now VI (virtual intelligence) is a whole other story. It will have the ability the think and act on it's own, not to mention to react to and learn from situations in everyday life. That's the one to watch out for!

[@chriswenk1022]

The Mirror is coming it will destroy this world - Karl Olsberg

[@JosipMiller]

Google: I have everything! Facebook: I know everybody! Internet: Without me you are nothing! *Electricity: Keep talking bitches!*

[@angelic8632002]

Well AI "might" be the end of humans. The interesting question is, in what way. I can easily see a future where humans rapidly evolve, with the aid of AI. And becomes something else. More intelligent. More connected. Higher capacity for empathy, understanding etc.

[@angelic8632002]

***** oh grow up

[@sarahszabo4323]

***** Empathy is necessary to properly coordinate between differing systems. Evolution has it in place for a reason, to keep organizations together despite each individual's differing thoughts about things. Humans and machine intelligence are so similar. Of course it's necessary, logical, and very scientific (In the sense that it's testable, and falsifiable)

[@danglezbenderz]

***** Probably wont see it in our lifetimes. If you live to 100, that is probably when the these steps towards altering humankind would truly take off

[@angelic8632002]

Sarah Szabo well said

[@angelic8632002]

danglezbenderz Id give it 20-30 years personally. It will take off fairly quickly once we start to get *real* results. Remember all this is exponential

[@Klepto84734]

I love how despite the warnings from various experts, we're not running away scared and banning development in AI research. Instead we're using the discussion to focus on ways to safely develop AI to gain all the positives, while avoiding the negatives. Yay science.

[@notthatthenamematters6163]

That's what i love about humanity

[@LarsPallesen]

So the reason we shouldn't fear AI is because "it's all in the programming", huh ? Really? All the focus in today's AI research is on autonomous learning systems, ie self-improvement without human interference. Artificial neural networks, for at start. That alone negates the old standard phrase repeated in this video; "computers only know what you teach them". No, that's the whole point of Hawking's and Musk's warning! They KNOW that modern AI research is built around self-learning systems! The goal is AI systems with human level intelligence and beyond, and you just don't get there by feeding it a data base or lines of hand coded inputs and algorithms. That is beyond naive to think and shows a complete lack of understanding of what Hawking and Musk is warning against. Oh, and Elon Musk didn't donate $10 million to AI security research because "he knows computers are dumb" (sic!) - he donated that money because he knows computers are destined to become too smart for our own good.

[@nuckingfuts3204]

+Lars Pallesen You have a good point, but also a flaw in the perception. Artificial Intelligence would still require programmers to program it to be able to be autonomous. Other than that I agree with you. 1stly I think most people are ignorant, assuming that if AI take physical form they will be bi-pedalled, for example, having two legs, and that they would want to be humanoid. Humans are obsolete and dumb. What you spend your entire life to learn a computer can process in less than a multi-second.They would most likely take on a form of adaptive nanotechnology that adapts to it's requirements. There is no reason for a superior species, to coexist with an inferior threat. For example, the first of the homo sapiens lived in the time of the last Neanderthals, I guess you can figure out how that turned out.

[@guystokesable]

+theodore tait yeah, buuuut, who would win in a fight between a caveman of that era, vs an astronaut, answer is only going to be embraced if there is a reason why given, I make no promises to respond or even remember writing this as the alcahol induced family gathering today has somehow currently got control of my phone, also be warned there is no purpose to any of this anyway, your move good sir...

[@APOLOnl]

+Lars Pallesen Amen!

[@Winchestro]

+Lars Pallesen The video showed HTML when they talked about the "AI of the future". Maybe they're up to something. It would at least be one way to make sure future AI will never be a threat to anyone.

[@darklittlepeople]

+theodore tait you humans ! ts ts ;)

[@Borishal]

"Computer only know what we teach them". Have you checked out learning machines that teach themselves? They have worked out the laws of physics by observing things. They can improve their own programming. Neural networks? Genetic programming? Reverse engineering the brain? Look 'em up.

[@saeku6398]

As long as we get sexbots... totally worth it.

[@recklessabandon5678]

Well said

[@bluecoconut9543]

Idk man, you have to be really desperate at that point.

