Fudan Intelligent Robot Learns To Fit In

martinsslaves writes "Shanghais' Fudan University have made some considerable progress in their endeavors in learning robots, with their new 'Fudan Intelligent Robot' looking a whole lot more polished than their previous Fudan-1 model. The goal is for the robot to learn new tasks by following voice commands, which the researchers hope could eventually allow the robot to help the elderly or become a 'good household mate' for families. Among other things, the bot can currently can plot out its own map of its surroundings and remember specific locations and, of course, change TV channels at your command (or even serve as a TV itself)."

Comments

Robot Succeeded where slashoters fail?

[psychicsword]

Looks like a robot succeeded where many slashdot readers failed... it fits in.

Lucy Liu bot make me a sandwich everyone else would.

Is it just me?

[doyoulikeworms]

Or is this "robot" not really very, well, good at anything? The description sounds like only a notch above children's toys today.

Could someone explain what's so story-worthy about this bot?

Re:Is it just me?

[psychicsword]

Could someone explain what's so story-worthy about this bot? It is for the jokes that comes out of it.

Re:Is it just me?

[JK_the_Slacker]

I could argue the same think about the neurons and synapses in your head. You'd be crazy to think that they could actually be capable of as advanced an activity as thought.

No, but seriously, the question becomes "Just how high a level of autonomy is required of any seemingly-sentient being before they are deemed capable of thought?" As with many things in computer science, it's all about abstraction; just how far outside the box are you sitting? I as a computer scientist may look at Asimo and say, "Boy that's neat, but it's not really thinking." On the other hand, I wonder just how many kids have watched Asimo walk with wonder in their eyes. "Mommy mommy, look at the robot walking around! He's waving at me!" I know that computers are only capable of doing what you program them to do, but several of my friends will swear that their computer hates them, or that their gaming console has it in for them. Were we to plop an Aibo down in front of a group of people in certain remote villages in certain third-world countries, how much would it take to convince them that it was alive?

Arthur C. Clarke suggested that any sufficiently advanced technology is indistinguishable from magic. I'd like to build on that and suggest that any sufficiently advanced AI is indistinguishable from real intelligence. The question is, how "intelligent" should we build them? In the case of a robot to assist the elderly, I'd say that accepting voice commands to perform common tasks is a pretty darn good start.

Re:Is it just me?

[sm62704]

I could argue the same think about the neurons and synapses in your head.

The neurins and synapses are in a 3d structure, and are electrochemical. Thought is a chemical/electrochemical process or getting drunk would be impossible. Your computer IS an abacus. Your brain, or your dog's for that matter, are not.

As with many things in computer science, it's all about abstraction

I don't argue that thought can't be simulated; I do argue that a flight simulator doesn't leave the ground and a neuclear explosion simulator doesn't leave any radioactivity.

"Mommy mommy, look at the robot walking around! He's waving at me!" I know that computers are only capable of doing what you program them to do, but several of my friends will swear that their computer hates them, or that their gaming console has it in for them

That behavior is called Anthropomorphism, "the attribution of uniquely human characteristics and qualities to nonhuman beings, inanimate objects, or natural or supernatural phenomena" and is perfectly normal. There must have been some evolutionary advantage to it, as it has existed as long as we have.

Arthur C. Clarke suggested that any sufficiently advanced technology is indistinguishable from magic. I'd like to build on that and suggest that any sufficiently advanced AI is indistinguishable from real intelligence.

I agree, which is why we need to educate people before they go and so something incredibly stupid, like passing "robotic rights" legislation.

Remote Remote

[orkysoft]

Among other things, the bot can currently can plot out its own map of its surroundings and remember specific locations and, of course, change TV channels at your command (or, as you can see above, even serve as a TV itself)."

So they finally invented the remote control for the remote control?

Yeah but

[dibblda]

It's probably made of lead paint and GHB....

cat threat

[xPsi]

The goal is for the robot to learn new tasks by following voice commands, which the researchers hope could eventually allow the robot to help the elderly or become a "good household mate" for families. Among other things, the bot can currently can plot out its own map of its surroundings and remember specific locations and, of course, change TV channels at your command (or, as you can see above, even serve as a TV itself). There are a lot of cats out there right now licking themselves acting aloof, pretending like they don't care, but really thinking, "there's no goddamn way I can compete with this thing."

The Perils of Voice Recognition...

[RudeIota]

Verbal input to robot:
"Put my java beans in grinder. Percolate water and coffee. Pour in the mug. Place the mug on my desk."

Robot output
"Put my human spleen in grinder. Exsanguinate daughter slowly. Pour in the blood. Place the blood on my chest"


... and this, my friends, is why I might wait for version 3.0.

Re:The Perils of Voice Recognition...

[Brad1138]

Maybe it runs Vista, it will ask "are you sure" about 50 times.

It may be intelligent, but...

[RockMFR]

Can it see why kids love cinnamon toast crunch?

Plagiarism

[clarkkent09]

I think it's obvious where they got the idea for the design of the robot. Just compare the two images, similarities are obvious:

http://file.shanghaidaily.com/News/Image//2007/2007-11/2007-11-08/20071108_337360_01.jpg

http://hubpages.com/u/49849_f260.jpg

Re:Robot Succeeded where Slashdotters fail?

[fastest fascist]

It's true, with a little effort, people learn to leave you alone.

Funny

[jagdish]

9 Comments moderated to +5, and all of them moderated Funny. Will someone post an insightful comment already?

They're made out of meat

[BlueHands]

I am personally surprised to find someone here on /. with such a different point of view. It is good to see that we are not as monolithic as we might seem.

However, if you agree that AI can become indistinguishable than it follows that you MUST grant them rights since there is no way to proved that any person you ever meet or have ever met is "real" - they could just be really advanced tech in a pretty package that cons you into thinking they are human. If there is no difference you can tell, what is the difference?

And for those few people who might read this and have not read the following short story:

They're made out of meat

is a classic and shorter than many /. posts.

Re:They're made out of meat

[sm62704]

That was great, thank you for that!

They talk by flapping their meat at each other. They can even sing by squirting air through their meat."

"Omigod. Singing meat. This is altogether too much. So what do you advise?"

The difference is that WE made the computers, and therefore WE know how they work. If you know how a computer works, you know it can't think.

Now, if you put me in a commercial flight simulator you could probably convince me that you'd flown me to Afghanistan, or even the moon if you could do something about that gravity thing. But a computer is only an abacus with two beads per wire, billions of wires, and mechanisms (nand gates, nor gates, shift circuits and the like) to manipulate the beads.

I would agree that, when robotics is advanced enough that you can't tell a humanoid robot from a human, you should grant it human rights until it is proven to be a robot. I'd hate to be confused with a robot!