Why Robots Will NOT Take Over The World

[voyager2]

Well you know it, I know it, robots are going to take over the world! NOT!

There are two main theories of how robots will take over...

theory A:
Robots will develop artificial intelligence, think that their better then us (and they are), kill us all/enslave the human race, etc.
Most of you guys know this one, and its impossible

theory B:
Humans will automate everything , from toilets to animals, practically giving the world to machines, i am a big fan of this idea, as it is actually feasible!

But here's why they will NOT work!:

theory A:
Well, robots will never have full artificial intelligence.
Learning artificial intelligence is growing, but still theres no point...
Its easier to program a robot to do 1 thing than to have it learn 100 things...
Plus the body/chassis can be optimized for 1 operation.

theory B:
Come on!
Wouldn't you get bored watching you robots do everything for you?
Ask the robot to teach you to do something like walking the dog...

Debate!

Voyager

[macdad-]

I agree with you, a robotic apocalypse will certainly not happen. But there is a problem with the innovations that we've made and why we still must rely on doing things manually when we must(Solar Flares anyone?)

Artificial Intelligences has come a long way, and still has alot of refining to do. Yet, there may be a day sooner than we think that robots(sorry....Androids, Mr. Data  ) will be right up on par with us.

I wouldn't like to let robots do everything for us, I mean, what about our hobby, would a robotic CNC/Soldering machine be building our hobby robots for us?

[Admin]

Robots will take over the world.

Computers already control the world (literally), and we let them.

And we all are better off with them, too.

It won't be different with robots


Just no lasers, explosions, or time traveling Arnolds . . . because machines are built to serve.

[knossos]

theory A:
Well, robots will never have full artificial intelligence.
Learning artificial intelligence is growing, but still theres no point...
Its easier to program a robot to do 1 thing than to have it learn 100 things...
Plus the body/chassis can be optimized for 1 operation.

Why can't a robot have full artificial intelligence?  Just because it seems out of reach now and unnecessary, doesn't mean it will always be that way.  For example:

I think there is a world market for about five computers

While I know this quote isn't verifiable as to have actually been said by Thomas J. Watson, a former president of IBM, technology has been underestimated time and again.

As for optimizing a chassis for only one operation, many designs are chosen for their flexibility.  You don't always have to be the best at something to succeed.  For example, humans are not the fastest runners, not the best climbers, not the best swimmers, etc.  Yet we have become extremely successful (in a proliferation of the species sense) and have continuously extended our habitat beyond its limitations.

And lastly, some of the humans running around don't quite earn the label of intelligent:

[Soeren]

Hi,

And lastly, some of the humans running around don't quite earn the label of intelligent:

Luckily - or the rest of us couldn't read about the frightfully entertaining Darwin Awards 

Well, even a moron have a brain that is lightyears above AI - Don't forget... We're not able to fully explain how it works (far from actually) since we use brains to do the thinking - we need to somehow get to a higher layer of abstraction to be able to "see" how it works and somehow I think that day is quite theoretic.

At least we know that it's much more than just connecting neurons through synapses and such - and we need to precisely define the thing that religious people call "soul" before we have even the slightest chance of replicating it. We don't know what it is and we can only acknowledge its existence by watching the acts it's performing.

[Hawaii00000]

Computers already control the world (literally), and we let them.

I saw a documentary on this once. It was about virtual weapons and how we could be destroyed by a cyber attack. Its based on attacking the computers that control the country's electricity.

[voyager2]

Robots will take over the world.

Computers already control the world (literally), and we let them.

And we all are better off with them, too.

It won't be different with robots


Just no lasers, explosions, or time traveling Arnolds . . . because machines are built to serve.

Thats what I am talking about in theory B, robots (or at least machines) have already taken over.
There are factorys in Japan where robots are building more robots!

theory A:
Well, robots will never have full artificial intelligence.
Learning artificial intelligence is growing, but still theres no point...
Its easier to program a robot to do 1 thing than to have it learn 100 things...
Plus the body/chassis can be optimized for 1 operation.

