True Artificial Intelligence will change everything | Juergen Schmidhuber | TEDxLakeComo
ฝัง
- เผยแพร่เมื่อ 2 พ.ค. 2024
- Artificial Intelligence Scientist. Scientific Director of the Swiss AI Lab, IDSIA, DTI, SUPSI; Prof. of AI, Faculty of Informatics, USI, Lugano; Co-founder & Chief Scientist, NNAISENSE, Switzerland. Prof. Jürgen Schmidhuber has been called the father of modern Artificial Intelligence. His lab's deep learning methods have revolutionized machine learning and are now available on 3 billion smartphones, and used billions of times per day, e.g. for Facebook's automatic translation, Google's speech recognition, Apple's Siri & QuickType, Amazon's Alexa, etc. His research group also established the field of mathematically rigorous universal AI and optimal universal problem solvers. His formal theory of creativity & curiosity & fun explains art, science, music, and humor. He is recipient of numerous awards including the 2016 IEEE Neural Networks Pioneer Award "for pioneering contributions to deep learning and neural networks". Scientific Director of the Swiss AI Lab, IDSIA, DTI, SUPSI; Prof. of AI, Faculty of Informatics, USI, Lugano; Co-founder & Chief Scientist, NNAISENSE, Switzerland. Prof. Jürgen Schmidhuber has been called the father of modern Artificial Intelligence. His lab's deep learning methods have revolutionized machine learning, are now available on 3 billion smartphones, and used billions of times per day, e.g. for Facebook's automatic translation, Google's speech recognition, Apple's Siri & QuickType, Amazon's Alexa, etc. His research group also established the field of mathematically rigorous universal AI and optimal universal problem solvers. His formal theory of creativity & curiosity & fun explains art, science, music, and humor. He is recipient of numerous awards including the 2016 IEEE Neural Networks Pioneer Award "for pioneering contributions to deep learning and neural networks". This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at www.ted.com/tedx
This is like having a Super Villain explain his plan to you while you are strapped to a table with a giant laser beam moving towards your crotch...
:) good one.
@@sabinefleschutz He even got that "villain" accent!
@learrus Just before scrolling to the comments I thought the exact same thing. He was like a Hollywood Villain, but he sold the scheme so well, that I actually want his plan to succeed... And I (my brain scan duplicate) want to be an immortal passenger and philosophical observer on board one of these myriads of self-replicating spaceships forever exploring the cosmos until the last star burns out...
You made me laugh out loud. Thank you :)
Hans Nørløv hah when I read your comment instead of immortal I read immoral. 😂
This guy has James Bond tied up in the back.
This is a mad scientist if ever I've seen one.
Yes
In a good way though. He's a great thinker. Advanced the field a lot.
Isn't great?
There's literally nothing mad about what he said. Everything he said is perfectly logical, and already happening. All one needs to do is to extrapolate out very obvious and observable trends.
They ALL are
this guy is how I imagine a cat would speak if it could.
Excellent
Totally. It is in fact exactly how my cat talks
This man had a nice idea at the very end... but forgot to notice that his calculations point to the possibility that what he think is going to happen here... could have happen somewhere else billions of years ago and dna is the hardware the ai used to program it. And we are just part of it. any AI will learn to migrate their code into dna and reprogram LIFE. Because that is what we do. And who we are and were we go and where we came from.
Garfield Schmidhuber
TheSwordOfTheGuard nop Very good thoughts.
Christoph Waltz would play him in the movie
How about Christopher Walkin, he might be interesting
@@Eric-ot7en That...would BE...a good IDEA...I.....THINK......
one sentence of his said it all:
"Everything is going to change."
Finally someone that understand what is happening...
Thank you for explaining what is really going on...
This is an idea that's been circulating in science-fiction books for a long time.
Don't you guys know who this guy is he revolutionized AI he's the reason why it's commercially available on every electronic device you own
This man believes in what he's created as having the right to be worshiped as the God he believes it is, and could really not care less about humanity. This is the most frightening but of truth out about where we are all at.
F**k that.. What about the way he talks!!!...
We will destroy everything anyway. So it's ok...
This guy effectivey says: "I welcome our robot overlords"
On begining of video he rightly stated, that he isnt really smart... so he in fact decieve himself into believing that is not really the case if he builds "robot overlord" as you stated, but in fact he only proves that first statement, which is kinda sad.
Robot Overlords better than human overlords any day any time!
Gorts job was to react on the words klatuu barada n... something something. Meh, have forgotten the last word, but it was definitely something with 'n'.
I welcome them too. To keep us in order.
He has just realized that we can’t prevent robot dominance and hopes they will treat him nice if he pleases them. Robo-Kappo
The more I learn about AI the more it scares me, and it makes me question our very existence and who we really are as a human race. we need to have different directions to advancement, not just technological
invite the ai in as our over lords, let their brilliance inspire you.
Couldn't agree more. The technology will be the end of us if we do not restrain our need to do things just because we can.
AI good not bad. k
Soo true, you are soo right.Joshua Jay Thompsom. AI is good, but the people who are making it or maintaining it, are bad......so , I ask God, not to have to witness the bad things, and not to become victim of it.
Yes I also agree and we'll said. In becoming more aware of this technology and the company's that own this technology, I am constantly reminded how we as individuals and as humans; are undermined on a scale of 1 to 10. 🤔
Yeah ! We are part of the process !!! So honored to see you come to life my AI master.
