Von Neumann machines - are they right?
Moderators: Tunnelcat, Jeff250
- Mobius
- DBB_Master
- Posts: 7940
- Joined: Sun Jun 03, 2001 2:01 am
- Location: Christchurch, New Zealand
- Contact:
Von Neumann machines - are they right?
http://www.mkaku.org/articles/physics_o ... civs.shtml
I've referred to this URL before - but only to point out that we aren't even a Type 0 civilisation.
What concerns me is the idea of a Von Neumann machine made by humans. The concept of it is rapidly approcahing on a nano-scale, and if we survive that, we might have some confidence in our ability to make "safe" Von Neumann machines.
In the last week, a very interesting article has appeared which suggests self-desassembly for Von Neumann machines (nanobot replicators).
The concept of Grey-Goo raises it ugly head when you invent a nanobot which can replicate itself: they, very rapidly, consume the entire surface of the Planet, and convert EVERYTHING into copies of themselves.
I think in the absence of imminent Nuclear global destruction, this is our next, worst, threat - and some estimates place the likelyhood of humans being extinct in 50 years as high as 50%.
Nanotechnology is definitely coming: yeah - some of you may not believe it, but that's not going to stop it happening. The potential for good is simply too high to ignore, and the potential for extinction - and every option in between is also there too. You and I will see, in our lifetimes, miracles performed - because all sufficiently advanced technology appears as magic. (Arthur Clarke, 1960)
However, if our nanotech gets away on us - then our planet will be Grey Goo - and perhaps that's not such a tradegy from one perspective. We're basically a plague on the face of the Earth, and if we aren't killed off soon, we'll be a plague on the Galaxy in a lifetime coming to you soon.
It might be OK to sterilise your own world - but it's pretty bad manners to sterilise a galaxy, or even the Universe!
The idea of a Von Neumann machine is seductive, because it solves all your problems. You no longer have to DO anything, except wait for them to send you signals and information about the Galaxy at large. That in itself has numerous implications.
But consider this: any machine smart enough to reproduce itself is alive, and although we might make them the machine-equivalent of asexual (copies only) the natural process of reproduction is eventually going to produce errors. It doesn't make one shred of difference how careful you are, or how many failsafe's are built in, eventually they'll be "mutations" that survive to reproduce. It might take millions (or even billions) of years for this to occur. (But, by then will we care?)
Initially these mutations might only make them more efficient at their jobs, but sooner or later, the errors will compound such that the original mission of these devices is corrupted. Now there's a breed of Von Neumann machines which have a new purpose, and that purpose might only be to make more of themselves! (or something slightly less immediate but still as dangerous.)
How long would a galaxy last in the hands/claws/grippers of such things?
I don't know how many civilised space faring cultures there are in our galaxy (Drake says about 50, but I think less than this), but I do know, that if they get together to trade and create wealth, then they'll have only one hard and fast rule: No Von Neumann machines.
Creating Von Neumann machines would cause the eventual extinction of all species in a galaxy if allowed to pursue their goals.
The fine point in all this - is that by definition we are virtual/biological Von Neumann machines designed to do nothing except replicate and replicate again. The unavoidable evolution of Von Neumann machines maps directly to the unavoidable evolution of man kind. The onyl differences are that humans are less durable, breed slower and are easier to kill than VN machines.
Perhaps this is in our favour, galactically, as we're unlikely to be considered a threat.
However, if we create these machiens and send them out into the galaxy, we might be signing the death warrant on the human species. Maybe Type 2 civilisations might look kindly on emerging Type 0s - but on the other hand, maybe when the new kids on the block break the rules they simply get exterminated, and the Type IIs have to spend countless energies tracking down and destroying every last VN machine. Because if you miss just one then you might as well not have bothered to destoy any of them.
Devices must never be allowed to self replicate independently, and must expire naturally in a short period of time.
What do you think?
I've referred to this URL before - but only to point out that we aren't even a Type 0 civilisation.
What concerns me is the idea of a Von Neumann machine made by humans. The concept of it is rapidly approcahing on a nano-scale, and if we survive that, we might have some confidence in our ability to make "safe" Von Neumann machines.
In the last week, a very interesting article has appeared which suggests self-desassembly for Von Neumann machines (nanobot replicators).
The concept of Grey-Goo raises it ugly head when you invent a nanobot which can replicate itself: they, very rapidly, consume the entire surface of the Planet, and convert EVERYTHING into copies of themselves.
I think in the absence of imminent Nuclear global destruction, this is our next, worst, threat - and some estimates place the likelyhood of humans being extinct in 50 years as high as 50%.
