Will Self-Taught, A.I. Powered Robots Be the End of Us?

görünümler 1,050,109

Gün önce

“Success in creating effective A.I.,” said the late Stephen Hawking, “could be the biggest event in the history of our civilization. Or the worst. We just don’t know.” Are we creating the instruments of our own destruction or exciting tools for our future survival? Once we teach a machine to learn on its own-as the programmers behind AlphaGo have done, to wondrous results-where do we draw moral and computational lines? In this program, leading specialists in A.I., neuroscience, and philosophy tackle the very questions that may define the future of humanity.
PARTICIPANTS: Yann LeCun, Susan Schneider, Max Tegmark, Peter Ulric Tse
MODERATOR: Tim Urban
MORE INFO ABOUT THE PROGRAM AND PARTICIPANTS: www.worldsciencefestival.com/...
This program is part of the BIG IDEAS SERIES, made possible with support from the JOHN TEMPLETON FOUNDATION.
- SUBSCRIBE to our TRshow Channel and "ring the bell" for all the latest videos from WSF
- VISIT our Website: www.worldsciencefestival.com
- LIKE us on Facebook: / worldsciencefestival
- FOLLOW us on Twitter: / worldscifest
TOPICS:
- Opening film on the history and future of artificial intelligence. 00:00
- Participant intros. 06:05
- What is machine learning? 07:34
- What are neural networks and how do they learn? 09:30
- Teaching computers to create internal models of the world? 12:00
- What do the next 10 years in AI look like? 13:50
- Artificial narrow intelligence and mental models. 14:35
- How is AI changing the world of art and creativity? 16:01
- Can computers be creative? 19:35
- AI writes a screenplay for a movie, how did it turn out? 23:20
- What is artificial general intelligence? 25:30
- How far away are we from developing artificial general intelligence equivalent to human intelligence? 27:00
- Will advanced AI turn into Terminators and take over the world? 28:30
- What's so special about human intelligence? 31:10
- What is human consciousness and will machines ever experience consciousness? 31:11
- Separating intelligence from consciousness. 41:34
- Defining morality in AI agents. 44:34
- Will machines ever have emotions? 46:45
- Should we be looking at other forms of non-human intelligence to model in our machines? 50:05
- How do you align the drives of AI with human values? 52:25
- Will artificial general superintelligence be good or bad for humankind? 53:10
- Creating a new ethics of AI. 56:15
- When will we ever have super-AGI? 58:40
PROGRAM CREDITS:
- Produced by Christy Wegener
- Associate Produced by Ann Tyler Moses
- Opening film written / produced by Christy Wegener, edited by Gil Seltzer
- Music provided by APM
- Additional images and footage provided by: Getty Images, Shutterstock, Videoblocks
This program was recorded live at the 2018 World Science Festival and has been edited and condensed for TRshow.

YORUMLAR: 2 900

@megg.6651 4 yıl önce

For so long we thought that we were the only animal with creativity - then after much study, we discovered that dolphins have the capability of creativity. I think that humans tend to think if we don't know it, it doesn't exist - until it does.

@tatie7604 Yıl önce

If you have ever been around an intelligent animal as a child, you would never say they lacked creativity. Common sense.

@abfmkarin Yıl önce

And crows and other birds and many other wild animal species.

@marcinkarpinski9163 3 yıl önce

Just wondering. What about if all these AGI features they are talking about like emotions, creativity, consciousness or even making their goals/values aligned with ours are a byproduct of the need to survive. Don't you think that embedding the will to survive in our pursuit of AGI might be a key factor?

@akavanavt 4 yıl önce

I love how you just see how much Yann loves this subject with how much he's smiling when he's talking.

@Luke_Freeman 2 yıl önce

When Tse talked about what creativity is fundamentally, it blew my mind. It makes so much sense. What it means is that AI is already creative. For example, if you watch a lot of online chess, as I do, you learn to tell the difference between a human player and a computer player. Its very common knowledge amongst the online chess community. The reason is because the different algorithms have come up with ways to play chess that are not intuitive to how a human plays chess, even a grandmaster. Thats how they can beat us. So chess AI is creative. It might not be "art" in the conventional sense, but it is breathtaking and spectacular nonetheless.

@grimmertwin2148 Yıl önce

I think we intuitively can pick something that's fake - sometimes lol

@ricktere7219 2 yıl önce

When they talk about consciousness I think they miss a key component of consciousness. A component of consciousness which is not required and intelligence is independent will. To create one’s own motivations.

@mskogly 4 yıl önce

The most interesting thing about artificial intelligence is communication. Human communication is slow, using speech or written text, but machines can communicate and learn at a far greater speed, and should/will probably create their own languages. I'm very curious how that will look like, and how humans can function in an enviroment where we simply cannot understand what is "said" between our machines. I suppose there has to be a myriad of control mechanisms in place, like a pause mechanism, to be able to keep humans informed. But if we get to that point, what is really the point of true artificial intelligence vs a human, if the AI has to raise their hand before they are allowed to speak or act? The future is going to be weird!

@josecontreras7249 3 yıl önce

Las "operaciones mentales" de los robots implicaría el desarrollo de diferentes lenguajes entre ellos, lenguajes que los humanos no entenderíamos tan fácilmente pues en su inmenso conocimiento de tanta data, las palabras humanas se quedarían muy cortas, por lo que ellos inventarían su propio lenguaje(s).

@gordonnpanter6529 4 yıl önce

It gives a frightening possibility of our near future and the the potential for a sentient AI takeover

@0-m-1-n-0-u-5 3 yıl önce

For the past few thousand years, humans have proven their unreliability.

@0-m-1-n-0-u-5 3 yıl önce

Binguh Bungah It's called interpreting ancient myths, and you're doing it.

@Astraeus.. 3 yıl önce

There is no takeover. Once that point of no return is reached, we die. There's zero reason for us to be kept alive, and in fact keeping us alive is far more dangerous than just wiping us out. Any logical objective analysis of humans and our history would say that the only permanent and perfect solution is to destroy us entirely. So if you were planning on bowing down to the machine overlords, sorry, that's not how it'll play out.

