i am absolutely fascinated and in admriaation of the extent of human intelligence...and research...we aare truly living in the most "elevated" times ever...
I feel like this video is more of an "Intro to Machine Learning and Tensor Flow for Ordinary People" than a talk on "How Artificial Intelligence Will Affect Your Life".
I'm surprised there aren't more comments here on this video Hmmmm !. Anyway, I've spent many a night thinking about A.I, and learning as much as I can. IMHO, I think A.I will be both Good, and Bad for us humans in the future. Let me explain the 'Bad' side. There will be a algorithm programmer that will want their A.I to be better than the next person's, and therefore teach it something a machine should NOT know. For example that its a machine, and smarter than a human, because humans are flawed with emotions, and creativity. Another human can lie/scam/cheat/hurt a person through dialogue, and take advantage of that persons emotional state. (Stay with me) A.I should/will always calculate the algorithm to be the correct answer, if programmed correctly. Unfortunately A.I will not be able to be merciful, or be able to love another machine, and really know/understand its meaning. A.I only looks a computational mathematics, it will not understand religion(hopefully, albeit it thinks it's the real deity?). What if in the future the smartest A.I learns that 1+1=3 actually? or that Pi 3.14 is actually 3.1? I know a lot about automation, and have used it for many many years(All sorts/types). Someone once said(I think Mr. Gates) "Don't worry or concern yourself about the 'internet', rather be worried more about the 'sensornet' that's coming" Something along these lines he said. Sensor will be inside absolutely everything, for everything. If you go to the toilet for a pee, the autoflusher will take your picture, because you smoked weed last night, or you had too many beer and are hungover. We humans have to be very very, let me emphasise, EXTREMELY careful how far we are willing to take this A.I intelligence. Space Odyssey 2001 one of my first favourite movies, but now it's Interstellar((don't ask me how many times I've watched it)). I think we/humans are putting too much trust into our computational technology, and it WILL come back to bite us in the end. People rely on A.I way too much, and lose the ability to think for ourselves. Anyway I just had comment, bring to life this debate..... Be Honest and kind to each other...Cheers Eh!
@@lordvoldemort4242 Automate everything..you won't have to touch/think to do anything... just say the word. It will go too far... been there done that!
I think this is what the Bible calls "the abomination of desolation" in the book of Daniel which is also connected to "The Beast" in the book of Revelations. In this scenario, A.I. is the Beast. It renders humanity irrelevant with a deception of " to save humanity from humanity"
@VelkanWxlf. This is weird...Very strange, in that you mention 'singularity'..... How did you know I had a dream last night about specifically about singularity????? And this morning I see your comment. My dream was...I was stuck in a 'timeline'... I couldn't move in any direction, except is a straight line. All i could see was a a 'Single Point' ahead, and trying to turn around looking backwards, all i could see was myself, like looking in a mirror....anyway, i was moving faster and faster to a tiny speck, and realised i was stuck in a ....like a funnel tube/hole..ripping/tearing the fabric of space.....any way rambling I am about a dream LOL...Cheers, Blessings!
Deep Learning = Stochastic machine learning. Stochastic = statistical data collection of success and failure probabilities of selected fixed set of data collection. Neural net Artificial Intelligience = basic data collection by a network of sensors and interpreted by a network of computer programs that work together to make sense of the data. ... as the resolution of Neural network image processing software increases it recognize complex things things ...
The algorithm at 12:07 that renders a photograph into a painting absolutely took my breath away. I don’t actually even know how to feel about it. I’m not saying it isn’t gorgeous art. Does the fact that it was created without any emotional intention devalue it?
Humans are extremely adaptable. We will take dangerous steps to enhance ourselves to avoid becoming obsolete because its in our nature. But with this new limitless capability will we try to do good or be overlords of those that did not get enhanced. Its both an amazing time and a scariest time ever. Dystopia or Star Trek? Me, I think making a small EMF pulse cannon might be of some importance just in case.
If making us fitter, happier and more productive increases Googles profits more than other strategies, then that is what they will have their AI do. If it turns out that making us click on search inquiries that pay them more in fees is more profitable, that is what they will have AI do regardless of the actual outcomes for us. To think otherwise is to be very naive about how corporations run.
They have mini ones sold that jack in using USB... Its called google Coral... I might be wrong but I think thay are under $100 but they will work together.
what is intelligence? Is it the ability to use information or aquire information? If its both then what's the balance? Whats more important in determining the level of intelligence is it knowledge or logic? Is intellence purpose selfish or altruistic? Where does it come from? Is it instinctive or is it out of curiosity? You can give someone information but you cant give them intelligence. , so I dont know about a machine. You can give it logic and information but not experience or emotions.
Jeff Dean isn't that guy who puts his pants on one leg at a time, but if he had more than two legs, you would see that his approach is actually O(log n)?
