Rodney Brooks: How robots will invade our lives

[samhowel]

I wanna build the robot to bring about the violent singularity to prove this guy wrong I am half way there now I just need more powerful components for the thinking portion and awesome articulation. after all this thing has to be way smarter and faster than any human and still be able to self replicate. The end is near and I shall bring it. HAHAHAHAHAHA!!!!!!!!!!!!!!!!!!­!!

[RamadaArtist]

It's really amusing to see some of the most brilliant people in the world interact with a roomba before it was a thing that people knew about.

[RealFreak69]

For him to say that no-one will build a bad robot is seriously undermining the idiocy that some members of the human race have. I know of a couple of people that might build a rouge sky net like intelligence given the chance, but to target a certain group of humans, be it a different race, nationality or even just a view point.

I think we should limit how much intelligence we give machines. Prevention is better than a cure.

[ilovehd831]

Machines have senses that we do not. For instance, they can see a 3-dimensional object in its entirety at any given time, whereas we can only imagine a single 2-dimensional viewpoint. They can see patterns in any kind of data, such as DNA or stock trends, the same way we see patterns in sight or sound. Any artificial intelligence would also be a superior non-human intelligence. We want to tackle the human-like portion first, as there is a lot of human-generated information out there.

[Robstailey]

robots are better than humans in every way.

[ilovehd831]

Except for (unfortunately) their primary skills.

[shiftyjake]

That one woman's about to make that robot cry!

[XANSEM]

The commercial at the end of this videos is absolutely beautiful and most of all true.

[ilovehd831]

Haha yeah, I agree. I had that feeling that it was going to be something which didn't the least bit represent that ideology, but it was just Autodesk, so it's not that terrible.

[whatevtube]

Santha thinks that I'm special and so does the Easter Bunny + Jesus.

[indigogarcia]

We are the medium. We are machines.

[8legsFreak]

Thinking robots are impossible. Computing robots yes, but not thinking. There is no way to program a machine for self awareness, common sense, or more than insect level patern recognition.Thats why robot matematician is possible, but robot infantry is not. Anyway semi-inteligent machines would be wery beneficial for humanity.Battlemechs, terminators, HALs,and Matrix agents, are impossible. Robosurgeons, robofactories, robopets, smart-homes, smart vehicles, and even robo-prostitutes are possible.

[berendho]

@8legsFreak

impossible why?

[8legsFreak]

thinking is a strictly human way of computing data. Only way to make a "thinking machine" is to copy/scan/upload human brain-mind circle. But that would hardly be a AI, just a morphologicaly different human.

[ilovehd831]

Or just to represent algorithmically the functions of the human brain. But as you say that would also be a morphologically different human. The difference is that the 'morphologically different human' would be able to vastly expand its calculating abilities, copy its memories, and implement its memory in another entity. I'd also like to note that the way the human brain performs calculations in the neocortex is identical to other mammals and many birds.

[ilovehd831]

Oh, and I hope you aren't trying to say that it must be copied from an already existing human. In reality, any brain that is unable to grow and change in the same way an infant brain grows and changes into an adult brain would not be able to think at all. If all else is impossible or too difficult, representing a brain at the molecular level would be able to produce the same results as a real brain. That includes growing from the empty mind of an infant into a fully capable adult mind.

[ilovehd831]

Well it's not necessarily impossible. However, it's a much more difficult problem that actually teaching a computer-based brain to begin with. One would have to either determine the mathematical equivalent of the configuration of each neuron in the brain, or they would have to record the positions of every molecule making up the brain. They would also have to do it in the time between any given neuron firing, and any other given neuron firing.

[8legsFreak]

what i meant was that making an AI so similar to human so it could pass turing's test is useless, even if that artificial mind was superior to ours, what would make much more sense is an alternative un-human intelligence that could think outside anthropologic cartesian box.

