Computers' 'Common Sense' for object recognizing

http://www.sciencedaily.com/releases/2007/10/071017174328.htm

A Google labs item used to improve object recognision. Maybe Aiko can use it in the future.
Maybe artificial common sense can be the key to improve not only image recognision, but also a hability to enable the machine to make inferences in a bank agency ( for example) or to perform tasks in a better way, so the owner will not have the need to explain everyting in a order ( for example : to prepair a coffe in a new coffe maker), because machine can make inferences from past expericences making coffes in an old coffe maker)... and so on...

What do you think about computers able to have some kind of common sense???

Gilda wrote:http://www.sciencedaily.com/releases/2007/10/071017174328.htm

A Google labs item used to improve object recognision. Maybe Aiko can use it in the future.
Maybe artificial common sense can be the key to improve not only image recognision, but also a hability to enable the machine to make inferences in a bank agency ( for example) or to perform tasks in a better way, so the owner will not have the need to explain everyting in a order ( for example : to prepair a coffe in a new coffe maker), because machine can make inferences from past expericences making coffes in an old coffe maker)... and so on...

What do you think about computers able to have some kind of common sense???

Common sense is going to require a lot of programming and/or learned behavior. One place made a robot that learned by trying how to do basic tasks, and it took many tries to figure out how to stick a cup in a dishwasher.
Common sense is going to put something like the above into the realm of 'that was easy.' But, oh man, will common sense in a robot be useful.
Thanks for sharing.

Are you reffering to Genetic Algorithms?

I didn't think about it.

With common sense I was just refering to habilities like recognizing objects and functions in a faster way without need of being re-teached everytime I'd buy - for example- a coffee maker ( wich the only difference is that the old model had on/off on the left and the new is on the right)...

People will need easy to deal robots not one that will stop working just because they forgot to teach the robot were the new on/off coffee maker buton is...

Gilda wrote:Are you reffering to Genetic Algorithms?

I didn't think about it.

With common sense I was just refering to habilities like recognizing objects and functions in a faster way without need of being re-teached everytime I'd buy - for example- a coffee maker ( wich the only difference is that the old model had on/off on the left and the new is on the right)...

People will need easy to deal robots not one that will stop working just because they forgot to teach the robot were the new on/off coffee maker buton is...

I guess both would be needed to solve the problem. My coffemaker broke, and got replaced by one that looks so far from a coffeemaker that the machine wpuild need a way of figuring out what it is. Perhaps some identifiers of objects could be made, but then an object that shares similarities but isn't the intended item could be damaged by improper usage.
Maybe we just need a robot that can identify items based on an image of them, as well as over writing a previous idea (the last coffeemaker broke, therefore, I do not need to recall the specifics of its appearance.)
Maybe all we need to do is make a robot that can read the manual and identify it from the drawings there.
Then again, it wouldn't necessarily be able to identify the coffeemaker in another house, and would need to learn how to work it (but then again, most people usually need to ask how to work the appliances when you are at another house and helping in the kitchen.)
An interesting conundrum.

It seems Japanese guys made something to help in the coffee maker issue
https://projectaiko.forumotion.com/t332-japanese-invented-a-robot-that-can-think-by-it-s-own#3011