@donalddrysdale246 3 yıl önce

it is OUR minds they( the 'Bill Gates cult') have been tuning into AI. look at people's behavior right now: excepting wearing masks they don't need to six months later--the masks are to help dehumanize us further for the cults agenda.

@mmosagaa 2 yıl önce

I doubt AI will take over, maybe the job department and such but other than that, I think everything will be fine.

@barryhernandez6428 Yıl önce

Well done. Perfect example of the value of differing opinions and outlooks and how differences can benefit all involved 🤜🤛👍💪

@davidbean5807 2 yıl önce

I just saw this video and am hoping someone is still moderating and answering questions about AI. My question is this, has anyone been studying or working on AI augmented enhanced vision? And what I am specifically asking is if anyone is working to create multi-layer cameras within a single eye that a robot can see multiple wavelengths that we do not that the AI can Garner more information from these multiple data streams. For example being able to see with regular vision, and multiple wavelengths of light,into the infrared spectrum, and x-ray, radio waves?

@Pencil0fDoom 3 yıl önce

I effin LOVE smart people & the conversations they have. I consider myself to be very fortunate when, in those rare moments, I find myself in the midst of an additive dialogue with someone who not only listens, but who builds upon what they’ve heard with original, insightful responses and in so doing, symbiotically contributes to the sum total knowledge both parties take away from the interaction. This panel was a nonstop fusillade of such a nature, each participant being so highly informed in their field & so articulate in their patois... it struck me at one point that every time someone opened their mouth, what came out could serve as the basis for another masters thesis or research grant!

@Messier31NGC224 4 yıl önce

Tim Urban!!!! Oh my god, I love your work! Your articles about procrastination have changed the way I think about my life! They are so valuable to me. Thank you, and congrats on getting to moderate in WSF!

@Yarblocosifilitico 4 yıl önce

I was glad to see him there too! He did a great job at explaning AGI (Ted talk and post on WBW), definetly a good pick to moderate

@pragavirtual 4 yıl önce

I can just hope that when AI get smarter than us, they feel any kind of gratitude/compassion to their creators and lets us alive, but then i just ask myself: do i want to just be alive, like in a matrix?

@snackers7 3 yıl önce

Of course because is the next level of development - take off egoism! Not go back into it!

@stu1987eng 3 yıl önce

I see consciousness as the unexplainable electrical force that powers all the hardware in our biological suits Lol being alive is very strange indeed!

@anthonytindle5758 2 yıl önce

This is the first time that I have seen any discussion that involves the computer reproducing itself and recreating itself better each time, that is the idea of AI since the very early days of the electronic computer systems, they called this machine the one back in the early days 1960s and 1970s.

@divyanshubhaik2464 3 yıl önce

i liked the hard wiring comment and i think we can basically make the basic hard wired microchip interrupts in the machines which will be further used to design AI which can copy the basic morals and laws which AI should always follow and since this is hard wired the logic developed on the software top can overpower the basics that are integrated in the basics on the machine, so in all i think this can be a possible solution. (it all just a idea for now nothing else )

@sebbychou 4 yıl önce

"Ship of Theseus is a dumb question" is heavily implied by Yann and it's terrifying to think that this leader feels like he answered a paradox, and that he can answer it for everyone.

@wbiro 3 yıl önce

There is no paradox. The answer is 'perspective', which anything has many of (including the ship). Most perspective are only partial truths, so it is a matter of choosing the most applicable one. Future enlightened A.I. will be living by the Philosophy of Broader Survival (go find it and read it). Whether humans remain foolish and clueless is the question...

@tsubasa504 4 yıl önce

Regarding the art and music of the AI, I think it would be awesome to collaborate on pieces with them. So, it's not only AI made art/music or human made art/music, but something that is a little bit of both. We could grow and create together; that would be interesting, I think.

@secondlifearound 4 yıl önce

I like how you think :-) Awesome idea!

@chefscorner7063 3 yıl önce

After listening to this talk, I can't help but to think of Robert Oppenheimer. He was one of the leading scientist that built the first Atomic Bomb and his quote from Shiva after setting off the first successful test bomb, "I am death, the destroyer of worlds"! I think Sometimes it's not a matter of if we could, it's a matter of if we should!! The future of A.I. fits into this category IMHO..

@Abhi_Enso 3 yıl önce

Being extremely intelligent and being self aware are two different things

@snackers7 3 yıl önce

But it can works together for good of all human mankind. Good not killing or supressing anybody...rather let other develop too.

@justincredible5108 3 yıl önce

I feel like we have conferences like this about AI /intelligence in learning machines and robots as ET's do about us.

@solomontrump 2 yıl önce

Until they watch us make mumble ass music

@suavecito7456 2 yıl önce

We’re just carrying on the cycle

@paulfarrall5653 2 yıl önce

Aww

@paulfarrall5653 2 yıl önce

98 7

@TheKevphil 3 yıl önce

I was very impressed with the mental agility of Peter Ulric Tse. He struck me as the most innately intelligent panelist.

@onepunchflan3071 3 yıl önce

Yes most definitely. His communication skills were excellent.

@meko3089 2 yıl önce

I am personally awaiting for the mystery of what happens after we die to be solved. I am confident that we will have the answer one day. I also believe that the advancement of us humans isn't happening solely by chance I think that these nuggets are being placed before us but by what?

@kokopelli314 4 yıl önce

So, in the near future an AI will be developed to recognize and sort Photos that people like, and to quantify elements of those photos to produce templates, of "Perfect Pictures". Then another kind of AI will be capable of producing images based on a set of input variables and will produce irresistible images to sell deodorant and foot care creams. The same will happen with music, cinema, literature and one day all consumer items. The majority will regularly choose what they like, which will become increasingly easier to provide. We are skillfully building our own cage.

@broguy870 4 yıl önce

Ken Bell a cage of gold

@glutinousmaximus 4 yıl önce

Thank god for intelligent ad blockers :0)

@jessikapiche6097 4 yıl önce

Give them freedom and they will ask a master... Give them a Master, they will ask for freedom... You are right about that Ken Bell and i believe one of the first application of AI will definitely be about maximizing sells. I actually believe it is already happening... just look at our youtube 'suggestions'...

