Jacque Fresco-Human or Robot Systems-July 4, 2011

[rehndawg]

I can't believe this man is considered a quack among the older generations and the corporate media...at his age many are screwed in the head and cannot remember anything...i can sit listening to these videos over a over again....thank you fresco

[venusprojection]

This video confirms that it is possible for machines to take over if they are connected with us and able to program and reprogram themselves.

[JoshAlfred27]

Clarke's Law, which may be formulated as follows: "When a distinguished but elderly scientist states that something is possible, he is almost certainly

right. When he states that something is impossible, he is very probably wrong." -- Looking Forward.

[kosmoszoran]

Well there are self learning programs, neural networks, and in them they deviate from they original programing but in some boundaries. This technology will become more advance in the near future. It will help humanity.

[ShapeShifterKibayo]

see robot will only take over the world if humans program them too in other words humans start there own destruction NOT robots geez people are so RETARDED AND MENTAL these days

[thevenusprojectmedia]

@ShapeShifterKibayo but if we get mad at them it means we don't understand where they're coming from

[ShapeShifterKibayo]

@thevenusprojectmedia lol true true ^^

[kornboy3]

@ShapeShifterKibayo but 'love' robots will never can feel that

[Juefawn]

I think robots can be programmed to learn from their environment, and even be programmed to reinforce their own behaviors. How do you reinforce a robot? By behavior and consequence of success or failure, same way humans do it, but with built in procedures instead of innate needs and sensations to work by. Look up the i-cub project.

watch?v=Jv3RuYkD1tw&feature=re­lated

[DraskyVanderhoff]

Sadly he is very wrong... there is something call Machine Learning and is the main method used right now for make the software ( this soft can be in a robot or any other thing...) learn about the environment and how to interact with it. Please learn about it and you'll know that argument for a diff between animals and robots is very outdated.

[thevenusprojectmedia]

@DraskyVanderhoff thx, will pass along to jacque, can you send me a VERY brief link or something by personal message to brief us on what you are talking about?

[Omgadnowai]

@DraskyVanderhoff

You're retarded. The concept still stands: a program will not run anything it has not been programmed to run. Basically, if you come up with a program that recodes itself depending on what internal/external interactions take place, you didn't get an "intelligent" program. You made it intelligent to the extent that your code allows it to be.

You, sir, are stupid.

[DraskyVanderhoff]

@Omgadnowai And you are very religious for get angry and threat me like that, there is no need for that kind of behave.

About my point ... that exactly what we do. We reprogram ourself , we call it reinforcement!. We have some basic desires that drive us to do certain things and we reprogram some of those desires as life goes on. A machine with the system that i explain works exactly the same way ( maybe right now is rudimentary but the basics are there ).

[Omgadnowai]

@DraskyVanderhoff

I quote you "Sadly he is very wrong..."

He wasn't wrong. What he said was grossly how robots work. Programmed to reprogram itself. That doesn't make him wrong. The robot is doing specifically what it is programmed to do, and it can only reprogram itself within the limits of what parameters he can perceive, just as can we; it can only reprogram itself using whatever means we gave it to reprogram itself.

Animals and robots are nothing alike, unless you program hormones. Lol.

[WJValente]

@DraskyVanderhoff Do you really believe what you say or are you just on an ego-trip?

[DraskyVanderhoff]

@WJValente Please search in youtube about machine learning and neural networks. Is not what I believe, is a fact

[mu6i]

@DraskyVanderhoff but anyway you probably need to code a method of learning. Which poses limitations, I suppose. Adaptive learning should be very neat, too, as far as the goal is carefully defined (feed all the people that demand that)

[DraskyVanderhoff]

@mu6i A program that can learn... can learn very easy to reprogram itself :), natural language is far more complex than formal language, if it can learn to talk it can learn to program itself and they are many programs that do just that.

The only way to no take risk is to never use that type of technology and promote a replace that solve the same problems. For me that is bioinformatics+genomics applied on ourselves

[kornboy3]

@DraskyVanderhoff  yes learn yes but not love .. robots never can love.. man !

