Japanese Scientists Unveil Thinking, Learning Robot

[Del-Toro]

Though leaping into that cliché would be the obvious route for this article, I will instead point out the myriad opportunities this technology presents.

Fine, if you won't, I will.

-Oh Japan, we can always count on you to expedite the apocalypse.

-We have the tech to make a robot walk and run, self acting targeting systems (Predator UAVs, ect) and now and OS that can learn. Japan, you have everything you need, so my only question now is: where's the Gundam? Are you waiting for an appropriately whiny kid to pilot it? I know a few who might work, just ask.

-God help us if the robots learn how to love. If that happens, we're one bad robo-relationship away from the machine wars.

 

[themerrygambit]

Why are we making robots to try and do our work for us when there are hundreds of thousands of people out of work around the globe... Didn't we learn from all the various artificially intelligent robot movies out there that this isn't a good idea? seriously.

 

[Use_Imagination_here]

Think that's impressive?

http://www.youtube.com/watch?v=RNKLuXUh3M4&feature=related

That is a robot telling a man he "loves" he is afraid of leaving for another country. The future's knocking on the door.

 

[themerrygambit]

See I'm all about giant robots that humans pilot... that would be the shizzle! but not crazy self aware learning types.

 

[Farther than stars]

For those wondering about the end of the world fears, The Animatrix, the animated prequel to Matrix: Reloaded, laid out a very plausible path for this.
They became smart enough to take over our menial jobs and humanity was freed of it's labors.
Until the robots realized the gap. Or as Agent Smith put it, "Until we started doing your thinking for you". They didn't outright attack. So there was no violence or contempt by either side at first. They just tried to do their own thing. But free of the limitations of flesh, their community output and worth grew much faster, and gained the ire of humans. Which snowballed into bad things.

Worst case scenario; of course. Still plausible; yes. So yah, this is a little creepy and foreboding.

True, but as I pointed out before, the same goes for nuclear fission and genetic engineering. Sure, the possibility's always going to be there, but heck, if we've managed to survive our own malicious nature so far, I'm not worried about that of any other "species".
And of course, there's coming to terms with the finality of things. I know that I have the odds of like 96.1 billion to 1 of dying some day, so what difference will it make if it's a machine which does it or whether I die in a car accident?
Come to thing of it, at least in the former I have hope that the miracle of life will continue to exist, with its roots in Earth and with humanity as part of its ancestors.

 

[Twilight_guy]

Amazing how the thread immediately fills with snide remarks about Skynet and robot uprisings. It's like watching puritans calling "witch" because their milk turned sour. Silly Escapist posters, you're just showing your ignorance. Silly posters.

It's interesting to see where AI currently is and how it advances but this just seems like a large scale demonstration of basic machine learning. Deciding what order to do things in is a basic AI problem and doesn't seem like a huge leap. Maybe they just didn't explain this that well but I think the AI used in Amazon to predict what your want to purchase is more impressive then the robot that decides to drop a bottle to pick up ice. Maybe they just made it look simpler then it was. You don't pull out the old neural net to do simple tasks after-all.

 

[Tragedy's Rebellion]

It would be awesome to have intelligent robots around and I think this fear of retaliation from them is irrational, if they can learn true emotion why would they try to destroy us? We haven't destroyed each other, completely anyway. It would be like glimpsing into a different world and it would be fascinating to talk to them about philosophical ideas and life in general. I've been thinking about this too - what if they learn to how to make other robots? Will they care for them and teach them all they know? I want to ask them so many questions, but patience is a virtue after all.

 

[Silenttalker22]

It's like watching puritans calling "witch" because their milk turned sour.

It's actually not even remotely the same string of logic, seeing as one is following plausible patterns and one isn't. Silly poster.

 

[Farther than stars]

Think that's impressive?

http://www.youtube.com/watch?v=RNKLuXUh3M4&feature=related

That is a robot telling a man he "loves" he is afraid of leaving for another country. The future's knocking on the door.

To be honest I think a robot actually learning something is more impressive. Sure, the mere mimicry of human expression like the Hanson robots do will have them pass the Turing Test, but it doesn't mean the robots are actually feeling for themselves.
I know you're not implying any different, but you're wording makes me feel the need to reiterate that this robot has been told to say that he is afraid by extracting information from humanly-compiled database, not because he's drawing a conclusion from a neurological response.

 

[Farther than stars]

It's like watching puritans calling "witch" because their milk turned sour.

It's actually not even remotely the same string of logic, seeing as one is following plausible patterns and one isn't. Silly poster.

Woah, that's a very rash statement. I believe it's perfectly OK to use a metaphore, since that's what it is; it's using the basic overlap of symbols to indicate similarity. That doesn't mean that each of its branching symbols need to match up. Indeed none of them have to.
Also, since I've condemned this fear as being irrational, I'd like to know what your "plausible patterns" are. And going back to the metaphore, you have to remember that plausible means a whole different thing when you have less/different information at your disposal (i.e. the 14th century).

 

[Jaime_Wolf]

Very, very impressive. And definitely a big step in the right direction (though it's worth noting that this step has been taken before - this is just another iteration of a very long-running line of research in AI).

But bear in mind that these pretty much inevitably have incredible limitations. Typically, they have to start with some certain knowledge base and, as mentioned, they aren't very good at trying things and extrapolating future solutions from failure (either it's gained enough knowledge to do something or it hasn't).

The biggest problem, however, is always going to be giving it instructions. Truly robust AI should be able to learn languages and, until we hit that point, we've still got a long way to go. There are tricks that can give language-like behavior, but no one has yet managed language comprehension even close to the ability of an average human.

Edit: Just noticed that they're using very strongly color-coded objects with extremely distinct geometries. That's also a pretty big limitation.

 

[Pinkamena]

It gathers information from the internet?
Ok, not to worry guys. It'll just discover my little pony and it'll be rendered hamrless.

 

[ZRendZ]

Fuck yeah, my dream of meeting a real life Megaman X is coming closer
brb while i fangasm...

 

[USSR]

snip

Well my bad for having a bit of fun with this.

 

[Twilight.falls]

This is why I'm going to implant my mind into a robotic shell. That way I shall be spared.

Honestly, this bit of news is insanely awesome.

 

[Reaper195]

Why always Japan with the technological leaps?

Because everyone else just wants to war with each other, and have no interest in actually furthering......anything at all.

 

[chronicfc]

I for one welcome our new robot overlords.

Yes, I have always preferred the robots and their giant death camps to our "free will", robots make society much more efficent.

 

[zehydra]

Useful, and amazing, but people ought to remind themselves that it is not self-aware, and that self-awareness is impossible with an Artificial intelligence.

I too believe in the existence of the human soul, but don't you think that it is unscientific to deny the possibility of self-awareness in machines? After all, neither you nor I can provide actual evidence that it isn't just a neurological construct which allows thought.
And what about animals? I argue that quite a lot of them are not actually "self-aware", but are definitely capable of complex tasks and even learning, as the sciences of classical and operant conditioning have taught us.
Come to think of it, if you believe in evolution, then you must also believe that self-awareness came to us during a stage of neurological development, which poses the question whether or not robots could evolve to become self-aware.
Like I said, I too believe in the existence of the human soul, but that makes it all the more important that I should take into account the possibility that synthetic life forms might exist. Because if I don't, I could potentially harm something which has the capability to feel without even realising it.

It is paradoxical. You cannot have AI that is self-aware, it will be REAL intelligence, not Artificial intelligence in that case.