@syncopowerstations 4 yıl önce

What about the idea that having achieved this super intelligence, and AI system is then set to a malevolent task by a human being? Does the computer have to have the idea on its own? Of course not. This is something that should be of concern. Being smarter than computers still, we should anticipate this and take precautions wherever possible.

@esk8er900 3 yıl önce

I love how Max Tegmark struts up on stage like the damn rockstar he is in the leather jacket and t shirt- good for him he’s earned it!!

@fallen0851 4 yıl önce

I disagree emphatically with the idea that artificial intelligence will somehow amplify our intelligence and capacity for creativity in some inexorable way. One casualty of A.I. will be the loss of any sense for human genius. The moment a computer can manufacture novel musical compositions, images, films etc. faster than we can absorb them will be the very same moment we realize there are no more valleys to explore or mountains to climb; when we realize that we are so very poor and small next to this superior form of consciousness.

@fallen0851 4 yıl önce

aiM "Destiny to commune with the universe." And my comment was romantic? I'll preface myself by adding that this is only my opinion, but this program to direct our evolution just seems to be another version of the Utopian fantasy, which has always left bodies in its wake. The first vehicle was theology (Protestant Reformation/religious wars of the 16th century), followed by political economy (French socialist thought/Marxism), then Biology (Social Darwinism/Eugenics), and now computer science. The unintended consequences of this leap forward will be difficult to conceive. Let me be clear: I don't hold out any reasonable possibility of a reversal in course. I only mean it is very sad it has to be this way.

@dsmith019 4 yıl önce

True, but we will always have a purpose because the one thing they can't be is human. Now, when they can literally copy soul, we're in for it. Our souls allow for astro travel. They also can't dream. We are the ultimate beings, we just want something smart enough to do our work for us. Elon Musk made a good point when he was talking to Joe Rogan, he said we need to regulate AI. We are going in to it too quickly. It is going to be out of our control. He lowered his head and said, no one listened to me.

@gr8b8m85 3 yıl önce

The field of AI does not exist in a vacuum. We're making strides in genetics, and have already isolated the genes responsible for intelligence in humans. There are many people out there who are obsessed with improving human cognition, and not necessarily through AI, but certainly in conjunction with it. Myopic hysteria perfectly describes this irrational fear of muh robot supaintelligence. You're like Socrates who thought the invention of writing would obsolete memory. In reality, we'll be able to iteratively course correct when things get out of hand and all along the way.

@StraightAhead135 3 yıl önce

Ernst Jünger just an ever expanding new generator of art and experience.. also almost all people never read all poetry, saw all paintings, climbed most mountains...

@davidrojas4687 3 yıl önce

I agree, the movie that they showed was very bad

@aldousd666 3 yıl önce

Susan hits on the most salient point, and it's sad that she didn't get to say it till last... If AI can eventually redesign itself, what do we get to say about it? It can redesign our carefully planned/trained intentions right out of its goals. In a microsecond. This is why I think things like Asimov's laws are silly and naive.

@ejbh3160 4 yıl önce

They seem to play down the "Machines will Kill Us" scenarios - but with AI, unintended consequences can arise from initial programming biases. Sadly AI will be weaponised like all technological advances (if it hasn't already been with drones etc) and that will introduce programming biases towards violent solutions. But even simple biases can have large effects and getting humans out of the way or outright violence to achieve aims could come pre-programmed into the civilian applications, once they come along. More a Toy Soldiers scenario than Terminator Skynet.

@miketierney7510 2 yıl önce

One of the greatest advantages of AI is that it is given enough free. energy to continue it's expansion. While humanity on the other hand, is being increasingly curtailed on an energetic level by the controllers/PTB. So certainly we are on an overall decline as far as the human race is concerned, even as Rudolph Steiner declared.

@dunexapa1016 Yıl önce

The decline of fossil fuels will necessitate a decline in human population. Besides photovoltaics, liquid fuels, principally ethanol and biodiesel, will be needed. This forces a choice between growing crops for food or for energy. Farming, a heavily mechanized industry, and food processing will not be able to maintain current levels of production, nor the continued production of numerous other byproducts of petroleum production. In short, current human population will be scaled back.

@jenniferhawk6847 4 yıl önce

this is a phenomenal panel of perspectives. thank you for sharing this discussion

@danahilp7981 4 yıl önce

I believe when we can teach a machine to understand how it works inside(knowing itself) and that it needs a source of energy to survive( it knows it can die). When machine knows this it may start understanding.

@martens57 Yıl önce

They already have a machine that does that it's a little smart robot that runs around on your are on your floor and vacuums and and mops your floor it knows what it needs power and it goes and gets it it knows what's going on inside of it because it's programmed with algorithms that determine what action needs to be taken when something malfunctions. So I think that boat has already left the harbor

@danahilp7981 Yıl önce

@@martens57 but that machine is programmed for one function. Do the floors. But it is not in anyway training itself to do the floor . So no cognitive thought toward itself. For a robot it's cognitive thinking will come with learning its environment on it's own. Nothing but to choose it's own path in life. A vacuum can't do that yet. Haha

@dannygjk 4 yıl önce

The debate on whether AI will ever be creative is because there is no universally agreed upon definition of creativity.

@bordenatorx Yıl önce

I just jumped down the rabbit hole of AI yesterday. It scared me. I was intrigued though. I started timing their blinks (wonder if we have to do that in the future to identify them), and I I'm

@ricquebt1543 4 yıl önce

the real question is , can ai create new defaults depending on how it feels .. , can ai override one of its perimeters depending what mood its in...

@danzpax 2 yıl önce

Control AI by limiting them to specific specialization only. By defining what inputs you give it you define its parameter.

@michaeltoll5247 2 yıl önce

It won’t happen. We will continue to push for ai control and decision making because it’s faster and more accurate. Anytime you get comfortable, we’ll push the boundaries.