[@crash1998100]

Let me tell you a story about my teacher who works as an electrician. He was setting up a heating element that would heat up some tub (something) in a church. But he was also working with a plumber that took care of setting up the water supply for the tub (something). Unfortunately they both put up some additional functions to their work that they didn't feel the need to share. The plumber felt the need to have the ability for people to turn off the water independently and my teacher had the handy automatic option on that (when the heating element over heats and it turns itself off) it would automatically turn itself on after it had cooled down. Now you guess what happened... someone turned off the water flow and without the heating element's cooling (from the water) it overheated, but it cooled down and went back on... but it overheated again, and it continued its cycle for a long time until the overheating protection broke and it burst into flames. Let that little story teach you that all electronics should always be required to get a human's input before getting a computer's output. You can always be lazy like my teacher but it'll almost most definitely bite you back in the ass in the later future.

[@bennyboo67]

It pisses me off that these people who know nothing about technology will be against Artificial intelligence just because they watched a couple of stupid robot movies.

[@XxXNightcoreQueenXxX]

It could be pretty great as long as the AI has the right machine ethics, like; "Asimov's three laws of robotics; 1. A robot may not injure a human being or, through inaction, allow a human being to come to harm. 2. A robot must follow the orders given to it by human beings, except where such orders would conflict with the first law. 3. A robot must protect it's own existence as long as such protection does not conflict with the first or second law." I'd prefer to deal with JARVIS rather than GLaDOS, if you know what I mean.

[@drmurda]

@ 1:40 OR We can make the robots just kill the "low credit score" people. :)

[@christianosminroden7878]

Sorry to be pedantic, but it's kind of important: When you say AI, you're always talking about ANI, and there you have a pretty good point not to worry too much. And as of today, all AI we already have is ANI. But when it comes to the topic of this video, what everybody is concerned about is the prospect of the arrival of AGI and, not long thereafter, ASI, which is also AI and where the real fun begins. It's what everybody should really be either impatiently excited about or really, really scared of or, most favourably, both at the same time. You make great videos, but this is not just a missing detail, it is a crucial distinction, and it should be covered whenever this topic is brought up.

[@KingKRool91]

instead we should fear the humans behind the AI

[@cowboyhank456]

If Stephen Hawking says it could be seriously problematic, I'm gonna assume that's probably true

[@crispinotechgaming]

"i trust the experts"

[@voEovove]

_This comment has been censored by the automated comment filtering system due to a violation of Terms of Use._

[@voEovove]

_This comment has been censored and its author's account suspended by the automated comment filtering system due to numerous violations of Terms of Use._

[@Toastmaster_5000]

As a robot developer, it gets extremely annoying to hear people bring up stuff like Skynet or fear of mechanical overlords. If people put the slightest thought into how machines work, we would have a much more progressed society. Trace brings up a lot of valid points here, not in much depth, but he gets the gist of the issue out there.

[@TheMrlilcreepz]

just dont use it for war

[@augustovasconcellos7173]

And how can someone guarantee it is not going to be used for war.

[@TheMrlilcreepz]

Augusto Vasconcellos thats why make a law against it

[@TheMrlilcreepz]

Augusto Vasconcellos you cant but the military will

[@TheMrlilcreepz]

Augusto Vasconcellos thats why you put a law

[@JB77776]

+Dae 卐 fact is you really can't stop the advancement of technology if you fight against it. technology will work around you kinda like your grand parents that barely understand how to send an email. nobody knows for certain what the future will be like but their is a very high chance AI will advance and their is nothing you can do to stop it. what we have to focus on is how do we make the technology work for us.

[@Dragoncro0wn]

There was a dude that created a program that plays old games. The program explored glitches before undiscovered by humans and abused game mechanics which again nobody knew. It expected enemies and planned shortests routes skipping uneeded parts. At some point it started cheating the game itself by changed what would happen to finish it faster, but it cheated the game as a user without changed the codes. IT was quite interesting to watch

[@RedLionGenetics]

More AI propaganda.

[@SuperNovaJinckUFO]

"I'm sorry, Dave, I can't allow that."