Nanotechnology is definitely coming: yeah - some of you may not believe it, but that's not going to stop it happening. The potential for good is simply too high to ignore, and the potential for extinction - and every option in between is also there too. You and I will see, in our lifetimes, miracles performed - because all sufficiently advanced technology appears as magic. (Arthur Clarke, 1960)
However, if our nanotech gets away on us - then our planet will be Grey Goo - and perhaps that's not such a tradegy from one perspective. We're basically a plague on the face of the Earth, and if we aren't killed off soon, we'll be a plague on the Galaxy in a lifetime coming to you soon.
It might be OK to sterilise your own world - but it's pretty bad manners to sterilise a galaxy, or even the Universe!
The idea of a Von Neumann machine is seductive, because it solves all your problems. You no longer have to DO anything, except wait for them to send you signals and information about the Galaxy at large. That in itself has numerous implications.
But consider this: any machine smart enough to reproduce itself is alive, and although we might make them the machine-equivalent of asexual (copies only) the natural process of reproduction is eventually going to produce errors. It doesn't make one shred of difference how careful you are, or how many failsafe's are built in, eventually they'll be "mutations" that survive to reproduce. It might take millions (or even billions) of years for this to occur. (But, by then will we care?)
Initially these mutations might only make them more efficient at their jobs, but sooner or later, the errors will compound such that the original mission of these devices is corrupted. Now there's a breed of Von Neumann machines which have a new purpose, and that purpose might only be to make more of themselves! (or something slightly less immediate but still as dangerous.)
How long would a galaxy last in the hands/claws/grippers of such things?
I don't know how many civilised space faring cultures there are in our galaxy (Drake says about 50, but I think less than this), but I do know, that if they get together to trade and create wealth, then they'll have only one hard and fast rule: No Von Neumann machines.
Creating Von Neumann machines would cause the eventual extinction of all species in a galaxy if allowed to pursue their goals.
The fine point in all this - is that by definition we are virtual/biological Von Neumann machines designed to do nothing except replicate and replicate again. The unavoidable evolution of Von Neumann machines maps directly to the unavoidable evolution of man kind. The onyl differences are that humans are less durable, breed slower and are easier to kill than VN machines.
Perhaps this is in our favour, galactically, as we're unlikely to be considered a threat.
However, if we create these machiens and send them out into the galaxy, we might be signing the death warrant on the human species. Maybe Type 2 civilisations might look kindly on emerging Type 0s - but on the other hand, maybe when the new kids on the block break the rules they simply get exterminated, and the Type IIs have to spend countless energies tracking down and destroying every last VN machine. Because if you miss just one then you might as well not have bothered to destoy any of them.
Devices must never be allowed to self replicate independently, and must expire naturally in a short period of time.
What do you think?
If you are WTF Mobius is talking about, AKA to all Dues Ex fans as Universal Contstructors.
If anything, we'd wipe ourselves out before we get to a point of being a threat to other planets.
self-expiry is a good idea.
this discussion also presumes that nanites would lack intellegence enough to able the adhere to the three rules of robotics.
If anything, we'd wipe ourselves out before we get to a point of being a threat to other planets.
self-expiry is a good idea.
this discussion also presumes that nanites would lack intellegence enough to able the adhere to the three rules of robotics.
You're going off the assumption that there is even any other civilizations in the galaxy as advanced or more advanced then us. Since we have no evidence of this it is all purely speculation.
How can we follow the rules of the playground if there is no playground rules? If you get my point. We've been given no guidelines to adhere to, so its safe to assume the guidelines are our own to make for now.
As far as nano-technologies, it really isn't any of your concern Mobi, because I don't think the technology will be able to be what it is intended to be until even long after you and I are dead. 50 years puts me at about 72 going on 73 and though I hope I live that long there are no garuntees. Either way we really don't have to worry about it right now because by the time it gets to the point where its a serious possible danger we won't be around to care.
Besides, we program them, they would operate under our control. I would hope the people who created them would have the sense make them do only the objectives they're programmed to do and nothing more. Like, being injected into a human being, fix whatevers wrong with the person then be extracted and disposed of.
Either way it will all come down to a matter of money and personally I don't think the technology put into building that kind of machine will be cost effective. It isn't now and with inflation the way it is, I doubt it will be in the future.
The other problem is how will they generate the materials needed to replicate themselves not to mention replicate organic material all synthetically? (because for the moment their primary uses would be medical, atleast from what I've read).
Lots of little problems for a little machine to even get started.
- Will Robinson
- DBB Grand Master
- Posts: 10124
- Joined: Tue Mar 07, 2000 3:01 am
Yeah Topher, it would be pretty cool, but look what happened in Freespace. We screw up one conversation with the first sentient species we come across, and *POOF*--14 years of galactic war. Kinda makes you think, doesn't it...
Mobius, I have little to no idea what you're talking about. What exactly is a Von Neumann machine? I have neither the time nor the will to pick through that entire article. Also, calling your own species a "plague" may be the sign of a mental health issue. Seek professional help immediately.