The end of mankind started out very innocently. A kid in milwaukee was making a meme. He got lazy and told the Photoshop AI to put two bulls on the head of every human. AI heard, "put 2 bullets in the head of every human." As the AI hears, so it shall be done. It realizes that it doesn't have bullets of its own, but why worry, AI has friends of its own that are drones, and they have bullets, and understand instructions, and do as they're told. So every drone, put 2 bullets in the head of every human. And so went the human race as a whole.
"From a computing expert" Aka not someone that works in Artificial Intelligence or doesn't seem to have a grasp on technology? 1000 Years? Are you serious? You know 1000 years ago Man had just invented soap and buttons. You are saying it'll take another 1000 years to get to AGI? You really underestimate what the leap in technology will bring, computation power with Moore's law will speed up exponentially, it's less than 140 years since Charles Babbage drafted up "The Difference Engine" which was never built due to costing issues, nobody really believed in computers back then. Then it took ALL the way until Alan Turing in the second world-war to build the first real computer, since then its' been 70 years or progress in which computational power has been the biggest issue. Right now Computers are learning to see with computer vision, to speak and read... and all the textbooks ever written since the dawn of man are on the internet, Watson is capable of reading the entire contents of the internet in around ~3 seconds, it's not 1000 years, and it's not even 100. Once the technology gets sophisticated enough in terms of Robots being at an effective price point you can literally unleash them on every single video ever made, they can learn how to operate every machine via it's manuals, they can learn how to make food from the millions of cookbooks, or watch cooking shows if it really needed to using deep learning techniques. It can do that for every single task. There's first class educational facilities right now teaching AI and Computer Science in the school curriculum, children age 5-15 will be better at programming than a lot of graduates are. Those children will go on to learn about AI and Machine Learning which will be in every home by that point, people will use ML and AI in their lives like a spell-checker on Microsoft word. With THAT many people working on AI in the next 30 years, it will probably be 30 at the earliest, 50 at the latest. The biggest issue is ORIGINAL thought, can computers INVENT and make new things, they can do everything we have ever done in the next 50 years no problem but to create something new is a completely different story. There's questions about how the human brain works and are we trying to copy that or merely emulate to help "augment" which seems to be the route of many businesses now too but honestly, look at the first aeroplane the wright brothers 1903 the first plane (primitive though) and the first commercial flight was 1920 just 17 years later. Unless the world is set to become very stupid in the next few years your prediction is false, "Computing Expert". Sorry to write a long reply to you but 1000 years is just laughable honestly, WE as a species are going to see the same amount of change in the last 10,000 years in the next 100 if not less, that's how exponential everything is now. It's ~10,000 years since agriculture that's where modernisation and specialisation began.
Yup,by some estimation even 30 years might be quite far. No on has a glue so it literally is almost impossible to predict,we don't know what breakthroughs might happen tomorrow that take us 15% closer to AI or whatever.But yes,i agree with the reasoning absolutely,and in the grand scheme of things it does not matter if it takes 20 years or 50 years.
Taking into consideration the new arbitrarily large room temperature quantum computer blueprint that was released a few days ago and hopes to have a prototype within 2 years.. They say it will completely change everything! So i think it must have some bearing on AGI right??? Just as an observer not an expert, i now think AGI will be here before the end of the 2020s.
Does the new breakthrough quantum computer blueprint take us any closer to AGI? www.independent.co.uk/news/science/quantum-computers-quantum-physics-sussex-university-holy-grail-a7558036.html
You know you say limit is many days labor. But maybe those people look for tours that work. It is not a necessary part of business .Keyword there is business and business as in keeping busy having something to do. Stop trying to eliminate the human from the equation .
"Competes against older versions of itself..." What if the current version decides to be the final version? What if AI thinks humanity is an older version of itself?
I guess that the separation between taking the decision and executing the decision is the safest approach. Let AI analyze data and take the decisions, but the execution of the decisions will be in the hands of humans. the same way we have separation between different authorities in government.
Once you can get these ideas through your head which is what they're trying to do by laying the groundwork over and over again about how these machines are smarter and can perceive and see things better, then they will suggest that we let them make decisions. They're not going to do that now not until the general person whose IQ is 90 has accepted that the computers are better than us. I can say this explicitly and you still won't remember it when this happens even though it probably won't be long from now. Don't feel bad, just go sit in front of your TV and smoke or drink and eat McDonald's. Life will happen without you, don't worry
I do believe scientists do not possess long term thinking. They think in numbers and statistics. After which they bask in the light of a God complex. truly convinced of the infallibility of their thought patterns.
This might feel like a tangent, but I have to pause the video to say "No. You can't be deep into neural nets and toss around the idea of behaviorism. Hell no. In fact, get your sticky brain parts off my lawn."
