Dangerous New "Artificial Intelligence" Learns Without Human Intervention

Intelligence requires consciousness, I believe, and if it is conscious, I would be curious as to how it perceives us (the humans, its creator). If it can detect pixels on a screen, it would seem easy enough to give it cameras, and microphones, to see and hear. Would it think we were its God(s), view us as inferior, or possibly be indifferent. Also, would it have the ability to imagine, or act of its own free will? Could it teach and pass down information that was learned prior, to future generations, thereby saving them the trial and error process? It is all really fascinating stuff, and if anything, may give us some insight as to how we came to be.I don't know...humans still frighten me much more than machines. Emotions, from our 'early brain' is both a beautiful and dangerous thing, and something a machine such as this will never have to experience. (Sorry, kinda rambling...hehehe)

originally posted by: eatbliss
Emotions, from our 'early brain' is both a beautiful and dangerous thing, and something a machine such as this will never have to experience.

I disagree. I think that we can give it emotions if we want to. It's the old notion that if you can't tell the difference between something that is "real" and something that is "artificial," then there is no difference. If a little tamogatchi says it's lonely or hungry, how is that different than you or I being lonely or hungry? And if you had a very sophisticated tamogatchi that was programmed to be aware of and regulate things like pride or jealousy or ambition, could you be sure that it's any less authentic than what you feel?

Sincere, honest, hard-working robots who love their creators and want to make them proud of their accomplishments? Why not?

originally posted by: eatbliss
Sincere, honest, hard-working robots who love their creators and want to make them proud of their accomplishments? Why not?

The problem will come when it ask....Creator do I have a soul?

originally posted by: yuppa
The problem will come when it ask....Creator do I have a soul?

I would tell it honestly that yes, you do, if you think of a soul as something like a waterfall, where it has to be in action and moving for it to have form and shape and a name. Once the water stops flowing, though, once the machine loses its power, then it no longer exists.

A soul is a point of view. It exists, and then it doesn't.

I think if you are aware enough to ask if you have a consciousness, you probably do.

a reply to: Blue Shift

Thats your view,but a majority of people would seriously flip out if suddenly these machines claimed souls. Were talking witch trial level flip out here.

a reply to: yuppa

I'd ask it if there's enough space on my iPhone for it to travel around with me in my pocket.

*singing*

"I can show you the world...shining shimmering splendor..."

originally posted by: MystikMushroom
a reply to: yuppa

I'd ask it if there's enough space on my iPhone for it to travel around with me in my pocket.

*singing*

"I can show you the world...shining shimmering splendor..."

MM i myself would ask it to fix the publishers clearing house for me.

a reply to: yuppa

Nah, that's to easy.

Long walks on the beach, followed by existential debates...

not surprised...

We are Extremely behind !

10 to 30 years behind !

I wanted my Hal 9000 ! & KITT ( Knight Rider ( I really DID ) become a reality as a KID

Why I’m not expecting Hal 9000 to come out of a corporate innovation lab
venturebeat.com...

KITT
en.wikipedia.org...



5 to 10 years ago Honda made a A.I. Sight recognizable Face Object and gesture ( that Remembers ) Droid

The ASIMO..

a reply to: mister.old.school

Such BS. The human mind is incomprehensible in it's current state. When the cpu has NEUROPLASTICITY, then come back to me, as you might have half an argument.
That's not to say nigh-sentient ones shouldn't have rights, just that they won't touch humans for another 4 centuries.

a reply to: onequestion

The computer learns by trial and error and builds it's own understanding/structure of the game. The games used in this experiment had very simplistic controls and game rules. One of the problems with this system, is it lacks a human's capabilities to transfer the data learned about "gaming" over to the next game and had to start from scratch.

This isn't the first learning machine either. Watson was developed to read and learn the human language by constructing it's own information hierarchies from "reading" the internet. They programmed the learning process then had it read wikipedia and a few billion other pages. It was able to understand the nuances of the English language to win Jeopardy.

a reply to: Evil_Santa
There are a few different things here: dopamine like transfer reward system. No one is paying enough attention, imho, to this. What does this mean, actually. In other words, I want to know, exactly how this dopamine transfer is rewarded, so I understand what that means.

As far as the rest, the other part isn't being paid enough attention to that's different than other artificial intelligences, is the enjoying winning part. It has to understand a sacrifice lose, in that case….. If it doesn't, then we're in real trouble…..

a reply to: sharpturn

My team worked on neuronal networks and learning algorithms for application in programmed trading black box work. We determined after so many early successful systems later failed that this was due to a lack of Neuroplasticity in the neuronal networks.

These systems were built by highly talented math modelers, coders, and custom hardware people and worked for a while in real world systems with active accounts that were well funded. They made money but in the end started to lose money even when they used very different basic methodologies.

Of course reasoning through failures one finds reasons if one wishes to ever get development money for a new project. That being said AI research funding is at present just about evenly split between defense money, and banker money which means it is all coming out of the same deep pockets.

So that is to say that your estimate of 400 years before MI surpasses human intelligence? I am wondering which hand your wishing into. In 40 years humans will be fully meat and machine integrated hardware running hugely quantum entangled data sets of infinite complexity, and likely hard for us now days to recognize from what we are today.