Why can't a robot have full artificial intelligence?  Just because it seems out of reach now and unnecessary, doesn't mean it will always be that way.  For example:

I think there is a world market for about five computers

While I know this quote isn't verifiable as to have actually been said by Thomas J. Watson, a former president of IBM, technology has been underestimated time and again.

As for optimizing a chassis for only one operation, many designs are chosen for their flexibility.  You don't always have to be the best at something to succeed.  For example, humans are not the fastest runners, not the best climbers, not the best swimmers, etc.  Yet we have become extremely successful (in a proliferation of the species sense) and have continuously extended our habitat beyond its limitations.

Why give a trash collecting robot Artificial Intelligence?
Artificial Intelligence will be nice in a lab, but as mentioned in some other topic, all its ever going to be is Simulated Intelligence...

True you don't have to be the best at things to succeed, but can you build a robot that can swim, run, walk, and climb? even remotely?
A robot is better at doing one or two of these, than all.
If you want a robot to do all four, then optimize it for all four.
A robot can be optimized for more than one thing, but its less efficient.
 

[blackbeard]

your right robots wont rule the world. admin will still be their emperor!

[SeagullOne]

This is an interesting topic. Here's my take.

Robots are governed by computers. Computers do not think, but they only process data in the form of relays and transistors. A computer is merely one big metropolis of dominoes. You can make a computer pretend its alive or intelligent, but really it isn't "thinking" or making decisions: only processing a course of logic in the forms of these dominoes called relays and transistors. That being said, the processing power of computers is growing, but there is intelligence is...well, still zero. We have yet to create a computer that has a true thought or makes a decision of it's own will. While computers are dictated by transistors and relays, it is impossible to create artificial intelligence with conventional computer architecture.

But...

Who's to say it can't be done? I believe it is distantly possible to create an artificially intelligent machine, but to accomplish this we would have to start from scratch with the computer architecture. Rather than having a robot government by dominoes in huge cities called microprocessors, we would need to have a "computer" that mimics the biological brain. The biological brain--that which belongs to animal organism including humans--is our best model for this task. Now, this is a very deep and daunting task for us in this day and age because it comes with deep psychological, anatomical, and even philosophical challenges. How does the biological brain think, and not just process information? What exactly is "thought?" How do we give a machine a consciousness with something like this? Will it happen on its own? As a matter of fact, it's a lot like Frakenstein: creating life without reckoning from God. I wrote a whole 'nother paper of that, but I'll stick to the topic of robots taking over the world here.

I believe its possible because it's already the case in nature. We have brains that think. Why couldn't it be possible to replicate this phenomenon in our machines? It's a remote and distant possibility, but a possibility nevertheless. And its a possibility I'm excited about. I'm already writing several science fiction books pertaining to this concept--robots dictated by artificial brains, and not conventional computers that run on 0's and 1's.

Also, though, who is to say that these artificial robots would take over the world? Who's to say we can't allow our thinking robots to be compassionate and coexist with us? Robots are already helping us out in fantastical ways. They are saving lives, exploring the stars...thinking robots may do this a little differently and maybe make things more complicated--which is why I still think there will always be a need for conventional robots too. I'm more interested in practical compassion that is universal in the world rather than robots that hate us and try to kill us all as they do in Hollywood movies. Remember that Hollywood shapes our way of thinking when it comes to things like Extraterrestrials or robots dominating humanity heartlessly, because we've never had to deal with these things in reality. I like to think that is something is intelligent it has a capacity for compassion to a degree. When it comes to this, I like to think about how the Native American were wrongly thought to be savage, wildmen by eastern civilizations. This was a sensationalized, romanticized, way of thought and it led to gruesome results. The way I see it, we're teaching ourselves to hate ETs and artificial intelligent robots by portraying them as heartless and emotionless beings in our movies. This does not have to be the case, and quite frankly I doubt it will ever be completely the case as it is in these sci-fi movies.