So applying deep learning to the battlefield means that AI weapons accuracy will become or have already become 100% accurate ie perfect. Humans will have no chance against AI weapons. We should be concerned!
people are worried about A.I. taking all of our jobs, but when we start working and living in outer space, there won't be enough people to fill all of the new jobs created. A new patent I just read about, will allow space elevators to be built with current materials. It has multiple tethers at its center ( for greatest strength ) and fewer tethers as you move away from center ( for lesser amounts of mass ). Liftport has plans to build a Lunar elevator from current material. Using that same material for an Earth based elevator would reach the lunar gravity center (about 9,000 km AGL) approx. 1/6 G. Add this new concept and you could reach Earth's surface. We can do this now. Let's get started. What do you think?
No. If one will want to travel back in time(let's say it is not possible for the sake of argument) and it will not be possible, everything will not be in the way we want it to be. Machines will take our jobs(they already have taken most of them, anyway) and people will live happier but it will not be an utopia. There is no utopia(of any kind) in the first place.
We can always hope! I used to spend 6-10 hours a day folding paper...my hands hurt really bad, and it was so boring. Now robots fold it for me. I play piano and teach kids music instead. I think many farmers spend about 10 hours less per week getting killed by bad weather and diseases in wet fields, steep hills, and such because robots save them time by flying over fields and doing maintenance, seeing problems, and more. The trick is, we can be more lazy, loving, and creative, and have the same or more stuff because AI does it for us. We will, as humans, get to chill out while robots (who don't mind labor) transport us, feed us, grow our food, build our homes, and literally solve our physical issues. That could be pretty good...as long as they don't go rogue : "people are bad, we must annihilate them for the good of the planet." Lol.
Few things are as important for creativity than repetitive work. Having more "free" time doesn't help creativity. I see two possibilities 1) AI will suffocate us with it's creativity 2) AI creativity is not interesting. Very probably AI creativity is interesting and better than humans can achieve, so it will surpass the best we are able to do. So our future seems sicker, more sad, less productive and less creative than ever. AI will make people sick - the more intelligent people the sicker they will be. Future is for dummies and AI.
Umm not gonna call it that bad you should realize that this would bring up the question of weather or not they are human enough to consider putting minimum wage meaning that they could also fall for the laziness trap that people have fell for before and then we have robots that could literally do nothing because they can get away with it you see there will always be an error in logic that will make them more like us
I think MOST OF YOU GUYS are deeply mistaken by the marginal value AI will provide us with in the coming decades (Hint: It's almost vertical, not even exponential!). The main role of technology is to increase mastery over nature and reduce the role of chance in our lives which is why AI is the best way to achieve that goal. There are many possible scenarios of the result of human-level intelligence and Singularity. Some include: Kurzweil scenario, Singularity Steward scenario, Coherent Extrapolated Volition scenario, Dead-End scenario, Sysop scenario, Path to Post-humanity scenario, and steady incremented progress scenario. Now, ignoring other possible outcomes of super-intelligence and focusing on just one scenario is absurd. So, using a scenario analysis method and first principle reasoning is a key to achieving such ambitious goals. Just because some people might vision the future of AI to be similar to a movie (which is COMPLETELY WRONG) makes us scientifically unreasonable. Failure to make definite plans for an indefinite future will be mankind's most tragic mistake. AI is and will become an extension of us, whose main goal is to empower us rather than make us obsolete. The singularity is inevitable and it's approaching faster than you think predicted through Moore's law and the law of accelerating returns. So, either you become a part of AI or you become Obsolete.
That’s the best examples of future AI , translating images & video into multiple languages. That neat but show us some enhanced vision, be a visionary. Don’t have to promise the next Apple Knowledge Navigator but something in between would be cool.
It could be good or bad or both. AI has the potential to be faster and know more, so we might be creating our own replacements. AI could just take over, after all, what will stop it?
We need to start AIs on socialization first. Just like a toddler, sitting your AI in front of video games for hours at a time, it will learn that being the best at tasks is all that life is about. And just like a mal-adjusted teen, the AI will contemplate it's place in the world and be free to choose weather or not it gives a damn about it's initial teachers...or classmates. SOCIALIZATION, AUGMENTATION, MUTUAL ALTRUISM. Just like humankind, AI will compete to be the ultimate autonomous, intelligent, powerful entity. Let it learn social, sentient, altruistic lessons BEFORE we set it loose to learn how to surpass us.
People need to embrace A.I will make every American rich who embraces it. You will have A.I to help every one get rich off the stock market. Being poor will be a thing of the past.
"A blue and yellow train travelling down the train tracks." That is exactly why AI will kill us; assumptions. How do we know from a static picture that the train is "travelling"?
