Fake Banner
    The Intelligence Paradox
    By Steve Davis | March 4th 2012 02:14 AM | 86 comments | Print | E-mail | Track Comments
    About Steve

    gadfly: noun (1) fly that stings horses and cattle. (2) (derog) annoying person, esp one that provokes others into action by criticism, etc....

    View Steve's Profile
    I would like you to consider carefully the following comments made by Gerhard Adam in the discussion that followed a recent article on artificial intelligence.

    Gerhard’s contribution is a lesson in the benefits of disciplined logical thought. Please read on:

    “Intelligence" didn't just "wake up" one day. Its presence is visible from microbes up to the highest organisms. The notion that if you just cobble together enough pieces and intelligence will emerge is simply magical thinking.”

    And in response to another comment; “It just seems that intelligence is being viewed as some arbitrary "add-on" to biology. Like it's some feature that is "out there" and has nothing to do with the organism in question.”

    And to another: “We already know that rudimentary intelligence exists at the microbial level. The ability to communicate and exchange information establishes a baseline of intelligence that isn't present in any other chemical systems. Even something as seemingly trivial as "self-identity" is present and allows bacteria to distinguish between their own kind and others. In addition, we can see how such behaviors change with colonization [i.e. colonial animals] with increasing centralized control over the constituent behaviors of the individual cells. Invariably this control requires extensive ability to map the organism's domain of cells and provide regulation, control, and feedback to ensure that the entire "colony" operates as a cohesive unit. Individual cells are no longer independent units, but are "answerable" to the controlling mechanisms of the whole.

    From this we can readily argue that a significant amount of data is directly related to sensory feedback mechanisms which are employed to facilitate the mapping provided by "central control" to ensure the survival and reproduction of the organism. Increasing sophistication creates an environment where direct sensory input can be abstracted into interpretable elements that can then be used in lieu of direct sensory data.


    However, the inescapable component of all of this, is the distinction in biology that all "intelligence" is oriented around the survival and reproduction of the organism in question. This is what creates the motivation and impetus necessary to act and to respond to situations [which in turn are mapped back to the central control mechanism].


    While we may use fuzzy terms like "instinct", those are fundamentally meaningless because they imply a kind of "hard-wiring" which doesn't actually negate or promote anything. Instinct isn't "hard-wiring", it's the "default" state. The longer-lived the organism, and the more diverse its circumstances, the greater the need to be able to acquire new information, modify behavior, and learn how to survive in different situations. This is what would largely be responsible for increased brain sizes and sophistication in "higher" animals.


    Intelligence is biology, because at the end of the day, you can build any kind of machine you like with whatever degree of sophistication technology allows and it can excel at its assigned task. But until you can build a machine that gives a damn, it is nothing but a set of human induced rules.”


    You might say that this is interesting, but hardly groundbreaking, just Gerhard being Gerhard. But you’d be wrong.

    With those thoughts Gerhard was a whisker away from solving one of the great problems – what is intelligence?

    The quotes in their entirety are useful, and should be kept in mind when considering the argument that I’m about to present, but let’s pick out the outstanding points.

    “Intelligence" didn't just "wake up" one day. Its presence is visible from microbes up to the highest organisms.”

    “We already know that rudimentary intelligence exists at the microbial level”


    “...the inescapable component of all of this is the distinction in biology that all "intelligence" is oriented around the survival and reproduction of the organism in question.”


    “Intelligence is biology,...”


    What makes these points outstanding?

    The fact that intelligence exists at all levels from  micro-organisms through to blue whales, tells us that intelligence is a feature of life itself, similar to the other recognisable features of life; growth, metabolism, reproduction and homeostasis. But those seem to me to be in a different class altogether to intelligence. They are what we might call tangible features; they are relatively easy to observe and measure.

    Does this leave intelligence in a category of its own? Not at all.

    Here is the point Gerhard almost stumbled upon; if “intelligence is biology” as he said, (correctly I believe) then intelligence is life.

    I can hear the howls of protest already!

    “But you maintain that cooperation is life.”

    Indeed I do.  (At this point those who had trouble accepting that life is cooperation should have a stiff whisky, because they’re in for a rough ride.)

    There is no contradiction here if we consider the possibility that cooperation IS intelligence.

    Can that be demonstrated? Certainly!

    Let’s consider that possibility in light of Gerhard’s comments.

    He said “Intelligence didn’t just wake up one day.” The same for cooperation.
    He said “Intelligence exists at the microbial level.” The same for cooperation.
    He said; "...intelligence is oriented around the survival and reproduction of the organism in question.” The same for cooperation.
    He said; ”Intelligence is biology...”  That is the case also for cooperation.

    But this is still a bit tenuous.  It’s not much of an argument unless the relationship between cooperation and intelligence can be widely demonstrated in the natural world. Which it can.

    The great Russian geographer Peter Kropotkin made the important point in Mutual Aid – A Factor in Evolution, that as a general rule the most social animals are the most intelligent animals. An observed increase in sociality generally goes hand in hand with an observable increase in intelligence. What is sociality but cooperation? The higher the level of cooperation, the higher the level of intelligence.

    But differing levels of  cooperation/intelligence are not just seen between species. Cooperation becomes more complex and advanced as we move from cell to organism to community, and intelligence follows exactly the same pattern on that pathway also.

    The correlation between cooperation and intelligence is so close and so consistent, that for all practical purposes we can assume that they are the same concept. Intelligence is cooperation.

    It seems incongruous to link the two, but their dissimilarity is an illusion created by natural limits to the way we think. Those limits to our understanding are due to the point Gerhard made that intelligence is all about survival and reproduction. Humans have not been subjected to pressures that select for the ability to contemplate abstractions such as intelligence. Our understanding, wonderful though we might think it to be, has been limited by our biological history.

    It might be a little difficult to see the equivalence of intelligence and cooperation, but if we accept the reality that all life forms, from the first cooperating molecules to cells and organisms and communities are based on group activity, then we see that any initiation of cooperative activity anywhere, anytime, immediately creates a new living entity, a new community with an expanded vitality and intelligence. That intelligence and vitality did not come from nothing. The input was cooperation, so intelligence is cooperation.

    But you won’t let go, will you.  You’re determined to hang on to the idea that intelligence is complex or mysterious.
    So here’s an example of the creation of a community with increased cooperation/intelligence.

    The astounding and rapid advances that took place in the 20th Century in the field of genetics were the result of a global sharing of information and hardware among laboratories. This raised level of cooperation effectively equated to a raised level of intelligence as was proved by the results. As soon as one laboratory cooperated with another, the intelligence of the new unit was higher than the sum of the two separate units. And as the number of participating laboratories grew and the cooperation developed, so did the intelligence. 

    The cooperation created the intelligence, because the cooperation was the intelligence. At this point you might argue that the higher level of intelligence followed the cooperation; that it was a consequence of the cooperation. But that is not the case because you can’t make something out of nothing. As Gerhard pointed out, “the notion that if you just cobble together enough pieces and intelligence will emerge is simply magical thinking.” The intelligence did not simply emerge out of nowhere; it grew as an integral part of the cooperation. It’s not magical, it’s not mysterious, and it’s not complex.

    In fact, it’s as simple as the old saying that two heads are smarter than one. So why are two heads smarter than one?

    It’s not because two is greater than one, after all, if the two don’t cooperate they achieve nothing. Two heads are only better than one if they cooperate.

     The more cooperative the two, the smarter they become. Because intelligence is cooperation.

    Comments

    vongehr
    As I said before, Gerhard Adam deserves an award for his commenting, in fact, he really should turn some of his comments into articles [instead of his articles, sorry ;-) ].

    So thumps up for Gerhard, but you have disappeared down the rabbit hole with pseudo questions like "what is life" and "what is intelligence". One of the most important ingredients to good science is terminology. One first defines the terminology (By "intelligence" we mean in the context of this work ...) in order to ask meaningful questions, then one can go on to enlighten. All else is hogwash like "intelligence is cooperation". No it isn't, cooperation is cooperation.

    Steve Davis
    So it's thumbs up for Gerhard, but I've gone down the rabbit hole?
     Yeah, that makes sense!
    Dear oh dear Sascha, I thought you would have learned your lesson here, but you've trotted out the same non-argument.
    Make a useful comment or none at all!
    Huh. Is this St Gerhard's Day or what?. I think Gerhard has had enough adulation for now: if he gets any more, his head won't fit in that big white hat of his  :)

    Sorry if I seem a little crabby - this is not a good day for me: already I find myself in agreement with Sascha. Oh woe, woe, and thrice woe!

    Actually, I think you do Gerhard a disservice by saying he equates intelligence with cooperation. If he comes forward and endorses you, then so much the worse for Gerhard. We have two separate words with two separate meanings. There is no more to be said.
     
    Of course what you really mean is that they are closely connected. But until automated computing was invented we only knew of one kind of intelligence, namely the kind that has evolved on the biological substrate. The wide variety of organisms that show intelligence may confuse us into thinking there are lots varieties of intelligence, but in the end they all have to survive or perish, there is no third option.  As you say, that puts a severe constraint on the sort of intelligence that can emerge - at least in the early stages* of evolution.