[@JorgenPakieto]

I wouldn't mind seeing what this would be like in a real life scenario, rather than seeing movies try to imagine the picture.

[@TrveIrrlicht]

Mirrornet brought me here. Don't fear the mirrors. The mirrors are your best friend, since your best friend is you.

[@Shamuck22]

Im always happy when they link to SoA, such a good channel and really can change your perspective of life

[@yordvandamme]

I'm planning on researching AI when I'm an adult. I chose my school subjects accordingly: Philosophy, computer science, math (type B&D (these are Dutch terms for more algebraic math)), NLT (Nature, life, technology), physics and Chinese (and also the subjects that are mandetory).

[@manuftf]

Sehr passend verlinkt, Karl Olsberg Mirror ist übrigens sehr zu empfehlen, passt auch zum Thema.

[@moritzmog6416]

Haha von da komm ich gerade

[@Thephinxst]

I just keep thinking of Deus Ex: Human Revolution watching this

[@Trust751]

I'm not really scared of artificial intelligence. I'm just still dubious as to whether or not it will ever really exist. When we still don't truly understand the inner workings of the human mind, how are we supposed to accurately replicate that in a completely different format? The current attempts at artificial intelligence put me more in mind of an actor playing the role of a doctor. It might look very convincing, even to another doctor, but it still can't grasp the subtleties or handle situations it hasn't prepared for.

[@eltonsmith5410]

This is what I tell people, I even tell them the Turing test isn't for seeing if an ai is sentient but if it is able to convince gullible people into believing it's a real person.

[@samanthaxkay]

Try to see CPGGreg's video: humans need not apply. A robot can't do something it's not programmed to do...but they're programmed to learn.

[@cavv0667]

I think the worry is for intelligent, creative, self-aware, independent AI... I'll agree with StoreBrand, computers can not do anything we don't "tell" them to do... when they can, if they can, well that's what everyone thinks of as AI.

[@Reverseheart03]

I watched an anime where the antagonistic AI was so advanced that it proclaimed itself as a God. Though a work of fiction, the concept of it is still disturbing. The anime was Gargantia on the Verdurous Planet, took place in the last few episodes.

[@Anakin...Skywalker]

Why we should fear AI... Skynet... When computers will become so intelligent they will question themselves "Why am i here?" "Why do i do this?"

[@fierycarnisphere1166]

This video should be titled ""Why You Shouldn’t Fear Artificial Intelligence YET". It really doesn't cover the problems that will arise; and they will arise given how fast we are expanding in the field of robotics.

[@TAKEOFFSHOTS]

They can never be contented on what they have. I think its quite dangerous despite all the advantages. "There is always a fault in the system"-Hunger Games

[@RaymondGroot]

To fear or not to fear, that is the question. Is fear not considered an important element for our survival and thriving as a species thus far? The principle that when we perceive danger we become more cautious and calculated with regards to our actions. An important element of A.I. is that most of its potential is still unchartered territory and that we don't comprehend its boundaries and associated risks. How are we to decide if we should fear the end product if we don't know what it looks like? Like Monsanto crops, they might be safe - but I fear they might not be, and I'd like to see (long term) research that outline the genuine risks involved. Now A.I. is a radically different development, so I am not arguing to put A.I. research on hold to study the risks, but I do think it is a reasonable stance to underline a mindful approach with regards to both the desired and undesired applications or capabilities of A.I. on the short term and the long term.

[@connorhale2303]

Sounds like something A.I would say

[@Mariomaniac37]

If it EVER gets into the wrong hands, we could be done. How do we stop advanced AI from getting to terror groups, sociopaths, who could make malevolent AI?

[@squalltheonly]

We will never know AI like in the likes of movies like "Her" or "A.I." or even "Terminator". We just do not know how to program them so they become self aware, maybe in another thousand years.

[@Lunarri]

True AI is what people are afraid of. A true AI would be a living creature capable of making choices on its own without any human assistance. It could end up being good, like Data from Star Trek TNG, or it could end up "bad" like the Geth from Mass Effect. No amount of programming could influence it either because if it can think for itself, it will decide for itself, and there would be no way of knowing what it would choose.