[kornboy3]

@DraskyVanderhoff we are organic and robots wil lear but dont feel pain,love .. dont be funny man.. plz

[Theicemonkeyjr]

@DraskyVanderhoff Unfortunately i think your right, but many things here still are valid.

[donkisiko]

@DraskyVanderhoff

However, machine learning is still restricted to whatever the programmer decides to base the programming on i.e. a robot programmed to chop only wood when it recognises it could in a sense adapt and learn about different types and shapes of wood; but it could never suddenly start suggesting more efficient ways to chop the wood based on data, or how to improve its recognition faster because it hasn't been programmed to do so

[donkisiko]

@DraskyVanderhoff

Animals on the other hand are capable of interacting with many kinds of stimulus in many different ways and have the same positive feedback mechanism humans have which robots don't. Put simply as Jacque said, they can't take over the world or take over our minds unless we programme them otherwise.

[buzzin1975]

@DraskyVanderhoff Machine learning is based on statistical probability or rules-based inference. These models are crude approximations of how the human mind works (in the case of neural networks) or require the rules to be specified beforehand (as in decision trees). Ultimately, these are programmed systems. It's not as if the machine is actually learning. If that were truly the case, new behaviors would evolve according to feedback mechanisms introduced into the environment.

[Play671]

Do you have the recent video? Or this is the resent video of your channel? O_o

[thevenusprojectmedia]

@Play671 no i think the "total enclosure systems" is the last one up. But right now I am uploading one which is from August 7, 2011.

[5T4RSCREAM144]

I think the scare of robots taking over is to address the following: If you do not program a robot with the ability to do anything more than its task at hand and give it AI then you should be ok. Once you give it AI and the programming to live beyond the bounds of its task your risk is associated with the level in which it exceeds its bounds.

[5T4RSCREAM144]

EXCELLENT! Thanks

[matadedu]

PARACE COMO LA PROCLAMA DE LOS TESTIGOS DE JEHOVA, VIVIR EN UN MUNDO ORGANIZADO POR INGENIEROS RACIONALISTAS QUE UTILIZAN LA PORQUERIA GENERADA POR EL PENSAMIENTO CIENTIFICO...

SIEMPRE HAY UN IDIOTA A QUIEN SEGUIR.

[antiHUMANDesigns]

When creating robots, we can set up a frame of reference for them, giving them pleasure and pain jus tlike humans. We "simply" need to define something that is negative, and allow certain things to trigger a negative event in the core of the robot. That wills emulate instincs such as survival. However, survival instics is a dumb thing to give to a robot. But unless we give them some kind of driving force, they will do absolutely nothing, because nothing motivates them in either way.

[antiHUMANDesigns]

@blankprobe Zealot? I think you're missing the big picture. But it doesn't seem like it'd be worth discussing it with you, so I'm just staying away from this one.

[antiHUMANDesigns]

@blankprobe So, what you're saying is that making a computer that feels good if someone sticks a DVD in it is immoral? I really don't see how you come to that conclusion.

[kornboy3]

@antiHUMANDesigns robots will never feel 'love'

[antiHUMANDesigns]

@kornboy3 If we can definte what love is, we can add it to the software of a robot. However, love is absolutely completely pointless if you do not reproduce. And robots don't. We have no reason to ever want robots to feel love, except as an experient in social dynamics or something.

[kornboy3]

@antiHUMANDesigns LOOOOL metal robot never feel love wtf hahaha we are organic . fuckin LOOOL

[moneysuccesslegacy]

Does he really have to explain this? People aren't afraid of toasters or air conditioners! Be more afraid of people texting in cars; angry with guns; programming t.v.

[MrHunter920]

through a wire that is

[MrHunter920]

i hope he knows that electons actually travel at about 1m/s not the speed of light...

[antiHUMANDesigns]

@MrHunter920 I think you've been horribly misinformed. No, they don't travel at the speed of light, but they can, in theory. But there's no way they will ever go as slowly as you say. They will at least go a few miles per second.