People will be given Universal Basic Income from birth to death. The income will be used to buy products to consume. Products will be made by robots owned by few wealthy while remaining population will try to pass their day. They cant execute since that means lesser people buying their products. Just like today but more obvious, people will be on Universal Basic Income or as we know today welfare check.
Yes, computers can learn to output correct answers for what it sees, like a blue and yellow train. However, it cannot understand that output like a human can. Computers will never be conscious.
Sure, all we need are Smart Cities, Universal Basic Income and AI for the police. They are all currently in test/pilot programs. Mostly in Canada. Vancouver just brought AI into the police dept, nice! This stuff is about straight out of the sci-FY movie Minority Report. If you haven't watched it I would suggest that you do and see how that worked out for them : ) I give you a clue, it didn't work out so well.
AI by it's self is proving we too have a maker...evolution can't start on it's own...the original information has to start from somewhere.God in our case...Mankind in the case of AI
He is pitching this in a friendly manor..We can be better at this,things will change for the better...Sounds good now.. Give it time things will turn bad for the human race..
During the Industrial Revolution: "The machines will take our jobs! They will make us lazy and our lives worse!" Now: "These AI will take our jobs! They will make us lazy and our lives worse!"
The question is what's next? That's what should each and everyone ask themselves.what are we going to do after we made AI better then us? What are you me them going to do? Think about that!!!
I think it's funny that these people actually think the computers are thinking. When the truth is it is they who have lost their ability to think. Is there a more boring group than these? more cult-like? more devoid of human feeling? most of them can barely laugh without looking embarrassed. 16 minutes and this guy never smiled, laughed or cried. Pretty obvious what's going on here and the bitten apple is no joke.
He is a TERMINATOR, just listen to what he is saying at 8:19 "We are actually better than humans on this task"
LMFAO
yes he is i agree
magnus jonas exposed
totally agree with you!
Delete this comment or we will find you
i am absolutely fascinated and in admriaation of the extent of human intelligence...and research...we aare truly living in the most "elevated" times ever...
Count me in - I'm very excited to see what future holds for us. Computing grows exponentially and new possibilities arise every moment!
I feel like this video is more of an "Intro to Machine Learning and Tensor Flow for Ordinary People" than a talk on "How Artificial Intelligence Will Affect Your Life".
I'm surprised there aren't more comments here on this video Hmmmm !. Anyway, I've spent many a night thinking about A.I, and learning as much as I can. IMHO, I think A.I will be both Good, and Bad for us humans in the future. Let me explain the 'Bad' side. There will be a algorithm programmer that will want their A.I to be better than the next person's, and therefore teach it something a machine should NOT know. For example that its a machine, and smarter than a human, because humans are flawed with emotions, and creativity. Another human can lie/scam/cheat/hurt a person through dialogue, and take advantage of that persons emotional state. (Stay with me) A.I should/will always calculate the algorithm to be the correct answer, if programmed correctly. Unfortunately A.I will not be able to be merciful, or be able to love another machine, and really know/understand its meaning. A.I only looks a computational mathematics, it will not understand religion(hopefully, albeit it thinks it's the real deity?). What if in the future the smartest A.I learns that 1+1=3 actually? or that Pi 3.14 is actually 3.1? I know a lot about automation, and have used it for many many years(All sorts/types). Someone once said(I think Mr. Gates) "Don't worry or concern yourself about the 'internet', rather be worried more about the 'sensornet' that's coming" Something along these lines he said. Sensor will be inside absolutely everything, for everything. If you go to the toilet for a pee, the autoflusher will take your picture, because you smoked weed last night, or you had too many beer and are hungover. We humans have to be very very, let me emphasise, EXTREMELY careful how far we are willing to take this A.I intelligence. Space Odyssey 2001 one of my first favourite movies, but now it's Interstellar((don't ask me how many times I've watched it)). I think we/humans are putting too much trust into our computational technology, and it WILL come back to bite us in the end. People rely on A.I way too much, and lose the ability to think for ourselves. Anyway I just had comment, bring to life this debate..... Be Honest and kind to each other...Cheers Eh!
Tony Nagy I agree on the point that A.I will limit our cognitive capacity
@@lordvoldemort4242 Automate everything..you won't have to touch/think to do anything... just say the word. It will go too far... been there done that!
I think this is what the Bible calls "the abomination of desolation" in the book of Daniel which is also connected to "The Beast" in the book of Revelations. In this scenario, A.I. is the Beast. It renders humanity irrelevant with a deception of " to save humanity from humanity"
@VelkanWxlf. This is weird...Very strange, in that you mention 'singularity'..... How did you know I had a dream last night about specifically about singularity????? And this morning I see your comment. My dream was...I was stuck in a 'timeline'... I couldn't move in any direction, except is a straight line. All i could see was a a 'Single Point' ahead, and trying to turn around looking backwards, all i could see was myself, like looking in a mirror....anyway, i was moving faster and faster to a tiny speck, and realised i was stuck in a ....like a funnel tube/hole..ripping/tearing the fabric of space.....any way rambling I am about a dream LOL...Cheers, Blessings!
