What is the State of Artificial Life Today?

Hello,

Back in the late 1990s, one section of my web site provided information 
about Artificial Life, including a set of Java applets that demonstrated 
cellular automata, autonomous agents, rule-based systems, and other 
concepts.

As time passed, the applets suffered from Java bit-rot, the articles 
became a bit dated, and my focus shifted to high-performance computing.

Rather than abandon the topic, I've decided to reinvigorate my site with 
new applets and updated articles, in conjunction with my recent 
presentations of practical genetic algorithms for software analysis.

ALife appears to be in the doldrums; along with "Complexity" and 
"Chaos", the field appears to have faded from the limelight. The Santa 
Fe Institute seems to have shunted ALife to another organization; many 
of the old web sites are gone.

But ALife is certainly not dead.

I note that the 9th International Conference on Artificial Life will be 
held in Boston, Massachusetts, September 12-15th 2004.

While I'll explore the web in search of information, I was hoping people 
here could provide some insight into where ALife stands today, and where 
the most exciting new research is taking place.

Thank you in advance!

..Scott

-- 
Scott Robert Ladd
Coyote Gulch Productions ( http://www.**--****.com/ )
Software Invention for High-Performance Computing

Back in the 80's I programmed Conway's "Game of Life" and displayed it on a
Univac monochrome monitor. I decided to "play God" and "changed the rules",
hoping that "by luck" the display would come up with a pattern that 
would duplicate itself.
To make a long story short, in high speed my display looked very much like
"chaos" in a cloud chamber where elementary particles/symbols would keep
colliding and mutating - with no "complexity of life" in sight.

What I needed then and, imho, what we need now, is a _mathematical 
formalization_
of biological processes: mutation, replication, evolution, heredity, etc...

If the "researches" still continue to be "experimental", I'm afraid that 
"complexity of life"
would be nowhere in sight, and the state of Alife/AI would be much 
dimmer that what
it has already been.

---Nam

Scott Robert Ladd < XXXX@XXXXX.COM > writes:


From my perspective what has changed is that researchers are
realising that the standard gradualist neo-Darwinist perspective
central to the original ALife exercise and GA techniques is simply
inadequate theoretically and practically to generate 'novelty',
differentiation or hierarchy, i.e. to deal with what has been called
'irreducible complexity'. In other words microevolutionary
processes have patently failed to account for macroevolution
after very extensive experimental testing spanning decades...

Attention is now switching to co-operative/synergistic techniques
such as 'compositional evolution' and these are showing great
promise, although few of the 'popular' applications seem yet to
incorporate such ideas (it is perhaps too early). There has
also been a move from calling it all 'ALife' towards what is
termed 'multi-agent systems' (MAS) or similar and this is more
concerned with modelling real issues (again usefully) in other
academic fields - rather than as stand-alone specialist scenarios.

I list various current techniques on my web pages, try looking
through 'Whatsnew' perhaps to see some recent additions:
 http://www.**--****.com/ 

This whole evolution thing is proving a lot more complicated
than 'modern synthesis' evolutionary biology views (including
the 'central dogma') makes out - but we still don't need a 'God' ;-)

I myself think that the values of the organisms are central to
evolution in practice, i.e. their 'purpose' means that evolution
is not random but directed (locally, not globally) - but of course
that subjective view is even more 'heretical' to the objective
survival-is-all 'believers' and it hasn't caught on much yet !

Best,

Chris Lucas
CALResCo
 http://www.**--****.com/ 
 

CALResCo < XXXX@XXXXX.COM > wrote or quoted:


Isn't 'irreducible complexity' bullshit terminology made up by
god-fearing folks in a desperate last-ditch attempt to find
intellectual support for their faith?
-- 
__________
 |im |yler   http://www.**--****.com/ @XXXXX.COM   Remove lock to reply.

Scott Robert Ladd < XXXX@XXXXX.COM > wrote or quoted:


My money would go on the movie and computer games industries.  They are 
both putting lots of effort into making realistic artificial creatures.

Lots of good work is obviously being done - though it might not be
quite what most alife folk would like to see being done.
-- 
__________
 |im |yler   http://www.**--****.com/ @XXXXX.COM   Remove lock to reply.

"realistic artificial creatures" are just "realistic-animated" 
creatures: a far cry from
real artificial creatures. [Hint: a real artificial creature is an 
_automaton_!]

Tim Tyler < XXXX@XXXXX.COM >  writes:


No it isn't (the term was contrived by Behe, a scientist).
It refers to the situation whereby a fitness landscape
contains a fitness saddle, such that bridging the gap
between a local optimum and higher optima requires
more than one simultaneous mutation, each of which
are individually less 'fit' than the previous state and 
would be 'selected against' as neo-Darwinists insist.
The 'gradualist' dogma is incapable of bridging that
gap. Such saddles are ubiquitous in epistatic systems
(with rugged landscapes), as Kauffman showed.

