Slashdot Mirror
Babybot Learns Like You Did
holy_calamity writes "A European project has produced this one-armed 'babybot' that learns like a human child. It experiments and knocks things over until it can pick them up for itself. Interestingly the next step is to build a fully humanoid version that's open source in both software and hardware."
From TFA: "The goal is to build a humanoid 2-year-old child," explains Metta. This will have all of Babybot's abilities and the researchers hope it may eventually even learn how to walk. "It will definitely crawl," says Metta, "and is designed so that walking is mechanically possible." Not a bad goal at all, and if it's open source they can't cheat by promoting a specific goal such as walking in the software. Reminds me of Prey where they couldn't figure out how to get the nanomachine swarm to fly so they let its AI "learn" how to do it on its own.
Making you think you're crazy is a billion dollar industry.
baby says fp!
Aren't you afraid this poor open source robot will get exploited by the other robots, or do the proprietary robots have something to hide? What kind of insults can we expect? Your father was a code monkey and your mother got her card punched by a UNIVAC!
babybot? robocub? fire your marketing people!
may mean that such machines can never become as intelligent as us
They don't know and they're playing with it. Have they even seen the Matrix??
Common sense is not so common
A one armed baby bot? That's disturbing on so many levels.
Philosophy.
LOL When "babybot" goes to grap the ball watch how fast he gets his hand out of the way!
Obviously babybot doesn't know it's own strength! LOL
Common sense is not so common
I for one, welcome out new Babybot overlords!
So this bot is going to lie in its crib, thrashing its arms and legs, screaming at the top of its lungs, until someone picks it, gives it a full juice bottle, a cookie and walks it around trying desparately to amuse it?
"Academicians are more likely to share each other's toothbrush than each other's nomenclature."
Cohen
A fun project, and potentially a good step on the road towards human-like intelligence. However, the "2-year-old" remark is again one of those far-fetched promises that is a loooooooooooooong way off. Making a robot-arm play with a rubber ducky is one thing, letting a robot understand what a rubber ducky is, is quite another. Making a robot crawl is one thing, but letting a robot crawl with a self-conscious purpose, again is quite another.
Fortunately, one of the researcher in TFA admits that 20 computers with a neural network on each is no replacement for a human brain. But the 2-year-old remark follows later, and is evidently entered as a way to generate funding. It sounds cool, but it is not what the result of this project will be. I assume the researchers know this all too well. Or perhaps they have no children of their own.
The story mentions that the AI is made using neural nets.
I think it's amazing how such simple data structures can generate such complex behaviour.
In case anyone is interested, there's this pretty easy to understand tutorial on neural nets here:
http://www.ai-junkie.com/ann/evolved/nnt1.html
Fortunately, one of the researcher in TFA admits that 20 computers with a neural network on each is no replacement for a human brain. But the 2-year-old remark follows later, and is evidently entered as a way to generate funding. It sounds cool, but it is not what the result of this project will be. I assume the researchers know this all too well. Or perhaps they have no children of their own.
Think of how Social Services could use something like this if it can act like a 2 year-old. Do they want to make sure you would be a good parent? They'll give you the robot for a week and based on the data they can then tell if you can be trusted (obviously assuming the robot is unhackable, or at least knows if it was hacked). If that doesn't generate government funding then I don't know what would!
How long until it learns how to frag?
Is this the offspring of Data and Tasha Yar?
That which does not kill me only postpones the inevitable.
This project was born from an engineering approach to the problem of what is consciousness. This is with this problem in mind that the European engineers designed BabyBot. And their experiments, while promising, don't solve entirely the problem of the definition of what is consciousness. So they're now designing new robots like the iCub. Read more for additional details and pictures of BabyBot and its successor, the iCub robot.
I wonder what happens when this bot discovers that it's a physical object, and can try and manipulate itself.
(... yeah, baby robot masturbation... but no, seriously...)
From TFA:
"Everything about it will be open source, including the hardware, so anyone can use it in their own work," Metta says.
I'm unclear on this concept. Do they mean off the shelf commodity parts? Blueprints so that you can machine the parts yourself, if you have a lathe? Or is open source going to become a euphemism like "five finger discount"?
Seriously, what is Open Source Hardware, if it's not just a sorry misuse of a buzzword?
It's not offtopic, dumbass. It's orthogonal.
'Where am I ?'
However, just suppose, and then suppose, and then suppose...
So far, we can build computers that can simulate brain cells. There is nothing stopping us making a computer that has a similar complexity to the brain. We will have to mimic the strange mix of part-design, part randomness that brains are. Or maybe we can just throw more computing power, and stuff the brain doesn't have, like the ability to back up and regress. Sooner or later - probably later is my guess, but who knows? - we are going to come up with something that shows intelligence, and probably has inteligence.
African grey parrots are kept as pets. These are said to be as intelligent as a two-year old. Some of them can understand sentances from a vocabulary of hundreds of words. They don't progress much beyond a two year old. And they are Not Like Us, so it's OK to keep them in cages. Apparently. Hmmm.
One day, someone is going to make something intelligent, and then turn it off, and there will be an outcry. Is anyone doing the thinking on the ethics of making it before making it?
To get people interested in cyborgs or androids we must make them look human. We should start by making furry versions of the Aibot dog or whatever it's called.
It learns by trial and error, eh?
How many dead babybots does it take to learn to use Windows?
- RG>
Hey pal, this isn't a pleasantforest, so don't waste my time with pleasantries!
It doesn't need diapers and doesn't cry during the night. Put a second arm on it and tell me when it hits the market, I'm buying one!