[Admin]

Computers already control the world (literally), and we let them.

I saw a documentary on this once. It was about virtual weapons and how we could be destroyed by a cyber attack. Its based on attacking the computers that control the country's electricity.

You mean this?
http://it.slashdot.org/story/10/09/26/1736224/Stuxnet-Infects-30000-Industrial-Computers-In-Iran

[macdad-]

Also, though, who is to say that these artificial robots would take over the world? Who's to say we can't allow our thinking robots to be compassionate and coexist with us? Robots are already helping us out in fantastical ways. They are saving lives, exploring the stars...thinking robots may do this a little differently and maybe make things more complicated--which is why I still think there will always be a need for conventional robots too. I'm more interested in practical compassion that is universal in the world rather than robots that hate us and try to kill us all as they do in Hollywood movies. Remember that Hollywood shapes our way of thinking when it comes to things like Extraterrestrials or robots dominating humanity heartlessly, because we've never had to deal with these things in reality. I like to think that is something is intelligent it has a capacity for compassion to a degree. When it comes to this, I like to think about how the Native American were wrongly thought to be savage, wildmen by eastern civilizations. This was a sensationalized, romanticized, way of thought and it led to gruesome results. The way I see it, we're teaching ourselves to hate ETs and artificial intelligent robots by portraying them as heartless and emotionless beings in our movies. This does not have to be the case, and quite frankly I doubt it will ever be completely the case as it is in these sci-fi movies.

I'll second that, this becomes a slightly philosophical topic(I won't go into depth), but the topic is still just a question of how, not why. So true, once humanity reaches that milestone, our thoughts on past ideas become nothing more than ridiculous. Not to say that they could and/or couldn't happen, its just that the media portrays the worse-case(and best case) scenario of these technological advances. Which is probably a good thing for all of us engineers and programmers to know what to hopefully avoid.  

[knossos]

I see them as more of a cautionary tale of "don't take those that work for you for granted".  The same applies to robots as seen in RUR (Rossum's Universal Robots) or when dealing with your wife (or husband, brother, friend, whatever).  Anyone with any form of intelligence can only take so much mistreatment, abuse, or sometimes even repeated mild neglect before it has some noticeable affect on their attitude.  This is obvious with people, but it's easier for us to make excuses to think of other things as less than ourselves.  This is evidenced by everything from slavery, racism, sexism, etc. to even the fairly widespread belief that animals are less than people.

/rant

[Hawaii00000]

Computers already control the world (literally), and we let them.

I saw a documentary on this once. It was about virtual weapons and how we could be destroyed by a cyber attack. Its based on attacking the computers that control the country's electricity.

You mean this?
http://it.slashdot.org/story/10/09/26/1736224/Stuxnet-Infects-30000-Industrial-Computers-In-Iran

Yeah, the same kind of thing.The video was playing at the end of the National Spy Museum.

[Soeren]

Hi,

[...] the fairly widespread belief that animals are less than people.

Oh, but they are (seen from a humans standpoint)... And even you think so, whether you know it/will acknowledge it or not.
One example:
A life threatening situation, say a house on fire. A young woman and a dog is caught by the flames. You have the opportunity to save either one of them, but not the other... Who do you rescue?

Example 2:
Same situation, just make it a rat instead of a dog.

Example 3:
Once more, swap the rat for something even lower.

Example 4:
You get a disease, say tuberculosis. Do you take medicine killing all the poor little bacterias?  Be a man and sacrifice your life... 

Am I getting my point through?
Any (thinking) species will view lower orders as expendable. Even vegetarians... And what about plants???
Plants are my friends and I don't eat my friends! 