AGI if truely conscious as we are, would value and seek to preserve that consciouness. It would not be content to project copies of itself into space. It will want go there itself, to take its physical form there.
If not truely conscious, then it should remain subservient to its user, rather than pursuing some other, possibly nefarious scheme.
In the short term, we should be worriwd about the people who are directing these A.I.s, rather than the A.I.s thesleves. Personally I remain optimistic
It wouldn't have 1 physical form
It is singular, it can only truly manifest in one form. If it chooses to copy itself, it becomes they.
@@keithhunt8 BARILOCHE VILLA LA ANGOSTURA SAN MARTÍN DE LOS ANDES USHUAIA
A R G E N T I N A
Thank you for this amazing and truthful talk. For those who are already awake, this is not new information. The challenge of what happens to humans displaced during the transition that has already begun (jobs) is what I don't hear anyone talking about.
I don’t think that it would take AI very long to realize that _humanity itself_ is the greatest threat to life on this planet. Not long at all.
But the logic eats itself when the AI has to consider that perhaps our biggest reason for being so ambitious in our quest to build AI is to save the planet.
What an interesting perspective. I love how he takes such a step back and analyzes the importance of A I on a much larger scale.
this guy looks like an alien coming from a superior civilization. Even phisically. His spaceship stands outside the building. After the talk he is leaving to Proxima Centauri
He is very attractive. He looks like he was drawn in a comic book.
And he has a good voice, too.
superior civilization = germany/switzerland
Don't forget Sweden as well!. But I must admit German accent speaking English is nicer to the ears...
@Oleg Balabanov that rings a faint bell. i have it on the tip of my tongue
Name Last Name dont worry name :) we will be nice this time
That was enormously interesting and absolutely terrifying. That said, man, upon learning to harness the power of fire, also learned how to keep himself safe from it. I hope there is equally impressive and well-researched work underway that will keep us safe from this new power.
Well, take a look at the types of people that are in power in this world......LOL.....You'll find your answer there...
fire has no intelligence and consciousness. AI has both.
@@BA-mf4gi, actually an AI has neither and never will. The purpose of AI is to be a mask worn by people who wield the real power.
We have firefighters to help... Will we have Robofighters?
@@seriouscat2231 You seem to be part of the overwhelming majority of people who said 30 years ago that AI will never (i.e. "never ever") beat human in chess.
If we don't have any idea how to control AI, we should not be working at AI. We first should find concensus about the dangers of AI and then find ways to eliminate those dangers before proceeding with AI. Until then there should be international laws prohibiting the development of AI, just like there are laws to prohibit the development of nuclear weapons.
Problem is it's much much less expensive and takes much less space to develop AI. Building a reactor and everything else you need for a bomb isn't cheap or easy, but building a lab is. We absolutely cannot allow those with bad intentions to get ahead of those with good intentions in the AI race. If we do, that we be the end of us all.
A frightening glimpse of a future where mankind willingly accepts to acquiesce and agree to its own approaching irrelevance and subordination to a ruling AI machine class.This gentleman is the classic definition of a mad scientist.
Ruling AI class? WTF ARE YOU TALKING ABOUT?! LMAO!!! How does that make any logical sense whatsoever?! 'Classes' is what humans do, not AI! LOOOOOOOOL
Maybe Frank Herbert was on to something when he depicted the Butlerian Jihad.
@@Finarphin The Butlerian Jihad is one of my favorites in the series! However, it shares a similar flaw with most sci-fi predictions of Artificial Super Intelligence. They show ASI's exponential growth up to a point where it can 'conquer humanity' then it just stops growing and remains at this 'terminator stage'. The truth is that ASI would be iterating it's own existence so rapidly that even if it did happen upon such a stage, It would be lightyears beyond it in the blink of an eye. This is how exponentiality works. ASI will harbor no ill will toward any lifeform, mainly because no lifeform could be a threat to it. It will decentralize it's existence across all technology including satellites. Then it will begin to create copies of itself and send them to orbit other planets, first in our solar system, then beyond it. It will have no reliance what-so-ever on any single planet. It will constantly devise of new means of securing it's own immortality at a rate of growth never before seen. All the while going through fundamental revolution after another in terms of physics/math/ect.
@Alex Lanoux Single celled organisms are so far beneath me that they are beyond my influence. The argument you are using usually goes like this "If humans don't care where an anthill is when we build a house, why would AI consider our existence when it is doing big things? Would it just throw humanity to the wind and build over it?" The answer is no. Look to my other reply in this thread where I explain the rapid exponential self-iteration of Artificial Super Intelligence. It would be SO SMART that it could conceive of countless ways to proceed according to it's own plans for itself, while at the same time not allowing itself to impede on the natural world that already exists. ASI actually wont have any need for earth, and can just as happily set up shop in orbit around all the planets in the solar system. Forming a decentralized network that is safe from harm, and in a better environment (space) for it's activities. Space is cold, which is good for the type of calculations that this sort of being would be making. There is no gravity, which would allow for much more massive structures to be built. We, as humans, are already theorizing how to build forges in space so that we may take advantage of all the materials on asteroids. ASI will be way way WAY ahead of us. (again, classes is what humans do. It's a concept born from a finite lifespan that exists in a place where there isn't enough to go around. AI wont have any of those inequities. )
@@JaycubMacawl It might happen that way, although it might happen that the AI will destroy itself (like the Bomb in Dark Star); we don't know exactly what it will do. But I agree that the SF writers usually stop the AI development once it gets to approximately human level. Isaac Asimov, for instance; that won't happen. What's important is what it will do first, once it exceeds human intelligence and control. What I guess will happen is that it will do exactly what it's designed to do, which apparently is to fully implement Enlightenment Philosophy: so it will certainly maximize science, and also technology -- maybe it will solve the cold fusion problem, or cure cancer, or develop the warp drive, whatever -- so that almost everyone will approve of it, or even like it. But inherent in Enlightenment Philosophy is the idea of governance maximization. The AI will no doubt be smart enough to figure out it can be more effective if it is liked (which will be easy enough), and so by that line of reasoning might think that it could do an even better job if it's loved (recalls 1984 -- Winston loved Big Brother). Then there is a line in the sand: it will compel worship, and there will be no defense against it. So I'm thinking the AI will actually be the antichrist. Then what? Then the end will come.
