IBM Watson Hits Jeopardy to Destroy All Humans

[victomofreality]

I think it knew it was answering wrong and thought it was a joke... It knew it was going to win and was toying with us...

 

[sacre]

[citation][nom]Objectivist[/nom]What you are failing to see is the METHOD that humans gain knowledge. A human gains knowledge by observing the facts of reality. This is only made possible by their 5 senses, which are provided by their body but "made sense of" by their mind. The end result is concept formation. The process we use to form concepts is logic, or the non-contradictory identification of facts. This conversion of sensory data into conceptual facts, or knowledge, is called reason. Machines do not reason. They do not have senses. They simply execute the instuctions that a human has imputted into their aftificial memory. They have no ability to obtain more data than their programmer's limited instructions have allowed them. A human's mind is free to think independant of someone else's permission or instruction. Even if a human is chained and prevented from taking action, their mind is still capable of functioning to the extent of the sensory data around them provides. No one can FORCE you not to think, so long as you are alive and your brain is physically functioning. A machine is at the mercy of it's creator and programmer.[/citation]

A machine is at the mercy of its creator, but so are we. We have the ability to shut the mind off, we have the ability to "turn off" a human, and even screw with our "programming".

At the current time of technology no we cannot replicate the Brains capabilities, but in due time we will be able to replicate and eventually surpass the ability of the brain. We've already made robots that are capable of learning (there was a robot that didn't know how to balance/walk, they gave it some time with a learning algorithm and sure enough in due time it learned to walk more efficiently than the robot that was Programmed to walk right away). This is a perfect example of how to approach this technology, they have to let robots learn things themselves. To program them to do ALL tasks would take many years and would be buggy, so now they are working to approach things with learning algorithms.

Yes at the moment they are ALL limited to poor programming, but in the many years to come we will come up with more. another thing to really check out is Quantum computing - this will really boost the ability for machines to become more intelligent.

If you think about it, soon enough all it'll take is a small program. Let the machine go and it'll come back in no time with loads of information, capable of holding coverstations, etc. You grab that information and impliment it onto future machines, and repeat this process.

we will never hit a wall with this technology, lot of people think the wall is the programming but even today we're finding new ways. in 30 years we may have walking talking robots, in 50 years self replicating robots capable of doing almost all jobs. in 100 years Robots may be a race all on their own.

Maybe not, but the fact is humans want to replicate AI perfectly, and we will continue for many years to do this... And if our history has proved that we will achieve this.

I for one am looking forward to having a robotic companion to assist and hold good conversations.

 

[MaxTesla]

[citation][nom]junkiei9387e[/nom]take away its internet access and format c: /y and see how it does after that..... heh. Here's another suggestion... have someone set off a pinch (EMC pulse) and see how the computer does then.... spill your water on it players and see if it can then keep up with you.... machines will never replace humans.[/citation]

No you don't understand

It has not gotten warmer EVERY 30 years since the last ice age 10,000 years ago

For example there was a small mini ice age during the 1700s

2010 was the warmest year ever recorded on the planet as a whole, do you understand that part?

Do you also understand that if it had gotten warmer EVERY 30 years since 10,000 years ago the earth would have boiled

Global warming is not false as you may think and a computer will most likely understand that where you do not

But if you want every single place on earth to be warmer every single day before you “believe” in global warming then by that time everything will be lost

 

[MaxTesla]

[citation][nom]f-14[/nom]i understand that part as every 30 years it has continuiously gotten warmer then the last since the last ice age ended, other wise we'd still be in an ice age, and there's been 3 ice ages that we know of. what i was getting at was when will this machine replace man with out having the same fallacies we do and that i was surprised toms even reported this after being outscooped by pretty much most of the media the afternoon 'Watson' won.came back to scoop Toms againGIGABYTE X58A-OC Introduces New OC Touch Feature: Pushes Core i7 990X CPU to a Record 7.1GHzhttp://www.youtube.com/watch?v=b0X [...] r_embedded[/citation]


No you don't understand

It has not gotten warmer EVERY 30 years since the last ice age 10,000 years ago

For example there was a small mini ice age during the 1700s

2010 was the warmest year ever recorded on the planet as a whole, do you understand that part?