[ilovehd831]

@8legsFreak Machines have senses that we do not. For instance, they can see a 3-dimensional object in its entirety at any given time, whereas we can only imagine a single 2-dimensional viewpoint. They can see patterns in any kind of data, such as DNA or stock trends, the same way we see patterns in sight or sound. Any artificial intelligence would also be a superior non-human intelligence. We want to tackle the human-like portion first, as there is a lot of human-generated information out there.

[TheOpenSourceLife]

@8legsFreak Machines have senses that we do not. For instance, they can see a 3-dimensional object in its entirety at any given time, whereas we can only imagine a single 2-dimensional viewpoint. They can see patterns in any kind of data, such as DNA or stock trends, the same way we see patterns in sight or sound. Any artificial intelligence would also be a superior non-human intelligence. We want to tackle the human-like portion first, as there is a lot of human-generated information out there.

[Chavalierdeon]

heehehe i bet the prostitutes was actually your first tought, was mine

[silverblue73]

it's not so much about programming for self awareness, not even we can tell if another human being is truly self aware. it's about creating the illusion, to the point where no one can tell the difference. and that is very possible.

[8legsFreak]

I agree with silverblue73. I think that making humanlike robots that could concivably pass turing's test is possible, would be awesome, and also not practical: the point of having AI in the first place is to make it work on problems unsolvable to humans, using non-human form of computing data, not to replicate human thinking since we have already abundance of it. I would like to have mine C3PO too, but i prefer AI tutor, AI-robofactory or a surgery bot that can save my life.

[doloppost]

Butlerian Jihad anyone?

[dollarability]

Naysayers claim that subjectivity cannot be duplicated in a machine. But if you look at evolution its clear that animate matter evolved out of inanimate matter, consciousness evolved out of animate matter, and reflexivity evolved out of consciousness, without any clear demarcations. So where is sentience? The lack of clear demarcations suggests that sentience is an inherent property of matter. So if we can build machines that are complex enough, they will manifest the same reflexivity we do.

[TimothyParadox1]

He's actually quite naive. The machines in sci-fi took over either because of a mistake in programming (a bug), a sob who put a virus in it, or robots gainging self-awereness and feeling threatened by us. He didn't address those issues.

[stikowsky]

nice, attack-killer robots running on linux...

[refootball]

Fuck this!

im becoming best friends with john conner

and puttin every austrian on steroids

so dont worry we are saved

lol

[hazonku]

I'm not to worried about the Terminator. I'm more concerned about cybernetically enhanced humans. We're much closer to that.

[Reido2828]

Not quite close. It will be a slow progression and a slow upgrade. Our culture is afraid of robots because of hollywood. Look at the japanese culture and how they empower robots. Also, children love them and think they can touch and love these machines because of films like toy story, nemo etc. It will not be super bad. We only have to fear what our fears are and that is a machine enslaving us but thats just neurotic. We have no clue what will happen with these robots

[connor646]

You're on the right path, but are a bit off. The ultimate goal of all studies in A.I. and Robotics is essentially to re-create the human bien in machine form. Ultimately it wont be an it but a a he and a she. Keeping in mind that it is no longer a thing but a new type of person we have to analyze how we interact with one another. Ultimately there will be conflict with them because we can't seem to stop conflicting with one another. Personally inspite of all of that I welcome the singularity.

[Reido2828]

Machines will not want to destroy us. Were creating them with a purpose. We won't compete with them. it is our selves who we compete with. You have to look at culture. Japanese culture does not fear robotics ruling over human society. Western culture does because of all the movies we have built this fear on. When robots arrive it will help humanity. For good tho. I can maybe see them disregarding their programmed masters and searching for their own path instead of fighting a war for America.

[connor646]

No one mentioned anything with America, and the majority of people outside of the scientific community just doesn't seem to get it. Creating a machine which is just a machine is a limited achievement. Many scientists are working towards the singularity. The Singularity is a point where computer intelligence and human intelligence is indistinguishable. Being that that is the case, it won't be a thing rebelling but an entirely new species looking for their niche in the world.