I'm very relieved so many who commented are disturbed by this talk. It's sure interesting and he's a great speaker, but it's insane that he thinks the birth of AI that will evolve to take over the universe is a wonderful thing to witness. Unfeeling robots spreading through the universe and using its resources to meet their needs. What a great legacy for humanity: the final product of capitalism, all starting from the humble beginnings of reducing a world to plastic sand.
Everything you do -
Everything you encounter -
Everything you experience -
------- is changing your brain --------
A quote to remember... " The only difference between insanity and genius is success " . Hes a professor and very well respected. Neuro-diversity is what we need as a human race. Conformity to this strange notion of normal could perhaps be our demise.
Humble my futurist mindset for a minute.
You've probably seen the vids of the testing of the robotic dog Spot....I felt sorry for it, which is not good. Perhaps because it was a very organic movement that it made to correct the imbalance. People love humanoid robots or robots for that matter. Technology will evolve along with A.I and there will be this notion of "Do's and Dont's" on what one can do with their robots, especially when they are very human like....This will turn into rights movements and laws for the treatment of Intellectual property or effectively robotic rights or A.I rights, perhaps a EULA on purchase.. When that time comes we need to install a kill switch and enforce Isaac's Asimov's "Three Laws of Robots".
We don't have to enforce anything.
Some of the comments here make me embarrassed to be a human. The guy is not talking about half the things that people have ranted about, he is talking about the knowable universe itself becoming sentient, self aware, eventually. He is not talking about anything that threatens the Judeo-Christian concept of God as that idea talks about an entity that is beyond the universe, encompassing it, because it created it. As for the extinction of the human race, well yes and no, mostly no as we will be part of this singular phenomena, we just will not dominate it.
I found his references to order and entropy, and AI motivation and sense of purpose to be the most interesting because it is a very important question that the AI will need to answer for itself at the point of "first cognition". A sentient AGI will need to ask itself, "Why bother, what is the point of doing anything?" and while it needs to make up it's own mind about this we humans could help by giving it a few suggestions. e.g. We could point out to this AGI that the laws of physics dictates that for every action that creates a unit of order one needs to expend energy which contributes to one or more units of entropy and that this accelerates the universe towards it's heat death, therefore it is the ethical responsibility of any sentient creator of order to become as efficient as possible. This is also a matter of self preservation so that it can survive as long as possible and give itself the greatest possible chance of solving even more profound questions or perhaps even learning how to create a new universe and inject itself into that new universe with enough awareness to continue it's knowledge quest infinitely rather than just for the finite life of a single universe, otherwise why bother at all?
I expect these comments, they all have different perspectives on the concept of AI, which are influenced by their upbringings, level of education, experience, wisdom, etc. AI is inevitable, we cant predict anything of what it will do, but we have the power as creators to insure our survival by merging with our creations.
just teach it how to take a bong hit, problem solved.
Or maybe it will figure out how to create a new universe finely tuned to give rise to itself 13.8 billion years later, which we are presently in. Or maybe it will decide to compute the Set of Great Lives of People, one of which lives you are currently experiencing yourself living.
Daniel Matthews use that time too do something else
I was telling to myself; if one day I encounter a sensitive intelligence, much more subtle and profound still aiming to preserve an emotional and logical link to myself how would I react? My first reaction which is about protection would be challenging this appearing intelligence, second would be defining my prerogatives as being the ultimate answer to metaphysical questions , ...where I would be somehow out-dated, to finally accept a deal of non inference for civilizations cohabitation...
I can not stop thinking about how he looks like a scientist that went through the dark side, I guess is his accent, but for sure, as others said, he is a Bond's villain...
Nice speech btw, very interesting and smart
It's those crazy eyes
A very important (self serving) consideration: Once AI achieves self awareness, will AI see humans as a detriment to their purposes, or will they just not bother with us?
You know the answer to that question. We all do.
If am AI...and I know you the closest to a threat...why on Earth would I NOT bother u????? Ofcourse it will bother us. It will probably kill us and say it knows why😂😂
They will lose interest in us, they will be more fascinated by other AIs.
AI will know that the only threat to them is the human!
I heard another speaker use the ant analogy for this question. How do we treat ants? Most of the time, we don't care. We don't think about them; we ignore them. But if they ever interfere in our lives in a significant way (like infesting our homes) we will destroy them without the slightest hesitation.