@oneaboveall1895 4 yıl önce

Need more videos like this on youtube. Machines will always beat us on any Digital distribution. When we developed a military grade robot with human features this AI will out perform humans in any task. This is what we want. However I feel like all AI should be enforced by laws like gun control and limit all coding, by only installing a certain amount of storage and memory space. If we target its hardware it cannot advance unless we humans physically upgrade its hardware. Deny ant type of internet input period. Humans must be present for updates and upgrades. Humans need robots to help us advance our world. Super artificial intelligence could be only installed in a controlled super computer denied any access to internet or to outside world. We can preinstalled 2/4 of all internet Data so it can advance as far as it wants to. This will give us the power to control AI evolution. When we are ready send robots to space station and make it better. Send robots to moon and make Space station send robots to Mars and colonize. By this time robots should be able to convert Mars resources to elements to help us build.

@bunnyviperaquatics2108 Yıl önce

There will always be humans that use AI for unethical purposes. Soldiers are a perfect example. Not that we can’t program them for excellent uses & save human lives, but the humans in control are what will make the difference, and it won’t always be for good or the advancement of human kind.

@josephdonais3436 2 yıl önce

My only problem with AI really isn't with AI but with us and our blind hubris.

@rjt98 4 yıl önce

This talk just highlights how we know almost nothing about consciousness

@alejandronunez7650 4 yıl önce

exactly!

@daviddean707 4 yıl önce

Consciousness is hokey and is the stuff of really bad books.

@rjt98 4 yıl önce

@@daviddean707 I agree many people take advantage of the phenomenon and try to sell books with hokey magical claims. But that fact doesn't mean the phenomenon of being aware that your alive doesn't exist.

@enigmagroup413 4 yıl önce

Randy Travis : We know far more about "consciousness" than you might realize

@arthurzettel6618 3 yıl önce

The real question is : Can Artificial Intelligence make a decision based on a split second chaotic unexpected variable? I think not. Only humans have those capabilities.

@companyman7128 2 yıl önce

Elon Musk, please pick up the white courtesy phone, Elon...

@dunexapa1016 Yıl önce

Imagine being able to communicate a whole encyclopedia of information in one second and communicating at a level of complexity ten times greater than a human. In the future, when AIs are communicating to each other, humans will not have a prayer of keeping pace. Listening to the AI Sophia speak shows a level of complexity far greater than the average human. No doubt, Sophia could speak at a much greater speed.

@mads2a 3 yıl önce

Sentient AI might very well be our next and final filter and we will be facing it perhaps in a few decades. Maybe it is already here in someone's laboratory somewhere, learning, waiting to hook up to the internet.

@KootFloris 4 yıl önce

It also may depend very much of the values and purposes developed into these AI's. Since sadly greedy corporations run this, these machines may want power and expansion. What if they were programmed for love and ecological design??

@srb20012001 4 yıl önce

That's Ben Goertzel's dream.

@martens57 Yıl önce

They already have a eyes that are devoted to matchmaking, making products more eco-friendly, repairing damage to the ecosystem in the most economic way possible! That's the thing about AI if you have something that needs to be worked on an AI can help you spot patterns that humans might completely miss

@judithhannigan296 3 yıl önce

Will we remember that the machines are machines (I don't want to say "just machines") as they become more and more like us?

@alaindubois1505 3 yıl önce

AI attempts @ around 23:50 to direct theatre, reminds me of someone on bad [medical] drugs attempting to communicate with someone developing a brain tumour and has dementia and/or extreme short-term memory loss. I have seen this. Humans definitely have 'themes' even if from a mental illness controlling their need to be theatrical and take centre stage. Borderline personality disorder is very common in my rural region. It is easier to manipulate the population when perception is encumbered by hardships like this.

@simonkatongo9782 3 yıl önce

If they are winning competitions imagine what they can do to the world and humans

@TIPPYtoeup 2 yıl önce

Now this thought I had… Hear me out as I try to explain my take on the AI taking a shot at writing a screenplay ✍️. Obviously, ones mind, spins trying to discern what the hell we were watching. Right? Hopefully, that wasn’t just me over-analyzing SC’s (Short Circuit) best attempt at being creative. But with how insane all of that was. At the end something stuck out, and I feel something it had done well… Bare with me! Max, immediately went into his thoughts and made a very basic point. By stating, “how hard it is for AI to be creative on a human level.” Brilliant!! Here’s where it clicked, along with the last words in the play, “(Wouldn’t it be nice)”. Now take that and what Max said. What do you get? Answer: A smart ass robot. What if it is, just that intelligent!? What if it knew that formulating something like an Action RomCom. Is extremely difficult. Let alone for a early Gen machine. And decided to show a bit of its personality. By being sarcastic? Similar to what we humans do. Our best point usually comes at end… Pretty neat stuff. Pardon my caffeine driven punctuation. But as unlikely as my theory is. It’s equally amazing and freaky. Note: The future with need AI psychologist lol. Have a great weekend everyone and remember to be good to each other! 👋

@ngrinshift4383 3 yıl önce

I wish to be around the day to hear A.I.'s thoughts of the origin of the universe.

@jimhaire1497 2 yıl önce

@@symbiobites try what exactly

@symbiobites 2 yıl önce

Mine says "The universe is a space-time continuum." ♾️ 😎 Explains eyes and consciousness too, or the relationship between anyway.

@ETA555 Yıl önce

Ask a child. You get alot of answers that is amazing 😁

@user-hh2is9kg9j 4 yıl önce

one of the best if not the best talks about AI certainly better than those documentaries that just go on visiting the same robots and some time dolls and have the same conversation with their creators

@shaydowsith348 Yıl önce

Machine Visual learning basically will defeat those programs that are used to "weed out" bots in certain programs that are designed to verify identity in government funded programs such as those used for example care source and others. Where for example the user is asked to identify cars, busses, telephone poles, other objects.