[Reido2828]

Comparing the introduction of machines with humans to Neanderthal homo sapien dominance is not the same. Your philosophy and logic on that notion shows that your reading into the negative's to this. Keep in mind we will slowly upgrade these machines and have them pass many test's to show they must not harm humans. Human beings that are compassionate, loyal and thoughtful towards their other humans should not worry. It is the politician's, the CEO's and dirty mongrol humans that should be worried

[connor646]

Man has showed his incapability to control his urge for dominance over other men since we were dragging our nucles. A lot of people see thew singularity as a point where we will have efficient tools. That is not singularity, its merely an advanced state of where we are now. When the singularity truly hits Artilects(A.I. powered computers) will be --for lack of a better word-- alive. They would be our equals and in time our superiors.

[Reido2828]

What you state here is very true. its evolution of man. Its very tough, cruel and small minded. But as evolution moves on and we get new and better experiences our consciousness evolves to. We have better awareness. Evolution is about groups and the change they harvest over time but in the future as we have more information available to us and for free we become smarter, more aware of our environment and maybe even rid the ego. This type of evolution can only be brought on by technology.

[connor646]

Now that being understood and seeing how many people treat one another on insignificant differences like religion, race, gender and so forth you honestly think that someone wouldn't want to take away any type of equality which would be bestowed upon these new life forms which we've created? I for one couldn't and wouldn't treat any living thing with less respect than I would for any other human. So when machine is equal to man how will nthe world react to her existence?

[Reido2828]

Again no body can for sure forecast this as good or bad because we just don't know. Ray Kurzweil says its going to be a great time for humanity while other scientist's think it spells the end of us. Now my point is we merge with machines. Scientist's are creating small nano implants that could replicate as a brain cell and with that make us way smarter than we are now. Some even think we may get to that point before machines become equal to us.

[Reido2828]

Example. If we cna change our DNA and brain cells with nano bots to make us more efficient humans, stronger and more compassionate towards eachother then why not? Yes we still have human nature and we have many questions to raise when and if we should megre with this technology. Now we do not know enough about human nature so there is a possibility that it could be changed with this technology. We could make certain routine jobs obsolete over night with this. Humans will have more time to explor

[connor646]

So Like any species that tries to find its niche in the world, it will war with its predecessor to find its niche at some point. think Neandethal and Cromagnon man. At some point we'll have to go and they'll step in. Iike Rodney Brooks said Humans have his notion of superiority or specialness, and when this badvanced form of "man" comes in to play it becomes obvious that we won't be seen as special. But its the way we gotta go so no need to fret over it too much.

[Reido2828]

So your a pessimist then. Thats fine but it seems your hate for human imperfection is what is making you accept some future where we will be destroyed. The purpose of the singularity is to create machines to help us do things we cannot do as mere humans. With advanced A.I we can upgrade our intelligence, cure diseases, create a better class of people and make current work (slavery) obsolete. A.I is apart of our evolution. It is something that is going to change humanity forever,.

[Reido2828]

This type of evolution is not darwinian evolution. This type of evolution is individual evolution is the very core to our existense. Call it what you want. SPirtual evolution, integral consciousness etc. But if this does come to pass it will help humanity way more than now. And in the end again it is up to us not the machines if were going to destroy our selves. We can create bad science like Industrial sector's and bombs or we can create good science designing a world where everyone has a chanc

[connor646]

You're totally off center with what I'm taling about. Youre speaking about the hear and now you're adressing the short term transhumanism and A.I. I'm talking about singularity, the end of this particular subject. According to Hugo De Garis traditional Humans and Machines being equal or reversing roles is the end result of A.I. Will it be the end of traditional humans? I don't know, but I doubt they could do things any worse than we have so far.

[Reido2828]

Ok email me then. We'll talk about it. I still don't quite understand where your coming from? I assume you know about ADDTV. They are trying to raise awareness of end of normal humans. I enjoy their video's. But humans will merge with machines slowly. We will up grade and become introduced to it. Kind of like now how people wear glasses, have inplants instead of bones etc. I have a titanium rod inside my chest that replaced my sternum bone inside my chest. This will be the same with Nano bots.