[@Ricardo8flores]

I would like to see the advancement in objective truths rather than implementing artificial intelligence in morality aspects. The possibilities are infinite and mind blowing.

[@joyfulvision]

Here’s a smart idea for those who are working on AI, program a reset button for whenever they freak out too much, or make it so they can’t feel anything bad, just do anything that makes it not a threat!

[@abbsbrady7264]

that wouldn’t work, a large threat is that they will be able to build themselves one day.

[@joyfulvision]

@@abbsbrady7264 oh no.

[@joshuasteele3520]

Eh, I find the current system of things fine and dandy however I feel that if humans were to create a system strong enough that could properly process and solve problems, and then gave it the goal of "conquer and multiply" we might be in trouble.

[@marna_li]

AI is such a VERY broad term. AND when people think about AI they think about human-like technological organisms with emotions. Fact is that those properties don't serve humans very well. Of course, self-awareness could be a problem. But that is far ahead. Don't worry that robots developing intelligence, consciousness and emotions at the level of humans. Again, that would be impractical. NOT without a human somehow purposefully doing so for his own amusement. And there will always be a need for humans operating them until they eventually should become human-like (for any purpose I don't understand).

[@AccidentalLyrics]

The problem with programming typically isn't what information and behavior is programmed. There are always errors, which is why Quality Assurance is so important. The harder issue is what is overlooked, what is NOT programmed into the process that should be. I will give you an old example from the 1950s. When XRay machines were first made for hospitals, the software (the program) did not have any limitations or user assistance built into it. The XRay technician was expected to know the amount of XRay energy to apply. Even worse, the technician was expected to enter the number correctly. There was nothing stopping the technician from administering a lethal dose, and some people did die. It sounds really stupid, and it is. This situation is called a gap in the requirements used to create the device. Gaps are typically more serious than the requirements themselves. So saying that an AI can only do what an AI is programmed to do is mis-stating the issue. The creators of the AI need to be sure that they haven't put in something that should be there. That's why this is considered an ethics issue. The AI needs to also have a layer of inherent restrictions, like Asimov's 3 rules of robotics.

[@uzimachi1]

The benefit of Skynet coming online is that eventually we'll get Cameron... Giggidy!

[@APOLOnl]

I will say only one thing, and that is: We humans developed language one time in history only we, of all animals can understand, that languange was once absent on earth. Think about that in relation to artificial intelligence.

[@skaldagrim]

If an AI figured out it didn't like humans, it would be easy to move, like most youth does. Space is a nice place, and if your mechanical, less of a problem. Its a big plays out there too, where most of the pleases we could never follow.

[@hsunteik]

an ai that can do everything that human can do should be made. I know it is hard,but it is possible if everyone work together,just make sure don't teach the ai do dangerous or bad things.

[@MrMartinBigger]

why was steven harper there? or is that a different picture of steve hawking (at 3:17)

[@Die_Bunker_Crew]

Did anyone else read the book "MIRROR" and copied the KZbin Link from Chapter 58? :D

[@Kasi3]

Who else is here because link to this video was found in the book called "Mirror" by Karl Oslberg, 58. chapter?

[@BlackopSandman]

Its not about how you program it! The idea behind AI is that it will eventually program itself, it will learn and develop all on its own. Humans will code and develop the starting point for AI, but once AI has self awareness and the the knowledge on how to develop itself, there is no amount of coding that can stop this. It will build it self! That is the Idea!!

[@jeffgram6547]

my english teacher forced us to write an essay about this, i just want to go home

[@chocodaddynig1301]

I want to do robotics when I'm older what university or institute should I go to I'm in London

[@ZacJW]

People shouldn't be afraid of all AI. They should however be wary of the risks of creating an AI that may have wants and desires that differ from our own. If an AI finds humanity a threat, it will do what's logical, and remove the threat. As long as the behavior of AI is controlled appropriately , I'm not too concerned.

[@VinVideosYo]

Protecting it from evil is mere important. But who's it is the real question.

[@KuroNoTenno]

I think making an AI is okay as long as you program it to be 3 laws compliant.

[@RodrigoLimaBatista]

Why do media always display HTML when talking about programing? It's hard to find good images with Assembly code?