    But with computing we have a new substrate. There is certainly selection pressure on systems to "work", so, at some "unified level", biology and computing are subject to the same evolutionary laws. But at the level of practical implementation, a brain is natural and is the product of billions of generations of natural selection; a computer is designed with a purpose that is for all practical purposes decoupled from the Darwinian origins of its designers. There's no point at all in defining intelligence to be everything that biology has thrown into the pot.  That's precisely why we talk about artificial intelligence - it's an artifact.
     
    Finally I would make a plea that if you are going to employ reason at all, that you use logic rather than some sort of rhetoric based on making pretty patterns with words
    He said “Intelligence didn’t just wake up one day.” The same for cooperation excretion
    He said “Intelligence exists at the microbial level.” The same for cooperation excretion
    He said; "...intelligence is oriented around the survival and reproduction of the organism in question.” The same for cooperation excretion
    He said; ”Intelligence is biology...” That is the case also for cooperation excretion
    Hence intelligence is excretion.

    By the way, who exactly is saying intelligence is something mysterious? That bit of silliness was put to bed by Gilbert Ryle In _The Concept of Mind_ more than half a century ago. I doubt whether anyone who has survived more than a day of exposure to science20 would even consider it an issue any more.


    *   Footnote:  I have argued elsewhere that reasoning has, in our species, reached a stage where it can break free from its biological roots and has already done so in the field of mathematics where the theorems of Turing and Godel formally prove the universality - and the limitations - of some forms of reasoning. This however, is a recent development, way beyond basic intelligence. 
     


    vongehr
    Ha Ha Hi Hi - your comment here and especially over there (didn't even know that post) made my day - love it mate.

    Year 12, month 3, on the fourth day, 1234, that's easy to remember, St. Gerhard's day came to be.
    Thor Russell
    Excellent! You really need to define things properly or you cannot really say anything at all.
    Thor Russell
    Steve Davis
    Well well, aren't these interesting comments!
    Two very intelligent people have waded in to deflect readers' attention from what I've said and direct it to their own favourite hobbyhorse! And no warning lights went off in their heads to alert them to the possibilty that an article on intelligence should be treated with great care?
     Even though I made that point in the opening sentence?
    What are they thinking?
    I'll tell you what they are thinking. They demonstrate one of the points I made, that our history hinders us in thinking of two things at once, or to put it another way, in seeing patterns in the natural world.
    Thor Russell
    If you say one thing IS the other, then you are talking about only one thing.
    Thor Russell
    Steve Davis
    If there is one thing we mere mortals can glean from quantum physics it's that the universe is not as it seems. So it's extraordinarily ironic that the two most strident critics of this article claim to have an interest in or expertise in quantum physics. But instead of giving qualified support or reasonable criticism, they have both stamped their feet, said the universe is just as it seems, and what's more, words only have one meaning because we like to keep things simple. Well done guys.
    vongehr
    Steve, please, calm down, what is up with you; I have known you as a reasonable guy; don't grow old and silly that fast. What has the support for anti-realism by quantum mechanics to do with avoiding pseudo questions? Your argument is now down to the level of "quantum is weird, all is not as it is, thus all who don't like what I do just take the universe as it seems"! That is how one sells new age crystal necklaces to emo girls (and may even assure a top article on a science site on the internet).

    Quantum physics is not necessary to tell you that the world is not what it seems - science teaches that lesson in many ways. The lesson with quantum mechanics is the one you refuse for some strange reason: Pseudo questions like "what is existence" are nonsense. Words are tools to think; you need to sharply define those, or you will produce mere hogwash. Terminology is to make a difference in the first place, to distinguish fruit and vegetables so you can say something like "tomato is a fruit and not a vegetable" at all. It is childish prescience to get hung up on the discovery that fruit and vegetables are the same ("intelligence is cooperation"). They are not because we distinguish them.
    Steve Davis
    Well Sascha, there's not much more I can say to you, except to get philosophical about it.
    I can point the moon out to you, but if you keep looking at my finger you won't see the wonder of it.
    And can you do me a favour?
    Don't ask me to sharply define what I mean by "the moon".
    An Intelligence may require cooperation between its constituent parts (whatever those happen to be), but I don't see how intelligence and cooperation can be synonyms.

    A pocket watch requires co-operation between it's constituent cogs and springs to function as a system - but this cooperation does not demonstrate intelligence of the system in any widely understood sense of the word. (I won't make references to the watchmaker!)

    That would be the two most strident critics out of, let me see now, oh yes, three in total. All of whom are saying the same thing.
     
    Well, I haven't seen any rebuttal coming from you, just a load of surly ad hominems and vacuous sophistry. There's only one thing to do.
     
    Goodbye Steve. Enjoy talking to yourself.

    *Plonk*


    Steve Davis
    Don't you dare hang up on me Derek Potter!
    And just when we were getting to know each other!
    They're a fickle lot these science bloggers; no commitment to long-term relationships. 
    Thor Russell
    OK let me have a go at explaining things. I don't plan to win any competition for conciseness or being pithy, thats already been done, instead I will try to raise things for you to consider in the hope that you will realize there is a lot more out there than you have thought about.

    Firstly defining terms.
    To start with, for a term like "intelligence" there is a commonly accepted understanding in the general public's mind and science 2.0 audience about what it means. It is slightly different to each person, and although there are similarities there are also differences and some people will have different ideas on what it is. In order to start talking about intelligence without being misunderstood you first need to explain the concepts that fit your understanding of it and those that you do not, or think aren't important. So what are the important qualities of intelligence that are included in your definition and what aren't? Surely you include qualities such as pattern recognition, adaptability, but not ones like excretion! To say something meaningful about intelligence you need to explain how these concepts are related, just saying "intelligence is x" will not do it. I want to draw your attention to an existing comment on this blog:

    "An Intelligence may require cooperation between its constituent parts (whatever those happen to be), but I don't see how intelligence and cooperation can be synonyms."


    Unless I am mistaken this is the feeling of the vast majority of readers of this blog. Yes cooperation is important for intelligence but cooperation is important for any system, so cooperation has no special place in an understanding of intelligence.


    Let me give an analogy:
    Suppose you have to describe a car, and you say that just like intelligence and life always being related, a car and its engine always go together. You then say that because they always go together they are the same. Hence


    "Car is engine". 


    Now this tells you nothing about either a car or its engine, is wrong and can lead to no further understanding of either cars or car engines. "Intelligence is cooperation" to me makes as much sense as saying "car is engine" or "vegetable is cucumber". Yes a cucumber is a vegetable, but saying vegetable and cucumber are the same is nonsense.




    Secondly intelligence itself:


    You yourself make the comment:
    "or to put it another way, in seeing patterns in the natural world"
    Yet you do not think to relate "seeing patterns in the natural world" to intelligence!!
    Subconsciously at least "seeing patterns" is part of you own understanding of intelligence like many other peoples. 


    So among several qualities of intelligence that any discussion needs to address are pattern recognition, adaptability, and understanding. What does it mean to recognize a pattern, adapt to changing circumstances or understand something? 
    There are many wonderful and surprising things about the brain that you should consider. Did you know that if you put on glasses that make the world appear upside down, you can adapt after a while until you see it up the right way again, then if you take those glasses off, the normal world appears upside down! With further training you can then immediately adapt to both the glasses off and on, with the world being up the right way in both cases. To me and I am sure many people this is a demonstration of intelligence, illustrating adaptability being an important part of it. Just saying "cooperation" cannot yield any insights in this situation.


    Finally I urge you to consider the functional view of intelligence. (structure rather than biology) I don't know why Gerhard rejects it, because if you want a simple explanation of intelligence this is how you will get it. Such a view is widely accepted in psychology, I know because my dad is a psychology academic and I have had conversations with many researchers of course. If you reject functionalism you will get easily tripped up by thought experiments and have to use ridiculous terms like "simulated intelligence" compared to "real intelligence" when comparing two systems that behave in the same way, and have the same internal structure. You are also forced into an irreducible complexibility type argument that can yield no further insight. Coming back to the car, even though an engine always goes with a car, you can definitely study an engine in isolation. Life uses intelligence, like it uses hearing, vision or a heart and you can learn things from studying it without always talking about life and purpose, like you can learn things about the heart by studying heart cells in a dish. 


    http://plato.stanford.edu/entries/functionalism/ 
    http://en.wikipedia.org/wiki/Functionalism_(philosophy_of_mind) 




    So please reconsider and start anew, don't play word games with cooperation, but think about the commonly accepted qualities of intelligence, including pattern recognition, adaptability, and understanding and see how you can relate these concepts to the more basic ones of internal structure and signal processing.

    Thor Russell
    Steve Davis
    Thor, thanks for giving the article proper consideration; you and MiCro are a breath of fresh air.
    But I’ve got some points to make. (You knew I would!)

    You said “I hope you will realise that there is a lot more out there than you have thought about.”
    What did you mean by that? That I think I have given the last word on intelligence? Far from it. Don’t mistake a confident presentation for a delusion of omniscience. You know by now that anyone presenting material here without a confident attitude will soon be mincemeat.

