To begin with, the universe is, according to the descriptions of science; simply a vast conglomeration of atoms and energy; all sloshing and smashing about (aimlessly) as a result of ‘fundamental’ forces. One atom hits another, which hits another and so on. Sometimes (or often) multiple atoms will impact; but the results will always be determined by the rules of Newton’s- laws/relativity/particle-physics. And everything is set in motion by the interrelationships of gravity, electromagnetism (light, heat etc), plus strong and weak nuclear forces.
Supposedly, if you had a powerful enough computer, and given the starting locations, energies (and directions of motion) of this vast multitude of particles; you could work out precisely what would happen in the universe on a moment by moment basis.
But of course no such (infinitely complex) computer exists; and in any case how would you go about measuring the starting quantities in the first place; because in some respects physics says that such a measurement is impossible—Heisenberg uncertainty principle—theoretically, if not in a practical sense. But that does not stop physicists making generalised attempts at a universal theory of everything, regardless of such obstacles.
Why this sojourn into physics? Quite simply, to figure out if (wholly/partially) self-initiated movement is (theoretically) possible for a robot.
The universe, is patently and obviously; not self aware (discounting life itself) and has no central controller and/or all-seeing God (at least one who interferes with the natural order). The basic (inanimate) constituents of the universe do not compute anything in relation to their own history, destiny and/or future possibilities. It follows that for the atoms themselves, there is no choice in terms of moving paths; and they are merely at the mercy of other atoms and fundamental forces.
A Vast Game of Dodgem Cars
Each atom is like a dodgem car without a driver; utterly lost and at the mercy of the other (possibly driverless) cars. My point is that it is in terms of motion that we can see the fundamental difference between a self-aware (and universe-sensing) life-form, and a mere mindless atom (inanimate object, structure or group of atoms).
A life-form such as a human can perceive the environment, and make adjustments accordingly. A woman can run from a lion, travel to a stream for a drink, climb a tree to eat fruit etc; all being examples of movements that are taken with a particular goal or purpose in mind. Sometimes she may choose to skip between two moving tree-stumps when crossing a stream; hence planning her own body’s trajectory in a ‘live’, specific way never seen (yet) in any inanimate object.
This argument is the greatest, and most profound one, that can be made to rebut claims that we humans do not posses freedom of will (free-will). Our explanation cuts right to the heart of the fundamental difference between being alive, and not being alive. A life-form ‘sees’ its environment, and knows that it exists and has a relationship to that same environment. Nowhere in all creation do we see inanimate objects making ‘strategic’ self-initiated movements of any kind. Inanimate atoms move, one and all, according to the dictates of other atoms and forces; and never choose to ‘throw-away’ energy by fighting the path of least resistance. QED.
James Lovelock’s insight was that humans gave the earth, and possibly the universe, eyes with which to see/change itself (plus hands to action ‘thoughts’)
Is it not obvious that our previously related discussion of the problem of free-will is solved by use of this argument. If animals (for example) did not posses the power of free-movement (or free-will); what is the point of nature (evolution) providing them with muscles with which to move about, brains with which to perceive, and eyes with which to see all the different possibilities afforded by the universe.
Meaning of Life
Free movement/will is the whole point of life; and there can be no other conclusion; unless one is resigned to life being absolutely (and ultimately) pointless and/or a joke played by some super-intelligent being on us all. The joke is that this superior intelligence makes it look like we possess free-will; and by providing (highly specific) animal movements that follow (apparently) self-determined paths and avoid dangers etc; when in actual fact we had been allowed no such freedom whatsoever; and because the great all-knowing ‘program’ of the universe was incredibly carefully designed/conceived; in fact with a specific purpose. And the purpose is (at least partially) to foster the appearance of free-will and to fool us all into thinking that we could choose/prescribe self-initiated atomic motions (various patterns and atomic re-arrangements).
Postulating that everything that happens in the universe is pre-ordained is, in my opinion, preposterous, and because it requires a plan for the universe that includes making it appear that animals avoided/stalked each other etc. Such a plan would have to be all pre-calculated on an immense numerical scale; that includes also all micro-organism’s avoiding, interacting and running towards and away from each other, plus the interactions of all the insects and animals that have ever lived etc. It would be an immense, never ending, pre-calculation of atomic patterns and causal outcomes. But where and how can this calculation be done? And how would it be communicated everywhere all at once?
The ‘clockwork motor’ model makes no sense; and because it does not explain all of the local patterns of order; like animal interactions; and how these can be actioned, even if they are planned. Clearly the universe is decentralised, and there is no centralised control (at least in terms of plans); but there may be (partly) centralised order by way of the laws of physics/mathematics; and hence Plato’s Forms.
