Self-replicating robot

This idea came to me a few years ago - I'm no tech so don't expect this to read like an Apollo spacecraft's manual!

The idea is this - you would create a simple robot with the ability to replicate itself at 50% it's original size. Once replicated the ‘replicant’ would break it's creator down and take the necessary materials for creating another relpicant at 50% it's size and so on.

In theory this process could continue indefinitely until you have a robot the size of an atom - and a lot of left-over bits and pieces!

Of course there would have to be some sort of programming involved so that the robots stop replicating after reaching a certain size. Additional programming would also be necessary for the replicants to be able to interact with electrons etc at the atomic level.

Powering the robots is also an issue...

Any thoughts?

PS - Please excuse the theft of the term ‘replicant’, I realise we aren't talking biorobotics here...

While I understand your basic theory, my question would be this..

What would the materials need to be, to be able to acheive this 50%?

I would think that if the same size parts were used, that the resulting "copy" would be the same size as the original.

Unless I am missing a key point somewhere.



edit to add...

Self replication would be possible in a more normal way though, if the robot were programmed to be able to create or otherwise obtain the necessary parts for the construction of the next, and so on.

Now what if you coupled this with true "Learning A. I. "

That would be an awesome feat. Of course the military would most likely seize this for their own uses.

I'm no genius either, so whether it would be truly feasible is questionable.

[edit on 2/6/2007 by Mechanic 32]

We first need to create a robot capable of replicated itself regardless of size. Pulling that trick off we could then go and try to use robots to build smaller robots. It's basically the same process, with some parts removed of course, as the one used since the dawn of the industrial revolution.

That is: Engineer/Tinkerer/Inventor Invents Measuring device > Scientist Uses Measuring device to make sense of the world > Engineer Uses knowledge from previous step to invent better manufacturing techniques and processes > Scientist investigates new invention for any anomalies, if any are found a new measuring device is created to refine capabilities and we're back to square one again.

This process of innovation has been speeding up and such a device capable of replicating itself making itself more advanced with each replication would push the pace of progress through the roof. It would be like living at the dawn of the Aviation Age, or the Dawn of Industrialization... except compressed to such an extent that many of us alive today has a good chance of seeing it... It could work... it may take a while but it could work...

Originally posted by Mechanic 32
While I understand your basic theory, my question would be this..

What would the materials need to be, to be able to acheive this 50%?

I would think that if the same size parts were used, that the resulting "copy" would be the same size as the original.

Unless I am missing a key point somewhere.

The robot's 'program' would have to give it the necessary instructions to physically break the materials down by 50% each time.

edit to add...

Self replication would be possible in a more normal way though, if the robot were programmed to be able to create or otherwise obtain the necessary parts for the construction of the next, and so on.

This is basically what I meant. The necessary parts would have to be bundled with the first robot.

Now what if you coupled this with true "Learning A. I. "

That would be an awesome feat. Of course the military would most likely seize this for their own uses.

I'm no genius either, so whether it would be truly feasible is questionable.

Agreed!

Originally posted by sardion2000
We first need to create a robot capable of replicated itself regardless of size. Pulling that trick off we could then go and try to use robots to build smaller robots. It's basically the same process, with some parts removed of course, as the one used since the dawn of the industrial revolution.

It looks like there has been some progress in this area. Self-replicating robots

This process of innovation has been speeding up and such a device capable of replicating itself making itself more advanced with each replication would push the pace of progress through the roof. It would be like living at the dawn of the Aviation Age, or the Dawn of Industrialization... except compressed to such an extent that many of us alive today has a good chance of seeing it... It could work... it may take a while but it could work...

As long as it doesn't go down that old The Terminator route...

[edit on 6/2/07 by thebox]

Good thread, btw.

I like threads that make you think outside the box.

(pun intended)

you guys need to have a sit down with Mrs.conor

Oh dear. I hope I'm not dissillusioning anyone here but...

If you were to create a machine 50% it's size each time, each time you create a new machine, that machine would have to use different building materials. The reason being that large machines are better suited to use cruder materials giving it better tensile abilities. As you get smaller, the materials you use would have to be stiffer and denser, otherwise you will reach many points of which the material expands and contracts far too easily for use.

This being said, if you took care of that, the next step is understanding that different sized resistors, transistors, inductors etc act differently. You would have to account for that variation each time aswell.

Operating environment : Smaller machines would require greater sheilding from EM disturbances.

Once you hit the nano-scale level, you will have to start thinking less electronically and more chemically. Most nano machines are simply too small to use transistors etc. You have to start thinking mechanically for its functionality, and chemically for its reactionary propulsion and construction.

Essentially, each size of machine would have to be a whole new design.

Beyond that, if you've gotten that far... now you have to ask yourself... why did we waste all those intermediate steps of creating machines, when you could have used the first one to create the smallest one?

Also, with those machines sole purpose being to create another machine... what would be the end purpose? To say you did it? I suppose thats good enough. Nevermind.

Put simply, it was just an idea.

Oh crap... i've just had a horrible flashback to Stargate and the replicators....