    You said “To say something meaningful about intelligence you need to explain how these concepts are related.”
    Well, I thought that’s what I had done. What was the article about if not the relationship between life, cooperation and intelligence? You might have also noticed a link to related articles. Regular contributors need readers to refer to those, particularly if a run of articles are on the same subject, because themes can’t be repeated over and over; the nitpickers would have a field day. (Well, that happens anyway!)

    “I don’t see how intelligence and cooperation can be synonyms.
    Again, that’s what the article was about, again, it might be necessary to read the related articles, but more importantly, if you disagree with my argument that’s fine, but you need to pick it apart piece by piece. Actually, I’m surprised that no-one has targeted the intelligence/life concept in detail. At least you made an effort.

    But now I must get really critical. Your counter-analogy of car/engine to my life/intelligence was a shocker. Cars and engines do not always go together as you say, so what has that to do with life and intelligence which do always go together? Actually, you began that angle by talking of life and intelligence but in the course of the story it changed to cooperation and intelligence.  So you’re a little confused there, unless we assume that you have subconsciously accepted my proposition that life is cooperation is intelligence. :)

    But let’s look at the life/intelligence thing a little more closely.

    There’s a school of thought now in quantum physics that the foundation of the universe is information. That the universe is comprised of bits of information. If that’s the case then it’s undeniable that life forms are no more than entities that interpret and use that information. The underlying essence of that interpretation and utilisation is intelligence. So life is intelligence.  And as life is cooperation, then intelligence is cooperation. See, it’s not hard at all, is it?

    So when you said “Life uses intelligence” you got it wrong because you did not go to the trouble of first defining life. It’s not life that uses intelligence, it’s life forms. Remember what I said in an earlier article about the careless use of language?
    Thor Russell
    Oh dear, you don't seem to have got my points. Firstly saying "So life is intelligence.  And as life is cooperation, then intelligence is cooperation." quite quickly gets meaningless. 
    If everything is the same, then the only thing you can end up saying is just "universe, universe, universe ..." or 
    <your favorite deity>,<your favorite deity> ,<your favorite deity>  ...
    Insight comes from separating concepts as much as combining them. 


    Please consider having intelligence without life. It is the opinion of a large number of psychologists etc that this is possible. I know what your definition of life is, this doesn't affect you considering this. Just denying that life and intelligence can be separate does not make it so. Absolutely agree that information could be the basis of the universe, but that doesn't in any way mean that information giving rise to life could not be separate to information giving rise to intelligence.



    Once again, please consider how adaptability, pattern recognition and understanding and what they mean. They are obviously separate concepts.
    Saying cooperation=life=intelligence=pattern recognition=the universe and everything is precisely the same as saying nothing at all.


    Thor Russell
    Steve Davis
    "Firstly saying "So life is intelligence. And as life is cooperation, then intelligence is cooperation." quite quickly gets meaningless."
    We'll have to agree to disagree on that one Thor, I think it's full of meaning.

    "Insight comes from separating concepts as much as combining them."
    It can do, but there comes a stage where you enter the realm of reductionism beyond the point where analysis is useful, as Ernst Mayr said of gene-centrism.

    "Please consider having intelligence without life."
    Why would I want to? When you give some evidence of the value of looking at that, I will consider it. 

    "Absolutely agree that information could be the basis of the universe, but that doesn't in any way mean that information giving rise to life could not be separate to information giving rise to intelligence."
    So give us the evidence.

    "Saying cooperation=life=intelligence=pattern recognition=the universe and everything is precisely the same as saying nothing at all."
    Somehow I don't recall actually saying that!
    Thor Russell
    Ok on regarding intelligence without life what are you going to do if non-living things i.e. machines display ever more intelligence, insist that it is "simulated intelligence". What is a good reason for not considering this? If intelligence can exist without life, then you are guaranteed to have an incorrect view of it by rejecting such a possibility. I would say the onus is on you to prove that intelligence cannot exist without life, and just saying "intelligence is life" won't do it. That is just defining it out of existence.
    I know you didn't say that cooperation is everything, but I am worried that your next blog posts will go something like "consciousness is cooperation" "chemistry is cooperation" "physics is cooperation" "existence is cooperation" etc. 

    But more importantly you have still completely refused to consider pattern recognition and its role in intelligence. Keeping on relating intelligence to cooperation and life I think has caused this. 

    Finally and perhaps most importantly is where can this view that intelligence is cooperation get you? Surely it must explain the features of intelligence/psychology to be useful. So if you think you have a unique and insightful view of intelligence, how do you apply that to psychological phenomena such as the upside down glasses I talked about, or confirmation bias, inattentional blindness, negative priming, or the vast wealth of knowledge we have in the psychological literature? If your approach is useful, it should make understanding of many phenomena easier, enable you to easily come up with testable predictions and theories rather than just look like word games. After all science is about predictions and testable hypotheses.
    Thor Russell
    Steve Davis
    "...on regarding intelligence without life what are you going to do if non-living things i.e. machines display ever more intelligence, insist that it is "simulated intelligence"?
    Yes.
    "What is a good reason for not considering this?"
    Because simulated intelligence is what it is.
    "I would say the onus is on you to prove that intelligence cannot exist without life, and just saying "intelligence is life" won't do it. That is just defining it out of existence."
    Not at all. The equation now becomes "artificial intelligence is artificial life." You added an extra factor, a machine, to your scenario so I'm entitled to add an extra to mine.
    "I am worried that your next blog posts will go something like "consciousness is cooperation" "chemistry is cooperation" "physics is cooperation" "existence is cooperation" etc."
    Stop giving me ideas Thor!
    "you have still completely refused to consider pattern recognition and its role in intelligence."
    Pattern recognition is certainly relevant to intelligence, but that question is not relevant to this discussion.
    "where can this view that intelligence is cooperation get you?"
    As I said earlier, I'm not trying to have the last word on this. All that an individual can hope for is to add a little to the general understanding.
    " how do you apply that to psychological phenomena such as the upside down glasses I talked about,"
    In that instance how was the adjustment made? By the cooperation of various organs and nervous system. 
     "If your approach is useful, it should make understanding of many phenomena easier,..."
    I think that if there was general acceptance of the cooperation/intelligence view of life it would throw light on many seemingly intractable problems, but this is not the forum to discuss that.
    Thor Russell
    Ok, while this topic may not be a forum to discuss intelligence and how you think it relates to cooperation, if you cannot demonstrate such insights, and stop thinking about intelligence further, it will just seem like word games to me.
    Saying cooperation between elements of the nervous system relates to upside down glasses is not really true, the eye behaves in exactly the same way, its the brain that adapts, the brain would adapt in just same way if the image came from a non-living camera. Your claim that cooperation is related to intelligence can and must be judged on what insights/theories/predictions it can give. I am not convinced it can give any.
    Thor Russell
    Tony Fleming
    There's a historical debate allied to this question of yours and it is 'how do you differentiate living and inert forms of being?' This is probably the more basic or primitive question. Your concept is to have a psychological element to life. What I'm saying is that we do not need a psychological system for life to be present; that may come much later along the evolutionary tree. When a person detects, perhaps subconsiously, someone looking at them from behind, they turn around and spot the person. What is behind the effect is the way the person 'feels' the photons that are being emitted by the viewer, a lot like an antenna receiver receiving a signal. Now while a stone cannot turn around when looked at, it will nevertheless 'detect' the signal and probably reflect the signal thus giving the appearance of a stone to the observor. So what is life? Somewhere in the evolutionary past there was a time when 'life' did not react psychologically but we still call it early life. Sometime in the past, your 'psychological life' sprang into being. I would think that across the Galaxy or even across the Cosmos, there are 'pre-psychological' forms of life evolving as we speak. Perhaps there are even some 'post-psychological' forms of biology 'out there'. So 'cooperation' between cells, between biological entities of say green algae, of many different kinds formed an early form of 'intelligence' that sought to replicate in order to achieve survival. A fundamental element of life.
    Tony Fleming Biophotonics Research Institute tfleming@unifiedphysics.com
    Steve Davis
    Exactly Tony.

    Thor, you said "Your claim that cooperation is related to intelligence can and must be judged on what insights/theories/predictions it can give. I am not convinced it can give any."

    My claim that cooperation is related to intelligence was not plucked out of the air, I constructed an argument, I gave evidence for the claim. If you want to publicly disagree with the claim you should start refuting the evidence, instead of just putting up a counter claim.
    Or go write your own article.

    As you're keen on psychology you could start with the motivation behind the desire to create artificial life and have it accepted as "real" life. Is there a "god-complex" at work there, do people with this fascinating fetish hope to experience some surge of omnipotence? Why are they not content with the prospect of creating artificial life?

    Thor Russell
    Yep I'm going to write my own article, I have already promised Gerhard this also, so you will have the chance to disagree with my interpretation.
    Thor Russell
    MikeCrow
    I don't see cooperation between microbes as a sign of intelligence, In higher order life it would be an attribute of intelligence.

    I think my 2 statements are the minimum states.