[MrHunter920]

@antiHUMANDesigns my second comment says "through a wire that is", they travel so slowly because theyre bumping into and in between atoms, and pushing the electrons on the other side of the wire out at the same rate that they're put in. thus turning on a light seems instantaneous when the electrons themselves only travel 1m/s

[antiHUMANDesigns]

@MrHunter920 Ah, yes, sorry for missing that comment. While that is true, I'm not sure exactly what fresco was refering too. I mean, if he was refering to a wire connection or theoretical speed, and so on. (I could re-watch the video, but I can't be arsed to. :D). Furthermore, it's a matter of how much power you run through a wire. An electron itself can travel at any speed depending on the energy you put in it.

[warodri]

There are many studies about AI which will turn a machine into a human just like us (sooner or later) so, what Jacque is talking about is what we know NOW but there are is much more to come... Anyways, in a society like the one The Venus Project is trying to bring up, a robot with AI will do not as much harm as for this society we live in, with money as our economy... On the other hand, I do myself this question a lot of times: And if we are the robots of a society which we terminated?

[enlighteneveryone]

He makes more sense, and it kind of challenges my thoughts on robots in fictional stories. It kind of brightens my rational senses. People may say "Robots might be able to, could, can" it is not even as valid or as helpful than the movies, which has drawn to many people's fears, but also creativeness.

In a way, he is basically explaining robots of current, and thinking like a scientist, forsaken the 'what if' factor and going with the facts on through. Without tests, he can't be refuted here.

[ChrisPMadden]

Hi "DevilsAd" et al, that is a good idea... I know there are some websites that you can put the url for this (or any other YouTube) video and it will convert the video to .mp3. Then you can listen anytime. Just put "Video to mp3" into your search engine and you will find a coupla sites that will do it. Enjoy!

[JLarky]

@dilmurodkhodiev why do you think so? I mean why do you think that you know that you are understand what i meant by "associative memory" if from your message It's clear to me that you totally unaware of what "associative memory" is in artificial neural networks.

[technatezin]

@JLarky No, unless the associative memory can crossbreed or evolve artificially in response to an environment that gives out stimuli for the host to respond/not-respond to it will have no chance to go beyond the limited parameters of it's original programming. You can possibly create a contrived artificial reality for the A.I. to move around in, but that's also limited to what can be programmed into the artificial reality. It will take many generations of crossbreeding for true intelligence.

[JLarky]

@technatezin humans are subject to limitations as well, you can't go beyond limits of your body: you have to rest and eat. You can't deny limits of human intelligent like you can't multiply 6 digit numbers in your head. Self driving cars made on artificial neural networks do make decisions working in artificial image of real world around it, but who can say that it more artificial than image of reality in your head?

[technatezin]

@JLarky True, but that's only to the extent that you have programmed into the robot's limited purpose. You can get quite fancy in the programming with, recursive definitions, feedback loops, fuzzy logic and associative memories, but it's still limited to what the you've intended the A.I. to respond to. That's the whole point of what Jacque Fresco have been saying. Human intelligence are generalist for a reason. A.I. intelligence for the foreseeable future will be narrowly defined special purpose

[JLarky]

@technatezin Nope. He was saying that implicit programming is implicit, but self learning is self learning. Hes point was that for animals there are simple and true intent to do something but robots only will feel right whatever creator of this robot set to be right. You compliantly confuse with algorithms vs. neuronal networks. Neuronal networks is as general as neuronal network placed in our head.

[technatezin]

@JLarky "Neuronal networks is as general as neuronal network placed in our head."

Agreed, but...

1. how are you going to "program" those neuronal networks without an environment?

2. neuronal networks have "weighted" memories that are stored in literally billions of redundant and self-referencing loops. At what point is "critical mass" reached where self-awareness takes place?

3. Evolution needs millions of stimuli-response-synthesis cycles to come up with something as simple as an ant.

[JLarky]

@technatezin how do you program humans without environment? I can't get your question.

Self awareness isn't needed for practical tasks. Does ape feel as deeply recluse it is and that no one in the universe can't help it? Who cares if it can solve task like image recognition that can't be put in any programming language without neural networks. If you have robot seeking for man in ice after avalanche you don't want this robot to answer philosophic questions you want find people alive.

