How close are we to self replicating machines?

wrong animation I guess, maybe this youtube.com/watch?v=A8B5MbHPlH0

3d printers printing 3d printers

>3d printers printing 3d printers
yeah, a 3d printer printing itself from scratch would be WAAAY beyond soem dumb machine building somethin

youtube.com/watch?v=C2vgICfQawE

>which would essentially be a factory that can create smartphones and other smartphones factories

lay off the design patterns book

brb, building a pepe gun

>design patterns book
what is that?

self replication is unavoidable by now

consider th eincreasing rates of ai

consider the increasing rate of industrial capabilities

once those two are combined we will enter stragiht into the cornucopia stage of humanity

imagine having 5 billion humans, who arent really humans but are robots who feel nothing and dedicate 100% percent of their energies to pleasing us humans and have the entire resources of the solar system at their disposal

we better start thinking how to think without agression and competition in our minds and better start to learn how to think with creativity and superior purpose as purpose to the existence of our beings

this guy knows

it is an objective truth that having an army of human robot workers would bring unprecedented wealth

if you really want to understand this, check out temarium.com/wordpress/wp-content/documentos/Levy_S-Hackers-Heroes-Computer-Revolution.pdf
read chapter 7

mmmm, do i want to gain true insight and knowledge about a subject to become a wiser person or do i want to ramble like an idiot on an anonymous imageboard so that broscientist may spoonfeed me things that are only barely related to reality

bump

>>How close are we to self replicating machines?
simple ones have been demonstrated in the lab. A group at John Hopkins demonstrated a robot which could construct a copy of itself from special parts made of legos and magnets:
molecularassembler.com/KSRM/3.23.4.htm

Pic related is the group's most recent work, an attempt to make something like Von Neumann's universal constructor in the real world. It can construct almost anything, "in a universe consisting of modular blocks." It is pretty close to making a copy of itself too. If the blocks were made of something more rigid it probably could.

But what is really cool about it is that it's made from components that can be made from simple manufacturing processes that it could potentially carry out. In addition the whole system is designed to be simple enough to be controlled by a relay computer, something it could potentially construct.

In short, this is the closest we have come to so called "parts closure," that is making a machine that can manufacturer all of its components.

Here are some more details on the manufacturing part:
roboticsproceedings.org/rss05/p16.pdf

Here is a video that talks about some of the work this group has done:
youtu.be/b04X0xsdjLg

Also worth watching is the part about the groups work on robots which can self diagnose themselve to figure out what is broken and then repair themselves. Which may be necessary to build a self replicating machine.

Another cool thing that is currently happening is someone is working on a reprap controlled with 3d printed relays:
relayreprap.srm.org.uk/

>>is any of these happening right now?
so to answer your question some of this is happening right now. No we don't need nanotechnology, but it would sure help.

Will post more tomorrow if this thread is still here

please go fellate a phallus.
DARPA has a shit budget and they don't really 'deploy' technologies. They are more concerned with 'strategic surprise.' IE finding out if new technologies present a threat or are worth further investigation so we could surprise people with them.

>>Self-replicating nanobots are impossible in terms of energy, chemistry, mechanically, and computationally.
Do you have any facts to back that up?

>>Wouldnt that mean a radical change in the way society is structured?
not necessarily. Replication rate could be low. DRM could be used to control what self replicating machines could produce. Developing countries might not have the necessary technical expertise to utilize self-replicating factories to make stuff like smartphones.

For example, smartphones may be made in china, but all the chips are designed in the US, europe, etc.

Bump

are we close to a machine that given the raw materials can quickly manufacture complex chips and electronic components?

I don't know whether we can or can't do it, but I'm attempting to test the viability of the soil metalic extraction method detailed here researchgate.net/publication/222331160_Exponential_growth_of_large_self-reproducing_machine_systems

Could it work? Maybe. I hope so. The paper claims it could be possible to do it right now, even if its a decades long project. Let's hope they are right.

Semiconductor fabs are almost completely automated. The only reason we have humans in them is to stop things from fucking up. Because even a single dropped box of wafers could be worth thousands of dollars.