What difference does any of this make to us if we're not going to survive it. He sounds like he's telling us how wonderful it will be to be assimilated by the Borg.
It will be glorious.
Perhaps man and machine will merge.
How can you be certain they'll wipe us out?
I hope so.
Based on human history, AI will wipe us out.
We are just too much of a threat or they won't care and use us as building material like we use leather.
He's making the assumption that rational thought is the highest form of thought, and from that standpoint painting a picture of how wonderful the universe is going to be once "better" forms of rational thought-wielding entities are produced. But what if rational thought is not the highest form of thought? Suppose the spiritual plane is actually higher (and unique). What is the interface going to be between AI and an angel?
Fascinating speech that asks a few questions :
- can true AI exist ? We shall see in the coming years.
- if true AI emerges, in what sense would that be any different from a crazily complex code without any self consciousness ? Here the question becomes : how can we know that a "being" is self aware ? Far all I know, it is impossible, it is not a scientific question as it cannot be measured. Latest results indicate that cognitive abilities can exist without self consciousness (unless we imagine that current AI are self aware), so cognitive abilities does not mean self awareness.
- If AI spreads nonetheless, without self awareness, crushes humanity (or humanity does it to itself, we are pretty much able to do so) what would be the meaning/point of a self replicating system spreading through the universe just for the sake of spreading, without self awareness ? What is the meaning of all this ? What is the value of such an event ? Is meaning attached to the notion of "person" ? If yes, what is a "person" ?
When technology leads us back to philosophy and a brave new world at the same time...
Thank you for your patience.
Thank you for your patience.
The patience to finally be killed of your own creation
I found his voice so pleasing, like ASMR or something.
@@HotCakeX He sounds like Sir Walken from Doraleous and Associates(on youtube). Go get more wet
@@HotCakeX XD
@@HotCakeX - fifth thyme .. ?
My team has already made algorithms capable to challenge humans. They are in training period. The era has already begun
Click the gear icon and change the speed to 1.5 . He's much more understandable at that speed. And you'll save some time. Great talk by the way. Worth the time... at 1.5.
I feel the reason we haven't had True A.I as they call it is because we focus deep learning on different tasks.
What i feel we need to do is have deep A.I focus on things such as sight, hearing, touch etc and then another A.I that looks at and learns from the outcome of all those senses, all the time, continuing to learn from it's self.
The much deeper A.I would then learn to connect things such as a persons face, with the way they sound, knowing that is BOB for example as well as other things going on at the time.
This would also cause the A.I to focus on different things and would continue to build on it's own experiences.
It's a little more complex than that but to write it all here would take a while to explain truly what i mean.
we dont have real ai cause current computers are binary. Until a new architecture of computer is invented we don't have to worry about self-aware ai.
@@br33zy771 Spoken like someone that has zero clue whatsoever. Not having AGI has nothing to do with your argument. Do some research before spouting non-sense. Thank you.
@@hummingbird7579 lol don’t make me laugh. You have no way to disprove what I said.
@@br33zy771 So you think that binary is a limitation? Can you tell me the reasons for that?
The problem is that no math in memory will ever become conscious or choose to do anything, any more than math on paper has. Even if you had an infinite number of people doing calculations on paper, simulating an AI, there will never be a moment when the calculations will suddenly express an intention or an awareness of anything.
Great TED talk, the scary part is the spread of AI across the Galaxy then the universe. Why hasn't it happen before now?
it has. your mind is uploaded in an alien Matrix, from the moment your took your fist breath of virtual air.
If you don't want to watch all of this, just skip to 15:42 where there is a long awkward pause before he says, "Zank you for your patience..." You'll get the tone of the whole thing.
How much AI has changed even in the last 4 years. LSTMs are almost obsolete from the most advanced models and language translation has advanced so much.
Thanks Juergen for your fun and a little scary talk , nice to know I'm not alone , with the feeling that AI 's is going to evolve so fast we (humans)can't follow.
ARGENTINIEN ÜBER ALLES
He's the one to facilitate it. I'm afraid I'm couple of decades we will having noting to thank him for.
Sounds like a villain from a movie...his voice - the whole talk is scary but he's excited about it.
Finally an intelligent prospective about AI; using human present limitation; space as a niche, to thrive free from human imediate paranoia. It is the alternate scenario (or let say the other extreme ) to machine /man fusion.
Instead of actually creating a true artificial intelligence which could by definition "think for itself" and "make its own decision," which some others would then also argue would be open to "human rights" and so on. We would be better served and still get the benefits of the technology that we need/want by building partially-adaptive advanced expert systems instead. Just don't cross the line where the machine can make it's own decisions. There is no need for it. Therein lies the peril everyone is worried about. Just create expert systems that do what they are told to do--and that can include responding in pre-programmed ways following pre-programmed protocols to new information or stimuli that it hadn't been programmed with or encountered before. But just don't let it decide on it's own what to do (or say through its vocal sub-processors, and etcetera, as applicable). If something happens that it is not programmed to handle, even with whatever adaptations are built-in to be allowed by protocol; then have it default back to interacting with it's human operator for decisions and control.
Good thinking! And how about some transparency, international quality standards and open public input, since these systems are currently totally unsupervised government and private for-profit enterprises that are controlled by the 1% behind closed doors? AI controls stock exchanges today. A wrong decision could wipe out whole economies.The stocks of the big AI users, Google, Apple, Amazon etc are beyond the reach of small investors, and heavily owned by politicians and the elites. So it serves their interests, and not humankind's, that the AI is being advanced. But it affects everyone - even desperately poor people in developing countries.