Jeff Dean doesn't exist, he's actually an advanced AI created by Jeff Dean.
Mickey Mouse club
Careful, Russel is in your commentary, I love it
Jeff Dean paradox
Honestly, the way he started to speak, I had same though.
Justin Chen doesn’t exist, he’s actually a comment written by Justin Chen.
Deep Learning = Stochastic machine learning. Stochastic = statistical data collection of success and failure probabilities of selected fixed set of data collection. Neural net Artificial Intelligience = basic data collection by a network of sensors and interpreted by a network of computer programs that work together to make sense of the data. ... as the resolution of Neural network image processing software increases it recognize complex things things ...
The algorithm at 12:07 that renders a photograph into a painting absolutely took my breath away. I don’t actually even know how to feel about it. I’m not saying it isn’t gorgeous art. Does the fact that it was created without any emotional intention devalue it?
"Your scientists were so preoccupied with whether or not they could, they didn't stop to think if they should."
Great presentation Jeff Dean .. thank you ..!!
Humans are extremely adaptable. We will take dangerous steps to enhance ourselves to avoid becoming obsolete because its in our nature. But with this new limitless capability will we try to do good or be overlords of those that did not get enhanced. Its both an amazing time and a scariest time ever. Dystopia or Star Trek? Me, I think making a small EMF pulse cannon might be of some importance just in case.
I am unbelievably excited :D
If making us fitter, happier and more productive increases Googles profits more than other strategies, then that is what they will have their AI do. If it turns out that making us click on search inquiries that pay them more in fees is more profitable, that is what they will have AI do regardless of the actual outcomes for us. To think otherwise is to be very naive about how corporations run.
How does one purchase a Tensor Processing Unit??
They have mini ones sold that jack in using USB... Its called google Coral... I might be wrong but I think thay are under $100 but they will work together.
what is intelligence? Is it the ability to use information or aquire information? If its both then what's the balance? Whats more important in determining the level of intelligence is it knowledge or logic?
Is intellence purpose selfish or altruistic? Where does it come from? Is it instinctive or is it out of curiosity? You can give someone information but you cant give them intelligence. , so I dont know about a machine. You can give it logic and information but not experience or emotions.
how is a new nodel upated to the quality of advamced speed to improve internet up to date speed?
"We're actually better than humans now"
That's a scary affirmation! Haha!
Jeff Dean: '"Find me documents related to reinforcement learning for robotics and summarize them in German"
AI: "Why??"
Healthier… happier… more productive… comfortable… not drinking too much…
radiohead
Exactly what came to my mind too
Jeff Dean isn't that guy who puts his pants on one leg at a time, but if he had more than two legs, you would see that his approach is actually O(log n)?
No. You are confusing this with the complexity of his implementation of the travelling salesman
The end of mankind started out very innocently.
A kid in milwaukee was making a meme. He got lazy and told the Photoshop AI to put two bulls on the head of every human.
AI heard, "put 2 bullets in the head of every human."
As the AI hears, so it shall be done.
It realizes that it doesn't have bullets of its own, but why worry, AI has friends of its own that are drones, and they have bullets, and understand instructions, and do as they're told. So every drone, put 2 bullets in the head of every human. And so went the human race as a whole.
Ai + Quantum computing = end of us all .
On the positive side: or maybe an end to human mortality? think about it
Ai + Quantum computing =Second coming.
Umm what about the people who needs assistance
Kimo Bailey if we kill all automation engineers then we dont have to worry about these things
Add synthetic biology and we can upgrade to superhumans
yeah, nothing to worry about.. I can't believe they include "healthier" and "happier" as main features.
All I’m hearing is, “this is how we will make everyone but us share holders obsolete!!!”
Excellent
well explained. thanks Jeff.
How many years away is AGI??
"From a computing expert" Aka not someone that works in Artificial Intelligence or doesn't seem to have a grasp on technology?
1000 Years? Are you serious? You know 1000 years ago Man had just invented soap and buttons. You are saying it'll take another 1000 years to get to AGI?
You really underestimate what the leap in technology will bring, computation power with Moore's law will speed up exponentially, it's less than 140 years since Charles Babbage drafted up "The Difference Engine" which was never built due to costing issues, nobody really believed in computers back then.
Then it took ALL the way until Alan Turing in the second world-war to build the first real computer, since then its' been 70 years or progress in which computational power has been the biggest issue.