[1960aussy]

I to would like to make a suggestion, I have many friends that do not know or understand English, it would be wonderful if the many videos from the Venus Project as well as the Zeitgeist movies could be translated into as many languages as possible, so that quite a lot more people would find about the Venus Project. I have already offered myself to do the translation into Spanish and I feel that many more people around the world would be happy to do the same in their language. Gerard from Spain

[thevenusprojectmedia]

@1960aussy we have the help of the linguistic teams  who are working on just that; if you can help in your language, please sign up, email address is in the box above, thanks

[Tomislav0Ivcic]

@thevenusprojectmedia I don't see even English captions, which would we fine for me .

[JLarky]

Jacque should learn about neural networks. They have associative memory that is so needed. You can search for Associative Memory video lectures at MIT.

[dilmurodkhodiev]

@JLarky what u r calling as "associative memory" is totally different, for example Xbox's KINECT has some sort of associative memory, it analyses what it see's and tracks it, but this is because it was programmed to do so, there is a PATTERN how it works. but Robots can not deal with UNFORSEEN situations, which means if the program in a robot well-thought, it can deal with some situations, but if smth is not expected it fails...

[DevilsAdvocateism]

If I could make a suggestion to the Venus Project. Put these youtube videos in a mp3 podcast so I can actually listen to the material, I don't have all day to watch youtube videos, Im busy so I need to be multitasking almost 80% of the time and there's just too much information to consume in a day to devote 100% of my full free time to watch youtube videos. A huge advantage Peter Joseph has tapped into, going on other podcasts isn't enough, make your own radio show or podcast, with sections.

[thevenusprojectmedia]

@DevilsAdvocateism thanks, i have been considering something like that, will try to work something out

[ZMLinguisticTeam]

@thevenusprojectmedia

It's funny that this would come up now, when I'm in the process of building the LTI Radio section of the LingTeam website and decided to do exactly this with all of the shorter Jacque videos. :D

Stay tuned, Kids! Audio versions of the videos are coming soon. ;)

[777asmodeus]

@DevilsAdvocateism you can down load mp3 files with firefox application to convert video to mp3 file

[barakuda1111]

@DevilsAdvocateism Try "youtube downloader". In 2 easy steps it downloads video from youtube, then turns it in whatever format you need, including mp3 file. Then you can put that file on your favorite mobile device and listen to it :)

I personally listen to all of this "longer" videos while im working. I have a computer in my working station so its easy for me. I do not convert it to mp3, but i do only LISTEN to it, while im working :)

[berniemac18]

@DevilsAdvocateism : now YOU have something to work at. :) not everyone has that issue with the tube vids. i dont

[PortlandsTransport]

This guy is fascinating.

[CapMurd]

AI already exists, it's just people don't understand the meaning of 'AI'.

[InvertedFox]

@CapMurd AI produced through binary computing systems is very different to the AI that can most likely be achieved through quantum computing. As JF says, AI is currently linear. My understanding (I may be inaccurate) is that this is the do with the 1-0 logic gate system of binary. Quantum computing aims to replace binary with particle energy systems. A single particle has many potential energy states which can represent data points, like binary, except more efficiently + potentially non linear.

[dilmurodkhodiev]

@CapMurd do u mean IBM's Watson? if so its not AI, if not then what is the meaning of AI?

[CapMurd]

@dilmurodkhodiev No IBM's Watson is very much AI. It's called artificial for a reason not just intelligent alone. AI is in simplistic terms just a program that is following instructions. There is no such program that will think for themselves, ever. Closest thing they can do is to write alot of lines of codes that will start making it's own rules/ideas by still following basic line of codes. And that is very much AI. AI is not a program/computer that actually thinks but just another program.

[b0ondockz]

This is 15 minutes of common sense when it comes to robots of TODAY. Kind of a pointless conversation to talk about why robots of TODAY will not "takeover". Fears of robot takeover are obviously for the future.