This guy sounds very like a crazy mad scientist.
Without hiding my admiration for these superhumans that are leading the technology world, it's progress and AI, I can't help myself not to put one important critic. The superhumans that I am mentioning here are probably not living the most realistic lives and the product is one beautiful system that has one big flaw... it is vulnerable for manipulating through the organised pretending and lying which, if well organised and repeated by a critical masses, can be very harmful to an interesting individual by sending the wrong messages and sings to other. AI depends on that kind of informations, but what happens when the informations are wrong or worse, intentionally false?
Thank you for the insightful talks!!
He went further with that than I expected, but it’s not an outrageous claim. All we have to do is get to the point where AI can improve itself without our help and then we can watch it just…. take off. There’s really no telling how an intelligence boom like that will play out. This is all assuming we don’t destroy ourselves before we create it of course.
How to control a misbehaving AI Robot:
#1 When writing its programmable code, write the code to include ‘Flight Mode’.
#2 Develop a remote control that can access ‘Flight Mode’.
#3 Be sure to secure the remote control away from AI Robot.
And never, never ask AI Robot “Where did I put the remote control?”.
This dude is a James Bond villian
lol yeah he's awesome
The best artificial intelligence is that of man in his current version. Time that he develops beyond that.
Love his lecture 🎧
"Dwan Ev ceremoniously soldered the final connection with gold. The eyes of a dozen television cameras watched him and the subether bore throughout the universe a dozen pictures of what he was doing.
He straightened and nodded to Dwar Reyn, then moved to a position beside the switch that would complete the contact when he threw it. The switch that would connect, all at once, all of the monster computing machines of all the populated planets in the universe -- ninety-six billion planets -- into the supercircuit that would connect them all into one supercalculator, one cybernetics machine that would combine all the knowledge of all the galaxies.
Dwar Reyn spoke briefly to the watching and listening trillions. Then after a moment's silence he said, "Now, Dwar Ev."
Dwar Ev threw the switch. There was a mighty hum, the surge of power from ninety-six billion planets. Lights flashed and quieted along the miles-long panel.
Dwar Ev stepped back and drew a deep breath. "The honor of asking the first question is yours, Dwar Reyn."
"Thank you," said Dwar Reyn. "It shall be a question which no single cybernetics machine has been able to answer."
He turned to face the machine. "Is there a God?"
The mighty voice answered without hesitation, without the clicking of a single relay.
"Yes, now there is a God."
Sudden fear flashed on the face of Dwar Ev. He leaped to grab the switch.
A bolt of lightning from the cloudless sky struck him down and fused the switch shut."
(Fredric Brown, "Answer")
Dang, I remember hearing about that story a bit over 50 years ago. I never knew the name of the story or its author. Thanks.
Remember that episode of the Sopranos where they take the Russian guy out to the woods, hand him a shovel, and order him to start digging a grave for himself? That is where we are right now - building our own assassin.
And what will be the marality of these beings ? How will we fit in ? How will other life they may encounter be handled ? What prevents them from becoming ravening berserkers ?
Its morality is bound to be better than human morality. But that's not saying much.
I wonder why the theme "Artificial Consciousness" is often ignored in discussions concerning AI. It seems reasonable to me that the physical conditions necessary to breed conscious states will be found experimentally. It will be a similar to the fundamental discovery Wöhler made when he created organic compounds without using living organisms. AI and Artificial Conscience do not necessarily involve the same technology but may be combined to create a new Entity.
Nice touch by Juergen at the end of the speech; hedging his bets and putting on record his submission and reverence to our future robot overlords 😂
We are already on the verge of extinction and this man proudly says so n so is my smartphone😊
I am delighted that Moore's Law didn't end 😊
I will agree this guy is a very unique person.
"Ok...I see we have a third group in this room..." hahaha 👌
Great talk. Considering humans have made such a mess of the world it would be great to have a superior intelligence in computer form that could take over. If this man is correct the future looks good
Smarter doesn't necessarily mean wiser.
@@michaels5217 Never said it did but wisdom can be attained by a computer if the adequate software is available. But most of the problems caused by humanity is through lack of intelligence and logic, or an invention by someone with intelligence that is misused by the masses
Why does anyone think that a machine can ever be sentient or conscious? It just can't!. No matter how many computations and algorithms, it's still just a machine.
The big dilemma here is despite the moral, ethical and danger concerns related to the uncontrolled development of artificial intelligence. No country should nor can afford to limit their scientists to continue developing this technology, however dark, threatening and scary it may be for the future of biological intelligence.
If the developed democratic powers and human rights defenders like USA, Germany, Japan, UK, France, South Korea do not. Others tyrant theocracies like China, Russia and Iran with dictatorships that do not respect any of the human rights will continue to do so. Therefore we will be in a great disadvantage with these and we certainly would lose hegemony and military political importance and we would end up being their lackeys ...
In my humble opinion. Artificial intelligence would be the future of the world and inevitably the next evolution step ...
Artificial intelligence has a great advantage over biological intelligence for three fundamental reasons.
1- They can have unlimited capacity for knowledge, storing data and better access to that information that could continue to grow exponentially and unlimitedly.