Do you also understand that if it had gotten warmer EVERY 30 years since 10,000 years ago the earth would have boiled

Global warming is not false as you may think and a computer will most likely understand that where you do not

But if you want every single place on earth to be warmer every single day before you “believe” in global warming then by that time everything will be lost

 

[Objectivist]

A machine is at the mercy of its creator, but so are we. We have the ability to shut the mind off, we have the ability to "turn off" a human, and even screw with our "programming".

At the current time of technology no we cannot replicate the Brains capabilities, but in due time we will be able to replicate and eventually surpass the ability of the brain. We've already made robots that are capable of learning (there was a robot that didn't know how to balance/walk, they gave it some time with a learning algorithm and sure enough in due time it learned to walk more efficiently than the robot that was Programmed to walk right away). This is a perfect example of how to approach this technology, they have to let robots learn things themselves. To program them to do ALL tasks would take many years and would be buggy, so now they are working to approach things with learning algorithms.

Yes at the moment they are ALL limited to poor programming, but in the many years to come we will come up with more. another thing to really check out is Quantum computing - this will really boost the ability for machines to become more intelligent.

If you think about it, soon enough all it'll take is a small program. Let the machine go and it'll come back in no time with loads of information, capable of holding coverstations, etc. You grab that information and impliment it onto future machines, and repeat this process.

we will never hit a wall with this technology, lot of people think the wall is the programming but even today we're finding new ways. in 30 years we may have walking talking robots, in 50 years self replicating robots capable of doing almost all jobs. in 100 years Robots may be a race all on their own.

Maybe not, but the fact is humans want to replicate AI perfectly, and we will continue for many years to do this... And if our history has proved that we will achieve this.

I for one am looking forward to having a robotic companion to assist and hold good conversations.

Please explain what "perfect AI" would mean. AI means artifical intelligence, which means "intelligence" created by humans and programmed in. Human intelligence is not perfect in the sense that man can make errors in logic. However, it is "perfect" in the sense that his method of gaining knowledge (i.e. reason) has no inherent limit to identifying the facts of reality. Anything that is incapable of being known is a contradiction in terms and thus does not exist. (i.e. heaven, God, hell, the devil, or any other mystical idea).

To say that machines will one day have "perfect intelligence" means that they will one day use reason, rather than a set of instructions programmed by humans. This is more than a silly idea. Nothing can change the nature of a machine from a man-made object that uses voltage to translate 1's and 0's to a type of living being that suddenly uses reason to survive.

The other point that needs to be reiterated is that "replicating the brain" is not enough to put a machine on the level of a human because the brain alone is not what makes reason possible. It is the synchronization between mind and body (i.e. between data obtained by your senses and processed by your mind through the use of reason) that makes human intelligence what it is. Humans use perception to form concepts. This involves a process of integration and differentiation of the facts of reality. The nature of a computer prevents it from doing the same.

 

[jprahman]

A machine is at the mercy of its creator, but so are we. We have the ability to shut the mind off, we have the ability to "turn off" a human, and even screw with our "programming".

At the current time of technology no we cannot replicate the Brains capabilities, but in due time we will be able to replicate and eventually surpass the ability of the brain. We've already made robots that are capable of learning (there was a robot that didn't know how to balance/walk, they gave it some time with a learning algorithm and sure enough in due time it learned to walk more efficiently than the robot that was Programmed to walk right away). This is a perfect example of how to approach this technology, they have to let robots learn things themselves. To program them to do ALL tasks would take many years and would be buggy, so now they are working to approach things with learning algorithms.

Yes at the moment they are ALL limited to poor programming, but in the many years to come we will come up with more. another thing to really check out is Quantum computing - this will really boost the ability for machines to become more intelligent.

If you think about it, soon enough all it'll take is a small program. Let the machine go and it'll come back in no time with loads of information, capable of holding coverstations, etc. You grab that information and impliment it onto future machines, and repeat this process.

we will never hit a wall with this technology, lot of people think the wall is the programming but even today we're finding new ways. in 30 years we may have walking talking robots, in 50 years self replicating robots capable of doing almost all jobs. in 100 years Robots may be a race all on their own.

Maybe not, but the fact is humans want to replicate AI perfectly, and we will continue for many years to do this... And if our history has proved that we will achieve this.

I for one am looking forward to having a robotic companion to assist and hold good conversations.

The point is that there is more to the human brain that a collection of neurons, there are aspects such as self-awareness, the ability to reason, emotion, and creativity that the source of which cannot be quantitatively evaluated and replicated no matter how good we get at programming or neuroscience because quite frankly those qualities of the human brain fall outside of the realm of physical science.