    And accordingly we already have non-biological life.
    Never is a long time.
    Gerhard Adam
    And accordingly we already have non-biological life.
    How so?
    Mundus vult decipi
    MikeCrow
    This is my definition of life:
    Living things respond to their environment, while going about attempting to reproduce.
    No more, no less.
    Based on this definition, a factory of robots (as they are used in industry now), making more robots in a symbiotic relationship with humans(who supply parts and whatever) is as alive as microbes that depend on other microbes(or humans) for needed environmental conditions.
    Never is a long time.
    Steve Davis
    Mi Cro, you have not defined life, you've defined life forms. This has been a sticking point all through this discussion. I'm not singling you out for criticism, we all do it, I even have to check myself sometimes. But life is the essence underlying life forms.
    MikeCrow
    These are IMO the minimum requirements for life. No other qualifications required.

    Life responds to environmental stimulus, and it reproduces.

    I think restricting it to anything else is to restrictive. It might be okay for life on earth, but we don't know if that will hold true elsewhere. And I think to make such restrictions shows our bias, not a universal requirement. And yes it does leave it open for things we might not think qualifies. Such as robot factories.
    Never is a long time.
    Gerhard Adam
    Sorry, I don't buy it.  The problem with your definition is that it overlooks the most important ingredient.  The ability to adapt.  Without that you've got a factory full of robots.  Nothing more.
    Mundus vult decipi
    MikeCrow
    Robots adapt. Granted it's mostly at our hands now, but I don't think it really invalidates my definition.
    How quickly does a microbe adapt?
    Never is a long time.
    Thor Russell
    Your robot factory meets the definition of life as cooperation, as that does not involve adaptation.
    However I think it makes more sense to think of the development of life as a process rather then a yes/no boundary. Otherwise you are forced to choose an arbitrary boundary in what may be a many step process. Consider if all the intermediate stages were still around for us to observe, we wouldn't be trying to ask "what is life" in fact we may not even have such a word with an apparently clear definition. To me asking when does life start is like asking when does a number between 1-100 become "big". 
    A robot factory that cannot adapt is at an earlier stage than a microbe, but is still is somewhere along the process.

    Cooperation becomes before what most people consider to be life (a word only means what people think it does, there is no timeless pure definition of it) because you can have cooperation without the passing on of information such as DNA. Your robot factory could continue to cooperate for ages without progressing any further up the life-like process because it cannot adapt. Because of that I consider a microbe to be further along the process, and your factory would get stuck and not be able to progress further without our explicit help.


    Thor Russell
    Gerhard Adam
    How quickly does a microbe adapt?
    Well, that's not really the question, but a population can adapt within a few generations.
    Mundus vult decipi
    MikeCrow
    You're right, it wasn't really the question.

    The question is would you consider a microbe living even if it's genes didn't change for thousands-millions of generations?
    Never is a long time.
    Gerhard Adam
    Your hypothetical makes no sense, since for that to occur, the genes must not be passing information from generation to generation, since nothing changes.  The more subtle point being that for genes to go unchanged means that there can be no variation even within the present population, otherwise change is introduced.

    Therefore with no information, the genes become a meaningless structure, so we would be justified in questioning whether the "organism" in question were alive.  If you want to speculate more than that, then we're back into "Skynet" territory regarding computers.
    Mundus vult decipi
    MikeCrow
    Actually I think it makes lots of sense, the average splitting of a microbe results in a replica of the parent. I'm not saying they don't ever change, just not very often(otherwise wouldn't they evolve into other species?).
    And if I read this right, that's what it's saying as well.
    E.Coli mutation rate is 0.26 x 10-9 per nucleotide * ~ 300,000 +/- 50,000 nucleotides.

    And I know manufacturing changes happen fairly often.
    Never is a long time.
    Gerhard Adam
    See reply at the bottom.
    Mundus vult decipi
    Thor Russell
    Here is my take on it:http://www.science20.com/thor_russell/blog/basic_intelligence_and_amazing_ability_neurons-87866


    Thor Russell
    MikeCrow
    Thinking about this, I came up with this.

    Living things respond to their environment, while going about attempting to reproduce.

    Intelligence is changing that response (by learning) in an attempt that might improve those odds.

    Now I know the pedantic's will want to quibble about this word or that, and no I don't mean that they only respond while copulating (if that's how they procreate), just that between birth and reproduction there's a lot of responding going on, blah, blah, blah......
    Never is a long time.
    UvaE
    The late Jacob Bronowski wrote,
    "Man is unique not because he does science (and you might add, not because he is intelligent), and  not because he does art, but because science and art (and human intelligence in general) equally are expressions of his marvelous plasticity of mind."

    Decades ago, he also reminded us that,
     No science is immune to the infection of politics and the corruption of power.
    Tony Fleming
    Actually this all raises a central point about mitosis and cooperation. It appears bubbles were a possible form of proto-life and proto-cell. Certainly not very intelligent but they did cooperate. Cells we know are fundamental to life and the many trillions of cells that form into the various neurons spread diversely throughout the body. So, what is intelligence? We see a progressive evolution following the Big Bang towards 'more intelligence' where mankind has a 'central nervous system' compared with say an octopus that has a very diffuse 'decentralized nervous system'. So are octupi any less intelligent than humans? They have an ability to change their skin for camouflage or to show emotion. We thus see an evolution in 'intelligence' from the cell to biological systems all on the back of the cooperation of the cell and the cosmology around. So what is intelligence? Is it a passing on of the large to the small? Big Bang to cell to mammals and other life forms? cheers Tony (the elephant in the room who will try to be 'nice').
    Tony Fleming Biophotonics Research Institute tfleming@unifiedphysics.com
    Steve Davis
    Enrico and Tony, thanks for your thoughts.
    Gerhard Adam
    Mi cro

    You're misunderstanding the nature of evolution.  The population is already more diverse than your robots.  As an example, consider antibiotic resistance in bacteria.  That trait is already present in the population, so when antibiotics are administered those that are susceptible die off, while those with resistance will tend to dominate the future generations.  The diversity [changes] is already intrinsic in the population.  When this is coupled with the ability for microbes to exchange genetic information your analogy collapses completely.

    In "higher" animals it is even more pronounced since no two members of a species are "exactly" the same.  This is precisely why diversity is important in biology and catastrophic in engineering.
    Mundus vult decipi
    MikeCrow
    No, I don't think I do.
    Yes, the population is more diverse, it's had 3+ billion years of evolution, robots a few decades, but there are still lots of diversity in the robot population. And I like Thor's point that you would probably consider Robots at a stage that's really prior to fully functioning single cell life.

    And I left out any reference to passing change on to decedents (which happens during manufacturing BTW), because in single cell life, the daughter cells are exact copies of the parent. So including that as a requirement would disqualify a lot of simple lifeforms.
    Never is a long time.
    Gerhard Adam
    They are faithful copies, but it would be erroneous to conclude that they are "exact".  That would stop natural selection dead in its tracks.
    Mundus vult decipi
    MikeCrow
    According to to my math E.Coli has 1 mutated nucleotide for every ~9,100 exact copies.
    I'm not sure if that'd equal 9,100 splits, or 9,100 generations.

    But it is plain that most copies are exact.
    Never is a long time.
    Gerhard Adam
    Sorry, but you're assuming that diversity or change is only a direct cause of genetic encoding.  The point is that even the same gene expressing the same protein doesn't produce identical results in every organism.

    That's why we may all share the genes that produce a circulatory system or a nervous system, but it would be preposterous to suggest that each individual's system would map identically to each other. 
    Mundus vult decipi
    MikeCrow
    True, but the mechanism of inheritance is DNA, not expression.
    Never is a long time.
    Tony Fleming
    There are elements that do correspond between life and robots including the individual nature of each 'copy', similar to mitosis. But they are being 'fathered and mothered' by humans, not robots.
    Tony Fleming Biophotonics Research Institute tfleming@unifiedphysics.com
    MikeCrow
    The first generation sure, but I think you could call ongoing manufacturing a symbiotic relationship.
    Never is a long time.
    Gerhard Adam
    Sorry, but it just ain't so.

    While you may think there are similarities with manufacturing and biological reproduction they are completely illusory.  The biological system is a self-contained mechanism whose survival and reproduction are bound together, which is what makes them "selectable". 

    You can build millions upon millions of robots, and without some means of "feedback" your system could never and would never evolve.  It's a completely contrived scenario which is simple manufacturing when its all said and done.

    The most sophisticated robotic manufacturing you can envision can't even emulate a bacterial colony.
    Mundus vult decipi
    MikeCrow
    Is feedback all that's missing?
    Never is a long time.
    Gerhard Adam
    Indirectly.  You'd have to be prepared to have 1,000 robots let loose and then have half [or more] be destroyed while the other have [with whatever variations were advantageous] use that advantage to create more, etc. etc.

    The problem we face with the robots is that we don't know what's advantageous without having planned and engineered it.  Therefore we're looking for external feedback, some sort of "stamp of approval" which is precisely the wrong thing to look for.
    Mundus vult decipi
    MikeCrow
    But isn't that almost exactly what happens with microbes?
    They don't arrange for their mutation, and the fitness applied is all external.

    The only thing we do for robots that doesn't happen for microbes is we design the next generation as opposed for it to be the roll of the dice in their dna. But then they're both tested for fitness in their environment.

    Look, it's obvious you disagree, so give me your 10-20 word definition of life.
    Never is a long time.
    Gerhard Adam
    The problem with your robots, is that you must realize that one cannot live someone else's life for them.  As a result, I'm assuming that you can see the difficulty in arguing that human engineering can be a replacement for self-reproduction [in machines].

