[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index] 

Re: machines vs. living systems (was Re: J. Craig Venter - frabrication: in stores now!?!)

Dan Fiscus wrote:
PS - I saw your website - it was swarm; and I was also born in Houston, TX! Small world... 

Yep!  I thought I remembered your name; but, I've met so many people I
often get confused as to who I've met.  Plus, my imagination does a good
job creating pseudo-visuals for the people I interact with online... So,
I'm continually in a state of curiosity about whether I've met people
before.  It forces me to be more open-minded than I might otherwise be.
[grin]

Could you design, create, imagine a machine that could truly surprise
you (open realm of behavior)? And can you imagine a life form that would never surprise you (closed realm of behavior)? 

Yes on both counts.  Surprise is, it seems to me, purely a subjective
phenomenon.  I've seen, otherwise intelligent, humans look at a cloud
formation or a trail of ants and think nothing more than "It may rain"
or "I hate ants", thereby showing no more curiosity or interest than a
machine would.  That happens to me _alot_.  (And, being one of my fellow
men, I'm _sure_ that I've done that at times.)  Granted, that's usually
a short window, however.  So, your use of the word "never" comes into
play.  Can we imagine a life form that would _never_ surprise us?  I
think we can because all it would take is for me to become non-curious
or non-interested.  If, for example, I became chronically depressed and
stayed that way until I died.  Then no life forms would surprise me
because I wouldn't put enough effort into thinking about (modeling)
them.  (I realize this may not be the direction you were headed... but,
my point is that "surprise" is purely subjective and not very useful,
really.)

As for creating a machine that surprises me, yes, absolutely!  In fact,
I do this all the time. [grin]  As a programmer (and even as I rekindle
my fondness for electronics), the machines I create surprise me all the
time, sometimes in a good way and sometimes in a bad way.

Lastly, a short quote from an article I just got today by Michael Conrad:
[...]
I can give the cite if interested... 

Yes!  I would like the citation, especially if I can find an electronic
copy.

Dan Fiscus wrote:
Autotrophic versus heterotrophic functional types - can you imagine an animal that can eat sunlight directly? 

I don't really understand this.  It seems like you're suggesting that
complex animals need complex food and, therefore, a distinction between
a machine and an animal can be made via distinguishing the complexity of
the food.  If that's the point, then I can catch a glimpse of it.  But,
it seems fairly simple to come up with counter examples like a gasoline
engine, which is a machine that is powered by things like oil and
gasoline, both of which are complex (in the typical english sense of the
word "complex").

But, any reasoning like that is a bit flawed, it seems, because a
machine is part of the extended physiology of humans or human society.
The artifact is, like the external kidney of a spittle bug, a mechanism
for processing energy.  The distinction between the gas engine and the
human that uses it is clear in some respects but not so clear in other
respects.  So, my response to this point (if it really is what you
intended ;-) is, given two objects of which you are not familiar, can
you judge them machines or living?  And how does auto/hetero-trophy play
into that measurement?

Male and female - why strive to create a male that could carry and deliver babies? 

Again, you're too cryptic for me. [grin]  I'm a bit of a literal person
and need things spelled out.  Is this a question about ploidity (is that
even a word?) where one talks about whether reproduction requires
multiple parents or just one parent?

Art/creativity and physics/determinism - would you want to be able to
 model, explain, formalize, predict the Mona Lisa?


Yes.  However, I would NOT want to automate the interpretation of the
Mona Lisa.  Art, artifacts, and natural phenomena are not interesting
because the processes by which they emerge is "alive" or "mechanistic".
They're interesting because of our subjective exploratory/exploitative
tendencies.  I, personally, find just as much interest in things like
planets and turbulence as I do in music or paintings.  That leads me to
believe that something intriguing like the Mona Lisa could easily be
created by a machine and still be just as intriguing.

Machine/controllable vs life/self-empowered - would you want a toaster or a jet airplane that asked why? or gave you back sas when you give it a command? Would you want to have to consider the feelings and rights of a space craft that you send to the edge of the
solar system (Voyager is there now...started in the 70's). 