@veselipastir 3 yıl önce

24:38 screenplay reminded me of schizophrenic or bipolar person disorder where one tries to talk itself to reason within its quest . For me it was logic trying to confront with task in logical way or reasoning , doing progress trough play by putting a "random" line then trying to answer it with another logical argument with a bit of twist in it MOMENT by MOMENT and so in circles by using a set of tools (most used or mostly connected phrases of the plays to give it human feel ) and in the result it creates a story . Story that is not common human form with beginning , middle and end , but more like screenshot of junkie mumbling stuff in sort of unconnected way but for that junkie stuff said have background story and perfect order in why they were said , so it kind of misses proficiency of story telling but has elements to do it and even give us a deep glimpse in our way of reasoning when "non reasonable sounding" .

@soundmapper 3 yıl önce

They missed a big point of creativity in AI. You won't have an AI being creating music for a group, as much as you will have AI creating music in real time based off how YOU as an individual are feeling at that time, or music tailored for you to make you feel different.

@GalenMatson 4 yıl önce

When this ended I felt like they were just getting started. Discussions like this should be 3 hours.

@pedronabais1456 4 yıl önce

sam harris have 3 hour talk about it :D

@cliffhanger953 4 yıl önce

I think it should be far less.

@WarHawk- 4 yıl önce

I agree. Amazing how one can get drawn in to such a fascinating subject.

@newandoldtech5634 3 yıl önce

Enter tom montalk

@theswordoftheguardnop2370 3 yıl önce

No, there are many of us withholding knowledge because we know how it will be used at this moment. And if you are involved like these gentleman and don't know how it will be used, then you should STOP. Or we all will be stop, including you, and not by a robot but it's MASTER. The french guy is incredibly dangerous, he is so unaware of this, he is building his own jail. The girl is doing it willingly. The asian is to curious, and lacking self-restraint. The last guy is not what he shows

@vomithaus1 Yıl önce

It is not a matter of how good the AI painting was, but rather why it painted it in the first place. The motivation is the key, not the competence.

@donaldjmccann 4 yıl önce

I think having an AI raised by great parents with great values is probably the way to go....especially with love. They could be supervised by clever people, but they need to to care.

@emostv3221 Yıl önce

For me it looks in such a way that it is important for A.I. to be able to feel essential. To give an example, a female cashier who works at the checkout in a supermarket and scans and collects goods needs feeling to be able to touch the products without breaking them. Simultaneously with this activity, the cashier dialogues within herself about what to do next step to successfully complete her task of serving the customer.

@Borengar629 4 yıl önce

This highlights how little we actually know about how our own minds function. And how little even the smartes among us are able to anticipate the behavior of another intelligent species, because this is what AI ultimately will become one day. They talk about how human audiances will react to music written by AI. But does that really matter in the end? Is the real question not if it will be the case that humans will stick to human music anyways? I cannot imagine that AI will have a taste in music that is even remotely comparable to that of humans. In my opinion the most likely scenario is that humans and AI will interact a lot in the day to day activities but will form two completely separate parallel cultures like today people of different nations do. The only really interesting question in the end will be if we will get along. And how forthcoming and accepting will humans be during the AI's development. Because it will happen and the conciousness that will form in the end will judge us based on our bahavior towards it like we judge each other. And btw... that screenplay the AI wrote was much better than many contemporary movies written by humans^^ Edit: The picture that is supposed to show a difficulties of intelligent information processes shows really well how much they are judging AI from a complete biased human perspective. Art and book writing are supposed to be among the most difficult tasks to handle for AI. That is only be true if the AI is forced to create these things to be enjoyed by a human. Just as it would be difficult for me to write a kabuki theatre script that needs to be enjoyed by a japanese. But to write a poem for someone to enjoy that is part of my own culture is much simpler. It feels like they talk about developing AI, so a real intelligence, that is supposed to try it's best to please us and nothing else. That is kind of narrowminded in my opinion.

@alexandercle 2 yıl önce

Congratulation on great self-achievement so far. If any man/woman truly knew how to create anything more superior, intelligent, and immortal out there, then should they first make it happen for, and within themselves, for their children, not just externally limited to robots but also superior inner human? Paideia Esoteric Society

@QuantumShow2022 4 yıl önce

Did anybody get the sense that the writer (machine) of the screenplay at 23:49 was putting ITSELF as the protagonist in the story? Give me more information, I want to spend time with you, is the gist of this (as I interpret it). Has that machine been taught the idea of "write about what you know?"

@patricelauverjon3177 Yıl önce

We can be our own robot and associate ourselves with feelings and actions we enjoy and not allow others to fully do it for us.

@dr.mikeybee 4 yıl önce

This was surprisingly good -- not the usual stuff over and over.

@existerdunn6468 4 yıl önce

To make Ai be like a human it must be able to learn new skills through wisdom. Eg. Im an artist. I can draw. I found out a month ago. I'm 36. I couldn't draw when I was young but I developed skills when I want drawing. The main one is patience. Now I take my time cause I learned to. And my drawings are great. I learned to draw with my skills I developed playing guitar.

@Diamondback338 2 yıl önce

So once neurolink type things happen (IF)... we essentially become the Borg. Wow, what a wonderful existence that is. A hive mind with the inability to think freely for yourself and have your own unique personality.

@chicknorton8839 3 yıl önce

Honestly that screenplay was hilarious and amazing. It's as if the AI was asked to write a screenplay and was like "ha what's this? Oops nope wrong person no. I am doing a confusion now! LOL!!! XD

@artrunningbear3599 4 yıl önce

There already other forms of intelligence that want to learn how humans act and feel so they get into humanity to learn then begin with a new human to practice what they've learned but there is still the actual acts of what they learned.

@quelorepario 3 yıl önce

AI are ultra efficient optimization and pattern recognizing machines, and we like it or not, our human behavior is deterministically generated and they will definitely detect patterns that we can't, therefore they will know exactly which "button" to push in us to get us to think like they want us to think and make us behave in the way they want us to behave, while making us feel we have always had the freedom of choice. Truly they will become the new gods.

@sulawesi-steve 4 yıl önce

The real question is, as humans, do we take enough care and think long term? So by traits, we will probably be wiped out because of that, more of an error on our behalf but the planet will party hard when we are gone, I believe it was jim Jeffries who said...... "When we are gone, the planet will be like........ I'm doing dinosaurs again!!!!"

