Self reproducing robot developed

[Balschoiw 0]

Scientists create self-replicating robot

Quote[/b] ]Scientists at the Cornell University in Ithaca, New York have created small robots that can build copies of themselves.

Each robot consists of several 10-cm cubes which have identical machinery, electromagnets to attach and detach to each other and a computer programme for replication. The robots can bend and pick up and stack the cubes.

"Although the machines we have created are still simple compared with biological self-reproduction, they demonstrate that mechanical self-reproduction is possible and not unique to biology," Hod Lipson said in a report in the science journal Nature on Wednesday.

He and his team believe the design principle could be used to make long term, self-repairing robots that could mend themselves and be used in hazardous situations and on space flights.

The experimental robots, which don't do anything else except make copies of themselves, are powered through contacts on the surface of the table and transfer data through their faces. They self-replicate by using additional modules placed in special "feeding locations."

The machines duplicate themselves by bending over and putting their top cube on the table. Then they bend again, pick up another cube, put it on top of the first and repeat the entire process. As the new robot begins to take shape it helps to build itself.

"The four-module robot was able to construct a replica in 2.5 minutes by lifting and assembling cubes from the feeding locations," said Lipson.

This is a pretty interesting step towards autonomy of robots that need to reproduce themselves i.g in large, to be built platforms or space stations while I also can think of them taking over the jobs of house building companies oneday.

I hope the don´t use Windows as OS though

On the way to mars:

[MachoMan 0]

To be honest I'm not impressed, it's pretty much child's play!

These robot don't create everything they need themselves, the parts are still made by others robots/humans, think about the chips ,etc.

Nothing to be scared of .... yet ...

It's primarily a PR stunt.

[InqWiper 0]

To be honest I'm not impressed, it's pretty much child's play!

These robot don't create everything they need themselves, the parts are still made by others robots/humans, think about the chips ,etc.

Nothing to be scared of .... yet ...

It's primarily a PR stunt.

I agree. This is probably closer to a script that reproduces itself than nanorobots

[theavonlady 2]

It's OK with me as long as this ability isn't carried over to OFP's enemy AIs.

[iNeo 0]

It's ok with me as long as they don't develop free will.

[marcusjm 0]

It's a question of scale so i am impressed. This is just the beginning of it all .

I think the Japanese are onto something similar.

Soon we'll have Terminator for real .

marcus

[Scrub 0]

I say "EMP 'em all and let the engineer sort 'em out!" lol  

I'll be worried when they can identify the raw materials and understand the method of turning them into logic circuits & proper structural materials. Long way off. Without understanding, it's just a script.

[andersson 285]

Humans = the missing link between monkeys and robots?

[Acecombat 0]

Kill it now , this is what the animatrix talked about .

[Ex-RoNiN 0]

I say "EMP 'em all and let the engineer sort 'em out!" lol

I'll be worried when they can identify the raw materials and understand the method of turning them into logic circuits & proper structural materials. Long way off. Without understanding, it's just a script.

You can script "understanding" too Would be one hell of a script, but still "understanding" as you put it.

Oh, and free will is an illusion, there are limits to it, its the same for computer programs, it doesn't exist, its all governed by some sort of law and thus there are boundaries.

Sure, you can push the boundaries, use soft-and hardware neural nets, use self-modifying FPGAs, but it is all described in the code that you create and thus there is a boundary there.

[sanctuary 19]

I say "EMP 'em all and let the engineer sort 'em out!" lol

I'll be worried when they can identify the raw materials and understand the method of turning them into logic circuits & proper structural materials. Long way off. Without understanding, it's just a script.

You can script "understanding" too Would be one hell of a script, but still "understanding" as you put it.

Oh, and free will is an illusion, there are limits to it, its the same for computer programs, it doesn't exist, its all governed by some sort of law and thus there are boundaries.

Sure, you can push the boundaries, use soft-and hardware neural nets, use self-modifying FPGAs, but it is all described in the code that you create and thus there is a boundary there.

As long the scientists have no forgotten the line

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">?(thatsenough) : exit

there is nothing to worry.

If they have forgotten that line, an EMP attack is required quickly.

Just hope they have optimised the loop to not cause lag too.

[Apollo 0]

Onward technocrats.

Sure ,they arn't creating the build materials themselfs ,but then they could create a fully automated material suply module that supply's the material and a automated material creation module maybe ,that combined with a material extraction module .... and afcourse an automated module cration module.

With the technological acceleration these days ,thing how much we have progressed the last 20 years in the field of electronics ,robotics ,software engineering ,and how much faster we are going to progress in the next decades.It won't take that long anymore before robotics can outperform most humans in most fields of labor ,pitty we still got the money system.

[denoir 0]

These robot don't create everything they need themselves, the parts are still made by others robots/humans, think about the chips ,etc.

Self-replicating biological structures (such as DNA) don't create the materials they need either.

Quote[/b] ]Oh, and free will is an illusion, there are limits to it, its the same for computer programs, it doesn't exist, its all governed by some sort of law and thus there are boundaries.

A lot of biochemical processes are controlled at the quantum level, and quantum mechanics is stochastic i.e not deterministic. One could argue that the "free will" is a consequence of that randomness.

There are however interesting experiments that seem to indicate that what we perceive as "free will" or as consciously making a decision is rather a process that comes after the actual decision has been made. There is evidence pointing to that our conscious thoughts are actually just the neocortex's way of explaining what just happened - after the fact. It is sort of an after action report that is used for structuring hierarchical memory.