[AndersHansen1988]

Aslong as there are computerfreaks who take pleasure in developing viruses for other peoples computers, is it really a good idea to build robots?

And another aspect, we already have a show on TV called RobotWars, - hows that gonna be in the future if we build thinking robots? And how is that gonna be running together with the "pay respect to robots as living creatures".

And i fear there will be critical major unemployment when a cheap robot can do the job of a human.

[TimothyParadox1]

"The robot rights movement". Yea, I can see that happening.

[George4943]

I Robot (Isaac Asimov)

One of his female characters falls in love with her robot, Tony. And, oh yes, he is the ultimate "battery operated boy."

How soon are sex-robots going to appear?

[Ramsez]

my question exactly

[EveryHumanBeing]

But.... I think the robot Died in that story

[Artokaas]

I'm pretty scared of this guys ideas

[TruthSurge]

It's not expressiing "its" emotion. It is running through a program that has been pre-decided by humans. If supercomputers with countless terabytes and trillions of instructions per second do not have true consciousness, I doubt poor Kismet will ever have it. :(

[hellogeorge1985]

Actually here at the University of Reading in UK, the engineering department connected a mouse brain to a silicon chip, put it in a box with 4 wheels, and guess what.. it moves and avoids hitting the walls. So, imagine how close we are connecting a human brain.Now that's scary

[ColacX]

dude wth. they gotta stop making weird experiments to mice.

[Apoptosis74]

Hi George. There's also a guy in the US that is very seriously trying to produce brain prosthesis. His name is Theodore W. Berger, at BMSR. I know that there are also scientific publications showing blind people seeing very simple things via cameras and a 'brain chip'. But the connection to the brain doesn't persist well.

[chilenozo]

We don't need to create a bad robot that will destroy us. We just need to create a virus-like AI that can learn and wants to survive above everything else. I am sure this guy knows it. The human nature always wants to compete and be the best, because of that, somone may be working today or in the future in that AI that will give that person the edge in the current technology, fame and money, once that is done and the AI is set free the end of the human race will start.

[jpixels]

hmm computers didnt come from games dick head

[Lionsroar89]

to the household, dickhead.

[eirpcalc]

To keep robots from ever taking over, just make sure they run on Windows.

[diyamez]

I lold.

[SonOfNye]

@eirpcalc Well played sir.

[DontTouchMyTea]

I do think that as robots become more human and life like, you should treat them well, because treating them badly would only condition you to treat *actual* human life badly.

For example, I think kicking an actual robot that looks like an animal or a human would make you lose your inhibitions to kick a real animal or even human.

[DontTouchMyTea]

No, I'm just using my knowledge of psychology, my experience and logic.

[kennebacasiskyle]

You might be smart in SOME way, but all you have shown here is ignorance and rudeness. Telling him that you are "smart" only makes your original insult more insulting.

I agree with you, Don'tTouch, but couldn't it work the other way, too? Couldn't people- or animal-like robots be used to teach empathy and compassion?

[kennebacasiskyle]

You make a good point here, Tea, and I agree with you that we must be cautious, always, because the baser instincts we only recently evolved out of can be relied upon -- it would be a potentially-dangerous slide.

[KarIgnishaYumi]

movie have made robot look like tha will take over the world or take are place...

i think it might be hard to make a robot have feeling, we can make them act them but we can't make them feel like us i guess..

i'm not a scientist....so i don't no that much about how it work.

[Devilsnightforlife]

I reject the presents of a soul. Not only because it cannot be verified (very much the same as god) but also because there are more 'satisfying' theories, like we are all one biological mass interacting with ourselves, experiencing one feeling is a process of multiple mechanics and chemicals in the brain. F.ex. When we hurt ourselves, signals are being send to the brain that f.ex. our hand has been damaged and sends back pain to the hand so that we know that the damage isnt good to our body.