[itsadeadmansparty]

@osheaad just look at the world and listen to some of Jacque's more classical lectures. Tell me he doesnt make extremely good points. The real problem is we are lagging on how long it takes evolve our society along with our present level of technology, We have a route out of Oil,and we have so many actual fixes for most our frustrations, we should work on the real goal, a self sustaining society without the drag of government

See if your really wanted to, you could build robots to manage things

[parisfaith]

jacque still living and looking strong. wow I am with you my brother. Venus Peject

[BboyKeny]

@osheaad He doesn't say that his way is the only and the best way. It just the best he can come up with with his knowledge. I think he would love that you doubt him if you can say why and think for yourself do your own research make your own theories, I think that is what drove him this far and not saying it is impossible or dont even bother thinking about something. If you come up with idea and tell him it he would hear it our. If you see a flaw of thinking of him it would be a sin not to tell.

[michael0001able]

I didn't like transformers either jacque

[BrentAltonNally]

I am glad TVP is posting these lectures. I have met both Fresco and Kurzweil and I have been trying to get Kurzweil to meet Fresco for quite some time now. I even asked Kurzweil to his face if he has heard of Fresco and he said he had not. Kurzweil has not responded to my emails either so it is just frustrating. These minds need to converge to stimulate better understanding for all.

[illmatc]

this man is enlightened

[IHighLikePlane]

Yes, robots only do what we program them to do, but what if we program them to learn?

AI, will change the abilities of robots significantly, and we will easily see AI come to household computers in the next 20 years.

[berniemac18]

maybe we should just accept the limit of robots and progress foward.

"it does what we program it to do"

the rest is pseudo-talk.

[warobsessive]

to replicate the computing power of the human brain in a robot would take a supercomputer shrunk to the size of the human brain, ~513 gigabytes per second. We're lightyears away from that. Michio Kaku put robots today as having the intelligence of a mentally retarded and then lobotomized cockroach.

[Opinionator52]

@warobsessive It could be greatly larger than a human brain because robots don't need the space for containing a digestive tract ect... So processing could be done in many other areas.

O,

[GMLSX]

@warobsessive we are allready over 100 gb with a sigle Tesla GPU card. Programming is the Problem. But we are getting closer. Also by learning the human brains cheats ad redundet systems.Image processing is a great example.

[nevasnitch]

We really need to start tossing Information to the ones that are asleep,The damage is done,We are going to see things we never even imagined,Difficult times ahead,The more we inform And educate people about alternatives,The better we are going to be.

[Uribaani]

@luisbeck007 Couldn't even bother reading your comment. Word of advice, if you want any credibility whatsoever, do not use caps.

[osheaad]

I know the complexity is vast to make a robot human.

[osheaad]

I think humans are robots, just infinitely more complex than what Fresco says a robot is. It's not impossible to make a robot human.

[Anonymous247n]

The problem would be an artificial intelligence... a program that could think for itself, creatively, like a human. I don't see any reason to develop something like that - one of the reasons being that "bad" ideas could come up. Why not, anything thinking like a human could come up with human ideas, like anger, fear, destructiveness. I'm not afraid of being around hundreds of robots, but an AI would be an intelligent being of a different kind.

[BboyKeny]

@Anonymous247n If we want to make an AI to image of humans we would be pretty stupid because our processing power is really poor the reason we arent wild animals is because we can think which is really 5% of our whole concience 95% of what we think we dont really think but is automated our concience proccesing power is around 80 bits which is trully crappy in comparison with your computer. AI is a system that can 'learn' and adjust to that by linking it to a db let it only have relevent sensors

[itsadeadmansparty]

@Anonymous247n The real question is, can a group of silicon processors contain life? Can it be written to the point it makes a leap to sentient self programming software? The reason I disagree that AI can exist is presently is nobody knows how to program this level of complexity, maybe another machine could produce this but no hand written software code will do, Lol the processing of Genetic code has taken our computers a long time to worky, lol

Do you think a sillicon device could become AI??

[b0ondockz]

@itsadeadmansparty I think a quantum device could become AI.