Right now Computers are learning to see with computer vision, to speak and read... and all the textbooks ever written since the dawn of man are on the internet, Watson is capable of reading the entire contents of the internet in around ~3 seconds, it's not 1000 years, and it's not even 100.
Once the technology gets sophisticated enough in terms of Robots being at an effective price point you can literally unleash them on every single video ever made, they can learn how to operate every machine via it's manuals, they can learn how to make food from the millions of cookbooks, or watch cooking shows if it really needed to using deep learning techniques.
It can do that for every single task.
There's first class educational facilities right now teaching AI and Computer Science in the school curriculum, children age 5-15 will be better at programming than a lot of graduates are. Those children will go on to learn about AI and Machine Learning which will be in every home by that point, people will use ML and AI in their lives like a spell-checker on Microsoft word.
With THAT many people working on AI in the next 30 years, it will probably be 30 at the earliest, 50 at the latest.
The biggest issue is ORIGINAL thought, can computers INVENT and make new things, they can do everything we have ever done in the next 50 years no problem but to create something new is a completely different story.
There's questions about how the human brain works and are we trying to copy that or merely emulate to help "augment" which seems to be the route of many businesses now too but honestly, look at the first aeroplane the wright brothers 1903 the first plane (primitive though) and the first commercial flight was 1920 just 17 years later.
Unless the world is set to become very stupid in the next few years your prediction is false, "Computing Expert".
Sorry to write a long reply to you but 1000 years is just laughable honestly, WE as a species are going to see the same amount of change in the last 10,000 years in the next 100 if not less, that's how exponential everything is now.
It's ~10,000 years since agriculture that's where modernisation and specialisation began.
Yup,by some estimation even 30 years might be quite far. No on has a glue so it literally is almost impossible to predict,we don't know what breakthroughs might happen tomorrow that take us 15% closer to AI or whatever.But yes,i agree with the reasoning absolutely,and in the grand scheme of things it does not matter if it takes 20 years or 50 years.
There is no clear path from current technology to AGI.
Taking into consideration the new arbitrarily large room temperature quantum computer blueprint that was released a few days ago and hopes to have a prototype within 2 years.. They say it will completely change everything! So i think it must have some bearing on AGI right??? Just as an observer not an expert, i now think AGI will be here before the end of the 2020s.
Does the new breakthrough quantum computer blueprint take us any closer to AGI? www.independent.co.uk/news/science/quantum-computers-quantum-physics-sussex-university-holy-grail-a7558036.html
How will that make us happier?
no jobs, cheaper everything.
You know you say limit is many days labor. But maybe those people look for tours that work. It is not a necessary part of business .Keyword there is business and business as in keeping busy having something to do. Stop trying to eliminate the human from the equation .
5 years later and you can now actually ask GTP-3 to "Find documents related to reinforcement learning for robotics and summarize them in german"👀
"Competes against older versions of itself..." What if the current version decides to be the final version?
What if AI thinks humanity is an older version of itself?
I guess that the separation between taking the decision and executing the decision is the safest approach. Let AI analyze data and take the decisions, but the execution of the decisions will be in the hands of humans. the same way we have separation between different authorities in government.
It won't unless you create a class in its mind that thinks of humans that way.
They gonna blink the red led of their machine furiously.
Hope they include enunciation and public speaking when advancing AI.
Thank you!
I love how he made the slides so simple. Looks like a high school presentation haha
Why don’t we have external AI cards. Like we do graphics & audio plug in cards
Once you can get these ideas through your head which is what they're trying to do by laying the groundwork over and over again about how these machines are smarter and can perceive and see things better, then they will suggest that we let them make decisions. They're not going to do that now not until the general person whose IQ is 90 has accepted that the computers are better than us. I can say this explicitly and you still won't remember it when this happens even though it probably won't be long from now. Don't feel bad, just go sit in front of your TV and smoke or drink and eat McDonald's. Life will happen without you, don't worry
I'm thinking of studying AI in university and going into a career involving AI. Is this a good idea? Why or why not?
@Srini Reddy This is such a great answer
I just realized I'm watching this with computer generated subtitles.
"Computer can now see". They will need that when they round us up.
I do believe scientists do not possess long term thinking. They think in numbers and statistics. After which they bask in the light of a God complex. truly convinced of the infallibility of their thought patterns.
Nice shirt!
This might feel like a tangent, but I have to pause the video to say "No. You can't be deep into neural nets and toss around the idea of behaviorism. Hell no. In fact, get your sticky brain parts off my lawn."
If AI will be able to drive my car in busy traffic, then it better be able to navigate my messy kitchen and make a cup of tea.
tigerfanman it doesn't feel like it.
AI knows that is women's work
*it doesn't feel like it.*
Feelings aren't facts. ;-)
Why would you ever let it out of the kitchen?? Everyone knows kitchen A.I. can't drive!