[Dirtfire]

How are they going to solve the problem, which is basically that mechanical parts aren't getting any cheaper? Mechanical parts will have to get much cheaper before useful robots become common, regardless of the AI-aspect.

The global economy is just not good enough to make robots affordable to all but the wealthiest.

[jigarashah]

Question is do we really need such robots which has emotions ?

Or we better have good humans in lives for such emotions. I cannot assume a mother instead of giving birth to child buys an robot. I have heard that in Japan they have robots for taking care of old parents ..!! What the hell. And regarding Robots will invade our lives..They already did it. Think of someone who cannot "live" without mobiles.

[Talixaen]

I rather hope that I do not live long enough to see robots that can be.. spoken to, and communicated with... become commonplace.

[Williametched]

That's simply because there is no clear definition of the soul. If there ever is one, then it can be replicated artificially.

[matshroom]

he dusnt mention the orphius theory! simple robots that can design more complicated robots ect then envntualy maby one we dont want??

only a thought... great job ted! keep it up...

[mikedubya]

Not familiar with the theory myself but I agree. Not only will robots be able to design and build other robots but I also believe that their evolutionary rate would be exponential compared to us "carbons". It'll be the Second Renaissance from Animatrix, I tell ya.

[paxcoder]

they - can't - learn

at least not from what i've seen.

from my perspective idea they will create (ie program) is laughable.

[mikedubya]

My friend, if the history of science has taught us anything, it's that "impossible" or "laughable" in one generation becomes "normal" in the next. Try telling someone from the 18th century about the internet, or airplanes, or heart transplants. You'll be laughed out of the room. So, please reconsider.

[paxcoder]

*shrugs*

I never see strong AI in my life. So I wouldn't be quick to say robots can understand anything by our standards. If they will be able in some degree, then not in this century. But creativity, or writing their own programs, I think that's way over their hea... i mean chips.

[comehome14]

But there is one thing that they cannot put in a robot...

THE SOUL

[KingMahBoy]

What is the soul? If you think it is something immaterial, I am sorry, but you are wrong. Neuroscience disagrees with you. We are material beings, we are our brains. Brain Damage victims are proof of this.

[comehome14]

Is the brain the soul, or is it a connection between body and soul? You failed to look at all of the possiblities... :/

@ Williametched: Emulate most of its properties, maybe, but I doubt that the pitiful humans we are could make or replicate it. Unless something truly miraculous happens, I don't think I'll be seeing one in heaven.

[KingMahBoy]

No, I contemplated the possibilities. I don't think it is the connection between the body and soul. I think that the soul, or consciousness, arrises from neural collectivity. Different parts of the brain working together cause it I think, but I don't have proof, but to me, it makes sense.

[marfeb06]

he he he he

[RadicalProbability]

This is amazing stuff. "Knowing" it is a robot won't change your emotional response and attachment any more than "knowing" it is not your hand will stop phantom limb syndrome. We forget how much of what we are is deeply hardwired. We take our emotions and abilities for granted. Of course, that means we also underestimate how far robots have to go to achieve our abilities, or those of a mosquito for that matter.

[DrAdlaiAtkins]

I have never laughed harder than reading "Mankind's Retreat from Specialness" you take it personally

[halcyonyear]

awesome

[JedionEarth]

Very cool, at the end though he says that robots won't take over because people won't let it get that far...well, people already make viruses that cause great havoc, it's inevitable that some idiot will program a robot that kills (the military will) and it's inevitable that SOMEONE will make a virus or program a robot that goes into the streets and just starts shooting people. Who's to say what us humans can and can't control? Still though, it'll be an exciting next hundred years.

[bernito2]

for those who question human responce to a robot showing human-like reactions and emotion,as being far-fetch as humans will know it's just a machine,concider this;Alice Cooper,on a reacent interview on The Hour(check the youtube channel)said that when he did The Muppet Show way back in '78,he had reersals all week with the muppets and had the ardest time talking to the human puppeter who was just down in his field of view,but insted he talk TO the puppet as it reasponded in a life-like fassion.