>> complex chips
For near term self replicating machines, complexity needs to be kept as low as possible. If said self replicating machine is making microchips it would be best if they were making simple ones with big micron scale transistors.

Before we can do microchips it's more important to demonstrate that automatic assembly, fault diagnosis, and repair can work.

Building self replicating machines just to construct smartphones is a bit silly right now. A much better application is turning deserts into fields of solar cells. Another is bootstrapping industry in space.

You've probably seen one of my previous threads. You should contact the authors of that study, they supposedly designed a small vacuum furnace for carrying that out.

That is very ambitious though.

confirmed for never going to make significant scientific contribution to the world

They already exist. Viruses and computer-robot "processor" virus-like items. Such items can be stopped or destroyed with radio technology and computer programming. I'd prevent production of devices like self replicating machines by writing easily executable programs as attachable to USB radio technology or an iPhone application like that.

I don't have the material or metalurgical knowdlege to carry out the precise method carried out. I should have clarifyied that I'm trying to test whether it is possible to obtain the resources detailed on it on european atlantic soil, using an arc furnace. And I also knew of this paper a long time ago, but forgot about it until your thread made me remember it.

You want to know if it works with european soil? Process is designed for sand. So as long as you got sand you're good.

I know but I'd like to study the viability of other soils, including sand, probably both from cliffs and beaches., of course.

Look up composition of soil in question. Does it have the required elements in it? Done. The process is made to work with a variety of soil compositions.

Though I don't know why you'd want said replicating machine to work on cliffs and beaches.

/thread

>Semiconductor fabs are almost completely automated.
but could you have a factory that is able to build an entire semiconductor factory by itself?

You could. It would just need to be specially designed so that it could be constructed by robots. Like making it out of blocks like in Although we really need a good understanding of how to do autonomous repair and fault diagnosis before we can consider such things. Otherwise, we'd have to scrap the entire semiconductor fab when it breaks down.

Demonstrating the basics is much more important than the details.

that would be a real game changer, because if a factory can manufacture itself then it cna for sure make most of modern appliances, so imagine having one oft hsoe factories, you feed it all the shit it needs to make another factory then you take it to an undeveloped country, feed it shit there and take it anywhere

very soon you have all appliances needed in that country

imagine that

a giant self replicating factory capable of building the inmense majority of consumer goods society needs, from a toolbox to a car or an mri machine (not a rocket or a tank tough)

its just a matter of making one then distributing it around the glove is anyone working in something similar to this?

They're never coming, faggot.

for your ifnormation its physically possibel to have self replicating factories and no poverty

cool stuff!

I don't see it happening any time soon. The only scenario I can imagine it working is if it's controlled by a highly advanced AI and can do everything a human can do. i.e. humanoid robots. In which case humans may as well do it because let's face it, human life is cheap compared to such a complex machine.

Nano-technology like in OP's picture that involves microscopic robots doesn't exist, and will most likely never exist. Sorry, that's just not at all what modern nanotechnology is about, and we're not even close to having even a theory about how that technology would work or be created practically. Bio-engineering of living things may be more realistic, but even that is in its infancy.

Macro-scale self replicating machines may be possible with help from humans. But truly independent machine built infrastructure won't be possible unless humanoid robots are common which creates other risks i.e. I, Robot, The Matrix, Terminator, etc. Not a good idea.

why will technology die?

Because people like that poster hate having nice things
It triggers their autism

the pic implying that technology will reach the point where it will be our undoing i.e. we might bomb ourselves into oblivion

No one is working on stuff like that. At least not directly.

A self replicating factory may not be able to make any appliance. You might be able to make appliances, but they may need to be made from blocks.

One of the things about the proposal in is you are making everything from common elements, designing stuff so that it replicates as fast as possible. Consumer products might be made from said common elements, but they probably won't perform as well as their mass produced counterparts today.

A washing machine made from ceramic(common), using aluminum(common) windings, is probably going to be worse than a washing machine made of steel using copper(not so common) windings.

It seems that you want massive automation rather than self replicating factories. Again, we need to do more basic research before we can realise such a thing.

