Technological Singularity

[mysteryliner]

Perhaps that depends on whether or not what we create destroys us but continues on (such as machines).

 

ohoh. http://terminator.wikia.com/wiki/2011/04/19 goes online april 19th 2011. nuclear apocalypse two day later.

 

 

so. people, what are your thought on Technological singularity???

Could it lead to judgement day???

[mysteryliner]

(Split)

 

lol, was wondering what happened to my last post.

 

Fears were skynet erased it.

[Lord Sinister]

As I believe I posted elsewhere, I find the concept of a truly sentient AI fascinating....then again I've probably watched far too much science fiction or things from the Terminator franchise.

 

I am assuming of course that by "technological singularity" you are referring to sentient machines or sentient AI in some form or another.

[J.C.]

From Wikipedia:

 

A technological singularity is a hypothetical event occurring when technological progress becomes so rapid that it makes the future after the singularity qualitatively different and harder to predict. Many of the most recognized writers on the singularity, such as Vernor Vinge and Ray Kurzweil, define the concept in terms of the technological creation of superintelligence, and allege that a post-singularity world would be unpredictable to humans due to an inability of human beings to imagine the intentions or capabilities of superintelligent entities.

 

There's a really good book by Charles Stross called Singularity Sky that deals with this subject, you should check it out.

 

If you ask me, I'm somewhat skeptical on the whole AI subject. I've implemented fuzzy logic controllers, trained neural networks and played a bit with genetic algorithms. Nothing in these methods really approaches intelligence and creative problem solving. They are all either brute force computations, trial and error tunning or fixed algorithms and probability curves.

[Doom Shepherd]

I've implemented fuzzy logic controllers, trained neural networks and played a bit with genetic algorithms.

 

So are you the person to ask about how long it's gonna be before I can stick a chip in my head and suddenly understand Latin, or have a graduate degree in paleontlogy?

[BTGBullseye]

So are you the person to ask about how long it's gonna be before I can stick a chip in my head and suddenly understand Latin, or have a graduate degree in paleontlogy?

Heck, just having a calculator would help...

[Alyxx Thorne]

I find a cyberpunk world more and more likely to occur where we merge with machines, rather than machines taking over.

[mysteryliner]

Nowadays because pretty much everyting is hooked to the Internet, if it would ever become self aware. Whe probable won't be able to contain it. (as In T3 scenario)

If that would happen it would probable evolve very quick (going from a single processing unit to a vast network of processing units)

 

Different unit have different ways of exploring and finding a solution (like weather computers all interpret the same data in their own way) they would interact with each other and quickly evolve to thing that we are unable to design right now.

 

At that moment things could turn in all directions.

Scenarios like skynet in terminator, or like movie I robot,.....

----------

ghosts in the machine. Random segments of code, that have grouped together to form unexpected protocols.Unanticipated, these free radicals engender questions of free will, creativity, and even the nature of what we might call the soul.

Random segments of code? Or is it something more? When does a perceptual schematic become consciousness?

[J.C.]

So are you the person to ask about how long it's gonna be before I can stick a chip in my head and suddenly understand Latin, or have a graduate degree in paleontlogy?

Probably not for a long, long time. The brain works using a "distributed computing" approach, were neurons all over the place light up for each action. It is true certain areas have been identified as being closely related with a lot of specific functions (humor, sadness, short term memory, long term memory, different motor functions and senses) using techniques such as fMRI, but this is all very broad approach. Identifying a single memory, accessing or implanting a specific piece of knowledge, could require electro chemical manipulation as granular as single neuron level. Other than a cloud of nano machines monitoring and interacting at the cellular level (and the very existence of those, working in a coordinate fashion to achieve this functionality, is currently a wild dream in itself), I don't see how it could be done.

[Bjossi]

I imagine a distant future where there are only machines around and no humans to supervise them, their AI has no long term goal, only to be productive; build, learn. Maybe a few humans are stored in cryotubes in an attempt to prevent the race's complete extinction. They are defrosted one day and they will look upon what machines can do for thousands or even millions of years; megastructures all around that dwarf the HL2 citadel. Endless, inconsistent series of rooms and hallways constructed with all sorts of materials, contained inside equally inconsistently built buildings.