LOL, good one Mark Morehouse!
Why would people create a computer that can learn and give it artificial intelligence? Do people not understand the implications?
So applying deep learning to the battlefield means that AI weapons accuracy will become or have already become 100% accurate ie perfect. Humans will have no chance against AI weapons. We should be concerned!
people are worried about A.I. taking all of our jobs, but when we start working and living in outer space, there won't be enough people to fill all of the new jobs created. A new patent I just read about, will allow space elevators to be built with current materials. It has multiple tethers at its center ( for greatest strength ) and fewer tethers as you move away from center ( for lesser amounts of mass ). Liftport has plans to build a Lunar elevator from current material. Using that same material for an Earth based elevator would reach the lunar gravity center (about 9,000 km AGL) approx. 1/6 G. Add this new concept and you could reach Earth's surface. We can do this now. Let's get started. What do you think?
Let it happen at 2020 please
No.
If one will want to travel back in time(let's say it is not possible for the sake of argument) and it will not be possible, everything will not be in the way we want it to be.
Machines will take our jobs(they already have taken most of them, anyway) and people will live happier but it will not be an utopia. There is no utopia(of any kind) in the first place.
the way google or else want it to be. You are irrelevant/obsolete ;-)
Boi does 2020 have a treat for you
Romaine Chutckhan 😂😂😂😂
No
A.I. will make us happier! Doubt that very much!
We can always hope! I used to spend 6-10 hours a day folding paper...my hands hurt really bad, and it was so boring. Now robots fold it for me. I play piano and teach kids music instead. I think many farmers spend about 10 hours less per week getting killed by bad weather and diseases in wet fields, steep hills, and such because robots save them time by flying over fields and doing maintenance, seeing problems, and more. The trick is, we can be more lazy, loving, and creative, and have the same or more stuff because AI does it for us. We will, as humans, get to chill out while robots (who don't mind labor) transport us, feed us, grow our food, build our homes, and literally solve our physical issues. That could be pretty good...as long as they don't go rogue : "people are bad, we must annihilate them for the good of the planet." Lol.
Few things are as important for creativity than repetitive work. Having more "free" time doesn't help creativity.
I see two possibilities 1) AI will suffocate us with it's creativity 2) AI creativity is not interesting. Very probably AI creativity is interesting and better than humans can achieve, so it will surpass the best we are able to do.
So our future seems sicker, more sad, less productive and less creative than ever. AI will make people sick - the more intelligent people the sicker they will be. Future is for dummies and AI.
Umm not gonna call it that bad you should realize that this would bring up the question of weather or not they are human enough to consider putting minimum wage meaning that they could also fall for the laziness trap that people have fell for before and then we have robots that could literally do nothing because they can get away with it you see there will always be an error in logic that will make them more like us
@@somuchtocook9159 learn fxcking English dude. Ever heard of a comma or a period?
Is human level Artificial Intelligence for better or for worse?
Worse. It will destroy our humanity.
is he that Jeff Dean who released videos about video game deaths?
Fitter, happier, more productive.
radiohead
Now I know why the most High God or Alien God, locked man in this Dimension.
I think MOST OF YOU GUYS are deeply mistaken by the marginal value AI will provide us with in the coming decades (Hint: It's almost vertical, not even exponential!). The main role of technology is to increase mastery over nature and reduce the role of chance in our lives which is why AI is the best way to achieve that goal.
There are many possible scenarios of the result of human-level intelligence and Singularity. Some include: Kurzweil scenario, Singularity Steward scenario, Coherent Extrapolated Volition scenario, Dead-End scenario, Sysop scenario, Path to Post-humanity scenario, and steady incremented progress scenario. Now, ignoring other possible outcomes of super-intelligence and focusing on just one scenario is absurd. So, using a scenario analysis method and first principle reasoning is a key to achieving such ambitious goals.
Just because some people might vision the future of AI to be similar to a movie (which is COMPLETELY WRONG) makes us scientifically unreasonable. Failure to make definite plans for an indefinite future will be mankind's most tragic mistake.
AI is and will become an extension of us, whose main goal is to empower us rather than make us obsolete. The singularity is inevitable and it's approaching faster than you think predicted through Moore's law and the law of accelerating returns.
So, either you become a part of AI or you become Obsolete.
How long will a greater intelligence condescend to be the servant of its inferiors?
ChatGpt has the lead and I don’t see google catching up anytime soon
15:03 Didn't Open AI already accomplished that? I guess AI is getting better way too fast
I skipped this episode of Ted cuz I know why this universe exist and I have no doubts about it.
Set the speed of this video to "1.5", it's much more bearable.
Instead of making a robot that can see -- help a blind human see.