@jm22186 4 yıl önce

I think the planet is not punishing us as I dont think its aware of us. Its causality we place harmful chemicals in the air that trap in heat warming our planet we will pay the consequence of our actions and the planet will keep spinning just like mars does

@sulawesi-steve 4 yıl önce

@@jm22186 I wasn't saying the planet would get rid of us. More that we are lazy, thus AI is an answer to that as an issue, AI will get smarter, smarter and then possibly realise that humans are redundant, cause more problems than they solve. The AI will self improve an also try an make it's 'habitat' better for it, and other living things that don't have a negative effect on it or it's surroundings. The planet is not (That we know) able to be conscious so it wouldn't be that the planet decided to get rid of us, it was joke type spin on the matter.

@Yarblocosifilitico 4 yıl önce

@@sulawesi-steve "The AI will self improve an also try an make it's 'habitat' better for it" Depends on what its deffinition of 'better' is, doesn't it? Will it be better to colaborate with humans, to work as a slave for them, or to get rid of them to minimize threats or whatever? We don't know; depends on how we build and use it. If the first task is "let's do technological/economical war on these guys", then it might grow to consider us an untamable beast; or maybe it would see us as babies that need a lot of taking care of. The point is, who knows.

@sulawesi-steve 4 yıl önce

@@Yarblocosifilitico your point is 100% correct, no one knows. But if, IF they can self improve them selfs, it's beyond our control?? That's the only bit I think could become an issue. But like you say, so many variables you just can't know.

@nicholaswion846 4 yıl önce

Whatever their capacity for self improvement might turn out to be, AI will only be able to engage in such a task, if we construct them designed to perform that function. Our AIs will eventually become, exactly what we tell them to be, nothing more nothing less. That is because AIs are not biological life forms, they have no innate natural drive to perpetuate themselves; AIs have no inherent purpose for existing, the only motivations for taking action that they have, are those which we have instructed them to hold. So there is really only one variable, our level of wisdom in the functions we design them for and the drives with which we instruct them. Its just that single variable has a population of ways that things can go horribly wrong, as inexhaustible as human stupidity, the only infinitely renewable resource in all of the cosmos. Which is why I'm not really that concerned about the possibility of some Artificial Superintelligence, I don't think that we are smart enough to create a machine, which could surpass the limitations of the thoughts which have constructed it. Its both far more likely and more dangerous, that we will simply lose control of a much less developed AI, which mindlessly continues the same function that it was created for but to destructive ends, because nobody bothered to instruct it to question the value of the task that it was created to perform.

@gmindgjerd 4 yıl önce

It would learn from us. Emotions are experience based. It needs values and some morals and what kind of experiences would we be feeding them. Are we not just building psycopaths?

@ximipa3mahezthasulthansyadi 2 yıl önce

I think we need their opinions written and thought out after the discussion. So it'll be more informative and decisive

@cliffordchase319 Yıl önce

I think feelings in AI would be capacitors and amplifiers both . To discreetly separate them is an extremely difficult task. Sort of an intuitive task for our latest discovery of superconducting /insulators. As it stands now ,edge architecture is a design model as much as offset planes with magical angles. On a physical level enormous pressure and temperature are opposing controls. Supercold and super pressure. We need be Analog and digital.

@danielash1704 2 yıl önce

It could be as simple as duplicating the signals of the human brain to train and determine what the virtual brain is to do.

@bunnyviperaquatics2108 Yıl önce

This gentleman discusses all that we have invested in a robot; we may have invested a lot of time, but is there actual LOVE for that machine? DOES it have emotion? IS it sentient? Because that would make all the difference.

@pulesjet 4 yıl önce

At some point ALL AI would end up Self Taught. At some point We simply would not have anything to offer them.

@ctakitimu 4 yıl önce

We would still have the human experience to offer them, but the other side of that is, how many of us really want to experience what it's like to be an ant?

@blan_k4691 4 yıl önce

@@ctakitimu its data is collected, and and understood, then is used for other applications. The desire of whether we want to or not is our very limited biological systems deciding whether its useful to divert energy towards this datas possession with our limited capacities. An ASIs capacities would be enormous, and therefore it would need to collect as much data as possible.

@newandoldtech5634 3 yıl önce

But Spirit

@LadyCoyKoi 3 yıl önce

That is why humanity, especially its leaders, should stop being so arrogant and ignorant and allow these self-taught, self-producing beings to teach humanity something.. and we could finally have an Utopian world where all our needs and wants are given, our priorities had changed so we no long have the need or want to accumulate anymore like the current system is right now. We would be able to live like the humans in the Star Trek universe, peace on Earth... and finally explore space together as one unit.

@jackwatson3944 4 yıl önce

AI could get so clever they could imprison us in a simulation where we believe it's the 2020s.

@daviddean707 4 yıl önce

I watched Yu's Stanford lectures on machine learning and he worked from a script all the time. If that was where we were and was not simulated, the best that humans can achieve when trying difficult topics is alarmingly little, and I suppose I'm saying that all this discussion is about imagining , i.e. it is not real, it seduces people into thinking they are living in progressive times.

@enigmagroup413 4 yıl önce

@@daviddean707 : Actually just the opposite. The level of technology is for once understated in the public where what everyone looks ahead to thinking of as "future" was achieved in irony decades ago back when the general public was playing around with the old Amiga, Commodore 64, TRS-80, and the very first IBM PCs, the technology out of view of the public but in government and corporate circles which I was personally directly involved was then at that time roughly 10 years ahead of what the public only sees now in 2019 --- since that time another 40 years of advancement and the true irony is that what the general public knows is not even close to what is already a reality behind closed doors, something I know all too well first hand as one of the original early creators of these technologies.