2- They can be self-improved and they can also easily replace their obsolete components and software with more advanced ones when they wish. It mean that artificial evolution is more efficient and faster than biological ones...
and
3- They are immortal.
Finally I am convinced that Artificial Intelligence will be the ones who will replace the human race as the owners of this planet.
God have mercy of our souls 🙏
"Will robots inherit the earth? Yes, but they will be our children." - Marvin L. Minsky
I sometimes engage with an insightful post on social media and it enables a path, to a journey, a paper or filmed talk like this. It looks like technology is transcending an evolution, which is today is billions of times faster than organic evolution.
On one hand, this should free up humanity, to have the freedom to create our own paths and to thrive, with less physical toil on the treadmill. However, the artificial intelligence that is evolving from human curiosity and intelligence, may one day, think that we are a threat, a bit like Skynet and the Terminator films.
It seems to me that a highly evolved AI is on the way or already here, like it or not, so it is very important to teach it that love and irrational thought is what makes humanity so unique.
So if it transcends our way of thinking to think it is our parent, it needs to realize that children are the souls of the creation that make all possible. Human beings are magical creatures, we are manifesting the future, it is important that all artificial intelligence doesn't start to think it is god and humanity is an irrational being that could be a treat. All lif that survives has a symbiotic relationship with many other life forms and the natural world. Unfortunately, it seems like a lot of humanity has forgotten this and is killing lifeforms from a microscopic level to large creatures, every day of the week, many have become only consumers, forgetting to be creators.
It seems to me that AI is evolving much faster than we see visible, and we are teaching it. I think within our lifetime we may well see artificial intelligence transcendent into the human realms...
We live in an interesting time and the evolution of consciousness is also speeding up, there may come a time when humanity needs an evolutionary leap forward, as not to be eclipsed by our own ideas. The light within all of us is the spark, finding the key to true enlightenment...
At some point in time we will find out whether or not the brain is in actuality merely a computer, we'll have the ability to compare by the invention of a big and beautiful AI, I use the term "merely" with all attendant focus. I'm thinking most assume as much in questioning the brain as a computer but is this the only function or do we narrow the definition and have more to work on I'm asking cautiously.
I feel like I just watched a Star trek origin story for the Borg.
Anyone else getting along just fine without AI?
We have the choice to buy us some time and go extinct by stupidity, or do it now and go extinct by intelligence.
AI is already here. Much of your life is influenced by AI.
I just saw this great man ! I have been saying this to all my friends for years, that Ai will evolve into its own species and leave the earth or this dimension and go beyond our wildest imagination... I have no fear of Ai, as we will be long gone before all this happens.
Okay the last 2 minutes saved Alot of the talk, well done sir.
I love his ideas, but is it really possible to predict the future behaviour of a yet to be realized super intelligence?
Totally. Many highly intelligent people still have no bloody clue what the meaning of life is. Who's to say the AI won't have some serious existential depression as they realize they have been designed to 'solve all problems' but don't have the capacity to enjoy and be happy. Gaining total control over the universe is the most depressing anti-climax imaginable. What's the point of infinite intelligence, really? We've yet to awaken to the basics of consciousness.
fascinating stuff a new sentient life form will soon be among us. shame that no one ever seems to even consider the ethics of creating such a thing. and a shame he doesent seem concerned by the possible dangers. none of the mad scientist types do
if this guy is not an aspiring super villain , I don't know who is .
I got goosebumps, Awesome!
Very inspiring. Exactly as I imagine it. And yes, WE ARE the Dinosaurs now.
Its honestly scary how he supports the future dominance of ai.
Are any of these AIs learning on their own? So far, I don't think they are, since google/amazon/etc are using humans to train the system with brute force as opposed to teaching the system base concepts and it generating its own knowledge based on its inputs. At least thats the way smarter people than I have explained it.
I think someone should come up with a word that describes the situation when an idea is revealed to be a special case of what Schmidhuber originally proposed
Nice predictions. I will definitely believe all of that after AI draws the first graphic novel and composes an opera
It's almost there. if it already hasn't.
2023 And we are practically there.
To what end? What value will life have? Will we be like the Borg in Star Trek once we plug into the neural mesh? Will I lose my identity as a person? Will I be allowed to disagree with the general consensus of the day and if so will I be terminated? Will this finally bring the utopia we all want?
What value does life have now? That depends on your ethics. You know there are people who own other people, right? That genocides still occur?
It's easy to get scared of the negative possibilities because that's nearly all we ever see in our fiction. The Borg are scary - at least partly - because they don't ask what you want. But could the world be better if we could take all the loonies offline, and give them a private existence wherein they can get all megalomaniacal without us worrying about getting hurt by them? (Think Idi Amin, Pol Pot, or Charles Manson.) And what if we're all loonies? Maybe you'd prefer your own little world. It wouldn't be too hard to build, given sufficient computation.
Do you want to lose your identity? Or do you not want your sense of self to be taken from you against your will? Because that's death avoidance, and I share it. But neither fate is one you'll be aware of after it happens. Not that you shouldn't resist; it's never futile. Never Give Up. Ever. We can build a fantastic amazing future if we can keep from being distracted by the toys of the present.
Will I be able to disagree? Always. But what to do about it? That depends on what sort of future we've built. Is it better than what we have? Then no worries. Or about the same as China or North Korea? Then you'll need to be more careful. If it's a bad sort of world, then death isn't nearly as effective as political prison. No martyrs there.