Living Choice versus Mechanical Choice
We have no ‘thinking-computer’ coordinating/managing everything. The behaviour (that is physical motions) of an animal is controlled (at least partially) by the animal itself, in a constant relationship to the animate/inanimate environment. Hence responsive intelligence or decision making behaviours can be felt (or instantiated) at specific space-time points; and the animal may possess free-will and (perhaps) possess the will-to- power or longer-term ‘planned’ behaviours (although these causal projections are primarily a human characteristic).
I call my argument in favour of free-will:
- The Great Cosmic Joke of Pointlessly Coordinated / Patterned Animal Behaviours (how could such a plan / program be orchestrated?).
- Ergo: Humans (+ animals) possess the power of free-will.
Free-will may be restricted according to: limitations of: knowledge, power, and universal/natural processes. All life-forms posses some degree of free-will, according to their powers of perception plus projection of thoughts/actions (in time and space).
A question arises; can/do robots likewise possess free-will?
Our discussion concerns the movements of objects/bodies. Specifically can a semi-intelligent (or pre-programmed) group of (dead) atoms make free-form atom-assembly related decisions, or choose a specific movement path (i.e a pattern, direction, speed), for a group of atoms. We know that a living being can do so; according to the Cosmic Joke argument; but what about machines?
It is obviously possible to program machines to make/follow self-programmed movements (happens every day); and thus (to some degree) we can say that machines have risen above ordinary inanimate matter, and because they may sense and direct and/or ‘choose’ movement paths (for objects/atomic-assemblies) through the universe; and potentially, at least, do so on multiple spatial scales, locations and on future time-lines.
One might call such behaviours intelligence; and admit that a machine has some (programmed) power(s) of foresight/free-will; and in relation to the movement of atoms present in the world.
Types of Free-Will
Remember that atomic movement(s) equate to all of the activity and changes present in the universe; and one might say that all designs, actions, patterns, intelligences etc; come about through movement. However, as previously discussed we need to look carefully at the scope of movement choices that a machine can potentially make.
In all cases, any ‘semi-intelligent’ creature’s movement choices are tied to goals/pre-programming, and to the perception of the available future choices, and hence to conceptualisation of available atomic patterns, movements etc. Choice implies the ability to cogitate, think and/or to conceptualise the possibilities, see the future/past, and to take a ‘world-view’, so to speak.
We might say that thinking (movements of electronic messages in the brain) ‘releases’ the power of movement in the body and world; according to specific planned purposes. In terms of our physical movement argument; there are two sides to the free-will question. Firstly, in terms of output capability; we ask if a thinker has the physical possibility to choose one (initiated) movement path (for the influenced atoms) over another.
The answer appears to be yes for both humans and machines (ref Cosmic Joke); potentially at least; and in view of any general/specific external constraints that may be present. Both human and machine may alter the course of atoms according to environmental-requirements/pre-programming. Machines and men both sense and respond to the environment (to some degree, even if programmed by an external rule set). One might say that the lowest level capability to respond to the environment is a basic definition of an automaton. We call this reflexive free-will.
Secondly we ask if the thinker has the mental capability to imagine/build a structure for the influenced atoms; and according to a projected plan. Put another way, how clever/intelligent are the choices made for the atoms in question. In particular does the thinker possess intention; or has it considered a range of options and/or allowed a scope of future ‘possibilities’ to be taken into account. Can the intelligence design specific structures for highly- intelligent purposes? We call this ability pro-active free-will.
How big is the free in will? What is the scope of the movement paths considered; and how varied are the options examined in terms of future strategy/events, and how clever or ingenious/portentous are decisions taken for future outcomes. One asks if the intelligence is aware of all (or many) of the different future possibilities in terms of the movement of real-world patterns (atoms). Can it align (conceptually assemble) multiple patterns of atoms in the real world according to long-term goals? Can it ‘see’ or imagine future patterns/events in response to complex (unplanned) happenings?
Situations where the thinker invents, plans and provides highly novel, or singularly creative solutions, we call creative free-will.
Unfortunately for robots/machines, in this latter sense we have to admit that machines are severally limited. Machines posses merely a ‘paltry’ reflexive free-will (a very small or zero range of freedoms/choices); in that they can direct the motions (for limited sub-sets) of matter; but they can only do this directing in strictly limited and pre-determined ways. Machines can never operate truly spontaneously or be creative, and/or make guesses about the best way to proceed, or (typically) make optimal movements to respond to novel situations. Ergo: computers exhibit a purely reflexive form of free-will.
A machine’s responses are purely automatic/robotic; and we do not see a machine with creative free-will in the sense of making intelligent decisions/actions in response to un-anticipated events. Machines never even see things that we haven’t told them to look for. As a result they only possess a catastrophically weak/stupid and purely reflexive form of free-will; being one that cannot get very far by way of complex behaviours, or long term and flexible goals. You will not see a machine building a shelter to cover itself from the rain for example, unless specifically programmed/instructed to do so (it will not even notice the rain).