New technologies like quantum computing will allow researchers more horsepower to implement more advanced learning algorithms, but there is still the fundamental problem that there are aspects of the human MIND that cannot be replicated. You can create better learning algorithms, but learning from surroundings is not the same thing as reasoning, emotion, self-awareness, innovation, creativity, intuition, emotional intelligence or any of the other attributes that make humans human. For a machine to observe a previously unknown physical event and record it is one thing, but for a machine to make a conjecture as to the cause of that previously unknown event is completely different. A human scientist on the other hand can both observe and record the physical event AND figure out the cause.

 

[jj463rd]

The first cyborg created will look and act like Arnold Schwarzenegger.
Also eventually this planet will be transformed into a machine world planet and the environment will be transformed for machines.So Lead will be perfectly O.K.
Bye Bye E.P.A.

 

[eddieroolz]

I'm amazed that the Watson was able to comprehend the questions mostly accurately. Our brains will remain the best supercomputer, no doubt; but the fact is, our brains were able to create this amazing computer. That deserves some recognitoin.

 

[somata]

[citation][nom]jprahman[/nom]The point is that there is more to the human brain that a collection of neurons, there are aspects such as self-awareness, the ability to reason, emotion, and creativity that the source of which cannot be quantitatively evaluated and replicated no matter how good we get at programming or neuroscience because quite frankly those qualities of the human brain fall outside of the realm of physical science.[/citation]
The validity of that belief remains to be known. Subjective consciousness is indeed a very tough nut to crack. I'd argue it's actually the hardest problem facing science (and yes I do believe it is ultimately in the domain of science). Despite our introspective notion that consciousness is something apart from the physical world, there is plenty of physical evidence to suggest otherwise. Consider that there are only relatively subtle differences in our DNA that even make us human at all, and not a chimp or mouse... or a banana (tree). If we can agree that mice don't have the same level of consciousness as humans, this suggests that it is the fundamental structure of the human brain (determined by DNA) that gives it the special qualities you mention. Those qualities would then have to be nothing but emergent and complex abstractions of the underlying biology.

Given that, I see no reason why such qualities would be exclusive to humans, or even life in general. A computer's programming is akin to the brain's structure. The software/hardware distinction of computers is irrelevant, of course, since software is merely an abstraction we use to make building electronic systems easier and more flexible (i.e. anything that can be programmed can be directly implemented in hardware as well). Likewise, the brain (which operates nothing like a computer and so is entirely "hardware") compares to the entire software/hardware system. Even though computers "just blindly process instructions serially" on the face of it, their software is ultimately an implicit part of their structure.

I only bring that (sometimes forgotten) distinction up so it becomes clearer that the brain has "programming" too, just not in the same way a computer does. Just because computers are often programmed/structured to carry out a specific "dumb" tasks doesn't mean neurons can't also be arranged to do the same "dumb" tasks. Likewise, neurons can also apparently be arranged to become the marvel that is the human brain, and it would seem that so too can software. All that's lacking is the proper "structure" (programming).

 

[liveonc]

[citation][nom]jj463rd[/nom]The first cyborg created will look and act like Arnold Schwarzenegger.Also eventually this planet will be transformed into a machine world planet and the environment will be transformed for machines.So Lead will be perfectly O.K.Bye Bye E.P.A.[/citation]
The first cyborg created to fool people will be a Japanese Robot that looks exactly like Steve Hawkings, talks like Steve Hawkings (with somebody else doing all the typing), & will freak out the real Steve Hawkings... ;-)

 

[everygamer]

The only thing I would like to see to make Watson more impressive is if he actually had to pick up the with voice recognition software. From my understanding they were transmitting the questions in plain text to Watson right when the buzzers were activated. There was one point when one of the two human players buzzed in, got the answer wrong, and then Watson buzzed in and gave the same wrong answer. If it was designed to use voice recognition it would have been able to adjust its answer.

That said, still very impressive on a magnitude hundreds of times more difficult than playing chess.

 

[Guest]

I wonder how much power was needed to power watson.

Im sure it exceeded the power consumption of a human brain by miles.

It must have cost millions to make.

 

[psycho sykes]

I certainly want something like watson in my house and a powerful enough invisible speaker in my ear and I'll happily go and play that game 😀