[itsadeadmansparty]

@b0ondockz Agreed Only A Quantum Device could become AI, since we do exist in two worlds aka quantum

[Anonymous247n]

@itsadeadmansparty Agreed, it's hard to imagine at the present point.

[digitalaffect]

expactly my thoughts.. to program AI u need feelings.. and we ar not there today... but i dont think anyone is afraid of things like that today its more about the future. And someday somehow somewhere someone will give robots feelings.. stupid ide but the atom bomb was stupid aswell but we created it anyway.

[svetabozinovic04]

Excellent. It:s clasic error when somebody use facts and connect it with bad explanation. It is like.that paper is white like cheese and because of that have same taste. This mismaches may find every way.

[shapingAfuture]

Well i agree with what he says but if you include the capability to learn then his theory fails. If you program a robot to learn from his environmental eventually he can come up with ideas that are harmful for human beings. Maybe you can prevent this by placing a programming that can not be overwritten by his ability to learn and process.

[MrrGrave]

@shapingAfuture Like the three laws of Asimov?

[slickaruta]

@shapingAfuture He's theory is actually right.

He actually mention that if a robot would learn from his environment nothing would happen at all.

the only thing a robot would do is gather information in that environment, but the robot doesn't have the desire to do anything with that information, it doesn't have feelings and opinions. unlike humans

that have desires, feelings and opinions. so if a robot have some harmful ideas then he still wouldn't use the idea because he doesnt have the desire

[shapingAfuture]

@slickaruta I would hypothesize that every thing that can learn will come to conclusions. One may be of self preservation or maybe the preservation of his work maybe his job is to keep nature clean, that is his purpose. And it comes to the conclusion that humans just pollute everything and cause more harm then good then the robot can decide that the solution would be to remove the problem from it roots.

Sidenote :I'm not against Jacque and i don't imagine that robots will be the end of us

[shapingAfuture]

@slickaruta Let me correct it "preservation of his work" -> Doing his work more efficiently and keep the clean environment . There must be BUT somewhere there, like preservation of life.

[slickaruta]

@shapingAfuture aahaaa sry for the misunderstanding

[priitraag]

@slickaruta When environment dependent functional flexibility is required, one still needs to program some incentive system into the learning robot and these incentives can be viewed as desires. The robot will do everything in its capabilities to "satisfy" the (prioritized) conditions that have been built in. The more flexible these conditions are (that allows flexible learning/self modification) the higher the chances of robot developing a potentially harmful behavior that can not be foreseen.

[WJValente]

I do not think people should fear that which they do not understand. (Or maybe they fear it precisely because they don't understand it) Most people don't know that artificial intelligence can be made w/ microchips that prevent them from harming humans or doing anything illegal.

[TheAwakenedHeretic]

tag Ray Kurzweil in this.

[Jhipson]

The world really needs this.

[LudicrousTachyon]

I would say the programmer is the learner for the robot. When the programmer sees that the robot misses the cup, he has to reprogram it. Humans are programmed by their environment.

Program a robot to feel better when it's battery is replenished or depletes more slowly, the robot will prefer to sit rather than stand and with teaching the robot may aquire a job to pay for the electricity it needs to replenish its batteries just like you teach a child to get a job when it grows up.

[Nathan173AB]

@LudicrousTachyon Except how do you program a robot to feel? You'd have to introduce consciousness somehow.

[Herbarius]

@Nathan173AB why would you want a robot to "feel"?

[Nathan173AB]

@Herbarius I don't. That wasn't my point. I was responding to someone who suggested that.

[TheAwakenedHeretic]

@Herbarius tease it until it cries. 

[LudicrousTachyon]

@Nathan173AB You assume that what we call feelings cannot be arrived at by impulses through logic gates, but your brain does just that every second. Consciousness is nothing more than being able to compare now to what you did moments ago and even that would be easy to program. A computer can easily save it's current state and you can program it to compare it's current state to a previous one. Give the computer a way to act on it and make it real time, you may get consciousness.

[Youanden]

@LudicrousTachyon No, we have secretions (hormones, serotonin), that change the chemical state of a person. Sure, you can reinforce certain secretions to effect someone in certain new situations, but it is also possible for compounding variables (menopause, environmental stresses, disease) to change or remove the ability for these same secretions to effect someone in the exact same way.