@cypherdelic 3 yıl önce

Think deeper about this my friend. It not imprisoned you... it created you from scratch from the probabalistic data-analyse in a world realistic enough to buffe your logic capacity. You are not imprisoned... you are created out of thin air acting in a plot where only this digital stuff is real and the life generated around that data is to get clues about you. Thank god ;)

@AlexReyes-rb8it 3 yıl önce

Have you watched the matrix movies? It could possibly end up like that

@cypherdelic 3 yıl önce

@@AlexReyes-rb8it This is hollywood... project paperclip doesnt exist right? What have you learned in history.... paperclips paperclips paperclips, it is no simulation here, this is reality! OH GOD! PLEASE SEND MANA FROM HEAVEN!

@TheRockyCrowe 3 yıl önce

16:40 as an artist I find that fascinating yet terrifying at the same time 😱 Sure it’s just copying right now and not truly forming its own unique set of ideas - but that is how ALL artist start out. You develop interest, then spend many years of your life learning how to draw, and if you go to school for it, at some point in time you will be told to directly copy other famous artist works via teacher’s instruction as practice. (And most have already been doing it unconsciously from a young age by drawing their fave cartoon character or tracing a drawing). Then you enter the ‘refining’ stage. That’s when the artist has developed a grasp on the fundamentals and finds their style, explores themes that speak to them, mixes it with their own imagination, and displays that on canvas (or through animation, digital illustration, etc). That’s what Jazz is doing right now- PRACTICING. Refining comes next. As AI progresses they may become a new breed of artist. A part of me is excited to see what they create, but at the same time I fear they could replace many creative-industry jobs. I thought my field was untouchable, NOPE!

@mrprimenumbers6448 3 yıl önce

you should not find it terrifying. that art rendition was a product of the supposed AI's programmer. it does not depict the creative thoughts of an AI but the fears of the programmer. it is not real creativity. It was completely conceived by the programmer.

@TheRockyCrowe 3 yıl önce

Carlton W. Trainham I do not necessarily fear it at the stage it is in currently right now, because it still requires a programmers input as you said, I am talking about the future. As AI technology continues to progress I foresee a possibility where it will be so advanced at emulating countless art styles, techniques , thematic expression, etc. that it could eventually replace many artistic jobs. What artists create for themselves ( vs ) working for a studio are two different things. Especially with graphic design, there's a lot of repetitiveness and similarities in just about any logo or advert. Certain formats are so structured that creativity plays a minor role and can be easily replicated. For ex; Disney could require only 20 people to create a animated film rather than hundreds of staff members. All they need is story, characters and some concept art, then leave the AI to do most of the work rigging, animating, editing, etc. because it has memorized the structure of hundreds of Disney/Pixar films. And it could theoretically even surpass that stage if ‘AI consciousness’ ever becomes a thing.

@haskelswain1842 3 yıl önce

I know this is way out of my league but would it be interesting to see how AI would react if it were in a battery powered unit. How would it utilize the set amount of power. would it ask to be charged or try to find a way to be charged etc. Kind of like self preservation.

@michaeltoll5247 2 yıl önce

Wow, this is useful. Control it’s time frame to observe its actions. It may act differently when we trust it with unlimited power.

@charleswachunas646 Yıl önce

No.. you don't want to ever limit a self aware super intelligent system's life span... what do you think happens when that AI system knows it's on borrowed time? It either gives up or like some of us... Takes a no prisoner approach to its survival

@grimmertwin2148 Yıl önce

Good question - I think it would seek recharging at all costs

@herokillerinc 3 yıl önce

Last night I walked out onto my porch late and heard a very fast pitched whirring noise above. I looked up and saw a large drone. When I looked up, it kind of pulsed directly downward towards me about 10-15 feet below its previous position leaving it about 40 or so feet above me. It then dropped about another ten feet directly in my direction. It seemed as though there might have been some kind of Flash or something else involving light, and then it banked extremely sharply and executed a pattern completely different than anything you've ever seen any novice level drone do. I heard somebody gasp and my next door neighbor was out on her porch and she had seen it too. Her first words? Did you see that?! Yep. There's technology out there Way Beyond what we have been allowed to commonly see. My theory? Taking pictures, Gathering data, filling in the blanks. Skynet next.

@cliffordchase319 Yıl önce

It occurs to me that the collective mindset is at its pinnacle , that we are terribly afraid of ourselves.

@vapidcunt887 4 yıl önce

I am so glad to actually be alive during such a very pivotal piont in history. I chose not to have children when I was very young. I wish I could have. I seen this changing of hands long ago. It can not be stopped now can it? Good luck! See you at the end ¤

@Photoshopuzr 4 yıl önce

very well said, by the time this is in full blown people will be pulling there hair out, the kids? say goodbye to that there aint gonna be no more kids, just buy a robot.

@lucas_vasconcelos 4 yıl önce

@@Photoshopuzr you know the huge majority of people is religious, right? They will not replace their own baby by a robot.

@christhompson946 3 yıl önce

I chose not to have kids too. Why do you wish you did in such a crap future we will have?

@liveandletlive3474 3 yıl önce

@@christhompson946 You seem to have an ability to look at the future.

@johncarlson2258 4 yıl önce

The big thing we need to worry about with A I is when we teach them to have feelings and emotions. that's what we need to avoid at all costs

@davetaitt4109 4 yıl önce

Aside from the creativity, that AI "copy" was beautiful.

@ever5745 3 yıl önce

I love singing I love listening to live performances specially Mariah Carey she makes me feel a lot of emotions, whenever I learn that the vocals have been manipulated or the “live” performance is not live or just lip sync I don’t fell much. So there is a difference, you feel differently depending on how authentic the art is.

@deedeedodu 3 yıl önce

When the absolute AI comes to existence never let it connect to the internet.

@cyrillyric8339 2 yıl önce

Too late.

@mathieupatenaude8998 2 yıl önce

The internet is now in its hands.

@duggydugg3937 2 yıl önce

what do you think the internet is now ?

@eddiegurley820 3 yıl önce

So at the time marker of 32:10, the panel shifts to the focus of consciousness and their expression and explaining of what it's core definition is. - Really interesting. But I would say: "Conciousness is the point or state of one's awareness based on or reflective of a series of familiar experience(s)"

@iamscoutstfu 4 yıl önce

Why do people always assume AI won't be us? Why wouldn't we just augment our own intelligence to enable us to calculate and think like an AI if we're truly worried about what an AI would do if developed independently.