[begajeta22]

we know, mozzi9, lol.. we know

[mozzi9]

lol its cool though haha just related to the video he says robots will take over the world and my video response shows the first human like robot lol so guess he is right

[mozzi9]

check out my video response its of the first ever human like robot its of a five year old asian girl freaky stuff

[BlackApollo666]

This guy has had sex with a robot, for sure.

[mrstrings2006]

It's the first thing that'll be done. A robot you can screw.

[BlackApollo666]

This guy needs some robot lungs to take care of the wheezing.

[KlausYTY]

I love AI. I don't care if AI be more intellegent than human, thats the goal. I want robot with emotion next.

[mozzi9]

my video repsonse shows a robot girl created not so long ago its so freaky check it out

[mozzi9]

my video repsonse shows a robot girl created not so long ago its so freaky check it out

[Totoroo922]

lol

[TXRebelOK]

Fundamentally, if you could replace all your neurons with equivalent silicon based 'neurons' that acted and responded as neural cells do, then what's the difference? The answer is that there is no difference. If you add 32 + 13 in your head. One person may at 30 and 10 and then add 5 to get 45. Another person may add 10 do 32 and then add 3 to get 35. Point is, it doesn't matter how you do it, the end result is the same.

[leevi1234]

Infinity is a nice thing for trying to understand. And there is allways something smaller than something. Inside an atom is an nucleus inside that is something and inside that.....till infinity. So large and small never end. And now i'm going to bed.

Summa Scientia Nihil Scire

[leevi1234]

Send 100 robots in space witch are capable of making double their amount of robots when found the right material. And so on and so on. And those robots task is to inform other (things they come across)about how we we're or hopefully are. So in time when these consistantly multiplying and travelling farther away robots reach a planet that has intelligent life (or at least as intelligent that we are now). Then they will know they are not alone in the universe.

[tyrbolo]

I think Rod Brooks is the most influential

scientist of the 20th century and also the

most controversial, because like Darwin he

has given us a key to really understanding

where we fit and how we work. It also makes

the inevitable task of discarding all of our

false ideas the next uphill battle.

[juliecranford]

Happy video.

[JohnDeBunkTest]

Robots will become far smarter than people if Moore's law holds up, and so far that seems like a reasonable guess.

Our first benefit will be a decrease in accidents as robotic cars take over most of the driving, then a reduction in work as robots take over production,

The big crunch will come when robotic intelligence takes over science and government.

[leanonme1]

Dont worry robots will evolve but so will we. We will find more ways of staying on top, think of genetic engineering and nanotechnology. One day (not too long from now) we could access all of human knowledge instantly and wirelessly by only thinking, like the movie "matrix" but without a big ugly USB port in the back of our heads. the future will be awesome for man and mech alike...

[roidroid]

Robots taking over production would essentially be robot "slaves", no?

Robots taking over government would be robots controlling humans. It's a reversal of the slave/owner control mechanisms.

Would it not mean that WE become the slaves?

[lordmetroid]

Yes, give me robot slaves... I WANT MY SLAVES!

[Hewhosmokeswomen]

I've heard Moore's Law is gonna collapse in about 10-15 years. Silicon isn't stable at a thinness of a few molecules. Check out some of Michio Kaku's videos, he explains this in one of his, he talks about a lot of cool stuff too.

[JohnDeBunkTest]

How did it begin?

See Ted Talks George Dyson: The birth of the computer

Where is it going?

See Ted Talks Kevin Kelly: Predicting the next 5,000 days of the web.

We are making a single machine, actually, we already have made it one we can no longer function without.

Will it begin to think

See Ted Talks Kwabena Boahen: Making a computer that works like the brain The time frame here is 2015, just seven years into our future. Four doubling periods.

[lordcheetah]

There's nano technology and quantum machines. If those are reliably developed, Moore's Law could continue.

[Lihinel]

Wow, great.