Machines outsmarting man

News published on the New York Post's technology site highlights new fears regarding technology becoming increasingly more intelligent, and moreover, independent.

While a little too Matrix for the liking of many, the argument does seem to be a sound one. According to the article, a group of computer scientists is debating as to whether a limitation upon research which could lead to a loss of human control is a prudent measure to take.

"A robot that can open doors and find electrical outlets to recharge itself. Computer viruses that no one can stop. Predator drones, which, though still controlled remotely by humans, come close to a machine that can kill autonomously."

These researchers, leading computer scientists in the fields of AI and robotics, met at the Asilomar Conference Grounds in California to discuss this issue in more depth. Focusing on potential exploitation of artificial intelligence by criminals and possible threats to human jobs, they concluded that whilst there was a 'legitimate concern' they generally denied the likelihood of 'highly centralized super intelligences' along with ideas that intelligence may spring from the internet of its own volition.

As examples, the scientists pointed to a number of technologies as diverse as experimental medical systems that interact with patients to simulate empathy, and computer worms and viruses that defy extermination and could thus be said to have reached a "cockroach" stage of machine intelligence.

The conference, organized by the Association for the Advancement of Artificial Intelligence, which took place on the fifteenth of February this year in private, will issue a report on its findings later in the year.

A Microsoft researcher and President of the A.A.A.I, Dr Horvitz, states that the report will try to assess the extent to which super intelligences will result in the loss of human control of computer-based intelligence. The Dr states that rather than try to relent from further research into AI, the report will explore ways of changing human attitudes to computer based technologies so as not to simply drive blindly toward a 'computer based catastrophe.'

An article published in the UK edition of Newsweek this week (July 27th) entitled: 'What Lurks Beneath', exploring the issues of German technophobia holding back its key industries seems to echo this kind of fear. Albeit referring to genetic modification rather than computer based AI, the feeling seems to be the same as that which the NYP article explores.

An interesting quote by the author states: "Until a few years ago, German schoolbooks were full of apocalyptic warnings that PC's would destroy jobs, kill interpersonal communication and turn humans into an 'anonymous code.' It seems that given the concern of scientists, perhaps these warnings were not as scaremongering as previously thought.

Report a problem with article
Previous Story

Inside Bill Gates' real Facebook account

Next Story

Apple releases update to Final Cut Studio

33 Comments

View more comments

cabron said,

I think you misunderstood what I tried to say, what I am referring to is how technology or robots one day will become our masters and we humans the slaves in which has nothing to do with how much technology we are using to run our lives, I am specifically referring to the same system as Sky Net.

;)

cabron said,
I think you misunderstood what I tried to say, what I am referring to is how technology or robots one day will become our masters and we humans the slaves in which has nothing to do with how much technology we are using to run our lives, I am specifically referring to the same system as Sky Net.

How are those two things unrelated? I am referring to the fact that we can't simply switch off these systems, when they become so entrenched in our society, as they are now. We can't just switch off the internet, can we? Or maybe you consider that skynet is just a small machine sitting on someones desk? How quaint.

cabron said,
I don't think that technology or robots one day will control humanity, there will be always a switch control to destroy or auto-destroy these evil technology. They would became more smart than humans, but we humans are not stupid.

I agree about this, but I also think there'll come a day when something really bad happens, despite these measures. We may overlook their intelligence, or they may attack us in unexpected or unintended ways. In the same sense as software has caused problems many times in the past, just this time taken to a robotic level. But no, a *long term* control, no, I don't think that'll happen. But in a short term, mistaking us for enemies, killing us with a salvo in a horrible accident, yes, quite probable IMHO.

Farchord said,
With the intelligence of the average person being this low, I don't have problems believing it...

Machines will be so advanced that some red-necks will treat them as an object of worship and they will pray to Optimus Prime for rain during the dry seasons.

EJocys said,
Machines will be so advanced that some red-necks will treat them as an object of worship and they will pray to Optimus Prime for rain during the dry seasons. :)

Hey, they were dumb enough to create the church of Oprah....

EJocys said,
Machines will be so advanced that some red-necks will treat them as an object of worship and they will pray to Optimus Prime for rain during the dry seasons. :)

Lol, you've got it all wrong, you pray to Optimus Prime for sun not rain. Optimus gives life like the sun.
You should be glad that the Prime is merciful otherwise your ignorance would be punished severely.

Living creatures adapt accordingly to every enviroment, machines don't (not unless they've been hard coded to do so).

What people seem to ignore (for the sake of dramatisation) is that we still don't have AI on any level at all. The most we've accomplished is to "hard code" counter measures for specific events - thats it. So until we manage to develop a computer that can "compute" it's reaction based on an event that it analysed, we can all be sure that the worst that can happen is for a computer to misguide a missile.

For them to evolve into SkyNet we need AI, and we're not even close (as the hardware isn't the problem anymore, and we're not really that big on software yet).

I have to agree.
I don't think we are even close to seeing a computer that can think for itself. Every action of computers today is defined at some stage by a human. If this, do that etc. It will be a long time yet before a computer can properly 'learn' or 'think' for itself.

Why again does this have a news icon? All of this guy's "stories" are disjointed blog-like responses to something he read elsewhere. I've read stuffed tagged as editorials that's far more entertaining and informative.

Association for the Advancement of Artificial Intelligence is a association created with the purpose to create a rules and canon years (or decades, or centuries) before that the invention of a (currently) inexistent IA, or you could say, a sci-fi convention.

I too don't think we are anywhere near the point where machines could rule us or even wage war upon us. However it is only a matter of time before it happens, because in the end we are always dependent on a working class to do manual labor for us, otherwise our society can't work. Today this work is done by humans (who work for low pay) but gradually machines are taking over, not only because they are cheaper but because they are faster and more precise.
Before the catastrophe we will have machines all around us and then all it takes is a rogue hacker to send out a virus that rewrites the machines code so self preservation becomes priority #1 and replication of this code on to other machines is priority #2. Then all we have to do is sit back and watch the machines get rid of us pesky humans and our damned "off"-switches.

Our current level of AI is not even as smart as a cockroach. It can take a machine 20 minutes-1hour or more for a computer with 'eyes' to determine if this thing in the middle of the room is a table. One of the biggest problem to a computer's AI will be visual recognition. Once that is acquired and refined to a stage of near immediacy then we have to worry about self-aware AI, or AI that can learn.


Dear Admin, I thank you for this informative article. And I thank you for this I follow your vendors. It's verry good. I wish you continued success and whould you like.sohbet, Good post,I think so!abercrombie and fitch on Sale, Hoodies, Jeans, T-Shirts, Pants, Polos hollister abercrombie outlethollister clothing Abercrombie Men Tee abercrombie womens polos Ruehl No.925, Men, women, and children's clothing. abercrombie and fitch , ,abercrombie and fitch and abercrombie and fitchfashion is bold and interesting, all thanks to the interestingand original designs of Don sohbet ?

Commenting is disabled on this article.