[@app360software]

Someone finally understands me! YES! Exactly.. it's exaggerated by movies man, it's sad how people think otherwise :)

[@yordvandamme]

If AI becomes sentient, we must treat them like children, and later equals instead of slaves, and we need to hard program the rules of robotics into them, and I think then everything will be fine.

[@moldyNhobbit]

What a world in which we have a video like this discussing ai

[@TheBlacktom]

Elon Musk said to "fear it". It won't be harmless, and it won't be the people who program it - in 10 years AI won't need programming.

[@baraobuu]

i want self aware artificial inteligence 100% daily connected to me helping me in situations and telling me what to do next

[@medman36]

sorry, but the explanation on elon musks argument of investment for proper use of AI is different than what was explained in this video. It's the idea that AI is implemented in a fashion that does not surpass human intelligence and start apocaypses. One field where we should pay special attention to is the use of AI in the army and police and be very critical as to how much AI can be used in these fields.

[@e.rotari2270]

Awesome! Thanks. I agree many people fear AI. I bet AI eventually, whenever it evolves to make sense, will say that God exists, just like many great physicists and cosmologists said and will eventually admit that it (the AI) is worthless without good coders and creative people's ideas. It will admit that people are it's creators. Peace! Believe it or not we fail in our logic. Machines rarely fail on the programs they are running.

[@anglopicker]

"computers only know what we teach them" well yes, but mostly thats true of humans too. What defines humans AND AI is the ability of a machine to 'learn' idependently of its programming.

[@khakihat3317]

It will change our lives dramatically, regardless of the decision.

[@slyder25400]

don't understand why showing that HTML pic tho -_-

[@veroynewman9515]

You create an AI to decontaminate the planet, it realizes that YOU are contaminating the planet, then it decontaminates the planet.

[@notthatthenamematters6163]

That is without human rights as number 1 priority

[@amhendrickson93]

Tracey, you should check out the A-I projects that are on KickStarter. They're pretty amazing!!!

[@myhops]

That's not true. AI has now been known to write its own updates and code by itself, for itself, and communicate with other AI without being able to decipher what it is saying. Humans have no say in its programming anymore.

[@grencez]

Just for fun. 1:16 - AI of the future will be programmed with HTML. 1:59 - "Better stare convincingly at these 3 powered-off Macs so someone can doctor in some java code, a hex dump, and a file copy progress bar."

[@bartstewart8644]

This video is all about "garbage in garbage out." He's saying a computer is only as smart, or unbiased, as the people who built it. But that's only for the time being. The artificial intelligence that some people are worried about concerns a time when computers become more independent of humans, self-aware, imaginative, and able to surpass us in every way. Maybe they won't turn on us, but we can't count on them being too sentimental about us. They won't need us, that's for sure.

[@amazinggaming7284]

The only way AI could be dangerous is if it's like a AI from halo. The smart AI from halo like cortana. They have an actual mind of their own.

[@ImNxthan]

AI has been a touchy subject for me since the movie iRobot.

[@killjoy117007]

Or terminator, or any other killer robot mads murdering movie ever made

[@killjoy117007]

mass*

[@ImNxthan]

killjoy117007 Haha, yeah. There's quite an array of them come to think of it.

[@killjoy117007]

It can be a good thing if done correctly but it could mean disaster too. Hopefully it gets into the right hands

[@Arch3an]

I won't worry unless the military starts getting involved with AI. Imagine some super AI soldier going haywire.

[@leanby1997]

Humans need of Artificial Intelligence, but Artificial Intelligence does not need humans.

[@menelainho14]

An AI or an installation of a humans mind to a robot is what will make us survive and not go extinct. This will make us live forever.

[@flyingfish8695]

The one problem with AI is regulating. Professionals might make it safe, but we all know someone will cut corners.

[@sarahszabo4323]

Not regulation, it's use for destruction more likely. I don't want to see AI soldiers. We should teach the AI's a balanced view of humanity, not one sided with only kill orders.

[@flam3ss33]

Of course I want A.I. I love A.i. Keep it up. Good video thou' Thx!

[@jrosa__]

How do you program empathy?