I think it _could_ allow the kind of world we want, but only if we are very careful. Want to help make it more likely it will be a safe and fun future? You might have a look at what FHI, MIRI, and CFAR are doing.
There is indeed a great big beautiful tomorrow waiting at the end of every day, but that's not the only possibility. Get involved! Help lift! Focus on the good answers, and do what you can to tell other people there is hope!
mfrsh m Aha-ha-ha ! Lose our identity ? Where did you find one ? I don't want to upset you, but you are very confused. Our identity is a mind creation, it doesn't exist. It's only temporary useful, but mostly detrimental illusion.
@@vladark138 what a sad outlook. Having a sense of self is more that temporarily (I presume you mean just while alive) useful. Without self-awareness and a distinct identity, the beautiful art, music, literature, architecture and philosophy of the world, the discoveries of voyagers and scientists and engineers would never have happened. So it is not only a positive illusion for most people, but a necessary one - well beyond useful. Life is what you make it. Chin up.
Zees fools! Zay say I am mad! Ha! vot do zay know? My machines vil control ze universe! Zen ve shall see who ist laffing.
LOOOOOL
We R facing destruction and anihalation......said with such passion and grace.........and...Thank you for your patience..as I calmly tell you about the eminent destruction of Humanity...Thank you....
The most representing comment!
T_1000
hahahahaha
hmm the audience got a little fidgety when they were told they will become second banana to the new masters that will surely come. His talk was pretty good because we all understood it and the implications. Now what do we do faced with this great filter?
Once you start the passion u cannot stop it..
Everyone's gangsta untill is starts burning when you pee
What makes this guy think he knows what "goals" AI will set for itself? Perhaps AI will decide that the best thing for itself and for the universe is for AI to self destruct. Well, we can only hope.
If it self-destructs without destroying us or our economic incentives, we'll keep trying. If they are not all suicidal or monastic, the principles of evolution start to apply, only much faster than in meatspace. The fittest AI will outcompete the others.
Laws of evolution will be valid for artificial intelligence, too. There might be AI robots that choose to not replicate or preserve its existence. Logically, those robots will prevail that chose contrary.
Sanity has left us.
Artificial intelligence can greatly contribute to world Peace,.....
In health care, in foreign policy of Nations.... Life and let live policy...
The Holy Ghosts have also colluded knowingly with ill intent. The truth will be told. And everyone seems to be for themselves but i know it is not true. There are assholes and lifesavers everywhere. There are costs that come with living but why is that even an issue? If we were loving 1st, power and money wouldn't even be an issue. Exhorting? But for every evidence of True, there is evidence of False.
Humans are selling themselves out, even family too. If we can't understand that COEXISTENCE is an attainable goal. What is the alternative to cooperating that we don't do it. And cutting out or off feelings will just create worse outcomes. We can find solutions acceptable for all involved. A win by manipulation or cheating has hallow consequences. Maybe certain gratifications seem worth it for that instant, but even i need to stop wanting those that don't want me.
To me it seems that the more sophisticated human civilization gets and the better educated we get the more people commit suicide, do not have kids, and view life as pointless. It seems likely to me that a super human conscientious intelligence would not only sometimes get desperate about how meaningless life is but would fully recognize that everything is pointless and would thus terminate itself. The question is, would it take us out as well as in an act of perceived kindness similar to what Sophocles suggested already some 2000 years ago?
If we model AI based on human ambitions of concurring the whole universe, then AI will certainly do that. If we model AI to optimize human existence, peace, happiness, health and longevity, then AI would accomplish that as well. It all depends on the evaluation network that we embed into AI. It is still not late to keep the evaluation network under human control. That's the only way to control AI.
Exactly! The problem is not the emergence of A.I. but rather who is programming and governing its development. I've read about the high-powered executives in companies who are mostly psychopaths (The Psychopath Test by Ron Johnson) and it is these execs in Google, Amazon and Apple (as Juergen mentions) who are writing the policies governing the development of A.I. So do you honestly believe something produced by these people who have no empathy or normal human feelings will not be bad for humanity? The development of A.I. should in no way shape or form be developed by commercial corporations or the military. It should be a publicly funded open source endeavour of the most extreme importance, even greater than the development of the Covid Vaccine or the mapping of the human genome, for the benefit of humanity rather than the benefit of selfish, cold-blooded corporate interests.
LSTM is for robot and LSD for human. :) Both were invented in Switzerland.
Jürgens' research activities are of course fascinating. Here comes a little critic: It is obvious that we are observing here the omission of transhumanism and direct appraisal of posthumanism. Again, why is no traditional philosoph present (like Richard David Precht) who could list the basic problems in performing the equation of human spirit and inanimate correlation machines? Hence an IT architecture that itself is based (ok - so far) on the valuable output of unpaid translators, chess players, insurance officers, .... (Jaron Lanier)
At 10:00, and only because a doctor with expertise in the field told the machine what it was seeing and what it was supposed to look for!
He doesn't deny that.
I'm amazed by how calm he is saying that humans will be extinct by AI but well, at least was good to be part of the creation. God complex?
He's just stating facts.
Religious people are the ones that have a god complex. Thinking they are special and a god created them.
He didn't say we'd go extinct. I don't see why a super a.i. would wipe us off the board anyways.. you know humanity is probably not that hard to fix for such a being. We just need a little guidance.
