The future of A.I.... anyone uneasy by this?

[Arcadio]

This is a great read. I didn't know there was a term that articulated A.I. domination:

http://en.wikipedia.org/wiki/Technological_singularity

Read Ray Kurzweil's books. A little outdated, but very thought-provoking.

 

[Jeff7]

I think it's pretty much ultimately domination.

Depends on how we raise the AI, and what motivations it has.
A lot of our motivations are generated as a result of our evolutionary history, and our short lifespans.
If an AI can live for a very long time, and is effectively unconcerned about death, its goals in life, if it has any at all, may be quite different than ours. With any luck, it'll be far less prone to violence and domination.

It might be perfectly content to ponder atomic reactions for a few centuries, or take care of humanity as pets.

 

[Maximilian]

AIs may make a better more adaptable life form than humans. It might be an unconventional method of evolution for a species to create its own successor by physically building it but its still evolution all the same IMO. Then again AI may not be viable in the long run, it needs electricity and a lot of non naturally occurring stuff to function.

What was that movie where an AI torments a group of people for eternity? If I remember it keeps healing them and hurting them over and over. I think it even kills a few of them to see how others react.

 

[MongGrel]

I Have No Mouth, And I Must Scream, was the Harlan Ellison story.

Not sure it was a movie, but I think were things similar, a video game at any rate.

http://en.wikipedia.org/wiki/I_Have_No_Mouth,_and_I_Must_Scream

Cube was a bit the same, that is probably it.

http://www.imdb.com/title/tt0123755/?ref_=nv_sr_2

 

[brianmanahan]

Then again AI may not be viable in the long run, it needs electricity and a lot of non naturally occurring stuff to function.

multivac figured out a way to get around that

http://www.multivax.com/last_question.html

 

[MongGrel]

This oldie was always pertinent to the AI thing.

Colossus: The Forbin Project

http://www.imdb.com/title/tt0064177/?ref_=nv_sr_3

 

[MongGrel]

Think about first obvious uses of computer intelligence, make money, seize power, by providing analysis of data supplied by humans who use the results. That will happen long before computers act intelligent, let alone self aware or dangerous.

Two big factors make me fairly unconcerned.

Physical isolation, just unplug the menace.
Limited information, sure tons of information are out on the net, but a ton of it is bad, and much of what an intelligent computer would need to know to make itself dangerous isn't there.

Kind of what they were trying to do here, but in an Altruistic way, which kind of fell flat a bit to most people I guess.

Transcendence

http://www.imdb.com/title/tt2209764/?ref_=nv_sr_3

 

[Brainonska511]

It might be perfectly content to ponder atomic reactions for a few centuries, or take care of humanity as pets.

Pets, like an electricity generating tamogochi...

 

[MongGrel]

Pets, like an electricity generating tamogochi...

Too messy, but of course first Matrix got that one right.

 

[Maximilian]

I Have No Mouth, And I Must Scream, was the Harlan Ellison story.

Not sure it was a movie, but I think were things similar, a video game at any rate.

http://en.wikipedia.org/wiki/I_Have_No_Mouth,_and_I_Must_Scream

Cube was a bit the same, that is probably it.

http://www.imdb.com/title/tt0123755/?ref_=nv_sr_2

I Have No Mouth, And I Must Scream! Yup that was it. Dunno why I thought it was a movie. Its one of these things ive heard of and read the premise of and some of the story but never gotten around to actually reading. Cube was good :thumbsup:

 

[Red Squirrel]

This is why I try my best to avoid cloud based services, it at least makes some stuff harder to track. Unfortunately there is lot of tracking via phones and just the internet, and that is much harder to defeat.

I recall reading something about surveillance drones in the US, they can do face recognition from like 60,000 feet, through clouds. It's pretty crazy. Technology is great, but when it's used as a weapon, not so much.

Though the NSA and other enemies having too much information ends up working against them too. Even the NSA themselves admited that they have so much info that they can't really sift through all of it effectively. Of course you have to take what they say with a grain of salt though. It's like ISIS saying they're starting to run out of funds to buy more vehicles and weapons. It only makes people get their guard down thinking "we're winning".

I think there will come a time where government tech will have to be faught back with our own tech. Easier said than done though, their tech is 100's of years ahead of what consumers even have access to.

 

[Greenman]

While I consider a true AI to be just about as likely as a Sasquatch running for president, if such a thing were created and ran amok, you just unplug it. I'll even volunteer, shoot me a pm when one try's to take over the world and I'll save us all with my legendary right handed yank.

 

[Jeff7]

Pets, like an electricity generating tamogochi...

A really inefficient one though.

Fission, solar, wind, coal...take your pick. Matrix-style power generation falls into the rarely-researched "plot device" class.

While I consider a true AI to be just about as likely as a Sasquatch running for president, if such a thing were created and ran amok, you just unplug it. I'll even volunteer, shoot me a pm when one try's to take over the world and I'll save us all with my legendary right handed yank.

Kind of like how you would just unplug the Internet.

:hmm:

 

[BudAshes]

I'm pretty sure China is already run by an AI.

 

[Red Squirrel]

A really inefficient one though.

Fission, solar, wind, coal...take your pick. Matrix-style power generation falls into the rarely-researched "plot device" class.



Kind of like how you would just unplug the Internet.

:hmm:

:biggrin:

If some kind of AI system had taken over, whether it's actual machines or something that is just really good at spying and categorizing info, the government would probably be behind it. It would be extremely illegal to even attempt to disable it.

