Brain On A Chip?

http://www.sciencedaily.com/releases/2009/03/090318090142.htm

Practical neural computers could be only five years away. “The first step could be a little add-on to your computer at home, a device to handle very complex input data and to provide a simple decision,” Meier says. “A typical thing could be an internet search.

Where could this go? Meier points out that neural computing, with its low-power demands and tolerance of faults, may make it possible to reduce components to molecular size. “We may then be able to make computing devices which are radically different and have amazing performance which, at some point, may approach the performance of the human brain – or even go beyond it!

Basically they're putting very small neurons on 20cm wafers and link them with synapses, they used analogue electronics to represent the neurons and digital electronics to represent communications between them(synapses).

Since the neurons are so small, the system runs 100,000 times faster than the biological equivalent and 10 million times faster than a software simulation. “We can simulate a day in one second,” Meier notes.

Now that's a bit scary,but just a bit... to have a robot that is 100.000 times smarter than us

Sound nice, but what are the costs..... to have something like that.
It probably another 25 years before the average Joe can afford it...
Too bad....

They said that in 5years this technology will be used in commons personal computers but to have a chip that calculates close to the human brain it will take a few more decades.

Like i said. The technologies for better human like robots comes from it self. All what is needed for human robots already exist, only the technologies is at this time not good enough to make a really good human robot. But i think Aiko can be the first one when she is ready, but i also think that this will takes still some another years. Aiko V4 or so.

One thing to look like a real human is her reaction speed. At the videos from Aiko she is still very sluggish, so she needs much more CPU Power to calculate all things faster and can react faster. Neural computing can be here a great help. When Aiko can react faster with humans, she looks more human for us, because she let us not the time to think about her as a robot when we interact with her.

[quote=/]
One thing to look like a real human is her reaction speed. At the videos from Aiko she is still very sluggish, so she needs much more CPU Power to calculate all things faster and can react faster. Neural computing can be here a great help. When Aiko can react faster with humans, she looks more human for us, because she let us not the time to think about her as a robot when we interact with her.[/quote]

Dude this sound like star trekking "Scotty we need more speed" , "But Captain she's giving all she got more then this and she gonna BLOW".
Did you know that the human brain need less then that to accomplish what a computer does with brute force,
Faster and faster has its limits not to say cycle problems of hit and mis in CPU cache, Your asking for trouble you cant cope with.
There must be a simpler solution to keep speed for important things then every thing.

She must react so fast like a human, to be like a human. When you ask Aiko something and you must wait too long for a answer, then she is just not naturally enough for us. And when she should walk and move fast, then she need a fast CPU to calculate the movement as fast as she can to control her mechanic for the movements without delay, because she must hold her balance very fast. Of course not only a fast CPU is needed for that.

Later this year, a new revolutionary product will come out to allow us to actually move things with our mind.
First it will be used on video game characters, but will be improved to move anything electronic that can be hooked up to non electronic devices.
A similar device is already out for the medical community, giving paraplegics mobility for every day tasks, such as moving a wheelchair, playing video games, even E-mailing friends on the Web.

BRAINGATE

What a world we live in!

One way to make her faster would be what I call "layered" memory. Things she does often or most recently are placed in a "tier 1" category, and things she rarely does are relegated further and further down to "tier 5". You could also program things in at the various layers. It's like this: You ask one person what 2*2 is. They say 4. That is tier 1. Ask a person what the capital of Tunisia is, and they would probably have to stop before they respond "Tunis." This is not tier 1, it is more close to tier 4 or so. However, if a person is an ambassador to Tunisia, it would be tier 1.

See where I'm going with this?

Could you please put that in simpler terms the average noob could understand?

Imagine a five-story tall building. You work on the top floor, but will occasionally work in the lower stories. The top floor is "tier 1" memory, things you do regularly or know in great deal. The lowest floor is "tier 5" memory, things you rarely do or know little of.

It's like I'm using an elevator every minute of the day, and the "Muzac" is always playing while I ride.
A bit maddening for me maybe, but I suppose Aiko and BRAINS can take it all in.

Nice metaphor.

HI

Memory in layer that is very interesting....
Logically, it should work better.


That would change my entire database structure...arrrr.
something to look forward in the future.
Good thinking....Konrad13.

You may want to experiment with it, start small. My idea for layered memory is definitely vague, and my assumption is a very large memory capacity to deal with the layers.

Also, my job usually has me doing one thing for hours on end: standing at a station (I am a cashier at a local store, and often am relegated to working the self check outs, a job that allows me to write little notes and allow my mind to wander far indeed). I constantly think about these things at work and jot notes and such down.