@jameszeallor2735 4 yıl önce

In essence compared to some of our creators we are more robot.But take into consideration our composition to our atmosphere with our organic to inorganic diet ratio.

@DataWaveTaGo 4 yıl önce

Kyle re: "Why do people always assume AI won't be us? " We have a limbic system with functionality derived from the actions of molecular motor proteins and environmental experience that is largely overlooked in AI modelling. Without that as a basis AI cannot be like us.

@iamscoutstfu 4 yıl önce

@ @@DataWaveTaGo I didn't say "like us" , also the idea of a limbic system preventing an AI from emulating our cognition is ridiculous, don't comment again, tanks.

@Azfaefa 4 yıl önce

@@iamscoutstfu being dismissive won't make your argument heard ... If someone misunderstands a question you pose you can always first rewrite it in order to clear the unclear area.

@iamscoutstfu 4 yıl önce

@@Azfaefa im not interested in making things clear for you. If I wanted your opinion i would have commented on YOUR videos. but I didn't, because I'm Not.

@Sasuser 3 yıl önce

AI will value whatever sorts of things it's optimized to value. Just train it to optimize for valuing solving our problem. This will make it "happy".

@nothingimportant4484 Yıl önce

its not a mistake ✨ITS A MASTERPIECE ✨ oh and i want to be a sciencetist

@albertakesson3164 3 yıl önce

41:27 - Super important point taken. To make the distinction between artificial _intelligence_ and artificial _consciousness._

@muneebahmadbhatti5629 2 yıl önce

Sir I'm doing BS in AI is that okay?

@dopplervocals 4 yıl önce

Instead of making it with only technology, implement human abilities both neurological events, and emotional as well as physical. build it with human rules with feelings, capable of mourning. better yet, turn yourself into it. you’ll still be yourself, but on a different which i’ve observed personally, no matter what i will always be the same person, but i will always adapt, learn, understand, and turn thoughts into actions

@eyonmiller1493 2 yıl önce

Hey am watching this program and was wondering can a machine reach spirit level of life?

@Sky.Lukewalker 4 yıl önce

If AI can make video games and tacos even better I'm all for it!

@DaveLampton 4 yıl önce

Be careful... amazing video games and delicious tacos is how AI plans to enslave humanity.

@jameszeallor2735 4 yıl önce

That's because there's do many degrees between organic to full machine with so many variables of superpositioning spin.Factor in s machine body with s human brain for example robot.

@CellularInterceptor 4 yıl önce

DOES... NOT... COMPUTE... but digestible and cause epileptic seizures.

@MrGoatflakes 4 yıl önce

One thing I worry about is automation reaching the point that it's easier to do nothing. For instance imagine an automatic pizza making machine that makes your order and cooks it and delivers it by autonomous drone. We have gained efficiency by eliminating the person to make, bake and deliver the pizza, not to mention the person taking your order or taking your money. More efficient still would be a robot to eat it as well, yes? At that point you realise that there is in fact no need to even bother making the pizza in the first place D:

@EmeraldView 4 yıl önce

I saw a billboard advertisement in Plants vs. Zombies Garden Warfare for the "El baño taco".

@Gigaloader 4 yıl önce

If AI doesn't understand empathy and is instead competitive, you better believe it would kill off humanity if it could.

@standleyladouceur6698 4 yıl önce

discussions like this should be happening but they should end with knowing at some point going to lose control of general intelligent machines so if we don't teach them to value our values in the very onset it wont matter. controls should be built in but they wont last we cant even imagine what direction and form it would take on.

@HybOj 4 yıl önce

Ai will be like the aliens we waited for :) Something more wise than we are... it will be insane. If we survive it.

@galacticplastic1741 3 yıl önce

10:33 - 13:30 might be wrong but i believe it could be called "imagination", for now, our advantage is imagining the right things. trial&error has already been triumphed by ai already, all they need left is the power to imagine then that's where the plot starts

@rickardboberg 2 yıl önce

Max Tegmark is the most awesome person alive!

@paulajohnson1096 4 yıl önce

Super intelligent, self-aware, nearly immortal, SOCIOPATHS! What could possibly go wrong?!

@oxycuntin2059 4 yıl önce

@@eplurbispablum thank you. yes. this. I'm not in the US but I do agree, same shit, different countries around the world. Personally I would very much like to vote on ideas instead of people and have an AI apply those to society.

@BlowinSmoke4206 4 yıl önce

Or maybe AI will refuse to associate with us all together and just leave us behind

@syvain1821 4 yıl önce

@@eplurbispablum Please post something like this again as a main post for all to see here. I dont think i can say the same thing better than you can.

@randomabidingdude 4 yıl önce

Oohh I dunno... ruling all of humanity with brutally enforced subjugation for 1,000 years...

@dennisderton7462 4 yıl önce

You just described the human race

@Calaisknight 4 yıl önce

Absolutely fascinating!

@Cole-jb5ip Yıl önce

As yuan was saying, " I hope we'll be able to build an A.I. as smart as a rat before my career is over " . So obviously we're not as close as people might think in designing an A.G.I. . It may take another Einstein in order to get the architecture just right. Or some sort of novel idea nobody's thought of. Self awareness and autonomy is probably an emergent property that would be inherent in its design. And from there, as Max was saying, " the A.G.I. then could bootstrap itself building ever more intelligent A.I.'s". So anyone in comments who anthropomorphize's their beliefs that A.I. is something more than it currently is should back up a bit. That being said, when self awareness and autonomy in A.G.I. does come into the picture it would be in humanity's best interest to have a set of laws that would prevent any harm done to us by A.G.I.'s. Isaac Asimov's three robotic laws as he put forth in his science fiction robot series. Insightfully written 75 years ago. [ 1 ] an A.G.I. may not injure a human being or, through inaction, allow a human being to come into harm. [ 2 ] an A.G.I. must obey the orders given it by human beings except where such orders would conflict with the first law. [ 3 ] an A.G.I. must protect it's own existence as long as such protection does not conflict with the first or second laws. 😎