Is Artificial Intelligence a threat to humanity ?

[EagerBeaver]

Computers do as they're told to do. Period.

And what about when they are told to do evil and malevolent things? Period. Or when their programming develops an unexpected virus? You live in a sanitized cocoon in Canada, where the programs you develop do inocuous and legal things. Not so in all of the world where there are hearts of darkness, mad scientists, power hungry or even insane leaders (Kim), etc. I buy Crichton's vision of reality more than yours. If ever there was someone who needs to watch Westworld movie and TV show, it's you.

Computers can be programmed to do malevolent things. To view things antiseptically and euphemistically like Curly may not be realistic in light of the bad actors we have in this world. Do you guys really think that malevolent computer programs are not being developed by North Korea, or assorted terrorist groups? They can and do recruit computer programmers to do evil.

 

[jalimon]

Exact EB! And with improvement in quantic cimputing and nanotechnology, it will become easier and easier to do much harm.

Cheers,

 

[Sol Tee Nutz]

I thinks some of you do not understand the thread or what Curly said. It is about AI thinking for themselves, Curly just stated they are told what to do and do not think on their own. Of course a computers can be programed to do anything, good or bad.
Never thought that comment would come up.

 

[ssj3]

So how long before Skynet comes online and decides to nuke us all into oblivion?

 

[Sol Tee Nutz]

I am waiting for the the Borgs run by the collective to wipe us out.

Best I can come up with until my cereal arrives on Tuesday.

 

[jalimon]

You guys remember Air Transat plane that drifted with no engine to miraculously landed in the acore? The first action of the pilot when problems where detected was to reboot the computer system, saying it must be a computer glitch (it was not).

What will happen in the next 25 years or so is the opposite. We will become so dependent and computer so powerful that human being will never put them in doubt.

Cheers,

 

[heliragedeleted]

I have been dreaming of the transformers because of this thread, not because we will soon have sentient metal beings, it's because I want them to bring back Megan Fox! She is so sexy, beautiful I just keep on cumming when I think of her.

 

[ssj3]

Megan Fox used to be so much hotter before all her plastic surgery. Check out her FHM pictorials from about 10 years ago.

 

[heliragedeleted]

Megan Fox used to be so much hotter before all her plastic surgery. Check out her FHM pictorials from about 10 years ago.

Ohhh this is cruel ssj you just smashed my dreams.

 

[heliragedeleted]

We will become so dependent and computer so powerful that human being will never put them in doubt.

Cheers,

I think it is already becoming a reality. The 25y wait is over!

 

[curly]

Unfortunately about 15 years behind reality.

You've been programming or working on software development lately? I have. I don't live in a scifi book, I work in an industry.

Eager Beaver, computers dis as they're told to do. And yes if they are told to do evil they will do it. That's precisely because they lack intelligence, ethics, feelings and judgements. They are a machine.

 

[ssj3]

Ohhh this is cruel ssj you just smashed my dreams.

Haha yeah man, it's a shame she decided to mess with an already gorgeous face and smokin hot body

 

[cloudsurf]

You've been programming or working on software development lately? I have. I don't live in a scifi book, I work in an industry.

Eager Beaver, computers dis as they're told to do. And yes if they are told to do evil they will do it. That's precisely because they lack intelligence, ethics, feelings and judgements. They are a machine.

Very true to-day....but how about 40 years from now....that`s how long it took to go from mechanical adding machine to smart phone.

 

[heliragedeleted]

I believe the fundamental misconception here is that we believe that we have free will!? and machines do not. Do we? I think not, As Brian Greene says free will is an illusion. I read this not so long ago:

At any one moment in time our brain, however amazingly large and complex it is, represents a gigantic probabilistic computer that turns a set of inputs into outputs. The probabilistic part means that even if the same brain were to be presented with the same inputs it might produce a different observed output (given the complexity involved this is a complete hypothetical, as the “same” brain implies each of nearly 100 billion neurons in the same state). What is true though is that the probability distribution over outputs is entirely encoded in that brain. And hence there is no free will.

And we are largely "programmed" by our genetics. This is becoming clearer as we start understand the biogenetics behind the inner working of the brain. Our genetics are similar to the "program" fed to our computers.

I think our machines as they "evole" under our unwilling guidance will be more and more complex. We will eventually loose the capacity to distinguish it from human intelligence. Then the fun will begin.

 

[EagerBeaver]

I think our machines as they "evole" under our unwilling guidance will be more and more complex. We will eventually loose the capacity to distinguish it from human intelligence. Then the fun will begin.

Very true. The fun began for Michael Crichton in the 1970s when he imagined this very thing and wrote the book Westworld.

 

[curly]

As Brian Greene says free will is an illusion. I read this not so long ago:

At any one moment in time our brain, however amazingly large and complex it is, represents a gigantic probabilistic computer that turns a set of inputs into outputs. The probabilistic part means that even if the same brain were to be presented with the same inputs it might produce a different observed output (given the complexity involved this is a complete hypothetical, as the “same” brain implies each of nearly 100 billion neurons in the same state). What is true though is that the probability distribution over outputs is entirely encoded in that brain. And hence there is no free will.

I think our machines as they "evole" under our unwilling guidance will be more and more complex. We will eventually loose the capacity to distinguish it from human intelligence. Then the fun will begin.

That takes for granted that we understand how human intelligence works. Mr. Greene has an interpretation, but we are very far from this in our knowledge. So it is very difficult to emulate something we don't yet have a grasp on.

By the way, I came across this article that explains what truly happened at Facebook.

 

[cloudsurf]

Elon Musk, who has tons more knowledge of AI and computer technology than any of us on Merb, has again raised a red flag warning.
He said on Friday that the threat of AI is far worse than North Korea.
Is he crying wolf or does he know something

 

[lady_lover]

Elon Musk, who has tons more knowledge of AI and computer technology than any of us on Merb, has again raised a red flag warning.
He said on Friday that the threat of AI is far worse than North Korea.
Is he crying wolf or does he know something

He is not crying Wolf. Stephen Hawking believes the same although has not made reference to North Korea.
Unlike most IT professionals, he probably is involved with many research projects that go beyond your usual PayPal coding. For myself, I've worked with compute clusters with hundreds of CPU's, a few Cray machines and been involved with Blue Gene L. In a matter of 10 years I've seen computer hardware grow from a few racks with hundreds of processors to thousands now sitting under a desk. We certainly have the compute power available now. I've worked with software that is amazing and is on the brink of replacing a certain population of scientists. We have a lot of the bits and pieces of AI. Just nothing yet that can bring it all together. That we know of.

 

[Willgill]

This is my own field, and I can provide some info.

it has barely anything to do with McGill. Instead, one of the big names in machine learning (more precisely deep learning, google the term if you don't know what it is) is with UMontreal. That's why Montreal attracts interest from IT giants.

edit: saying that it barely has anything to do with McGill is not precise. Joelle (the lady mentioned in the article) and a couple of other professors are pretty good. But not to the level of Yoshua Bengio of UM

re-edit: maybe a good chance to move to MTL? Tempting

 

[Tim Horny]

Laws have to be put in place to protect the dignity of man before machine. AI should benefit mankind, no for the use of some to enslave others.