In fact, I wrote a note today: "If I win the Powerball (a lottery where I live), create Yumecom Technologies with Le Trung." Yep, if I win the powerball, Mr. Trung, you will find a very rich supporter of your tech. If I'm sole winner, I'm estimating a $40 million payout after taxes, $8 million will go towards this new company! I'd give you 95% power over the project, so no "bending to corporate ideas" and such.

If I won the lottery, I think I'd also be willing to donate a good sum of cash so that Mr. Trung could progress the advancements of his own dreams (with no strings attached) I've accomplished all my dreams for this life, but it is invigorating to see someone else's come to fruition.

Shashinka Ichiban wrote:If I won the lottery, I think I'd also be willing to donate a good sum of cash so that Mr. Trung could progress the advancements of his own dreams (with no strings attached) I've accomplished all my dreams for this life, but it is invigorating to see someone else's come to fruition.

While I respect this, if I were to come into a large sum of money with the ability to finance Mr. Trung, I would want some say in what the money is used for. However, as I stated, roughly 95% of the project is his, to do with as he sees fit. I would only want the right to speak up, come up with ideas and such. Also, I want a PC-tan!

But yes, I would definitely finance Mr. Trung, but only on my (fair and flexible) terms if it is an exorbitant amount of money is to be used.

Konrad13 wrote:

Shashinka Ichiban wrote:If I won the lottery, I think I'd also be willing to donate a good sum of cash so that Mr. Trung could progress the advancements of his own dreams (with no strings attached) I've accomplished all my dreams for this life, but it is invigorating to see someone else's come to fruition.

While I respect this, if I were to come into a large sum of money with the ability to finance Mr. Trung, I would want some say in what the money is used for. However, as I stated, roughly 95% of the project is his, to do with as he sees fit. I would only want the right to speak up, come up with ideas and such. Also, I want a PC-tan!

But yes, I would definitely finance Mr. Trung, but only on my (fair and flexible) terms if it is an exorbitant amount of money is to be used.

I'm not a robotics engineer. Hell, I was barely qualified to work for a year and a half as an engineer technical for bp solar. (Engineering is not my background) so as far as to pitch ideas is not my place. My dream would be less about having say in the creation of the yumecom, but rather that of seeing Le Trung ability to see his own dream come true. I'm better suited as a philanthropist

I'm thinking about a new kind of business model for this sort of software development. I wouldn't wantto give software away free under a GNU license, but it would still take a team of people when there is no cash up front.

Any ideas?

Ross

Make a wish and it comes true.

When my parents said that to me the way they interpret that if you want something in life you must work hard for it.
A true wish that makes everything come true is your own passion and drive.
I still believe that, Being on this forum I seen a man that went against all odds and realize his wish to build something
And he got very far at it also, Yes it came in small steps, But those small steps became big because we believe in
what he does and support him with that, For me its like sharing that dream with him by my own wishes and I hope he
succeed.

The chance of winning the lottery is very slim. Would you really do those thing if you where a winner I think you guys won
something already you have so much ideas and interresting stuf from the internet even ideas from your own about yumecom
Its much more then waiting for the lottery will happen. And if it ever happens that you'll be a winner by luck let wisdom be your
guide.

And.

I'm still wanting a kind of sumomo but the doll on amazone does not seem the right android thingie i'm looking for another
exoskeleton. If only the head could move that would be ok.

Yes it's good to dream and work hard. But if you want to achieve something in the long run, it helps to have a way to generate cash-flow. If you put in so much time that it's not a hobby anymore, you need at least enough money to live on.

So what sort of business would make it work.

Ross

Well... you can always do a rental business like they do with the Der2 and Der3 over in Japan.

Ross wrote:
So what sort of business would make it work.

Ross

Business proposal 1.

Get an acquiantance to yumecom.

Wel androids are not that common to introduce them to the public it must be something everyone is already vermilliar with.
because they are different then robots or what the public persume what robots supposed to look like.

I was thinking about ''Mannequins'' androids that can move and pose with the clothes they wearing and present them, Even tell
something about the brand , Creators or comming trendset or what fits most with the combinations. This kind of yumecom can be
preprogrammed with what the customer wishes and is a friendly dress up to work with which plays a role in having a good product.