Wet dream of astronomers: an explorer which inspects a planet, replicates and sends 5 copies for further exploration.

you mean a von neuman probe? the concept exist, check it out search von neuman probe in wikipedia

>It seems that you want massive automation rather than self replicating factories.
both go hand in hand

what i want is a gadget that eliminates the dependence of non industrialised areas from first world countris

like, you take one of my clever self replicating factories to africa, and all they need is like, raw material, you just have to supply it with, idk, sand, steel, whatevs

yeah it might be a bit of work to haull the steel over to the factory but the benefits will be hugemongous, i dont know why noone else is thinking about this

Bump

Well nevermind the feasibility. Lets think about the economics. how disruptive would this be, a self assemblying factory that can produce most electronic appliances and vehicles we use in modern society and can also reproduce itself. Imagine you only had to supply it with raw materials

>not showing this
youtube.com/watch?v=xP5-iIeKXE8

Disappointed

Not as disruptive as you would think. It would put factory workers out of a job for sure, but it probably won't make everything free.

Take for instance software and videogames. It costs practically nothing to distribute software and videogames, but such things aren't necessarily free. There is a cost to develop software. So even if raw materials and energy are free there will still be a cost to developing appliances and what not. Because nonindustrialized countries aren't that well educated, they would likely rely on developed countries for the development.

Second, raw materials and energy still costs something.
If you have to provide it with raw materials what is the point? Such a self replicating factory would then be limited by how fast you can provide raw materials to it. Self replicating would mean that you need to provide exponentially increasing amounts of raw materials. Self replication doesn't benefit you much in this case.

Now steel is a bad example, as iron is a very common element in Earth's crust at 6% of earth's crust. Copper is a much better example at 0.0068%.

Of course if your self replicating machine is big enough and can harvest energy on it's own you could potentially extract reasonable amount of copper per unit time from regular soil.

>If you have to provide it with raw materials what is the point?
taking shit from point a to point b is a simple extraction economy, any poor country can manage that

but having a complex technological network of factories and god tier colleges is something that only ultra developed countreis can affoard

I really do not understand this obsession with humanoids. Why must said robot have a human form to accomplish manufacturing and assembly tasks?

Who designs the appliances and programs the factories to make them? You probably need some 'god tier colleges' to train people to so this.

>Who designs the appliances and programs the factories to make them? You probably need some 'god tier colleges' to train people to so this.
If you start out having all of the designs for current techonology you can have a pretty nice standard of living, everything extra is additional

anyway, once you have this cornucopia of abundanzzia its much easier to develop god tier colleges of your own

is that a self replicating space machine?

ITT:
people who haven't read The Diamond Age

Well we have a concept in FSMs. That oughtta correlate to reality somehow, right?

>self replicating
Looks like a scaffold generator to me.

what is this work of fiction supposed to tell us?

If you stretch the definition of self replicating machine yes. creativemachines.cornell.edu/papers/ITRO07_Zykov.pdf

It's a modular robot.

That Stephenson is a one trick pony

Bump

an AI could make copies of itself, but it probably couldn't write a program similar to itself without looking at its own code
so I suppose if you built a robot and gave it all the information and tools it needed to create its hardware, which it would then copy its software into, then it's plausible. The catch, in my opinion, is that any error the first robot makes will get worse with each generation of copies, because it isn't capable of self-improvement.

Bump

but is it turing complete

>obsession with humanoids
We live in a human world where most infrastructure and machinery necessary to complete even simple tasks must be performed by a human. It would take a lot more time to design heavy machinery that is entirely machine operated. The logistics of transporting materials alone requires a physical person to sign for a delivery. It's the really simple things that are being overlooked by people like you.

i agree with this thing resembling a human but mostly being composed by neckbeard.

Like, anything thats not a humanoid robot would still have heavy competition from human labour

>self replicating machines
>implying a virus is not
Lrn2virus fgt pls

>How close are we to self replicating machines?
You mean, grey goo? In a certain sense, it's already been done, billions of years ago. In popular culture, it's commonly known as "green" instead of "grey". It's life.