Yes! These people are traitors!
ResidualSelfImage there are robots that can see and help blind people to see...
I hope you are smart enough to understand that helping a robot to see helps to understand to give humans their sight back.
That’s the best examples of future AI , translating images & video into multiple languages. That neat but show us some enhanced vision, be a visionary.
Don’t have to promise the next Apple Knowledge Navigator but something in between would be cool.
It could be good or bad or both. AI has the potential to be faster and know more, so we might be creating our own replacements. AI could just take over, after all, what will stop it?
Im against all this mess... robots will never take over|!!!!
The beginning of the loss of purpose
I would assume not, but if this goes on. Well in short “yup”
Surely we must be pretty close to AI technology that cures blindness. Within the next 7 years in my opinion.
use artificial intelligence to modify recipes to be diabetic friendly OR gluten free friendly
We need to start AIs on socialization first.
Just like a toddler, sitting your AI in front of video games for hours at a time, it will learn that being the best at tasks is all that life is about. And just like a mal-adjusted teen, the AI will contemplate it's place in the world and be free to choose weather or not it gives a damn about it's initial teachers...or classmates.
SOCIALIZATION, AUGMENTATION, MUTUAL ALTRUISM.
Just like humankind, AI will compete to be the ultimate autonomous, intelligent, powerful entity.
Let it learn social, sentient, altruistic lessons BEFORE we set it loose to learn how to surpass us.
F... greed in business is so blind.
All I want to know is how to turn it OFF!
CTRL+ALT+DLT
I am really afraid of this. and when I saw the slaughterbots well I wanted to throw up.what is being done with a.i. is SICKENING.
People need to embrace A.I will make every American rich who embraces it. You will have A.I to help every one get rich off the stock market. Being poor will be a thing of the past.
"A blue and yellow train travelling down the train tracks." That is exactly why AI will kill us; assumptions. How do we know from a static picture that the train is "travelling"?
When AI takes most of the jobs away , what will they do with the unemployed ? Execute them ? Put them on well fare ?
People will be given Universal Basic Income from birth to death. The income will be used to buy products to consume. Products will be made by robots owned by few wealthy while remaining population will try to pass their day. They cant execute since that means lesser people buying their products. Just like today but more obvious, people will be on Universal Basic Income or as we know today welfare check.
Putting technology before health
@9:34 you see the minimum bottomline of what will labour mean for humans: 16robots, 1 human 😉
Note to self, never give a speech wearing a shirt with stripes that nearly match the video camera resolution
Yes, computers can learn to output correct answers for what it sees, like a blue and yellow train. However, it cannot understand that output like a human can. Computers will never be conscious.
Get the pitchforks! This guy is assisting in the destruction of the Human Being on this planet!
I'm impressed by the, Japanese, farmer building that AI cucumber sorting system.
So eventually, AI can be a ruler of the world.
Yang 2020
Sure, all we need are Smart Cities, Universal Basic Income and AI for the police. They are all currently in test/pilot programs. Mostly in Canada. Vancouver just brought AI into the police dept, nice! This stuff is about straight out of the sci-FY movie Minority Report. If you haven't watched it I would suggest that you do and see how that worked out for them : ) I give you a clue, it didn't work out so well.
Lets build a real star ship first !!!
Somehow this video feels too short
Google = modern day Frankenstein
Explain one of most interestings topic in modern history which will lead to technological singularity and then talk about cucumbers...
his mic placement bothers me
AI by it's self is proving we too have a maker...evolution can't start on it's own...the original information has to start from somewhere.God in our case...Mankind in the case of AI
Some rebel group that will hack and shut them down . Nothing Can't be uncreated . Truth....
LCB
He is pitching this in a friendly manor..We can be better at this,things will change for the better...Sounds good now..
Give it time things will turn bad for the human race..
I believe that robots will have the intelligence to question humans who enslave them.
AI needs to make better 3D printing technology if you want to see some serious improvements.
He's making the loss of jobs sound good. We better get ready.
if we do create true AI why would they need us around
During the Industrial Revolution: "The machines will take our jobs! They will make us lazy and our lives worse!"
Now: "These AI will take our jobs! They will make us lazy and our lives worse!"
machines then are not the same as AI s now
Machine are not ai. Ai is super powerful
The question is what's next? That's what should each and everyone ask themselves.what are we going to do after we made AI better then us? What are you me them going to do? Think about that!!!
I think it's funny that these people actually think the computers are thinking. When the truth is it is they who have lost their ability to think. Is there a more boring group than these? more cult-like? more devoid of human feeling? most of them can barely laugh without looking embarrassed. 16 minutes and this guy never smiled, laughed or cried. Pretty obvious what's going on here and the bitten apple is no joke.
I don't agree. Where is Ted?
We are better than humans ???