I don't know.  This is an excellent point.  But, my response would be
similar to the above... I would not want to have to persuade my toaster
to toast some bread for me anymore than I'd want to have to persuade my
arm to put the bread into the toaster.  But, my arm is alive and the
toaster is not.  Perhaps the distinction between "alive" and "machine"
is just a psychological convenience that allows us to exploit certain
systems to our ends without worrying about the ethics involved?

Tim Gwinn wrote: 
GR: What I don't understand is this distinction between "machine" and
 ... well, "non-machine".  What do you mean by "life is no machine"?

TG: Briefly, 'machine' refers to a sub-class of 'mechanisms' in the Rosen terminology. Mechanisms are systems all of whose models are simulable (i.e., Turing-computable). Machines are mechanisms which additionally admit relational descriptions. The class of mechanisms is in the class of 'simple systems', so machines are therefore also simple systems.

Excellent! Thank you. That took me directly to the relevant part of Life Itself. However, I have some questions about this both practical and philsophical.

To say that life is not a machine is to say that living organisms are
physical systems whose models are not all Turing-computable; in other words, they are not simple systems, but rather, Rosennean
complex systems.

I can't see how this makes a practical difference when dealing with natural systems. Use a gas engine as an example. There exist models of such an engine that are certainly simulable (effective). But, there also exist models that are not. For example, an engine consists (in full dynamic glory) of things like metal crystals, gasoline, fire, electricity, magnetism, oil, "sludge" that builds up on the spark plugs, rubber (or some rubber-like synthetic) in the belts, fluid vapor mixtures and dynamics, etc.

A realistic enough model of this system will not be simulable.

So, for the practical (non-Rosennean) use of the term "machine", this type of argument still won't help me distinguish a machine from an organism.

For Rosen's more formal use of the term "machine", it makes me think that "machines" are platonic and will only help in the world of formalisms and inference. The term loses meaning when applied to the dirty world of reality. Again, I can point to the toaster and arm example above. Will this definition of the concept of mechanisms and machines help me with a practical distinction between a toaster and a biological, but still enslaved thing like my right arm?

Or, will the Rosennean definition assist me in knowing, say, just how detailed (non-abstract) a model must be in order to prove feasibility? I can imagine that this definition of "machine" might yield a useful threshold for feasibility. If the blueprint I have of a new bridge is still simulable, then it's not detailed enough! I have to continue to iterate the construction of my models until it is no longer simulable, and at that point, perhaps my model is complete enough to begin construction.

Because the term 'machine' is used in a very specific and narrow sense here, this may clarify the distinction made here between
machines and living systems. If not, just ask some more questions. :)

I wouldn't want to separate the specific sense Rosen uses from the vernacular sense... or at least I wouldn't want to overly separate them. One of the reasons I'm interested in Rosen is that he managed to think systemically and tie vernacular concepts to the over-specialized and jargonal buzzwords we find so often in our hyper-specialized society.

But, with regard specifically to this formal definition of "machine", I would appreciate it if you could talk a bit about the boundary between a set of simulable models and a set of non-simulable models, when the models are all referring to the same natural system. This, I think, relates to why Gell-Mann wants to call it "plectics" instead of "complexity". The idea is to study where, on the continuum between simple and complex, a system might be. And if we could _quantify_ the size of the set of simulable models a system might have versus the size of the non-simulable models of that same system, then we might come to a measure of the systems complexity. (rather than always resorting back to the old binary "simple" and "complex" duality)

GR: To set a little more context for the above question, I'll add that I think of something like von Neumann's self-reproducing automaton as both a machine _and_ an artifact that attempts to tacitly demonstrate a common theme in Rosen's writings.

TG: As I understand it, there were two very different formulations of
"self-reproducing" entities discussed by von Neumann. Can you elaborate on the one to which you are referring?

I'm not aware of two formulations. I'm referring to "Theory of Self-Reproducing Automata". What was the other?

--
glen e. p. ropella              =><=                Hail Eris!
H: 503-630-4505                       http://ropella.net/~gepr
M: 503-971-3846                        http://tempusdictum.com