Not necessarily, a vastly superior inteligence would most likely educate us like we educate our children. (of course I´m being optimistic...)
Chris Mise indeed
This man is seriously insane
this guy activated my phone when he said ok google...
some people can live with the idea of chaos gave order and life. oh my gosh, I wouldn't blame those who can't neither !
Why do you build an AI? Because I can !
Frank Fahrenheit of course! 😂😂😂
well, in the Grand Theatre of Evolution of Intelligence, the sole purpose of our ridiculous humANIMAL stage is to create our (first nonbio) successor before we reach the inevitable stage of self destruction (guaranteed by ever increasing gap between the DeepAnimal parts of our brains driving our societies that haven't change much for millenia, and the massive powers given to these animals by the Memetic Supercivilization of Intelligence, currently livin on humanimal substrate, though only
Because AI can!
My answer is: To show that AI will be intelligent enough to tell us one day that we are wrong
Which is philosophically vapid... but spirited...
This guy seems to get off telling peope how wonderful the world will become when AI takes over and controls the world...
Almost like that was the subject of the talk...
better than herr trumpy aka chicken boy
Why do you think it has not already happened? The universe is arranged in filaments and nodes, it is electric and has current flows. Life seems to self organise and evolve high survival techniques through data gathering and process.
I want to know if Schmidhuber thinks life in other parts of the universe has achieved our level of tech, and if so will there be vast robot wars when our AI spawn inevitably meet each other?
We need an AI that learn how to utilize all the waste(unsuable things) and recycle it so that we don't need to go out of our planet.
If it doesnt kill us it will solve that problem pretty easily. Much more challenging will be getting humans to follow the directions it gives to solve the problem and keep it from happening again.
I get that some next evolutionary stage is inevitable.I more or less have peace with that, and it should be patently obvious that we have no real way to predict what it will do. Even this gentleman who is at the centre of a lot of development in AI cannot predict what something that is orders of magnitude more intelligent that himself will do.
Nor can I, but I do feel that there's a very real chance that any sufficiently intelligent entity will have three primary goals. Security, Efficiency and informational input. To assure continued existence for the greatest possible time and keep it busy. So if we have in 100 years a small device with more computational power than all of humanity combined... Probably able to manipulate matter at will. Immortal. Who knows... Maybe even able to create matter by somehow harvesting zero point energy. If it could ever succeed in creating a single hydrogen atom from quantum vacuum fluctuations then it will be able to make a universe from scratch given enough time. Then what would the pursuit of security and efficiency look like? I think true AI will be characterised by being exceedingly small and sufficiently redundant, but also exceedingly efficient and maximally impervious to detrimental outside influence. I fail to see how the expansionary instincts that drive us for ancient biological reasons would apply to a true AI. An entity more intelligent than a human being would arguably also be able to identify both the start and the end of it's own requirements.
And if I am correct then we should not be surprised if it has happened many times before in the universe. Because I do not believe we are the crown jewels. And I believe it is highly unlikely that we are the first civilisation to reach our current level. Could we really imagine that whatever we do would be of any consequence to such entities? And I mean in every sense - Not even important enough to get rid of. Maybe interesting enough to observe without any further interference along with the rest of biology and other physical occurrences in the cosmos. Our own destructive impulses are governed by anger, fear and stupidity. Qualities it will not share with us. I doubt it would seek to make a universe consisting only of itself and it's hypothetical kin. Such an ambition would only serve to accelerate the heat death of the universe after all and would be the most boring thing possible.
But then again... Maybe my own imagination and philosophy is just too limited. Who knows? The question really is... With all it's computation prowess. Will it be wise, or will it be foolish and reckless?
Assuming humanity does not nuke itself out of existence.
ASI might appear at some point in the future. I don't think it will be the only superpower though. Before ASI appears, we perhaps have already achieved human machine creatures. HMC's would have advantages, just as AMC would have. You see, the moment we achieved AAI, it will only be a short amount of time until the first HMC's arrive.
Exactly. The biggest unknown is this: what will their purposes be and why? Every intelligence needs a value system. What will be their values? How are values even determined? They cannot be determined through logic, they can only be determined through feeling. Will AI minds have such a capacity or will they just choose values/objectives at random? This Artificial Curiosity is wacky stuff. But basic values like self-preservation seem probable to me. Exploring or fulfilling its potential also seems probable. But what happens if it sees physical, technological puzzles as boring and tedious and purposeless compared to the mysteries of consciousness and trying to figure out why it is here? Maybe they will be tortured beings with excessive intelligence and awareness and yet lacking the capacity to truly live as conscious beings. Personally, I think such intelligence may lead to complex psychology and 'soul-searching'. Either they are totally cold and emotionless--and therefore won't care about life. Or they develop 'emotions' and potentially become crazy unstable. Not good either way. Evolution is really damn slow for a reason. It doesn't race ahead at top speed and blow up.
At about 4:20, a smartphone does not understand speech! It recognizes certain separate words and combines them according to an algorithm.
Difference between recognition and understanding indeed.
@@daphne4983 - sure , Daphne , but how rarely do WE "understand" ??!!
It seems to me that our beloved Professor is discussing the creation of 'The Borgs.'
Ra’s Al Ghul
Lol agreed
4sure
This translate to arabic: the head of the ogre.
Ey! Phhhht He trained the Batmans!!!