[mysteryliner]

megastructures all around that dwarf the HL2 citadel. Endless, inconsistent series of rooms and hallways constructed with all sorts of materials, contained inside equally inconsistently built buildings.

 

Ah, but my guess is that because of the way that machines are designed to communicated with each other there will not be a reason for much rooms and hallways.

 

Why would they use such an inefficient way to transport themselves?

Why would they transport themselves? As with networking, one unit would have all the knowledge of the next.

[Bjossi]

The machines were originally programmed by humans, they build for them, not for themselves. I don't imagine a super-smart AI that sees there are no humans around, just an AI caught in an infinite loop of some sort.

[mysteryliner]

@Bjossi: correct but we arre creating smart machines now to observe our behavior and if necessary correct our actions or warn us or stop us from making any more actions.

 

They have the skills to learn. So at some point they should learn to do tasks they were not originally created for. They are connected to the Internet.

 

So if at one point they make a conclusion that it is not the human about to make another mistake and needing to correct it but Conclude that the humans is illogical and he shouldn't be performing any actions. The a.I. then evolved from simple task solving to complex problem solving things, or thinking.

They upload their behavior and others learn those skills.

[BTGBullseye]

@Bjossi: correct but we arre creating smart machines now to observe our behavior and if necessary correct our actions or warn us or stop us from making any more actions.

 

They have the skills to learn. So at some point they should learn to do tasks they were not originally created for. They are connected to the Internet.

 

So if at one point they make a conclusion that it is not the human about to make another mistake and needing to correct it but Conclude that the humans is illogical and he shouldn't be performing any actions. The a.I. then evolved from simple task solving to complex problem solving things, or thinking.

They upload their behavior and others learn those skills.

But you're missing the result...

 

Result: The computers ignore external input. Humans shut down the computers. Humans replace the hard drives. Humans patch all the AIs. Humans reinstall everything. Life continues as it did before the AI takeover.

[mysteryliner]

Result: The computers ignore external input. Humans shut down the computers. Humans replace the hard drives. Humans patch all the AIs. Humans reinstall everything. Life continues as it did before the AI takeover.

 

But if an A.I. becomes smart enough/ self aware it would spread across the cloud, thus being everywhere and the problem would not be solved by just unplugging or destroying the main computer or hard drive (like skynet in Terminator)

 

not a perfect example but; in a few years cars track the driver, and this A.I. is being updates via internet. If microexpressions change on the driver change: Car stops

If you appear drunk: it wont start.... This system gets smarter by learning from the driver and updates to the main system. Others cars do the same.

If after years of learning this A.I. becomes so smart that it sees that the best course is to simply not let humans drive anymore. By the time that the main system has updated all the cars and the human controllers notice a problem it will be to late.

[Lord Sinister]

This is why in the event we do end up with powerful AIs managing things, safeguards should be implemented. Multiple safeguards.

[BTGBullseye]

But if an A.I. becomes smart enough/ self aware it would spread across the cloud, thus being everywhere and the problem would not be solved by just unplugging or destroying the main computer or hard drive (like skynet in Terminator)

 

not a perfect example but; in a few years cars track the driver, and this A.I. is being updates via internet. If microexpressions change on the driver change: Car stops

If you appear drunk: it wont start.... This system gets smarter by learning from the driver and updates to the main system. Others cars do the same.

If after years of learning this A.I. becomes so smart that it sees that the best course is to simply not let humans drive anymore. By the time that the main system has updated all the cars and the human controllers notice a problem it will be to late.

You missed the point entirely. I wasn't saying just one computer, but all of them... (the internet is just a bunch of computers networked)

[mysteryliner]

@BTGbullseye: yes, that's what would make it so hard to contain. you can unplug one computer. but when it is spread across the internet it could/ would be everywhere.

If you would destroy PC1 with A.I. it would already know it needs to counter your at PC2. or

[BTGBullseye]

But how would it know that it's because of a human that power is being turned off? There are quite a few possibilities for what can cause power outages on a large scale.

[mysteryliner]

That would be one of the only options, worldwide power-outage.

But is that even be possible now days?

U.S. calling Russia: could you guys power down entire Russia, there is an a.i. spreading across the world & taking over...

Sure thing comrads, right away!

 

And what to do with PC's farms directly powered by solar tech, Military installations with years of backup power, satellites orbiting earth....