    However, my basic definition of life would be something that involves cooperation, and the ability to obtain data from its environment, which it can use to respond and exploit its surroundings.  To me the exploitation aspect is a key element, since it describes the whole "purpose" for the organisms existence.
    Mundus vult decipi
    You are confusing an objective definition with a run-down of your favourite metaphysical assumptions.

    Let's play this through:
    My basic definition of life would be:

    something that involves
    (1)           cooperation
    (2)           and the ability
    (3)                  to obtain
    (4)                        data

     from its environment
    (5)           which it can
    (6)                  use
    (7)                         to respond
    (8)                       and exploit

    its surroundings.
    I have numbered eight obvious weasel words, there may be more. However, one thing is perfectly clear - when you then say:
    To me the exploitation aspect is a key element since it describes the whole "purpose" for the organism's existence.
    you are obviously aware of what your definition is saying. However you may not be so aware that, as it stands, it is pure woo-woo. God knows, quite literally, that if I were to bring my own metaphysical beliefs into everything I say here as freely as you do yours, I would - rightly - be shot down for preaching. I can only assume that attributes of "purpose" are self-evident to you. However they are not self-evident to most people. A lipase may have an obvious function in breaking down fats in the digestive system, but that is simply what it does, and where it finds itself, in your small intestine, fully accounts for the appearence of purpose. Otherwise you would have to attribute purpose to the continental plates: they are there to hold up the terrestrial biosphere. Your "job" if I may put it that way, is to break down the "appearence of purpose"into objective terms so that science knows what you are talking about. Otherwise "purpose" becomes a homunculus pulling the strings - not just a shorthand for a complex concept but a new entity in the system, one highly vulnerable to Ockham's razor.

    So how well does your definition fare?
    (1)           cooperation
    begs the question of purpose and intent
    (2)           ability
    implies choice whether to do something or not
    (3)           obtain
    in the context of data, implies very specific forms of interaction - have you done the maths to demonstrate the data flow from the environment? Or are you just interpreting what you see?
    (4)           data
    is quantified in computing, partially quantified in physics but starts again with different concepts altogether in biology. Have you defined how to tell whether a system is obtaining data or just being pushed around by physical events?
    (5)           which it can
    again - choice
    (6)           use
    again - implying purpose and intent
    (7)           respond
    again - begging the question that it is a purposeful response and not merely a cascade of physical microevents 
    (8)           exploit
    again - implying purpose and also implying a value system of benefits to the putative "living" system. Does the Ameglian Major Cow (who wants to be eaten) have different purposes from the one in the field who runs away from the knife, or is she not properly alive? Be careful how you answer: saying she is not properly alive because her desires have been artificially decoupled from her evolutionary past will require you to define "artificial" in a way that distinguishes the activities of homo sapiens from the activities of every other creature on the planet. 
     
    Gerhard Adam
    (1)           cooperation
        I don't understand why you claim it "begs the question".  Many organisms cooperate for mutual benefit.  It seems that part of the problem is your insistence on anthropomorphizing the words "purpose" and "intent" as only being legitimate definitions for human cognitive processes.  Your example:
    A lipase may have an obvious function in breaking down fats in the digestive system, but that is simply what it does, and where it finds itself, in your small intestine,
    ... explains nothing because it simply presumes that this activity is the result of a happy coincidence.  Yet, you assume these activities take place within the context of a digestive system, which clearly has a "purpose" in an organism.  That "purpose" was established by natural selection, since organisms that failed to have a sufficient degree of organization were at a disadvantage to those that didn't.  As a result, repeatability and predictability became a staple of existence and consequently the more sophisticated an organism's structure, the more "purposeful" its cellular division of labor became.
    (2)           ability
        I don't understand why you think "ability" implies a choice.  Locomotion is an ability.  Phagocytosis is an ability.  Again, these are repeatable and controllable within the organism and aren't simply arbitrary or random actions. 
    (3)           obtain
        Again, "obtain" indicates either sensory organs, or a mechanism for an exchange of "information" between organisms.  This has been demonstrated from bacteria up to multi-cellular organisms.  Whether it be the exchange of a molecule to "sense" the presence of other bacteria, or other signaling processes by which environmental data is acquired and interpreted.  
    (4)           data
        When a system's behavior can change based on the "data", then it is clearly not being pushed around by physical events.  "Choices" [even those that are 'hard-wired'] still represent choices.  No cognition is required to respond to the data.   However, your point also seems to suggest that organisms simply respond to "physical events" as if this is somehow some deficiency.  Why do you assume that even physical events aren't data?  What do you respond to if not physical events?  
    (5)           which it can
        Again, you're assuming cognition.  Any option whereby more than one outcome can occur based on the circumstances [even if it is simple instinct] constitutes a "decision".  In that respect it represents a choice [although not necessarily a conscious one].
    (6)           use
        Are you claiming that when an organism eats it serves no purpose?  Are you claiming that when an organism "hunts" food it does so without "purpose"?   
    (7)           respond
        Again ... there is nothing that a human being does that isn't "merely a cascade of physical microevents", so what's your point?  The fact that everything can be reduced to the point of absurdity?  
     (8)           exploit
        Of course it's a "value" system.  It doesn't have to be consciously aware of any "value".  Organisms don't spend their time arbitrarily eating everything they encounter.  They exploit that which has "value" to their way of living.  Do you think leaf-cutter ants dont' have a "value" system in their farming?  How about honeybees informing the hive of pollen sources?  When a virus takes over a cell nucleus, do you think that this is just a series of random coincidences that result in replication?  

    There's no question that many of these words are loaded with human meanings, and too often, they are explicitly tied to cognitive processes that aren't even true for humans.  In short, the problem occurs, because there is nothing you can say to describe an organism that can't be equally stated about human beings.  If we follow that line, we find that we can't legitimately use these words in any context, because they always imply something that can be contermanded by further reduction.  

    If we consider the origins of life issue, then we invariably have to consider that at some point, some chemical interactions occurred that were the precursors to what we now call "life".  However, these chemical reactions are insufficient by themselves, so at some point, the chemistry had to be capable of being contained and sustained.  

    Again, regardless of actual mechanisms, at some point the chemistry would've been contained within a finite space with resources being added [to keep it going] and "waste" being eliminated.  In addition, at some point, cells may have broken apart and reformed, potentially being the precursors to reproduction.

    In any case, we know [based on current organisms] that whatever combination of events was taking place, the first cells that were able to respond "predictably" and "repeatably" would have been more successful than those that couldn't.  

    It is this aspect of "predictability" and "repeatability" that represents "intent" and "purpose" in biology.  Natural selection would have favored organisms that could repeat past behaviors reliably, as well as those that could adapt to new situations.  Moreover, adaptation requires a stable base, such that the majority of well-established processes and systems are heavily conserved.  

    Now, for those that are hung up on these concepts as implying something metaphysical or even magical, we can plainly state that nothing of the sort is required.  The purpose of describing "intent" and "purpose" is to provide a descriptive quality to the organism that allows us to predict behaviors [just as with the concept of Dennett's intentional systems].  

    In that way, we can state that the "purpose" of the heart is to pump blood, without suggesting that hearts are somehow independent life forms that seek out such an objective.  A heart operates within the context of the entire body and is a subsystem that has been specifically selected to perform a particular function on behalf of the organism.  However, the net effect remains the same, in that there is an organ formed when cells differentiate, whose purpose is to provide the necessary power to pump blood throughout the body.  In some cases the organ may even have additional functions.  However, from a scientific perspective the concept of "purpose" allows us to predict that it is unlikely that we will find any organ [or organelle] within any organism that doesn't specifically fulfill a requirement or function necessary to help the organism live, survive, and ultimately reproduce.

    Similarly we can argue that an organism develops certain traits, and learns things that enable it to survive, because natural selection shaped this organism to possess the necessary means to "reproduce".  Clearly then there are numerous environmental variables that an organism may encounter, but it is sufficient that the necessary traits exist, such that an organism can exhibit an "intentionality" to its actions that will potentially ensure survival and [hopefully] offspring.  Again, we can use the concept of "intent" to predict that any organism that we observe will behave with directed actions in acquiring food, protecting territory [if that affects food or offspring], and exploiting reproductive opportunities.
    Mundus vult decipi
    #1
    Cooperation is more than things fitting together giving a result that makes an observer feel warm and fuzzy inside. I dare say some of Thor's systems do that, yet even he is unlikely to claim they are truly alive. Merely observing that it is possible to anthropomorphize a cooperative structure does not mean it is unique to life. It therefore it has no part in a "definition" unless you can show what is unique about "cooperation for a purpose" in living systems - that is absent in non-living ones. Such vague question begging was prevalent in first year science textbooks in the 60's when everything was woo.

    #2
    My objection to the word "ability" is that it implies a control system which can choose to deploy that function or not. Unless you can define that control system in a way that distinguishes life from non-life you have just pushed it into metaphysics. 

    #3
    There are different types of data. Data can be quantified at the bit level and the quantum level. The "data" which organisms obtain has not been scientifically quantified. You are using a undefined term. How does an organism seek out food? Dongboogles flow into the organism (3) and as everyone knows, dongboogles (4) help an organism find its lunch.
     