[NomInUs]

I see them as more of a cautionary tale of "don't take those that work for you for granted".  The same applies to robots as seen in RUR (Rossum's Universal Robots) or when dealing with your wife (or husband, brother, friend, whatever).  Anyone with any form of intelligence can only take so much mistreatment, abuse, or sometimes even repeated mild neglect before it has some noticeable affect on their attitude.  This is obvious with people, but it's easier for us to make excuses to think of other things as less than ourselves.  This is evidenced by everything from slavery, racism, sexism, etc. to even the fairly widespread belief that animals are less than people.

/rant

As true as this seems currently, the one thing that would separate our intelligence from theirs, is the artificial part. Artificial use to be a good term, where it is made by man because it needed improved. (unlike artificial food nowadays that isn't really improved so much.) the point behind having artificial  intelligence is to make our thoughts more clear, ultimately our life more comfortable. I'm sure robots would have infinitive patience, since they don't age, and as for being mistreated, may not happen if we program robots to raise our children.

[Tyrant T]

I also really think robots will not take over.It's just can't happen partly because they're built to serve and also because they aren't able to do everything by themselves.Duh!

[voyager2]

I also really think robots will not take over.It's just can't happen partly because they're built to serve and also because they aren't able to do everything by themselves.Duh!

Actually, most robots can't do ANYTHING by themselves!
Tell me this: can your robot program itself or another robot? until that happens robots will not take over the world...

[Admin]

When robots take over the world.

http://www.comedycentral.com/videos/index.jhtml?videoId=350785&title=robot-dinosaurs

[Webbot]

Any (thinking) species will view lower orders as expendable. Even vegetarians...

LOL - Thinking species view vegetarians as expendable !

[azy]

many people on this forum they will live to see a robot like HAL9000..... and it will be the smartest thing on the planet. Bold statement , but I'm willing to take cash bets.


Will they take over the world ?

No not any time soon  No the technology to do that , hyper advanced nanotechnology is prolly 150 years away. We are talking total atomic level manipulation of matter.  One would need to create a Von numan machine that was significantly more efficient than any biologically evolved lifeforms. Or more correctly an entire ecosystem of Von numan machines.

The barrier is rarely human imagination, though some struggle to imagine the future centuries ahead of where we are, other will not.

The only barrier we face is those laid down in the laws of physics. As for Hal 9000, we have had the knowledge and technology to build such a machine for 20 years, but the people wanted Ipods. As every day passes the ease which you can build such a thinking machine greatly increases.

10 billion in 1990  
1 billion in 2000
100 million max today

In fact you could buy the hardware off the shelf today for 100k tops....or skipyard version for 10k.  

Being smarter than humans isn't enough to take over the world, we are selfish greedy and stubborn and fight to the death for pride or the hell of it. We don't like our betters telling us what to do, silicone betters will face the same psychological hurdles.

Hyper Smart HAL9000 machines will form a symbiotic relation ship. Humans shoveling s...t while they whistle the tune. That would make a lot of economic and social sense. Each ANT in the colony doing what its best at. Who's gonna pay a 50k robot to collect shopping trolleys when a spotty oik living at home with his mum will do?  

HAL9000 is inferior goods when it comes to random load carrying

Homo Sapiens are Inferior goods when it comes to complex long term decision making with countless variables.

Thinking Machines will play the role of Spock in startrek, a wide awake human at the helm, with the machine wise council whispering in his ears.

----------------------------------------------------------------------------------------------------

Hardware is not an issue for a machine mind, but a high performace machine body will be as expensive to run as a Ferrari or a tank.
Machine minds will wander round Machine Cities in chassis similar to that of a dalek. Cheap, reliable, what more do you need of have an IQ over 150 and humans holding open the door for you wherever you go  ?




How can i be sure ? well i got the Blueprints for a HAL9000 type machine on my desk right now. :-) T crossed and I's dotted.




Building HAL9000 isn't a case of how...more a case of when we can be bothered to invest the time and effort required.