Mobius, I have little to no idea what you're talking about. What exactly is a Von Neumann machine? I have neither the time nor the will to pick through that entire article. Also, calling your own species a "plague" may be the sign of a mental health issue. Seek professional help immediately.
- Nightshade
- DBB Master
- Posts: 5138
- Joined: Sun Jun 17, 2001 2:01 am
- Location: Planet Earth, USA
- Contact:
<BLOCKQUOTE><font size="1" face="Arial">quote:</font><HR><font face="Arial" size="3">
What's to say we aren't the only advanced creatures, or at least the most advanced. That would explain why we haven't been visited by any space traveling aliens.
</font><HR></BLOCKQUOTE>
A better reason is that civilizations, being EXTREMELY rare and more than likely hundreds if not thousands of light years apart, haven't even heard us yet. Our first broadcasts have only reached out maybe 50 light years away...and those are very faint... Military radars would probably be the first real signals we're broadcasting (unintentionally) and those probably haven't reached more than 30 or 40 light years out.
There are even more reasons for not having seen ET or even heard from ET, but they're quite numerous.
I'm pretty sure they're out there...it's just that they're VERY VERY far away.
What's to say we aren't the only advanced creatures, or at least the most advanced. That would explain why we haven't been visited by any space traveling aliens.
</font><HR></BLOCKQUOTE>
A better reason is that civilizations, being EXTREMELY rare and more than likely hundreds if not thousands of light years apart, haven't even heard us yet. Our first broadcasts have only reached out maybe 50 light years away...and those are very faint... Military radars would probably be the first real signals we're broadcasting (unintentionally) and those probably haven't reached more than 30 or 40 light years out.
There are even more reasons for not having seen ET or even heard from ET, but they're quite numerous.
I'm pretty sure they're out there...it's just that they're VERY VERY far away.
- Mobius
- DBB_Master
- Posts: 7940
- Joined: Sun Jun 03, 2001 2:01 am
- Location: Christchurch, New Zealand
- Contact:
Note: Our radiosphere now occupies 40,000 cubic light years of our galaxy. If you plot the size of it against the Galaxy it is a NON-TRIVIAL portion of it - and easily visible at a sensible scale.
Your assumptions about future tech are dead wrong. In the first 10 years of this century, we will DOUBLE the total amount of knowledge acquired in the whole 20th century. 2010-2020 will see it double again. So, by 2025 we will have approximately 5 times the knowledge we had in 2000.
Make no mistake - technology follows the law of accelerating returns - and Nanites/nanobots (whatever) - if they are possible, will DEFINITELY be manufactured before 2025. The self-assembling replicator (VN) type may be as much as 10 years later - but I suspect 5 years later - because the first true AIs will be born at about the same time.
Yet another threat to human existence. We better make sure we get the first AI right - because it'll be the one and ONLY AI we ever get to create. (After that, the first-gen AI design generation 2 etc).
You and I are going to witness the Singularity ( http://www.kurzweilai.net ) and by that time technology will be advancing at a rate which is too fast for humans to comprehend.
This assumes of course we're smart enough to prevent our inventions killing us all - accidentally, or by design!
Your assumptions about future tech are dead wrong. In the first 10 years of this century, we will DOUBLE the total amount of knowledge acquired in the whole 20th century. 2010-2020 will see it double again. So, by 2025 we will have approximately 5 times the knowledge we had in 2000.
Make no mistake - technology follows the law of accelerating returns - and Nanites/nanobots (whatever) - if they are possible, will DEFINITELY be manufactured before 2025. The self-assembling replicator (VN) type may be as much as 10 years later - but I suspect 5 years later - because the first true AIs will be born at about the same time.
Yet another threat to human existence. We better make sure we get the first AI right - because it'll be the one and ONLY AI we ever get to create. (After that, the first-gen AI design generation 2 etc).
You and I are going to witness the Singularity ( http://www.kurzweilai.net ) and by that time technology will be advancing at a rate which is too fast for humans to comprehend.
This assumes of course we're smart enough to prevent our inventions killing us all - accidentally, or by design!
<BLOCKQUOTE><font size="1" face="Arial">quote:</font><HR><font face="Arial" size="3">Originally posted by Mobius:
Your assumptions about future tech are dead wrong. In the first 10 years of this century, we will DOUBLE the total amount of knowledge acquired in the whole 20th century.</font><HR></BLOCKQUOTE>
Fascinating how you can quantify "knowledge."
It's 2004. We're almost midway through this decade. Where are groundbreaking inventions equivalent to the personal computer? The Internet? Color television? The automobile?