    #4
    No cognition is required to respond to the data. 
    Correct. So how is this any different from non-life? What is it about "responding to data" in the sense of a cascade of microevents that is peculiar to life?

    #5 
    Covered by #2 no need to repeat the point.

     #6
    Are you claiming that when an organism eats it serves no purpose? Are you claiming that when an organism "hunts" food it does so without "purpose"?
    Of course. The key is the fact that you have put the word "purpose" into quotes. It's not an intent, it's a function with a value that you, the observer, impute to the system. Objectively the system simply *is*.  Function as a high-level description of how a system operates is not unique to living things. Function with high-level "purpose" that merits quotes is woo-woo.

    #7 
    Again ... there is nothing that a human being does that isn't "merely a cascade of physical microevents", so what's your point? The fact that everything can be reduced to the point of absurdity?  
    The question is not whether reduction leads to a point of absurdity but whether such reduction reveals a criterion for deciding whether a system is living or not. So far, nada. Busy little bees filling their honey pots so that all the queen's children will be fed is Life. Busy little robots gathering fuel so that others in the network can run effectively is non-life, artificial - something Thor might have made on a good day. Pardon me if I fail to see the difference.

    #8
    They exploit that which has "value" to their way of living. Do you think leaf-cutter ants dont' have a "value" system in their farming?
    Of course they don't have a value system. That is *your* anthropomorphic interpretation . You can certainly quantify the effectiveness of their behaviour in ensuring survival, but that is to beg the question - that the purpose of life is to survive.  The colony simply *is*. if you want to wax metaphysical, I will join you, but let's do it properly and say "God looked and saw that it was good". At the moment we're stuck with "Gerhard looked and saw that it had purpose."
     
    Of course I appreciate what you are trying to do - the rest of your essay confirms it - but your discussion about origins imposing evolutionary pressure and thus appearence of goal-oriented behaviour is simply question-begging. You can - and sometimes do - cheat by saying that that's what life is, so anything else with different origins, even if it does the same thing, cannot be alive.  Clones cannot be alive - they did not evolve. Synthetic "life" - even DNA based - isn't really alive - it did not evolve. If we run into little green men who really did form spontaneously by a 1 in exp(exp(exp(120))) chance (helped along by quantum consistency), "oh they're not really alive because they didn't evolve".  I hope you are prepared to argue your case with them - our radio transmissions should have annoyed a good few of them by now: no doubt they are already on their way.
     
    Thor Russell
    Hi,
    I wrote some stuff about what life is a while ago, but decided not to post it here as there has been enough written about life already. I have posted it in the link below instead, (too long to be a science2.0 comment and I can better keep track of it on my random site with no visitors) with my thoughts on your comments at the end:

    http://www.seekinglambda.blogspot.co.nz/ 
    Its not 100% polished but you should get what I mean if you are interested.

    Thor Russell
    Gerhard Adam
    Clones cannot be alive - they did not evolve. Synthetic "life" - even DNA based - isn't really alive - it did not evolve.
    I've never argued that clones aren't alive.  Synthetic life may be a different matter, depending on exactly what is involved.  Obviously genetic modification still results in an organism that is "alive".  Similarly Venter's synthetic life appears to be "alive". 

    The issue isn't whether these organisms evolved, but rather whether they are capable of evolving.  That's certainly a key element. 

    As a case in point, considering Venter's synthetic life, I would argue that this is a fundamental characteristic of this synthetic lifeform.  If it turned out that it couldn't reproduce and that the only way the organism could change was by introducing new genetic material in the lab...   at that point, I would say that it wasn't alive.  It might be a biological machine, but I wouldn't consider it life.


    Mundus vult decipi
    Thor Russell
    OK I will try agian and see if I can get somewhere  ...
    I agree with much of what Derek has to say. When you talk about the intentional stance and give examples, in spite of claiming to the contrary you slip into judging by your values and anthropomorphising whats going on, and such judgement then takes on a life of its own. When you talk about the "purpose" of the heart it makes sense and I would talk in a similar way if asked to explain it. I have no objection to using an intentional like stance, and in many situations it is best, however you can stretch it beyond breaking point and end up with a self-contradictory position.  Whether something is alive or intelligent cannot depend on your subjective definition of whether it has purpose. 

    So talk about hearts/bacteria is mostly fair enough, however the "intent" you start requiring in intelligence is definitely not the same concept in spite of what you claim and is where things get ever more wrong. There is the fatal logical flaw related to isomorphism as pointed out by Derek. There is also the perhaps not related too narrow definition of "intent" that leads you to reject unsupervised learning for some reason.

    You use physics to explain atoms, chemistry for molecules, the intentional stance for biology and simple intelligence but you need to at least accept a new concept to help explain complex intelligence. In the same way that you criticize people for trying to use genes to explain biology because it is too simple and low level, a similar criticism applies if you apply the intentional stance simplistically to advanced intelligence as you try to do. 

    Its well known that neurons do not obey the kind of intent guidelines that you suggest. Neurons that are not used say in the vision part of the brain of a blind person will not just sit idle because they now have no purpose. They will cast around, looking for something to act on. For vision the signals they end up acting on are to do with sound/hearing and touch (used to read braille). Your understanding of intent will only lead you astray, making you deny the significance of such behavior. There is no prior "intent" in such neurons to improve the hearing ability of a blind person, there cannot be. Also babies use unsupervised learning very heavily. They will pick up a new language without explicit supervision. Now you could say the "intent" of such neurons is to make themselves useful and figure out whatever signal comes into them, but why stretch such a word? The concept of unsupervised learning is pretty much essential to higher level intelligence. The way it happens is because neurons have the ability to "figure out for themselves" the structure of a signal they are exposed to and extract the useful information without being told in advance what it is or be given feedback on whether they have it right. Such incredibly general and powerful signal processing ability is probably essential for higher level intelligence, but you are obliged to deny that such a thing has significance in biology as you have done.
    This is in direct contradiction to your "evolution is smarter then you" principle also. Even our primitive computers can do it to a small extent, and it is useful yet you will have me believe that the much more powerful biological brain has not evolved to have such an ability! Of course it has.



    I get the strong impression that you want to reject the religious like explanation for things  but do not accept the consequences of your position. So "intent" gets stretched beyond its meaning to serve that purpose and provide the "magic" that you emotionally want but deny. Just like "cooperation" has in these series of articles. The unfortunate thing is that in a desire to seem consistent, the errors spread into a wide range of areas that shouldn't be related at all. Even creationists are not obliged to deny the importance of unsupervised learning.



















    Thor Russell
    Gerhard Adam
    Whether something is alive or intelligent cannot depend on your subjective definition of whether it has purpose.
    I think you're getting yourself into the quandry.  Obviously any living organism is "alive" and/or "intelligent" on its own terms.  However, the classification of that trait is a human gesture and therefore it is entirely dependent on a "subjective definition".

    After all, the question surfaces because of human interest in the distinction, not because living organisms need to be told.  I have no problem with the concept of a "purpose" being subjectively evaluated.  What else could it be, since there is no objective definition of life?

    If we were to examine something on another planet, what criteria should be used?  Replication may be assumed, but we can't presume that an organism must be capable of reproduction to be alive.  After all, the first cells were "alive" and couldn't reproduce. Certainly without reproduction we could argue that it couldn't evolve, but that still doesn't mean it isn't alive.

    Just to be clear in my use of terms, I'm trying to use "purpose" to describe a characteristic of functions or structures, while "intent" is to describe the BEHAVIOR [in terms of ability to engage in directed actions] of the organism itself [a collection of subsystems].

    Again, with the concept of "purpose" we can predict that when we encounter something in another organism, that it has a role in ensuring the survival and/or reproduction of the original organism.  Even if we don't know what the purpose may be, we can see that those that have it are "alive" and those that don't aren't [again based on whatever our subjective criteria for "life" is at that time].  
    In the same way that you criticize people for trying to use genes to explain biology because it is too simple and low level, a similar criticism applies if you apply the intentional stance simplistically to advanced intelligence as you try to do.
    My problem with the genes, is that they have been insufficient to explain the results, so it seemed too simple and low level to be particularly meaningful.  In other words, genes produce the "basic" organism.  However, it is the environment that will determine whether those "basics" make it through to the next generation.  Genes are incapable of making such a prediction, therefore it becomes impossible to argue that they are competing with each other.  To what end, if they can't predict outcomes?  In any case, the problem with the "intent" argument in advanced intelligence comes down to this.

    If we consider that my use of the term "intent" is used to describe behaviors that promote the organism's survival and ultimately reproduction, then we can predict that the behaviors that we can expect the organism to engage in are "directed" to promote that objective [to varying degrees].  In other words, this is the result of a syngergistic effect between natural selection and the environment/genetics.  In my view, this is a fundamental characteristic of "life".

    The problem I have with the artificial intelligence arguments is that if someone provides the "intent" artificially to a robot or computer, then it cannot be considered to be intelligent.  It is a simulation.  Why?