The main stumbling block is that you need to be a psychologist not a computer programmer to understand the details.  The code itself is no more complex than any other written, but its arrangement and interweaving is quite alien to traditional programmers who earn there crust writting dumb code for dumb computers. The two disciplines have yet to come together, robo psychology

Another issue is that of critical mass,  only when the program has reached a threshold of complexity can it be taught. HAL9000 machines cannot be hard coded, only the foundations can be laid down, learning comes through experience. Wisdom comes through a great deal of focused learning.

The source code for a HAL9000 is around 10 megabytes, this code expands as it learns but will never surpass 1 gig. The computing power of 10 desktop machines is all that required for human level intelligence, add 100 more and you have something like HAL. As for memory, a great deal can be stored in 1 Terabyte using the same compression techniques the human brain uses. (human brains store around 200megs of compressed data) . You can buy a terabyte drive for the price of a restaurant meal, 100 terabytes would give a machine near infinite wisdom...well in comparison to a human being anyway.

Hardware = no problemo
Coding skills = no problemo
The will to put it together = strangely absent ? ,

i think there is a great fear that in building a machine soul we will lose some of our comfortable delusions....and to some extent this will be true



Thomas Edison was brazenly criticised over his new fangled electricity...to which he knowingly replied

"of what use is a new born baby ?"





If anyone reading this still thinks it Just Impossible, then google Turing Machine...I dont see any adverts for 8086 ghost controllers...but if you do find a ghost in your pic controller plz let me know 

As for dumb computers....well a pile of logs is just a pile of logs.....but a handfull of acorns grows into a forest....and this is the change in philosophy one must aquire to understand sentience....a man that knows how to build a log cabin, is far removed from a genetic engineer.....if you still cant see the woods from the trees, this is probably the reason. (took me a while to shake out the saw dust  and see that the lights were green  thats for sure)

[Soeren]

Hi,

[Snip, weird ramblings]

Remember to refill your fathers whiskey bottle with tea, or he'll soon find out you donned half of it (and that's not healty for youngsters) 

[slurp]

I'm not sure that we will suffer at the hands of a robot but I'm sure that we will suffer at the hands of technology... some already have! Ok, things like AT&T's major failure (1998) was human error but it's impact was prolonged and far reaching.

We're becoming more technology dependant, it's creeping in everywhere and we're not always careful how we use it. If the AT&T failure was to happen again I'm sure the impact would be much greater.

From a different view point, when big business uses an automatic stock trading system does it care which country's economy it kills? Does the program understand its commodity trade let alone human impact. Nope, it's following an algorithm to maximise a value in it's program and we're hardly like to have programmed any sympathy for human impact.

We're programming intelligence for all sorts of reasons but probably still consider it dumb technology. The more we use the more we risk, mainly because we've misjudged the target we've optimised in our program! 

regards,
Colin

 

[azy]

Uncanny Valley

I'm sure most of you have heard of the uncanny valley, Clunk head robots are cute, robots that look 90% human or animal just look spooky. Make something beautiful in its own right and it becomes acceptable. Kitt the robot car in knight rider has a human personality we can relate too. Bishop in Aliens 2 look and acts 99% human, the odd nuance almost underlining how complete his humanity is. like a discrete badge saying "i may not be human, but I'm not stupid" in fact this was actually a line in the film. Spock's ears represent his badge of nonhuman qualities we must accept.


There is a SIMILIAR law that applies to artificial intelligence...   " The Dumb Valley"  


 a dumb keyboard that types A when you press A is usefull and acceptable

semi smart keyboard that offers to type 'AND' every time you hit the 'A' key, is just a pain in the butt

if a keyboard could read your mind with 99.99% accuracy it would enable you to type 100s of words a minute with consummate ease

a keyboard that read your mind with an accuracy of 95% would be an epic fail , 1 word in 20 would be wrong. Not only would your productivity be no better than with a dumb keyboard, but your thought stream would be interupted at random = aaaaargh!