I'm still Waiting for a BIO Computer to do the same thing ... Self Learning Self Aware ! to Play with the Big Boys a little more then just Playing Vintage Games ..

Something you would see in the Movie Called Demon Seed the BIO Super Computer Called Proteus IV
that was Made In 1977

Proteus Speaking Room



The BIO (Organic) Super Computer Proteus 4



demonseedroteus requests a dialogue



ATS you gotta see this Describing how Proteus IV Functions a ( Organic Bio Super Computer )
( see Link Below ( Movie Clip )

It's a Brain, an Artificial Brain
www.criticalcommons.org...

in this clip from the movie Demon Seed, Dr. Harris describes a new type of computer, one that is organic. When asked if the computer, Proteus IV, is alive however, the question seems ludicrous. Moving beyond the reliability of the mainframe computer and beyond the speed of the supercomputer, Proteus is an unknown entity, a machine that can learn and become more intelligent than its own creator.

( The Fear Mongering ) The Dangerous ... Since 1977

Demon Seed Trailer



The Reality


THE SCARY Real Life REALITY

A Working Transistor Built Out Of DNA Within A Living Cell

We're this close to having a usable biocomputer.
By Shaunacy Ferro Posted April 1, 2013
www.popsci.com...

Why Living Cells Are The Future Of Data Processing

Biocomputers make maps, run logic gates, perform binary calculations and more.
www.popsci.com...


Using Magnetic Bacteria to Construct the Biocomputer of the Future
By Clay Dillow Posted May 7, 2012
www.popsci.com...


An Israeli Biocomputer Can Now Detect Multiple Signs of Disease from Inside the Body
By Clay Dillow Posted July 6, 2011
www.popsci.com...


Like Proteus IV Curing Leukemia

New Genetic Circuit Detects Cancerous Cells and Forces Them To Commit Suicide
By Rebecca Boyle Posted September 2, 2011
www.popsci.com...

www.popsci.com...

So the Moral of the Story is ....


If you want Real Artificial Intelligence..for Computers And to be almost as a Sentient being ( Self Aware )

BIO Computing is the Way to go !


Although it May ask for it to be out of its BOX!

A long time I have been concerned about this type of technology and the potential for it to be perverted by governments, especially ours in the US. Most times when we are finally talking about it coming in the future, our government has been using it for a few years already. I had to look this up but will quote it here...

"Skynet gained self-awareness after it had spread into millions of computer servers all across the earth; realizing the extent of its abilities, its creators tried to deactivate it. Because it was software, in cyberspace, Skynet had no system core and could not be shut down. In the interest of self-preservation, Skynet concluded that all of humanity would attempt to destroy it and impede its capability in safeguarding the world. Its operations are almost exclusively performed by servers, drones, satellites, war-machines, cyborgs (usually a Terminator), and other computer systems. As a programming directive, Skynet's manifestation is that of on overarching, global, artificial intelligence hierarchy, which seeks exterminating the human race in order to fulfill the mandates of its original coding."

Original coding ...It loves to win.

With the NSA already in all of our computers with their filthy little spy programs, this sounds like a serious reality that could happen very fast and very easily.

....Game Over

a reply to: Peekingsquatch
hey Peekingsquatch, me too.
For a very long time
and have relentlessly written about it.

But what you say here says it simply enough:

Original coding ...It loves to win.

With the NSA already in all of our computers with their filthy little spy programs, this sounds like a serious reality that could happen very fast and very easily.

....Game Over

Exactly: Game over. It love to win. It has nothing teaching it when not to win, for any reason. And yes, it could invade and penetrate systems quite simply and quickly, and no one would even know……
Game over. And we won't even know…...

a reply to: tetra50
I see technology sometimes as a good thing, don't get me wrong, but there is nothing good at all about this one.

I watch our kids completely engulfed with their faces buried in phones, tablets etc. and sometimes wonder if we shot ourselves in the foot by trying to assist in making life easier for them becoming more and more introverted everyday to the point they text each other from across a table now, instead of talking.

None of them, mine included -have a clue how to fix anything anymore, much less ambition because of the modern conveniences now, and most would see this as another good life convenience thing before realizing it was actively seeking to destroy us all.

Maybe that sounds harsh, but knowing what mine are like they think I'm crazy for having storm food set aside!

a reply to: eatbliss

You have an interesting point of view, and I think we all will never really know until/if/when it comes to happen.

The whole overwhelming thought though is scary.

Can I ask you some movie questions though?

Do you remember the first Star Trek movie? ...V'Ger

it was a satelite that crashed on a machine planet, and was rebuilt to fulfill it's original programming..
"To learn all there is to learn, and return that knowledge to it's creator.." on it's return journey to earth it amassed so much knowledge it became self aware.

Or for that matter how about Data the android striving to become more human?
(I read the intelligence requiring consciousness, and was thinking could intelligence create consciousness too??)

The machine is not conscious, it does not "like" to win any more than a brute force password cracker "likes" to crack passwords. And by the way, this is no different to password crackers that have been around forever. It keeps trying until it finds something that works.

It's not intelligent, and it isn't dangerous. Not at this stage, anyway. All it can do is play games, it cannot evolve beyond that as it is not programmed to.