    Because at no level as this "intent" been developed or evolved because of actual experience.  It was simply placed there in its entirety, based solely on an individual's subjective interpretation of what would constitute "intent".  In other words, the robot/computer is behaving according to someone else's "intent" or interpretation of it.  In my view that's a big difference.

    Let me also be clear that this doesn't mean that a machine couldn't be built that could "evolve" and achieve it's own level of intelligence.  In that respect I have no quarrel [despite what many people think].

    My primary problem with this scenario is that for this to occur, it would truly have to be capable of "living" and "existing" independently.  It would make it a new species, in its own right.  It would not be an arbitrary tool or servant of humans.  Humans would have no intrinsic rights to determine what it does or how it behaves.  

    This is the conflict.  Such a machine would be completely useless to humans, so to pursue it is simply an abstract notion to be explored simply to see if it can be done.  This is why I assert that the type of artificial intelligence people are working towards is impossible.  The kind they want can never be truly intelligent, and the kind that would work, isn't the kind they want.
    Mundus vult decipi
    Thor Russell
    You don't seem to have addressed my central point at all. It wasn't related to artificial intelligence I have given up on trying to discuss that with you. At one stage you said that you didn't see the significance of unsupervised learning in biology. I have never heard of anyone who is aware of neural plasticity and the brains adaptability making such a comment. It seems a totally bizarre thing to think, why do you do it?
    If it is not to do with your "intent" philosophy, then why?
    I assumed it was to do with "intent" as you seem to define it because such an ability possibly doesn't fit with quotes like this:

    "If we consider that my use of the term "intent" is used to describe behaviors that promote the organism's survival and ultimately reproduction, then we can predict that the behaviors that we can expect the organism to engage in are "directed" to promote that objective "


    Brain cells that automatically adapt to the input signal and extract useful information without knowing what it means or what it is for are useful for the organism as a whole, but the section of the brain doing such a thing doesn't have a clear objective related directly to survival. However a brain composed of neural structures that can adapt in this way has a significant advantage over one that does not.
    To me this is a higher form of intelligence, to that found in say an insect where everything is hard wired. A system that can do more than just obey its inheritance and can automatically adapt its brain structure to the environment in such a way with little to no feedback surely should be considered more intelligent. 


    This is why I brought in the analogy with the genes. Just as they are too low level to explain biology so is such "intent" to explain intelligence where every module is assumed incorrectly to have a predefined "purpose". The environment affects how genes are expressed etc in analogy to how an intelligent system capable of unsupervised learning adapts to the signals coming in without explicit "intent" based instructions. Modules that appeared designed to do one thing can adapt and do another. You cannot understand what brain of a human is capable of without seeing unsupervised learning in action in the same way that you cannot just look at a genome and know everything about the life of such a species. 


    Put it another way, say you have a organisation where to start with there are fixed tasks for everyone. Now if the environment of the organisation changes, the tasks for people to do may change significantly. Someone who can adapt their behavior is both more useful in the organisation and considered more intelligent. Neural modules that can adapt to a similar changing situation are useful and can be considered more intelligent also. They do this by automatically finding work for themselves to do, and process it without needing specific "intent" the way you define it. The "intent" of such modules if you must use such a word is always the same: Figure out the pattern in the data so another part of the brain can make use of it. Are you still trying to deny such an ability exists or its usefullness?


    Getting back to "intent" it appears to me that you change your definition completely even in the same post:

    "However, the classification of that trait is a human gesture and therefore it is entirely dependent on a "subjective definition"."


    Yes I totally agree, however you do not stick to this at all:


    "Because at no level as this "intent" been developed or evolved because of actual experience. "


    This is totally not a subjective definition at all now, you are saying such intent or lack of it is an actual quality of the organism.
    This change is an internal contradiction in your position and has nothing to do with AI.



    Thor Russell
    Gerhard Adam
    However a brain composed of neural structures that can adapt in this way has a significant advantage over one that does not.
    To me this is a higher form of intelligence, to that found in say an insect where everything is hard wired.
    I don't have a problem with that point.  In fact, it could be reduced even further, but I don't see how this tells us much of anything.  It is simply too low of a level.  Knowing that cardiac muscles possess a natural rhythm and are capable of pulsating in unison tells us nothing of the structure called a heart.

    I'm not clear on what you mean by "hard-wired".  Since insect behavior clearly isn't robotic, then you might argue that it has fewer options, but you'd be hard pressed to demonstrate that it is explicitly "hard-wired".
    A system that can do more than just obey its inheritance and can automatically adapt its brain structure to the environment in such a way with little to no feedback surely should be considered more intelligent.
    No problem.  In fact, the argument could be made that when more sophisticated decisions need to be made on behalf of an organism, that is when we begin to see the development of nervous systems and brains.
    They do this by automatically finding work for themselves to do, and process it without needing specific "intent" the way you define it. The "intent" of such modules if you must use such a word is always the same: Figure out the pattern in the data so another part of the brain can make use of it. Are you still trying to deny such an ability exists or its usefullness?
    I think you're just misunderstanding the word "intent" as I'm using it.  The inner ear provides a means of letting us balance ourselves to stand and walk upright.  There doesn't have to be any "intent" displayed in such a system, since that is the "purpose" of such a construct in the first place.  The "intent" is behavioral [as I keep saying] and plays a role when we examine WHY someone stands up or why they go in the particular direction they do.  "Intent" has nothing to do with simply functionality.

    Similarly a human doesn't display "intent" when learning a language.  As you've indicated, the brain simply learns the rudiments and eventually the individual acquires all the necessary skills with which to employ language.  I don't dispute this method of learning, nor do I dispute its importance.  Again, this is a process that takes place unconsciously, so you could argue that it is "hard-wired".  It is a function whose "purpose" is to integrate language into our species.

    If you're simply wanting me to agree that pattern recognition is an important element in the mechanics of learning something, then that's fine.  However, our dispute was in the area of "intent" which I indicated was behavioral and you kept trying to link to other events.  Unsupervised learning is not behavioral.  Just as most sensory data being processed, there is little conscious involvement and in the majority of cases, there's little behavioral associated with it.  That's why I'm trying to be specific in indicating that "intent" is only used to describe the behavior of an organism [narrowly defined to support its survival and reproduction]. 


    Mundus vult decipi
    Thor Russell
    Well then it seems you do acknowledge the usefullness of such a quality. The inevitable conclusion of this is that if you take a module capable of unsupervised learning out of its normal physical environment and use it as part of another system whether artificial or not, it will still perform the same useful function. There is no other way. It cannot decide not to work anymore, and its performance can be bent to our or any other entities purpose. Intent type arguments will do not apply to it. Whether you call that "intelligent" is irrelevant, a module performing unsupervised learning would have undeniable usefullness as part of a technological system. 
    Thor Russell
    Gerhard Adam
    This is totally not a subjective definition at all now, you are saying such intent or lack of it is an actual quality of the organism.  This change is an internal contradiction in your position and has nothing to do with AI.
    I'm saying that the organism is a product of such a process.  Let me try and be as explicit as I can be.  First, this is NOT about defining life. 

    So, let's engage in a thought experiment.  Let's assume that we have the means by which we can completely program a brain.  In short, we can create a brain for any kind of animal we like, and it would be 100% accurate in that representation.

    Now, let's take the brain of a horse, and reprogram it to be that of a dog.  At this point, my argument is that the horse would not longer be considered a horse.  It would be a simulation of a dog, since it would lack the "intent" that made it a horse [throughout its developmental history].   No matter what the horse did, how well it might perform/behave, or whatever behaviors it might exhibit, it would longer be considered a horse, but instead it would be displaying the behavior of a simulated dog.  Simulated?  ... because it isn't a dog. 

    If you wish to consider that an actual quality of the organism then so be it.  I'm simply saying that you can't take an organism and separate it from its history.  Life is a continuum such that each organism is as much represented by its physical presence as it is by its historical ancestors.  Therefore it's behavioral "intent" is a characteristic in the sense that it is something that has been acquired over the generations as natural selection has refined each generation to be slightly better at survival/reproduction than it's older competitors.
    Mundus vult decipi
    Thor Russell
    Sorry, but I just don't think there is any sense in calling the thing a simulation of a dog. That is just an arbitrary value judgement as far as I am concerned. I don't accept that the history has the importance you assign to it and I think that using "intent" in this context is misleading and has nothing to do with the normal understanding of the word. You are talking about some continuous history which if you think deserves a separate concept should be called something quite different. While you may argue about the meaning of taking an organism away from its ancestry, you can certainly take modules of a brain away from its ancestry and they will still exhibit what I regard as intelligent behavior. You can then combine them in different ways to bend them to your will.If you just define it not to be intelligent because it is out of the organism then so be it, but that is just an arbitrary value judgement on your part and can give no insight on what can be achieved by such things whether in neurons or silicon. 