These yumecom mannequins can be used as tools to pose for new kind of clothing for coutiers and as a static model in most
shops. In time better yumecoms "M" series (M for mannequins) designed for various posing can do more. Because the customer can
have them with certain moves they are customed to or visioned to how they see sales benefits best to seduce customers in by a
yumecome M serie that say's how well it would fit or that its trendy no matter which size a person is.
Technically yumecom have standard software functions to recognise those features. Those things are recently possible.

Discovering the market.

From there the possibilities can go even further as consumers will know yumecoms as mannequins other types can be introduce
to different markets like caretaking for the elders , In hospitals to entertain kids with the robot. Entertainment industries, Sales events.
You name it.

For business it means costumers can rent a yumecom type to their bussines to provide themself a new kind and broad way of telling or
presenting their products. Based on what the yumecom does their are cost for renting and programming custom behaviours.

Well the this is one of some idea's.

WulfCry wrote:
Business proposal 1.

Get an acquiantance to yumecom.

Wel androids are not that common to introduce them to the public it must be something everyone is already vermilliar with because they are different then robots or what the public persume what robots supposed to look like.

I was thinking about ''Mannequins'' androids that can move and pose with the clothes they wearing and present them, Even tell something about the brand , Creators or comming trendset or what fits most with the combinations. This kind of yumecom can be preprogrammed with what the customer wishes and is a friendly dress up to work with which plays a role in having a good product.

These yumecom mannequins can be used as tools to pose for new kind of clothing for coutiers and as a static model in most shops. In time better yumecoms "M" series (M for mannequins) designed for various posing can do more. Because the customer can have them with certain moves they are customed to or visioned to how they see sales benefits best to seduce customers in by a yumecome M serie that say's how well it would fit or that its trendy no matter which size a person is. Technically yumecom have standard software functions to recognise those features. Those things are recently possible.

Discovering the market.

Well the this is one of some idea's.

Don't read this as being discouraging or anything, but maybe I am missing something in this part of the reply, but I'm not sure what benefit it is to downgrade a Yumecom to being nothing more than any garden variety animatronic robot seen at a local Chucky Cheeses or Disney theme park.

Anamatronics has been around for decades, but what sets the Yumecom apart is the AI, which would really be wasted on trying to sell clothes in a department store. It would not take very long for the Yumicom to be regarded as nothing more than a gimic.

In retrospect this sorta contradicts my last comment about rentals, but with further thought, the implementations and future applicable technology of the Yumicom should be for something greater than window dressing, or hock ad space.

In the end, the real goal of what the Yumicom should be can only be decided by Le Troung himself.

WulfCry wrote:

Discovering the market.

From there the possibilities can go even further as consumers will know yumecoms as mannequins other types can be introduce to different markets like caretaking for the elders , In hospitals to entertain kids with the robot. Entertainment industries, Sales events. You name it.

For business it means costumers can rent a yumecom type to their bussines to provide themself a new kind and broad way of telling or presenting their products. Based on what the yumecom does their are cost for renting and programming custom behaviours.

Well the this is one of some idea's.

Now, this is where I can see a Yumicom having a real impact on society. Using them to assist in the care of the elderly or children. However as far as entertainment and sales goes, that leads right back to my statements about anamatronics used in theme parks and restaurants now.

A good amount of research from both sides of the pond have shown how pets improve the lives of people of all ages, but recent studies in Japan take this a step further with the introduction of limited AI equipped anamatronic "pet" animals. A yumicom with the ability of enhanced interactive AI could be greatly applied in this manner to serve and care for the elderly or provide daycare.

Yes I agree its hard to imagine if its even possible for androids or yumecoms to break through
and get public excaptance without the idea that its also scary or borring as a machines.

People want something friendly they themself could get attach to even knowing it is not real.
animatronics seems so degrading for that purpose I agree.
Still seems the most friendly approach to convey the image of androids to the masses.
We ourself here at this forum and other robotics enthousiast already share .

Friendly based androids

The potential to the approach I thought was getting people know that yumecom is a friendly based android you seen it as somekind of
mannequin and now you know that they support in care taking to. Imagine kids at a school trip to seaworld, Lots of information about sea life you could let them punch on a keyboard at an computer for additionel information , Or you can have a yumecom which understand them clearly tell them about anything they ask . Its al about getting used to the image of a yumecom if it gets so far that is.

This is very difficult stuff I tried once to explain on this forum what A.i could be based on an android exoskeleton.
Most people dont care about robots but if that robot or android has something appealing in motion or portraying emotion
that is easy to recall as friendly then that is the succes of a trade in yumecoms.

I have also posted about that in what A.i could be as an attempt to explain it but not so succesfully