One popular experiment is to give a person two buttons (one on the left side ond one on the right) and a screen with a watch on it. The person decides which button to push and notes the exact time when he/she made the decision of which button to push. The electrical signals in the person's brain are measured during the experiment.

As it turns out the brain sends out commands to activate the muscles on the left or right side of the body a long time before the person thinks he/she made the decision.

[Colossus 2]

What Yahoo didn't have was the movie

Got it right here, just press the play button as usual.

Very flexible robots I must say.

[mattxr 9]

Why not script them with the 4 laws

A robot may not injure a human being, or, through inaction, allow a human being to come to harm.

A robot must obey orders given it by human beings, except where such orders would conflict with the First Law.

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

A robot must not evolve on its own acord to become a harm to the human race.

i made the last one up lol

[EiZei 0]

Too bad those laws can not be applied for military purposes and you know what this kind of technology usually ends up to..

[Balschoiw 0]

While Asimov created those hyphotetical rules for robots in the 40´s ? they are more than paradox in the context we can see them today.

They are not necessary for machines that only serve one purpose and don´t have the ability to choose from options at it´s own "will" as the machine is only built to repeat a pattern or a scheme all over again. The purpose it was built for.

Now the paradox situation starts when robots are invented that have the ability to choose their options, learn undirected and fulfill not only a limited purpose but are built to evolve. This is something in the making with biomaterial chips and adapting robots. They have the ability to judge on their own what structures they develope more or less to make them ultimately efficient and flexible for the owner.

Small example: A houscleaning machine that uses sensors to

determine the most efficient ways within a room and the structure of the floor. The ideal housecleaning machine will learn how the room looks and combine those data with new sensor datas, like a chair that has been moved and such. The robot can be more faster that way as he can rely on what he has learned and can combine it with new sensor datas that indicate a change.

As it is a purpose to make those kind of machines even more flexible, efficient and independant they need to learn and remember what they have learned.

We can find such in many aspects of daily life today, for example cars that remember driving styles and therefore adjust the settings of a car.

It´s comfortable for us but also offers a risk. As the process of learning is happening within the machine it´s not that easy to overwatch it. Today, it is possible, but when biomaterial like nerves get implemented into those structures it gets freaky as the chip itself will build it´s structure. Comparable to the fine tuning of the human brain as a baby.

I am not talking about thinking machines. I´m talking about machines that get used to their surrounding to work more efficient.

Back the the rules.

Let´s assume a paradox situation:

I set up rules for a robot which has only the purpose to kill humans, a war-robot :

1. A robot has to kill a human whenever he faces one

2. A robot must not obey human orders unless they support the first rule.

3. A robot has to protect it´s own existance unless this doesn´t interfere with rule 1 and 2

Now let´s take this into the daily life of a human killing robot:

The robot has a defect and needs to be repaired by a human service technician.

If the robot obeys the first rule (killing the serviceman) than he does break his 3rd rule (protecting his own existance). If he obeys the 3rd rule (getting himself repaired), he breaks the first rule. Now the robot could think...click..click...let´s get me repaired and then kill the technician. But would he get the chance to be repaired again ? If not (technician dead) he would have broken rule 3.

You see the paradox situation ?

So what for made Asimov such rules ? For machines that can determine on their own what to do. It´s hypothetical, yes, but it´s a thing that will come. The rules should guarantee that the creators (humans) will never be injured, hurt or even killed as a result of a roboters decision.

Fair enough, but to make you completely fuzzed there is a flaw in this rules:

A bad or harming action must be in the pool of decisions of a robot to be able to be prevented. Good actions are only to be distinguished from bad actions if the robot is capable of performing bad actions as they make him aware of good actions

The conclusion is that only robots that are allowed to perform bad actions are able to decide what is good and what is bad.

So "thinking" machines end up in a dilemma oneday. I hope that we can be eyewitnesses oneday and I really do hope that we don´t have to watch that all from a real deep bunker

All the above nonsense should only demonstrate that the 3 rules that Asimov wrote are not practicable with the developement of machines we see today.

Asimov even added the 0 rule later:

0. A robot may not injure humanity, or, through inaction, allow humanity to come to harm.

The robot is now able to ignore the orders of a human to follow rule 0. As humanity is listed above the single human individual they count more.

But who says that the majority is right ?

Anyway, I hope you enjoyed the trip. click...click

Read Isaac Asimov and Stanislav Lem books. They are great science fiction, in the direct meaning of science and fiction. Great reads.

[InqWiper 0]

Quote[/b] ]As it turns out the brain sends out commands to activate the muscles on the left or right side of the body a long time before the person thinks he/she made the decision.

Can that be because after the first time he decided to keep clicking the same way again and after the 2nd click which the person has already decided what to do about the person reassures themself they will do the same thing the 3rd time?

I really dont know, its just a thought, I think its hard to prove what the person actually decides. does the person decide to click the 2nd time after they have clicked or was that the 3rd click the person decided about?

I really dont think we have to worry about robots going mad and causing havoc unless someone hacks them (or they use MS OS)

[USSRsniper 0]

hehe self reproducting robots bad idea. Robots get smart, reproduce thmselves. One time they will get sick about people and kill them all  

[ozanzac 0]

Cool technology, however, it's merely anything more than a production line that builds a production line.........

I'll start to get interested, when a robot will be able to reproduce itself and perform other tasks, as well produce modified versions of itself (for tasks other than the sole purpose of making copies of itself), and learn to adapt and update the design of itself and its offspring over time.