    Thor Russell
    Gerhard Adam
    I didn't say it couldn't behave intelligently.  I simply said that it was a simulated intelligence. 
    You can then combine them in different ways to bend them to your will.
    That statement alone would argue against it being intelligent.  Now you might be tempted to argue that we do that with domestic animals and training animals, but I would defy anyone to claim that such animals are "bent" to our will.
    Mundus vult decipi
    Thor Russell
    Its clear that you have just defined intelligent to mean what you want it to now. The business about simulated intelligence vs real intelligence makes no sense to me. It just seems an arbitrary and meaningless value judgement.
    Saying that combining them to our will makes them not intelligent is simply defining intelligence first to mean something that cant be exploited by another system. Its tautology and word games not insight. One more advanced system can't exploit the lesser intelligence of another one is no different to saying that we can't exploit a machine to do our bidding. 
    Thor Russell
    Gerhard Adam
    Saying that combining them to our will makes them not intelligent is simply defining intelligence first to mean something that cant be exploited by another system.
    First, let's be clear on the terminology.  I didn't say "exploited".  I responded specifically to your phrase "bent to our will".

    The latter quite clearly indicating that the "intelligence" is question lacks the freedom to exercise it's own volition.  Without that freedom I don't see that you have anything useful.  It's just a big calculator then.

    Exploitation, on the other hand, implies the ability to interact.  Even the most negative interpretation implies a coercive power which would be necessary to subjugate a free thinking entity.

    Without freedom of thought, I question your ability to have thought. 
    Mundus vult decipi
    Thor Russell
    Call it what you like, a calculator does not learn, yet a system without freedom of thought could perform unsupervised learning and be technologically useful. 
    Thor Russell
    Gerhard Adam
    A calculator is technologically useful.  So what's your point?
    Mundus vult decipi
    Thor Russell
    A system capable of unsupervised learning is vastly more technologically useful whether you call it intelligent or not. Something that learns is intelligent according to any sensible definition of the word.
    Thor Russell
    Gerhard Adam
    Brain cells that automatically adapt to the input signal and extract useful information without knowing what it means or what it is for are useful for the organism as a whole, but the section of the brain doing such a thing doesn't have a clear objective related directly to survival. However a brain composed of neural structures that can adapt in this way has a significant advantage over one that does not.
    See, that's my problem, because you're attaching a significance here that simply isn't relevant to the discussion.  It's not unimportant in terms of the brain, but it is no more significant than talking about the contractile ability of muscles, or the impenetrability of skin cells, or the role of rods/cones in the eye.  None of these has any knowledge or objective directly related to survival. 

    We know that muscles can adapt and change based on stress and usage, yet we wouldn't argue that they understand something objectively related to survival.

    So, I'm not clear on why you think this one aspect of it is different?

    I've said repeatedly, that my definition of "intent" is related to the behavior of the organism.  Not a subsystem, not an organ, not a cell.  The entire organism displays behavioral traits that indicate "directed" actions that are geared towards its survival and reproduction.


    Mundus vult decipi
    Thor Russell
    No these things don't have knowledge directly related to survival, that is my point. I regard them as intelligence (if you don't then that is a different definition and pointless to disagree on) and as a consequence regard intelligence as not needing to be related to the survival of the organism. They don't understand something related to survival, but according to what i consider to be a sensible definition of intelligence are intelligent. Intelligence has to be related to some signal processing ability as far as I am concerned and built up from that. Defining intelligence as directly related to survival doesn't work as you then cannot explain why some organisms are more intelligent than others. Just like some organisms have more need for contractile muscles, some have more need for advanced signal processing/unsupervised learning. It does make sense to measure the intelligence of a module as far as I am concerned and if "intent" does not apply, then "intent" type arguments cannot be used to claim things about what can be achieved by using such modules as part of a different system.
    Thor Russell
    Gerhard Adam
     Defining intelligence as directly related to survival doesn't work as you then cannot explain why some organisms are more intelligent than others.
    Of course I can.  Intelligence is going to be related to the cognitive requirements of the organism in question.  Those that have more sophisticated "problems" will likely have more sophisticated nervous systems and brains.  Those that have fewer requirements will have simpler systems in place.  A nervous system and brain is NOT a requirement for basic signal processing.

    It appears that you want to employ a reductionist approach.  That's fine, but in my view, you've simply defined intelligence to be whatever you happen to point at, and it has simply become irrelevant in any discussions about organisms. 
    It does make sense to measure the intelligence of a module as far as I am concerned and if "intent" does not apply, then "intent" type arguments cannot be used to claim things about what can be achieved by using such modules as part of a different system.
    What constitutes a module?  Is the immune system intelligent?  What about cancerous cells? 
    Mundus vult decipi
    Thor Russell
    A module is something you can identify some kind of boundary around. The immune system could be said to have very basic intelligence, perhaps a network of cancer cells. The experiment involving taking rat neurons out of a brain and growing them was an intelligent module because they performed a simple task, but without relevance to an organism. Yet the system of neurons in spite of not being a complete organism would be more intelligent than a single cell. Likewise the section of a blind persons brain, that is not being used and casts around for something useful to do is an intelligent module. Its not overly reductionist as such a module is more complex than the entire nervous system of an insect.Intelligence and survival intent are two separate things and you can't understand intelligence if you deny that.
    Thor Russell
    Gerhard Adam
    Yet the system of neurons in spite of not being a complete organism would be more intelligent than a single cell.
    So complexity is your argument?  The more neurons the more intelligence?
    Mundus vult decipi
    Thor Russell
    Computing power/neurons is a necessary but not sufficient condition for intelligence. The neurons/system has to be capable of some degree of learning to be credited with higher level intelligence. Learning and intelligence have to be linked, you can't say that something that learns is not intelligent just because it is a small part of a system, and doesn't understand how it all fits together. Systems/organisms will always be able to learn some things better than others, making comparison always a bit subjective.
    Thor Russell
    Gerhard Adam
    So you agree that bacteria are intelligent [albeit at a  lower level and no nervous system]?
    http://www.sciencedaily.com/releases/2009/06/090617131400.htm
    Mundus vult decipi
    Thor Russell
    Perhaps to some degree, though obviously much less than a mouse
    Thor Russell
    Tony Fleming
    I agree Gerhard

    Individuation is an important element of evolution and replication. I see it in mathematical terms. An eigensolution is the 'particular solution' to a system of differential equations, and correspond to a person's base pairs, 3,000,000,000 (3 Bbp) whereas the 'general' solution is the human genome, the 23,000 genes. It is the base pairs that control the chemistry that allow the cells to grow, differentiate, and develop into a specific human being. A person's intelligence emerges froma smaller group of base pairs and genes. It has been a 4 billion year evolution before human intelligence stepped upon the Earth. It was probably 500 million years before microbial intelligence sought to replicate to achieve survival. In between there have been a whole time sequence of biodiversities, of shades of intelligence. Other creatures, like elephants, dolphins, octupi, whales may have the correct amino fatty acids with a similar central nervous system to compete with humankind in the intelligence stakes, not forgetting of course our cousins, the primates.

    I see intelligence as an evolved faculty that we get from our parents, environment and local cosmology, similar to the way suns have been used to supply the elements above helium.

    I have this beautiful image of life and intelligence 'out there' as we speak. A marvelous thought that we are not alone. Not scary (I hope) like H.G. Wells saw it, but cooperative.

    Tony Fleming Biophotonics Research Institute tfleming@unifiedphysics.com
    John Hasenkam
    The notion that if you just cobble together enough pieces and intelligence will emerge is simply magical thinking.”

    So he has never heard of the tobacco virus. Take the virus, disassemble it, place it an appropriate broth, it will re-assemble into an active replicating virus in other organisms. 


    I don't like the word "intelligence". There are what we call "intelligent behaviors" but the idea that there is some essential "intelligence" out there is too animistic for me. I know physicists, headed by Wheeler I believe(it from bit), believe that the universe is all about information but I have problems with that concept too. When does information transfer occur? Physically, through interaction. I'd rather study those interactions. That is life. 


    To say life is intelligence is to replace one mystery with another. The best definition of life is ... it is an electron looking for a home. Probably made tongue in cheek but at least its creator stuck to something we can measure and study. Life does not obtain information, it responds to environmental contingencies, as does everything else. Life cannot even be defined with respect to the single organism. For example, we are entirely contingent on all the bugs on our bodies, without them we're dead in a week. So if you are going to define life you are going to have to define all those interactions that make life possible. That is, you are going to have to study stuff you can measure. 




    Steve Davis
    Thanks for your thoughts John.
    To say life is intelligence is to replace one mystery with another.
    You're right, except I think I've done a little more than that here. In fact, I've not only said that life is intelligence, but also that life is cooperation, and I even went so far as to say that life is virtue here.
    I think I would draw little criticism if I said that "life is a many-splendoured thing."
    So if you are going to define life you are going to have to define all those interactions that make life possible. That is, you are going to have to study stuff you can measure.
    I think that's the same error made by Szostak and Trifonov. They focused on the outcomes of living systems, the measurables, instead of looking at the driver of the systems, which is cooperation.
    Life cannot even be defined with respect to the single organism. For example, we are entirely contingent on all the bugs on our bodies, without them we're dead in a week.
    Quite right, and I covered that angle in The Purpose of Life.
    And you're right to imply that a reasonable definition of life must hold true at all levels of life. I believe the cooperation explanation of life does just that.
    John Hasenkam
    Thanks Steve
    Co-operation is physical interaction! 
    Steve, you are... Correct.
    Keep pushing in this direction.
    ------
    Some say that love is the glue that holds the universe together...

    Steve Davis
    Thank you - I will !