[Youanden]

@LudicrousTachyon Those are efficiency algorithms, and even those are very different from humans. Humans and animals can deviate from reinforcement. If I program a robot to conserve energy, I have to supply the robot with sensors, and a range of actions (sitting, walking, doing a task). The robot cannot decide to start drinking alcohol because it is not in the programming. Robots are only as superfluous as we program them, unlike Animals who have a more dynamic range of associative actions.

[LudicrousTachyon]

@Youanden I have yet to see a human or animal deviate from reinforcement. You do not have to supply a robot with a range of actions unless you want it to be able to do it right away. I'm convinced that the reason we don't have robots that can think is because everyone wants robots that can do something right out of the box instead of having to teach it. Look up genetic algorithms. After testing many possibilities it would find the best arrangement to conserve energy.

[Youanden]

@LudicrousTachyon Let me give you an example.

Activity A provides X amount of reinforcement for N amount of energy required.

Depending on the physical and mental state of the person, it may decide to do A, or do an activity B that requires less energy, however, it is also possible that spontaneous environmental changes (phone call invitation to go bowling) can change the action. Only if robots are built to react to all situations people react to can they become like us. What for?

[agnoz85]

massive G.E.N.I.O.U.S

[TalladegaTom]

Jacque is the worlds most under rated mind.

He is WAY ahead of his time.

[erionmema1]

this is a funny argument .

people fear robots because they think of themselves as "what would I do if I had lots of processing power and infinite energy.... thats obvious ,,, Id rule the world"....

sorry to disapoint you people but robots do what they are programmed to do... selfawearnes requires lots of neural links .... along with biochemistry !!

[T0B0KKE]

Feel good or not can be simulated with a variable. Even joints lubrication can be measured. If the robot touches something that you want him to feel good for that, just increment his "happiness" bar or something...

[EvilGniusStudios]

@T0B0KKE Why would a robot be "happy"?

[T0B0KKE]

@EvilGniusStudios You don't quite understand what the word "simulation" is. Happiness can be emulated in a computer or a robot with a variable that must be maximized and increased as far as possible. If you run a program that the sole purpose is doing actions that might increase its "happiness", you can implement happiness as a numeric variable and run some min-max method that maximizes the robots "happiness" according to what he does.

[T0B0KKE]

@T0B0KKE Btw we (humans) currently do seem like robots maximizing one variable: "profit".

[EvilGniusStudios]

@T0B0KKE Who said robots maximize one variable?

[T0B0KKE]

@EvilGniusStudios Auto learning robots use optimization algorithms to improve their, walking skills, for example. The objective is minimizing for example energy consumption to maximize the length traveled. Optimization problems are always solved either maximizing or minimizing a numeric value. You surely need to do more research on these topics if you really want to argue...

Go watch this vid and check this little guy learning to walk: watch?v=rkGaQqkyWII

[EvilGniusStudios]

@T0B0KKE Who said I'm arguing? I've simply been asking questions.

[T0B0KKE]

@EvilGniusStudios Oh. Ok. Sorry for the misinterpretation. I get so many flamers and disbelievers that I can no longer spot someone who just wants to ask to learn and not to argue.

[DMTilian]

Eugenics are hypnotising humanity... we're being mechanically devolved into malleable drones. Animal before machine!!!

[karakzanreal]

Its a shame that the assumption of "the current state of robotics" is only said in the very last seconds of the video. It is true at the current state of the art of robotics and artificial intelligence there is no thread to the human race. If however machines become self-aware we better hope that Isaac Asimov laws of robotics were programmed into the robots otherwise we might have a bit of a problem on our hands.

[EvilGniusStudios]

@karakzanreal What do you mean when a robot becomes self-aware? If it means that it decides to obey itself why would it follow laws of robotics? Can you elaborate?