I suggest, for example, you read (carefully) the recent
PhD thesis by Richard Watson which outlines all this in
considerable detail: 
 http://www.**--****.com/ ~richardw/thesis/

It is remarkably stupid IMHO to deny problems simply
because the person who raises them happens to have
an unrelated belief ! I frequently find scientists have stupid
beliefs also...  Myself, I believe neither in God, nor in
neo-Darwinism as a self-righteous religious sect persecuting
heretics !

Science is about honesty, a quality too often missing in
this subject I find.

Chris Lucas
CALResCo
 http://www.**--****.com/ 

A scientist who supports Intelligent Design, that is.

Google: behe "intelligent design"

-- 
 __ Erik Max Francis &&  XXXX@XXXXX.COM  &&  http://www.**--****.com/ 
/  \ San Jose, CA, USA && 37 20 N 121 53 W && AIM erikmaxfrancis
\__/ God heals, and the doctor takes the fee.
    -- Benjamin Franklin

 Erik Max Francis < XXXX@XXXXX.COM > writes:



Do you imply that a scientist who does not believe as you
do is NOT a scientist then ? If so, that is just the sort of trash
that gets science a bad name and rightly. In EVERY
field of science there is vast disagreement (despite the
oversimplified claims made in text books). It is DISHONEST
to pretend otherwise, and scientifically and philosophically
INCOMPETANT to arrogantly maintain that one view is
'truth' and everyone else's is rubbish. For Stuart Kauffman's
view (for example) on the legitimacy of ID as science see:
 http://www.**--****.com/ 

Neo-Darwinism is clueless when it comes to explaining
actual form, it just ignores the development problems
entirely. ALife cannot afford to do that, we get nowhere
if we do, so we test the assumptions - and they fail
to generate the higher forms predicted by Darwinists.
In fact, standard neo-Darwinism tends to eliminate variety,
it doesn't create it !

In science, when a theory is falsified it is WRONG (or
at least inadequate). To still hold it as total 'truth' simply
degenerates it into a form of religious dogma (scientism)
- it certainly is no longer valid 'science'.

Which other explanations are required is an experimental
question, not one to be answered by 'beliefs' of any sort.
Some may find my essay on knowledge and theories of
interest here:  http://www.**--****.com/ 

Chris

No, I said that he may be a scientist, but he's in the Intelligent
Design camp.  You're the one who implied that, because he's a scientist,
he wasn't:

| Tim Tyler < XXXX@XXXXX.COM >  writes:
|
| >Isn't 'irreducible complexity' bullshit terminology made up by
| >god-fearing folks in a desperate last-ditch attempt to find
| >intellectual support for their faith?
|
| No it isn't (the term was contrived by Behe, a scientist).


That's right.  If someone has a theory that the world is held up by
strings, you need to take them completely intellectually seriously and
devote copious amounts of time to analyzing their ideas or face being
called dishonest and incompetant [sic]!

Being a "scientist" isn't a guarantee that your ideas are right or even
scientific.  As a parallel example, take Alexander Abian, who was a
respected mathematician until he went completely off the deep end and
spent the last years of his life seriously suggesting how we needed to
move Earth into a Venus-like orbit.  The scientific method is what weeds
out bad theories, not simply giving someone the "scientist" label.

There are scientific theories (many of which aren't even unfalsifiable,
and so technically aren't theories) that have been discredited over the
years.  Scientific Creationism, and now Intelligent Design, easily
qualify.  Scientists can certainly spend their time disputing if they
wish to, but there is no need to; they have already been soundly
disputed.

Suggesting that there is "vast disagreement" about the accepted status
of Darwinism and evolution is completely disingenuous.

-- 
 __ Erik Max Francis &&  XXXX@XXXXX.COM  &&  http://www.**--****.com/ 
/  \ San Jose, CA, USA && 37 20 N 121 53 W && AIM erikmaxfrancis
\__/ My life was better before I knew you.
    -- Edith Wharton (to Morton Fullerton)

CALResCo < XXXX@XXXXX.COM > wrote or quoted:


That is completely different from Behe's own definiton of the term.

To quote:

``By irreducibly complex I mean a single system composed of several
  well-matched, interacting parts that contribute to the basic function,
  wherein the removal of any one of the parts causes the system to
  effectively cease functioning.'' - DBB, M. Behe, p.30.

As endlessly pointed out, this definition fails miserably in its original 
intended task of identifying things that could not have evolved without 
intelligent design - since the implicit assumption that evolution can
only proceed by *adding* components is totally incorrect.

Maybe if you have an incorrect idea about what Behe defined
"irreducibly complex" to mean, that explains why you think
it is something other than bullshit.

Saddles in evolutionary landscapes don't indicate that intelligent
design is needed to cross them either.  That is because such
landscapes are usually not fixed - but instead are mutable.

