Are replicators real?

[Seaboe Muffinchucker]

Somersaulting MIT cube robots can self-assemble.

Seaboe

[deathil93]

That's far from an actual replicator, but cool nontheless. Kinda creeps me out as well for some reason.

[Type-8]

those things are awesome, but i dont see them conquering us anytime soon

[Energizer_Vs_ZPM]

No.

[Aesop]

Somersaulting MIT cube robots can self-assemble.

Seaboe

Ha, I did a presentation on 4D printing and self-assembly over the summer, and it definitely makes me think of the Replicators.

[Peterking72]

"OMG we need to kill them before they advance this technology even further!"

OK just kidding, but creepy indeed.

[tpn]

It is technically possible to create a Replicator, although with current technology it wouldn't replicate unless you programmed it to.

[hedwig]

I read articles like this and immediately think of replicators in SG1/SGA, and think noooooooooooooo.

[AtlantisStargate]

I don't think so, not yet anyways.

[tpn]

The closest thing we have to a Replicator right now is the cellular automaton "replicator B1357/S1357"

[tpn]

Replicators are real. some people have made self-replicating technology.

[Anja]

No, not in the way they are shown in Stargate.
The technology is one thing, the self-awareness and ability to act (not react) is another one.
Machines do what they are programmed to do - nothing more. The more advanced the program the more advanced the deeds - but that's not thinking of sth new by the machine.

[thekillman]

Machines do what they are programmed to do - nothing more. The more advanced the program the more advanced the deeds - but that's not thinking of sth new by the machine.

Obviously we aren't there yet, but don't you think it's possible to create a machine that's truly self-aware? not just a clever statistical program like our current AI? Self-awareness exists in humans, and we're essentially a complex chemical vat. Don't you think we can recreate that with code?


Secondly, the MW replicators were not self-aware. They were slaves to their own programming.

[Michael Jansky]

Obviously we aren't there yet, but don't you think it's possible to create a machine that's truly self-aware? not just a clever statistical program like our current AI? Self-awareness exists in humans, and we're essentially a complex chemical vat. Don't you think we can recreate that with code?

Research is ongoing in neurology and neuropsychology into exactly how the human brain produces what we call "consciousness". The question is considered very hard (hence it is called the "Hard Problem of Consciousness"), and in fact we are currently unsure whether or not a Turing machine (a classical computer) is capable, even in theory, of reproducing brain function.

However, in general, I would be inclined to agree that if consciousness can arise on a biochemical substrate (i.e., the brain), no matter how complex, then it should, in principle, be possible to recreate it on some other information-processing system. That such a system could be artificial does not seem like an obstacle in principle.

Nonetheless, in practice, I think we are still very far from being able to create an information processing system capable of what we habitually call self-awareness.

I recall seeing an estimate in some scientific text that maintained that a single human neuron (brain cell) has a maximum data bandwidth of ~1 Gbps. The brain has on the order of 10^11 (hundred billion range) neurons, and they are connected in a manner much more complex than the 2-to-1 connection of transistors (i.e., each transistor is connected to 2 others) in a microchip -- some brain cells can have dozens to hundreds of connections to other brain cells. No one has concrete numbers, but it appears that the information processing capacity of an average human brain is orders of magnitude beyond even the most powerful supercomputer (although humans can't calculate nearly as fast, that's not the most relevant measure here).

And raw processing power aside, the ways in which current computers can process information are way, way more primitive than the ways the brain can. That's probably a bigger challenge than getting enough binary switches (Moore's law continues to soldier on, and processing power is not currently viewed as a limitation to much of anything in the long term).

[Anja]

Obviously we aren't there yet, but don't you think it's possible to create a machine that's truly self-aware? not just a clever statistical program like our current AI? Self-awareness exists in humans, and we're essentially a complex chemical vat. Don't you think we can recreate that with code?


Secondly, the MW replicators were not self-aware. They were slaves to their own programming.

No, I don't think so.
You're right, self-awareness was the wrong word, but the reps did more than merely reacting according to their programming. They could change strategy and they 'knew' what to look for and how to manipulate machinery.