So say we all
Ok call me a scifi nut but who on /. isnt? But can you say Cylon? Him first we start with Babybot, then crawlingbot, then a Walking Chrome Toaster, then 12 new human like models. All beleiving that there creator is flawed and is now believing in our God or Gods pending your religion.....
It experiments and knocks things over until it can pick them up for itself.
You don't need an advanced AI to do that, the algorithm goes like this:
while(1) {
throw_toy();
while(!toy_is_back())
cry_loud();
}
I question:
What happens when machines reach human level thought speech or better yet surpass it? What then about us becomes obsolete?
Smile.
Don't Worry -- it's only the end of the human era.
Have anyone seen the video?
I have seen 2 (all?) of them and I have noticed that the bot had to rest his hand on the surface everytime he fails the task before attempting again. Why does it have to do that?
Also. At first I have noticed that the bot drops objects into the hand of the researcher. But later I have noticed that it just drops it in the particular place (second video, pile of objects on the right at the level of the babytable). I guess the reasearcher sticks his hand so the object drops into his hand to make the behavior of the robohand look like he specifically drops the object into hand thus creating a wrong impression. I would advise to make more scientific and less marketing presentations next time, so people could learn from it.
I do not believe in karma. "Funny"=-6. Do good and forbid evil. Yours, Oft-Offtopic Flamebaiting Troll.
"The goal is to build a humanoid 2-year-old child," explains Metta.
There is a far easier and more pleasant way to create a child.
Unfortunately, it requires 2 years, nine months, and three minutes.
If "disco" means "I learn" in Latin, does "discothèque" mean "I learn technology"?
-- had to. :-P
the next step is to build a fully humanoid version that's open source in both software and hardware."
You mean, one where the microcode for any processor included in it is published openly, and the masks used at the chip foundry are also openly published? Or if it's a FPGA 'Free Hardware' design, all design details of the FPGA silicon are disclosed, and all of the code for the FPGA development software is open source (good luck)?
>To get people interested in cyborgs or androids we must make them look human
There is another side to doing that; When something looks human, we are more likely to attribute human like qualities to its action. Anthropomorphism. Works with animals too, ie Aibo.
MIT were doing some great work on this, and social computing, at the MIT Media lab in Dublin before it was shut down. I was lucky enough to see some of their ideas in action.
Real shame to see them go, I hope the work gets picked up elsewhere.
Open Source Drum Kit, LPLC deve board - mjhdesigns.com
...does it have the memory of an elephant ??
throw_toy();
while(!toy_is_back())
cry_loud();
if (mom_leaves) {runsilent();}
}
Trust me. Robot or not, its the oldest trick in the book.
barack to the future?
Not a bad goal at all, and if it's open source they can't cheat by promoting a specific goal such as walking in the software.
Yes. AI scientists have a bad habit of making implausible claims for their creations. The open approach will keep them honest and is to be commended. At the very least, such a robot needs several types of learning functions including perceptual, short and long term memory mechanisms, concept formation, pattern completion, anticipatory behavior, motor learning and coordination, operant and classical conditioning, etc... Does anybody know what sorts of NNs and what learning principles are being used in this bot?
That baby would be tough on the birth canal.
I don't believe they'll truly make a human-esque robot until they can make it understand pain.
Sometimes a child needs to have a hand across his/her hiney to teach him. What if the bot touches a hot stove and melts the crap out of its hand - without pain it would not know the difference.
Let a robot go through that, and then they might truly begin to learn like a human being.
What else can happen when an unstoppable force collides with an immovable object?
you insensitve clod.
A disturbing number of murders have occurred in the LIRA labs at the Genoa University. Victims appear to have been strangled, but a lack of fingerprints makes identification of the suspect problematic.
Do you see what I did there?
I applaud their work towards an open-source model. The model this is derived from--aka "human"--has been closed source since its creation almost 6000 years ago. The copyright expired long ago, but its Creator is unwilling to open its source. Many people cannot find the Creator, and some even doubt He is still around to release the source.
The human model has proven difficult to reverse engineer. We need its source to help fix bugs. For example, it's susceptible to viruses in its current state.
So, I welcome the open-source model. It is a giant step in the right direction. I hope one day we can replace all closed-source models with their open-source equivalents.
A droid wich is able to RTFM and STW? It seems droids are now more intelligent than most humans.
Can you teach it slam doors when its angry?
But will it run Linux?
If they had picked some other boring name, do you really think the article would have, e.g., made in on /.?
But will Robocub want to play with its Wii?
Why did they give it Mick Jagger's lips and Keith Richards' eyes?
Computers obey laws just fine. Self-awareness is the hard part, on the other hand.
Well, if theyre going to open source this, maybe we could develop some kind of collaboration program, just like the one that SETI uses to search for ET. How about "Help us to grow our robot beyond his current age". Maybe the whole internet could become its brain... Watch out Google! Of course, it kinda resembles the Terminators movie AI... but what the hell, AI is the next step in evolution anyway.
This seems like a good idea. I've always wondered why AI researchers want to try making AI that begins its existance near the level of an adult, with an understanding of language, "commons sense", etc... I understand it in that language recognition is an important piece of the AI puzzle, but researchers who want to make a "human-like" robot seem to aim too high.
Even the human brain, extremely advanced compared to where we're at in the creation of intelligence, starts out nearly helpless and takes years to learn the basics. We can try to skip all that, and program our prior knowledge in, but we're most definitely going to miss some realtionships and other important sub-concious features of the way ideas and concepts are linked in our brains, and the end result is just not going to be as robust as an instance where the "brain" learned for itself and made its own connections in its own way as it learned the concepts.