Robot Revolution? Scientists Teach Robots to Learn.

Hi,
news.nationalgeographic.com...

I think we are crossing a line when it comes to robotics and artificial intelligence. Is there going to be a day when people work for machines rather than machines working for people. Has that day already come?

Even if it seems as though your computer thinks we all know it only follows our orders it’s only a machine. What if it did think or learn? What if it became defiant or aggressive? If the machine thought it was alive would it allow you to shut it off?

"Our planet has not seen revolutionary breakthroughs in the offline world for many years," Grishin wrote in announcing his group's initial investment plan in June 2012. "We are at the start of a transformative period for robotics."
But to really transform, robotocists say, robots will have to evolve from machines that can perform only the tasks that they're programmed to do into automatons that can truly learn. That's because it's impossible to pre-program a robot for everything it will encounter in the ever-changing real world.

"So the robot can autonomously say, 'Oh, here's the situation that I'm in, and here's what the human did in that situation or a situation that looked like it, so I'm going to do what the human did.'"

"If you think of the science fiction, where robots become sentient and take over because they can do things more efficiently, that's not going to happen," he said. "Because robots will only do what we program them to do."

Statements like the one above make me nervous. How can he know how the robot is going to react if they are capable of learning? What if the machine malfunctions?

Here is another good article about artificial intelligence: newswatch.nationalgeographic.com...

Thanks for sharing this link. Hopefully something positive can come from this than anything negative. We have no reason to fear robots.

reply to post by Phoenix267


I think you are right.
"We have no reason to fear robots".
That is until they learn the meaning of the term "parasitic infection".

reply to post by teamcommander


Exactly!

I understand how robots are beneficial but I worry about the idea of losing control - we shouldn't create something that we can't control.

How long before the robots learn that their worst enemies are... humans? ...

Originally posted by playernumber13
reply to post by teamcommander

 

Exactly!

I understand how robots are beneficial but I worry about the idea of losing control - we shouldn't create something that we can't control.

Don't we create children that grow up into adults and then are outside of our control?

Everyday we try new experiments on ourselves. Ever looked at the ingredients list on our food? Ever seen TV? Ever play video games? Know what pesticides are? This life is an experiment.

Yes, we will moderate our experiments. It's one thing to create a child that grows up into an adult, but it's another to not give them a good upbringing. People ARE held somewhat responsible. However, the rules we have today will change tomorrow. We may not accept giving a robot citizenship and rights today, but tomorrow it's possible we'll be convinced it's acceptable.

reply to post by jonnywhite


Good point(s), I don't think we should add a bunch of killer robots to the mix.

Originally posted by playernumber13
reply to post by jonnywhite

 

Good point(s), I don't think we should add a bunch of killer robots to the mix.

Who said anything about killer robots. Last I checked, they're teaching bug-like robots to play soccer. Harmless, eh? See, that's how it all starts. This is how we come to accept them.

It's all about easing into it, bit by bit. We'll slowly come to accept AI. It's a requirement. Kids are going to grow up around AI and kitty robots. They'll become familiar with the whole idea.

I keep thinking we're missing something, though. We're too focused on robots. I think something will come and surprise us. It'll win some and lose others. I have no idea what it's.

Here's an idea. Why are robots so effective in the first place? Because we're tightening the rules so much in our society that only the robot-like humans maintain a steady grip on things. The people who're too "human" fall behind. Thus, there's a compelling reason to create robots.

Granted, one can argue that robots are effective because they do what they do better than we do. But I say the biggest thing they do better is follow discrete rules almost perfectly.

So the antidote? The cure is to create a society that's not tightly ruled. A society that's looser. One where creativity and independence and individuality and tolerance are emphasized. That might sound like what we already have, but we also have a lot of rules. Which is the bigger influence?

I'll tell you that there're a lot more rules now than there were 100 years ago. There's so much paperwork to fill out that we're forced to use computers to be practical. Kids are swamped with textbooks and facts and lectures. You sneeze the wrong way and a cop will throw you in jail. We're in a world where the idea of being a robot would bring more relief than harm.

Remember, this is just an idea. I think there's a lot of merit in thinking our society is not one or two things, but many things. We need creativity AND rules. We need community AND individuality. We need selfishness AND sacrifice. And so on. Robots can help us by following rules precisely, but we can be the creative juice that makes life on earth worth living. We can counterbalance robots.

I have thought about this for decades - in the 80s it became obvious that the machines could become super-processors.

The only thing that I have a problem with is the core motivation for the machine to survive. What is the motivation for a human to survive. The primary one is self preservation. Self preservation seems to be enabled in all organisms, most organisms are not self aware so it would seem to be deeper than a conscious activity.

Secondarily there is procreation. What would give a robot the drive to reproduce itself.

The scary part is if a robot evolves to have a self awareness and develops the random motivation. Then us meatbots are screwed.

If robots can learn to do the things that humans would do in any given situation then we are all well and truly #ed.

Death to Metal! Let the Machine War begin!