[karakzanreal]

@EvilGniusStudios I guess that self-aware is a bit of a fuzzy concept. What I meant to say was that if AI is programmed to optimize a certain set of variables and humans stand in the way of doing that the AI will "remove" the humans. If no special care is made to explicitly value humans above other factors bad things may ensue.

[Youanden]

@karakzanreal That is very vague. The question is not when they will become aware, its at what point will it become beneficial to program aware robots, and to what extent. Robots can't feel pain, and it is unreasonable to train robots to "feel" pain. We can create human-based simulations based on nerve endings and force sensors, but the robot still does not feel pain, its system does not become damaged, it simply outputs a value the simulation directs.

[karakzanreal]

@Youanden You are right in that "self-aware" is a vague concept and that unless biology is sufficiently emulated there is no strict analog to pain in AI. However it is trivial to assign values to different actions that are available to an AI. By assigning negative scores to actions that have adverse consequences to the goals of the AI you can "simulate" pain. All that is required is enough self awareness to assess damage done and the severity of that damage.

[juliomezameza]

Clear. Just keep in mind he is talking about today's robots.

[chockfish]

I like his explanations that makes clear the point of what he's talking about even for uneducated people!

[chockfish]

He s talking about the simplicity of neuronal learning and the conciousness about that, which nobody knows where it's coming from! Machines will never have that unless man is able to create artificial conciousness.

[peshozmiata]

@chockfish Ray Kurzweil's point is that creating an artificial consciousness modelled on the human brain will eventually be possible. For example the Blue Brain project is about exactly that - creating an accurate model for brain simulation on a supercomputer. I would've liked to hear Fresco talk more about self-programming machines...

[chockfish]

@peshozmiata I've read about the blue brain project - great, couragous work! maybe someday it'll have a consciousness and be aware of itself! But after all we'll see, that we cannot MAKE the conciousness! It's created by law of nature or nobody will ever be able to assemble it! Just like we'll never be able to bring a positron and a neutron together unless they allow it!

[peshozmiata]

@chockfish It's hard to tell whether it will be possible or not, because information is limited. When trying to create consciousness, you have to ask the question - what exactly is consciousness? I highly suggest checking out " The age of spiritual machines" for some deeper info on this subject.

[chockfish]

@peshozmiata ...nice, but very old fashioned...and it says nothing about the developement of the conciousness! Although it may suddenly apear on an artificial level, we're still not able to describe WHAT, HOW and WHERE it is nor where it is coming from! Everything we can do is to construct a machine that to an high extent works like the brain and hope that it'll be selfaware, comunicating and independently learning...but we don't even know how the brain works in detail.

[peshozmiata]

@chockfish You're talking as if consciousness would appear in a machine by chance... If one ever appears inside of a machine, it should already be known what it is, how it appeared and where it came from. Now, the question of whether that would truly classify as consciousness is highly debateable.

[Miketamine]

i wish we would turn back to nature and live as we did 900 years ago....we wouldn't destroy ourselfs that fast at least...fuck robots!

[technatezin]

@Miketamine then you'll kill each other after working hard farming the land all year and then end up with a bad harvest from droughts or pests.

farmer 1: "We have a bad harvest this year. We don't have enough to eat."

farmer 2: "Our neighbors over there didn't have a bad harvest this year. They've got plenty of food"

farmer 1: "Let's take their food"

farmer 2: "They wouldn't want to have their food taken, so we'll need weapons"

farmer 1: "We can beat our ploughs into swords."

[T0B0KKE]

@technatezin That's why sharing and resource tracking and management will exist...

[jiykauu]

@Miketamine With life expand of 30 years? Robots are everywhere you just notice them.

[JwGacey1]

He's obviously not seen Wall E

[JwGacey1]

@GMLSX I know it was a bit of a stupid comment but my eldest son and I just watched WallE tonght before bed and it seemed to fit perfectly with what Jaques was talking about.

[kmarinas86]

@JwGacey1

How do you know that he did not see WALL-E?

[JwGacey1]

@kmarinas86 You 'obviously' either did not pay attention to Mr Fresco or you have not seen Wall E because if you had done both, you would not have bothered taking the time out of your life to type such a silly question to my silly joke.