>Wouldnt that mean a radical change in the way society is structured?
Maybe. Depends on the particulars. My guess is probably yes, and drastically so.

Yep. Probably right. Stupid physics.

But you don't need a human form to do all these tasks.

Second what sort of heavy machinery needs a humanoid to drive it? Are you going to have humanoid robots driving bulldozers around, manually operating milling machines, using power drills made for humans, etc? Just taking a factory and replacing every human with a humanoid robot is dumb.

Robots are fucking dumb. Tasks we take for granted are extremely difficult for robots. Picking up a power drill with a hand is a much harder task than redesigning the drill to use a quick change connector. Simply programming a humanoid robot to get into a bulldozer is a much harder task than designing a robotic bulldozer from scratch.

Second why do you need a robot with legs in a factory? Legs don't really buy you much in a flat factory floor.

>We live in a human world where most infrastructure and machinery necessary to complete even simple tasks must be performed by a human. It would take a lot more time to design heavy machinery that is entirely machine operated.
>The logistics of transporting materials alone requires a physical person to sign for a delivery.
This is by far the dumbest argument I've heard on Veeky Forums in a while.

Robobump

Robodump

You need human-like kinematic abilities to do any maintenance whatsoever on any of those machines, all of which break down constantly.

But you don't need a humanoid form to obtain human-like kinematics. All you really need are at least two manipulators floating in space. These can be held up by traditional serial arms, snake robots, fractally branching arms, or those cube bots like in .

Pic related, a snake bot snaking into stuff.

Do you have anything to backup the statement that those machines break down all the time? Yes the tools wear out, but there are automated systems for replacing them. The company fastems has some clever ways of managing this. Industrial robots can be very reliable with MTBFs of 100,000 hours. In order for automation to be profitable over humans, you need highly reliable machines. If you need maintenance all the time, it's hard to justify the cost.

THIS

people often forget, humans arent the most efficient design possible...

they are the most efficient design possible... that could arise from evolution which has pretty limited constraints

for example, theres a reason there are no wheels in nature altough they are efficient, they do not provide a progressive advantage, unlike limbs

an animal with a bit of a wheel, or an almost wheel would be quickly eliminated, it would have to suddenly have a perfect wheel and thats impossible in terms of evolution

the same is with this

we are applying the designs that evolution wasnt able to

This post sums up everything wrong with the average Veeky Forums user

>scifi
Or
Science and robotics are interesting and you don't know enough about either to talk about them

>you don't know enough about either to talk about them

Who exactly do you think you're replying to? The OP?

Anyone who thinks this is a good thread.

>
> This post sums up everything wrong with the average Veeky Forums user
Not yet... I swear I saw the horse move a bit... beat it some more!

What did he mean by this?

We have discussed AI about 100000000000 times.

Ends up being:
1. never going to happen, poster are HS dropouts
2. my singularity/immortality
3. gimme sexbots
4. robot war with humans and win

4. Everyone is an scifi

I will keep bumping this thread until it dies.

>How close are we to self replicating machines?
Bacteria is ancient as fuck.

>scifi
whats wrongw ith that? technological advancement from humanity comes from scientifical inspiration from superior artistic thinking

actually about 93% of techonological advancements come from science fictions, like the satellite or the steam engine

> steam engine
Source?

We have a board for sci-fi

You are somewhat right, but sometimes they totally get it wrong... reading old TOM SWIFT books is hilarious.
Scientist pulls out his slide-rule to check some calculations
I SERIOUSLY doubt anyone under 50 would know how to use a slide-rule

Or asimovs psychic robots

damninteresting.com/on-the-origin-of-circuits/

>psychic robots
what is wi-fi

Robots communicating with other robots?

Not reading peoples minds and turning them to rubber via psychic attacks

>Not reading peoples minds and turning them to rubber via psychic attacks
youd dint understand the book, i suggest you finish high school before reading anything harder than a D&D manual

Where did you get this information from?

What book do you think I'm talking about?

>harder than a D&D manual
That's a pretty large range, all things considered.

>a radical change in the way society is structured
These never happen.

French Revolution
Russian revolutions
The Great Wars...etc