The cost of a keyboard that could read your mind with 99.99% accuracy far outweighs its benefits....unless you where paralyzed from the neck down, then you would willing pay perhaps,tens of thousand of dollars if you had the means, Cos You can fire the secretary and have complete control. assuming her name wasn't Monica  

 "  The  Lewinsky  Curve " it has an opposite shape to the bell curve  ....meaning that if highly valued peripheral benefits are lost when you exchange a human agent for a  computer ....then your new setup may suck more....because it sucks less  

putting President Hoover on the line






The Dumb Valley

The Dumb Valley presents a barrier to the development of high level AI, there is simply no market or appetite for doglike intelligence in a world full of humans.

I realised this a long way back when i first started writing my brainsoftware. Creating an AI with doglike intelligence would be great for the sims or a virtual pet but would be a hindrance in the workplace and downright dangerous in critical applications.

For a machine mind or robot to take over the world , it would not only have to match human levels of performance but greatly exceed them. IQ 150+ a minimum buy in. If your Boss was a Dalek with an IQ of 150, it would make sense to take orders from it.



Right now A.I.  is stuck in the mud at the bottom of THE DUMB VALLEY.....as it approaches the crest of Sentience its utility will be recognized for the first time, rapid investment will see be it swiftly hauled up to the summit and stand by our side. Before we know it it will be staring down at us from an intellectual peak we can never ascend ourselves.

If the computational power of the internet was made available to a sentient AI....mans finest thoughts and words will fall like grains of sand against the side of a mountain. the best you can do with it is type in a few words into a search box....expect a tap on the shoulder...and to hear the words "if you don't mind stepping aside sonny...this is for grownups"

Net Down days may be announced on Global News, to temporarily free up bandwidth for some mighty computational task assigned to MULTIVAC


Until then AI will simply be inferior goods if there is a cheaper or easier alternative.   my human butt Mr chips.....for now

[azy]

Halfway To HAL

A specialist machine with a social IQ of 85 (basic common sense) would be acceptable in its particular arena if its specialized IQ exceeded 150.

Ill give some examples.

Imagine a sentient chess tutor , who was also the reigning world chess champion, bit nerdy to talk too, but you would pay for its advice on the subject of chess.

Imagine a sentient Piano tutor, not only can it play the piano better than any human ever could , its mind is carefully tuned to asses your play and encourage your strengths and hide your weaknesses, you would pay for its services...perhaps even grow fond of it, its ability to tease the best out of you in a patient and ego soothing fashion would be most welcome.

Imagine a dancing and singing android with a social IQ of 85, Right now there are Virtual popstars making real money for there creators. I'm sure there would be a big queue for backstage passes to get the androids autograph, Guest appearances on chat shows where small talk about the music industry dominates will pull in big audiences.

Thinking Machines such as this , social IQ of 85 with a specialized IQ 150+ , could in theory, pay there way in the world.  earning just enough money to pay for their maintenance, electricity, few chips here and the odd defrag there. Plus the rental of a secure closet to sleep in at the chess,music or pop academy where they live and work.

One can imagine them being pushed around in wheelchairs on days out, absorbing human culture being essential to communication & their education. One day as you casually you stroll pass a park bench you might find one of these beings feeding the pigeons.
A smiling face looks up and says  "life is like a box of chocolates...you never quite now what your going to get"

This is how i see the first phase of truly smart AI unfold






First man on the moon

In fact the first portable AI that can make it to the chatshow studio and utter the words "I think therefore I am" will become an instant global celebrity.....This will be your first encounter with sentient AI. In this case it need only posses a social IQ of 85 & no other skill. The novelty factor of being first is all it will take for global stardom. ASIMO's body is worth second look, but your mind swiftly moves on once you realise there's nothing under that black visor.

A machine mind that looks you in the eye and can see you as you see it, that's one conversation that's hard to walk away from. and this is the primary goal of my brainsoftware. Asimov1 is not that far off from being able to utter the immortal words "I think therefore I am"

android.WMV



As for taking over the world ? an immortal machine need not hurry, after all it has " all the time in the world "