Hill-climbing organisms can sometimes traverse valleys between two 
fitness peaks - by deforming the fitness landscape they are on in such a 
way that they can move between the peaks while continouosly moving uphill.
-- 
__________
 |im |yler   http://www.**--****.com/ @XXXXX.COM   Remove lock to reply.

Erik Max Francis < XXXX@XXXXX.COM > writes:


Not at all, but it seems that you don't claim Behe's not a scientist
but then go on to say his work is 'discredited', a priori - for no
better reason than he uses the dreaded 'God' word I presume
- since you conflate ID with 'creationism'. Strange. Many of the
points made by the ID camp are made also by GA people and
numerous other scientific groups - all discredited are they ?


Funny that physicists devote all that time to 'string theory' isn't it ?
It is well known that every major new idea is subjected to ridicule
by the 'establishment' - before being declared 'obvious' of course ;-)

If you don't wish to analyse people's actual ideas and compare 
the genuine problems they raise with your own theories then you
should not pass judgement on them should you ?


True, it works both ways - standard textbook theories are not immune !
I've no wish to discuss ID as such, it does nothing for me anyway.
Falsifying a theory doesn't mean you need to accept in toto any
proposed alternative... The point I made (originally) was that gradualist
neo-Darwinism is not good enough - it is the 'discredited' theory, 
not so much for what it says (which isn't very much) but for
what it ignores or glosses over (which is most of evolution) ;-)


Obviously it is not. There are many serious disagreements in the
literature, not so much about whether evolution occured (most
scientists anyway seem to agree on that) but as to the processes
involved. For example the self-organization school (of which I am
a part) deny that selection plays a major part in structure - a very
fundamental disagreement to my mind ! I have to ask "accepted"
by whom ? If you mean only 'believers' then that is tautologous. It
may have been the case that neo-Darwinism was "accepted" 30 or
so years ago. It is not today, times change and so does scientific
knowledge. I've mentioned 3 camps already, throw in 'punctuated
equilibrium', 'symbiogenesis' and we have 5. There are many more
around in science. I don't think 'vast' is disingenuous, the two
closest camps - paleontologists and neo-Darwinists were barely on
speaking terms for decades ! See Eldredge "Reinventing Darwin",
Wesson "Beyond Natural Selection" and even Maynard Smith
"The Major Transitions in Evolution" for more disagreements and
alternative ideas...

My own essay "Emergence and Evolution" sets out my position:
 http://www.**--****.com/ 

Chris

My post is directed mainly to talk.origins and CALResCo I suppose.
Dunno what comp.ai.alife is. Btw, is this an example of crossposting?
I'm new to the usenet...

*deleted some stuff*


I think it would be a useful thread to list different serious(ly
taken) alternative (or "modification-") theories to darwinian
evolution. I usually have lots of trouble understanding in what
situations these theories tries to provide an answer and what evidence
exists (real studies with predictions, tests, etc!).

I'm a student of plant biology (*gasp*) interested in speciation and
biogeography. So, what's *not* contributed to the fields I'm
interested in by +- neodarwinism, I ask myself. Key innovations may be
a spawn of species sorting? Is apomixis an intrinsic factor speeding
up speciation? Major gene effects and developmental mutants is the
closest we get to "macromutations"? These are all pretty watered down
versions it appears to me. Why do I not see more? Am I just ignorant
of the history of ToE perhaps?

Btw, is the fact that Fagus sylvatica leaves doesn't decompose quickly
but stays on the ground and in doing so altering its own environment
an example of self-organization? If so, why should that interest me
and not only ecologists? Or rather, why should it interest me more
than some other ecological property? (I willingly admit I don't know
this particular "school" so I may be completely off here)

ErikW

I don't think xanthian's rant deserves an answer. It sounds
more 'fundamentalist' than many recent religious extremists ;-)

Anyone with an open mind can read the views of experienced
complexity scientists on the issue I originally raised (not the straw
men burned in irrelevant asides ;-) Such views are not the views
of people with allegiances to either ID or neo-Darwinian camps,
but are those of thoughtful experimenters who have investigated
basic neo-Darwinian assumptions within their systems and report
the results. They include Inman Harvey's team at Sussex and
Jordan Pollack's at Brandeis, both doing cutting-edge doctoral
and post-doct{*filter*}research on these matters. Throw in work also
by Kenneth DeJong at George Mason, Chris Adami at Caltech
and Gunter Wagner at Yale and you can no doubt judge the
likely scientific quality of this research for yourselves.

The links are on my website. I'll say no more on this subject,
other than to mention that 'evolution' isn't the same as the
more limited and specific 'neo-Darwinism' ('the modern
sythesis'), and especially not the same as what has been
called ultra-Darwinism - don't conflate any of these.

Chris
 http://www.**--****.com/ 
Don't shoot the messenger...