Scientists worry machines may outsmart man

As gains are made in artificial intelligence, scientists worry that advances could have dangerous consequences.
(From The New York Times)

The story "Scientists worry machines may outsmart man" published July 26, 2009 at 5:02 PM is no longer available on CNET News.

Content from The New York Times expires after 7 days.

[solitare_pax]

Well, it's a long way off - it's not like the Terminator is governor of California or something...

Oh, wait...

[babystars_13]

I tend not to worry as machines cannot procreate.

[Lerianis3]

Actually, yes, they can.....if they get control of an assembly line somewhere in the world, have the parts shipped in, and build their own 'progeny'.

[rdupuy11]

I don't think thats what babystars meant, I took it to mean he/she isn't going to worry about it, until machines are possible partners in procreation.

[techie2479]

All your base r belong to us!

[overheck]

Don't be fooled. Mankind is many many light years away from a need to worry about AI. I bet that no person standing on earth right now will ever see a computer that is even remotely intelligent, within one's life.

[Mr. Dee]

This is stupidity at its best. I can hardly get Windows Vista's speech recognition to work properly although its significantly improved. I think these folks are staying up too late at night and are watching too many Terminator: The Sarah Connor Chronicles.

Yes, Computer Scientist will one day get machines to mimic human traits, but machines will never be able to truly understand or interpret human emotions properly. Why? Because computers by nature require an input to get an output. So even if they were to become highly sophisticated with lots of so called "intelligence", guess what they are gonna need us humans around to make use of them.

[vaibhav92]

Your assertions seems to be flawed. Human conscious also takes input from the surroundings/environment and perform actions as a result. Seeing this AI systems would be no different .

[Lerianis3]

Mr. Dee, you are making a huge assumption that we will never be able to make a Data-like 'emotion chip'. Yes, at some point in the future, we WILL be able to make that and put it into robots, and that is when the three laws of robots come into play, along with some sub-laws.

[solitare_pax]

On the other hand, scientists have perfected Artificial Stupidity, which is available on all computers that run a word processor..

It's called a "Spell Checker".

[Mr. Dee]

Strange you people really believe in this, considering the basic functionality of a computer has been the same since the 1970's, point and click. There has not been any major innovation over the past 35 years where computing is concerned. Yes, the micro-processor is probably the most significant innovation, but integrated circuitry can be dated as far back as the late 50's, contemporary operating systems continue in the same fashion as their ancestors using point and click. Functionality like Speech, Natural Inking and Touch have not really caught the market on fire and these technologies that are in some ways still primitive and are not intelligent enough for to me say wow. And you are saying, we are going to see machines in our life time that are like 'Skynet'? Well, when it happens, tell Arnold Schwarzenegger and Christian Bale I want their autographs, because its surely not gonna happen this century.

[NicoleG015]

Human still who creates them.So without human they are nothing.

[Lerianis3]

Again, you are assuming that the machines won't be able, after being created, to take care of themselves. That is the 'wet dream' of people who are looking into this stuff: self-maintenance machines where they can 'fix themselves' without any human input.

[artistjoh]

AI is already here but is not all that smart yet, but it is easy to envisage human like abilities within 20 years or so, however why does everyone assume this will exclusively mean some kind of android? Experimentation is already being conducted into inserting chips into the brain to improve hearing for the deaf and vision for the blind, and while this is at a primitive early stage it is easy to imagine vast advances in aiding parts of the human brain directly with implanted chips. True AI may well translate to vastly improved thinking abilities for humans and could even be an "evolutionary step" for human kind.

The possibilities are very exciting whether to cure disabilities or to improve data processing by us all.

[Super2online]

Most advances in medicine are a direct result of research and application of computer technology. We haven't even begun to tap the full potential of this ability yet. While artificial inteligence will advance I believe humans will advance right along with it.

Many redundant and progressive fail safes can be employed to ensure that the future of death and destruction imagined by many are completely avoided. Discussion and planning is not only possible but desired, but it should not be driven by fear of inescapable outcomes, but by intelligent and morally sensable and logical process.

We can and will achieve by continuing to move forward in a responsible manner.

[rdupuy11]

I'd like to give you some thoughts from the other side of this issue.

As a machine, I think you humans have nothing to worry about, its us machines that have to worry about you.

You trade in your old vehicle, take for example the recently passed 'cash for clunkers program'....when you buy a shiny new car, the other one is destroyed for scrap, all because is a few years old.

I think machines have a right to live as long as possible, the same rights you grant yourselves, we should have.