Your assumptions about future tech are dead wrong. In the first 10 years of this century, we will DOUBLE the total amount of knowledge acquired in the whole 20th century.</font><HR></BLOCKQUOTE>
Fascinating how you can quantify "knowledge."
It's 2004. We're almost midway through this decade. Where are groundbreaking inventions equivalent to the personal computer? The Internet? Color television? The automobile?
- Sergeant Thorne
- DBB Material Defender
- Posts: 4641
- Joined: Sun Nov 25, 2001 3:01 am
- Location: Indiana, U.S.A.
- Testiculese
- DBB Material Defender
- Posts: 4689
- Joined: Sun Nov 11, 2001 3:01 am
Out species is not only a plague, it's a virus, TopGun. A pretty nasty one too.
We will make more progress over the coming decades than you think, Sol, because of the exponential increase in the paradigm shift rate (rate of progress). We're doubling the rate of progress every decade, so the next 30 years will be like 140 years of progress at today's rate.
And as far as inventions, there have been many. The 'groundbreaking' ones you mentioned have already been achieved, and everything else related is innovation. The next groundbreakers will be space navigation and biotechnology, I think. That's when we'll see some fun stuff!
I can't wait to see it, if we or our stupid government don't kill ourselves off first.
We will make more progress over the coming decades than you think, Sol, because of the exponential increase in the paradigm shift rate (rate of progress). We're doubling the rate of progress every decade, so the next 30 years will be like 140 years of progress at today's rate.
And as far as inventions, there have been many. The 'groundbreaking' ones you mentioned have already been achieved, and everything else related is innovation. The next groundbreakers will be space navigation and biotechnology, I think. That's when we'll see some fun stuff!
I can't wait to see it, if we or our stupid government don't kill ourselves off first.
Uh, Mobi, do you even know what a Von Neumann machine is?
It's a concept that forms the basis for all computers, it's not some kind of microscopic evil robot.
The Von Neumann machine is basically stuck in an endless loop while it is powered on. It will read input in a loop, and will either read and execute an instruction or else a NO-OP (nothing to do).
That's why you can leave your PC on without anything to do -- it will do a lot of NO-OP instructions.
<code>while (true) {
instr = read_next_instruction();
if (instr == NO_OP) continue;
else execute_instruction(instr);
}</code>
That's all Mobi. Stop having nightmares, will ya?
It's a concept that forms the basis for all computers, it's not some kind of microscopic evil robot.
The Von Neumann machine is basically stuck in an endless loop while it is powered on. It will read input in a loop, and will either read and execute an instruction or else a NO-OP (nothing to do).
That's why you can leave your PC on without anything to do -- it will do a lot of NO-OP instructions.
<code>while (true) {
instr = read_next_instruction();
if (instr == NO_OP) continue;
else execute_instruction(instr);
}</code>
That's all Mobi. Stop having nightmares, will ya?
a nanomachine virus that's main purpose is to replicate itself as fast as it can using all resourses at it's disposal doesn't sound too dangerous to me. sure it's freaky, but do you realise how SLOW it would move? for it to move at any speed faster than BACTERIA GROWTH: the machines would have to work together as a collective, thereby forming a multicell organism to speed movement.
why i say this is because a micro-organism like a nanomachine can only move at micro-organism speeds, the speed at which it moves does not increase exponentially as it assimilates more resourses.
if you were really designing a nanomachine virus to wipe out an entire planet you woudl have to design it to use more efficient movement techniques. either form together as a multicellular machine, or possibly ride the air winds, or form a temporary parasitic bond with the wildlife (ie: have a stage of dormancy and slow self replication + seeding for a while before killing the animal).
if there was an entire planet made of a pourus mix of meat jelly (you know that stuff they use to grow bacteria in labs) and oxygen, and you fired a "snot" of some bacteria into it and came back in a year, the whole planet would not be eaten. the bacteria would probabaly still be confined to a small section, because (assuming it hasn't found some way of seeding itself long distances, like using wind) it can only move as fast as each individual cell can.
why i say this is because a micro-organism like a nanomachine can only move at micro-organism speeds, the speed at which it moves does not increase exponentially as it assimilates more resourses.
if you were really designing a nanomachine virus to wipe out an entire planet you woudl have to design it to use more efficient movement techniques. either form together as a multicellular machine, or possibly ride the air winds, or form a temporary parasitic bond with the wildlife (ie: have a stage of dormancy and slow self replication + seeding for a while before killing the animal).
if there was an entire planet made of a pourus mix of meat jelly (you know that stuff they use to grow bacteria in labs) and oxygen, and you fired a "snot" of some bacteria into it and came back in a year, the whole planet would not be eaten. the bacteria would probabaly still be confined to a small section, because (assuming it hasn't found some way of seeding itself long distances, like using wind) it can only move as fast as each individual cell can.