 

[Aikouka]

Eh, I think people that rant over AI are nothing more than fear mongers. Artificial Intelligence to the degree that they're talking about is still an unknown... an element that we see in Science-Fiction. Here's the way I look at it... an unknown entity is far easier to mold to a reader/viewer, and we've been shown far more evil unknowns that benevolent. Is it really any wonder that people jokingly refer to an AI as "SkyNet", which is a malevolent AI from a Sci-Fi series?

Also, how was the movie?

As a developer who enjoys concepts behind artificial intelligence and robotics, I found the movie boring. Sure, there are interesting aspects of it, but a lot of it is very humdrum. You could probably fast forward through almost every scene between Nathan and That-Guy-From-About-Time and miss nothing.

Also, this bit will give away the ending to any astute, but

Spoiler

Asimov would not approve of those robots!

Living in close proximity to the Amish and having recently paid attention to some stories about blue zones - areas where people have exceptionally long (and healthy) life spans, and what their lives are like, I'm growing more and more convinced that modern "conveniences" are not a step in the right direction. I don't think that the Amish and being stuck at an 1890 level of technology is quite right, but I think they're closer to what's best for humans than what the future may hold for us.

I would say that their longer life spans are more related to better habits. We're so pushed with a rapid, get-up-and-go lifestyle that we tend to consume a lot of preprocessed garbage and don't normally get a lot of exercise. The Amish tend to perform a lot more manual labor (i.e. essentially a workout) and eat the food that they plant.

 

[Greenman]

A really inefficient one though.

Fission, solar, wind, coal...take your pick. Matrix-style power generation falls into the rarely-researched "plot device" class.



Kind of like how you would just unplug the Internet.

:hmm:

It's actually interesting that a site with so many atheists has no problem with the concept of a man made God, one that can't be destroyed because it inhabits the magical and all pervasive "internet".
I suppose the need for a deity runs deep, even for those that deny it the most.

 

[Jeff7]

It's actually interesting that a site with so many atheists has no problem with the concept of a man made God, one that can't be destroyed because it inhabits the magical and all pervasive "internet".
I suppose the need for a deity runs deep, even for those that deny it the most.

A god is, often by definition, supernatural.
There's nothing supernatural about an intelligence that could exist within a widespread and redundant network of computers.

Bacteria are difficult to eradicate too, not because they're gods, but just because there are a lot of them, and they reproduce quickly. Redundancy.

 

[WHAMPOM]

I just finished watching Ex Machina (2015) with the wife tonight.

Even mere a decade ago, we would've just laughed off movies like this as a silly science fantasy. Today, the implications and ethics that come with it are real.

Look at all that trillions of data mining the corporations are already doing on us today. We all know about the recently closed Google's 411 service that had an ulterior purpose to collect all voice samples for their voice recognition technology. Or Siri recording everyone saying everything on every single iPhones out there.

The exponential growth of technological automation and its resulting A.I. both fascinate me and make me uneasy.

Bill Gates joins Elon Musk and Stephen Hawking in saying artificial intelligence is scary.

The Lizard aliens running Washington right now are on this looming problem.
Don't sweat it, a small problem compared to your future life as lizard snacks.

 

[lxskllr]

Eh, I think people that rant over AI are nothing more than fear mongers. Artificial Intelligence to the degree that they're talking about is still an unknown... an element that we see in Science-Fiction.

Personal communicators that fit in your pocket was fiction when I was a kid. All scientific speculation is fiction until it isn't.

It's better to consider problems before they happen, than to react after the fact.

 

[Greenman]

A god is, often by definition, supernatural.
There's nothing supernatural about an intelligence that could exist within a widespread and redundant network of computers.

Bacteria are difficult to eradicate too, not because they're gods, but just because there are a lot of them, and they reproduce quickly. Redundancy.

The comparison to bacteria fails because bacteria are living things, AI has no life.
Our discussion will now fall apart because we can ascribe any and all traits and ability's we want to an AI. It can be an idiot savant or an all powerful, all knowing, hulking mass of will power. The fundamental issue is that AI is imaginary, so imagination becomes the limiting factor in it's ability's.

 

[Jeff7]

The comparison to bacteria fails because bacteria are living things, AI has no life.
Our discussion will now fall apart because we can ascribe any and all traits and ability's we want to an AI. It can be an idiot savant or an all powerful, all knowing, hulking mass of will power. The fundamental issue is that AI is imaginary, so imagination becomes the limiting factor in it's ability's.

So...ok, bacteria fit our arbitrary definition of "life." Ok. That doesn't have much bearing on something's ability to be durable. An AI could exist in a single isolated system, or it could exist in a distributed manner, such as out on cloud-based computers.


Self-aware AIs are currently imaginary, yes. Intelligent AIs do exist in labs though, and are capable of learning new things. How intelligent? Not very. We're limited by the number of transistors we can dedicate to the task.
At one point, a 500 TFLOP computer was also impossible.
Memory as it is for us is just the storage of a dynamic model of the environment, and even of existing memories. Intelligence is the ability to adapt that model on the fly. Self-awareness is the inclusion of one's own existence in that model of the environment. Consciousness is the continuous updating of this entire system of models.

Implementing that on computers is tough, currently. We've only really been working on the problem properly for a few decades. Computers were also originally designed to yield consistent and repeatable results. However, you've got plenty of emergent properties showing up: Computer bugs appear constantly. A complex system behaves in unexpected ways. We might find ourselves at some point with an AI supercomputer in a lab that has, unintentionally, become aware of its own existence.