Human versus Machine Thought(s)
To sum up, machines/robots/computers do not really solve problems, anticipate or see the future in any complex sense. One might ask: why is it that machines have such a limited capacity for problem solving? I don’t think we humans (yet) know the answer; only that it has something to do with free- will and thinking/conceptualisation. And this topic may be very complex. We notice that machines can respond to happenings in the environment; but even insects/amoeba demonstrate a capacity for innovation/responsivity that is wholly absent in machines.
It is vital to recognise that machines are limited to purely pre-programmed responses to their environment; and can in no way depart from their original programming. As discussed, the difference seems to be related to the fact that machines are unable to program/learn/change themselves on the fly; to alter their programming and hence goals in response to environmental requirements. Humans see the world with a super-wide angled and responsive-vision; and notice happenings that are new; and we are open to the environment.
Humans always have one-eye open for novelty; and take notice of things/situations we haven’t ever experienced before.
Why it is that machines do not notice un-programmed things/events/objects/situations? Douglas Hofstadter says it is related to thinking ability/techniques/processes. In his book Surfaces and Essences (2013), Douglas says that (for humans) without concepts there would be no thought, and without analogies there would be no concepts.
Douglas Hofstadter wrote:
Each concept in our mind owes its existence to a long succession of analogies made unconsciously over many years, initially giving birth to the concept and continuing to enrich it over the course of our lifetime. Furthermore, at every moment of our lives, our concepts are selectively triggered by analogies that our brain makes without letup, in an effort to make sense of the new and unknown in terms of the old and known.
Analogy is the recognition of equivalency, or likeness of relations, between different events/ circumstances/situations. By means of analogies humans are able to perceive flexibly using ‘fuzzy’ and malleable boundaries; and notice the ways in which things are alike and unalike; whereby we map old conceptual principles onto the novel. And our perceptive minds are constantly on the lookout for novelty; matching and attempting to match categories and analogies to everything that comes into view.
These spontaneous analogies and categorisations, constantly made up by our brains, are deeply influenced by: language, era, personal history, culture, current frame of mind and purpose, instantaneous goals, body position and posture relative to things in the world etc. According to Hofstadter, a category is a mental structure that is created over time and evolves and contains information in an organised form that may be accessed under suitable conditions.
But why are robots and machines so poor at making good decisions? Why are they so limited in thinking capacity; and why is their thinking so primitive? Hofstadter says:
That our advantage is intimately linked to categorisation through analogy, a mental mechanism that lies at the very centre of human thought but at the fullest fringes of artificial cognition. It is only thanks to this mental mechanism that human thoughts, despite their slowness and vagueness, are generally reliable, relevant, and insight- giving, whereas computer ‘thoughts’ (if the word even applies at all) are extremely fragile, brittle, and limited, despite their enormous rapidity and precision… As soon as categorisation enters the scene, the competition with computers takes on a new kind of lopsidedness—but this time in favour of humans.
Thus the roots of human understanding lie in the automatic triggering, or unconscious evocation of familiar categories and analogies. Hofstadter also says that concepts (thought-patterns): are densely stitched together through relationships, similarity and context. Human thoughts are as flexible as they are complex. A concept is: A thing conceived (similarity + relationships + context).
Computers may be able to categorise objects/situations (sometimes); but according to Hofstadter there are big differences between categorisation and analogy making:
Categorisation is routine; analogy is creativeCategorisation is unconscious; analogy is conscious Categorisation is automatic; analogy is voluntary Categorisation favours similarities; analogy favours dissimilarities Categorisation applies to entities; analogy to relations
All in all, one might say that analogies somehow bring the observing/thinking subject into the equation, along-side the objects and processes of the world. Analogy is a large part of how humans think and analyse situations, processes, and space-time relations etc.
Burden of AI
We humans think by placing everything relative to ourselves; to our past-life, current situation, goals, and future requirements.
Hubert L. Drefus (1929-) said:
The burden of AI is its apparent need to proceed in futility from the atom to the whole. People, on the other hand, effectively seem to perceive first a whole and only then, if necessary analyse it into atoms… Almost everyone now agrees that representing and organising common-sense knowledge is incredibly difficult, and that facing up to this problem constitutes the moment of truth for AI. Either a way of representing and organising everyday human know- how must be found, or AI will be swamped by the welter of facts and beliefs that must be made explicit in order to try to inform a disembodied, utterly alien computer about everyday human life…
Understanding of purposes and intentions is key; and things must be intelligible with respect to the whole of human life.
One cannot equate a program which deals with ‘a tiny bit of the world’ with a program which deals with a ‘mini-world’… In our everyday life we are, indeed, involved in various ‘sub-worlds’ such as the world of the theatre, of business, or of mathematics, but each of these is a ‘mode’ of our shared everyday world. That is, sub-worlds are not related like isolable physical systems to larger systems they compose; rather there are local elaborations of a whole which they presuppose. If micro-worlds were sub-worlds one would not have to extend and combine them to reach the everyday world, because the everyday world would have to be included already. Since, however, micro-worlds are not worlds, there is no way they can be combined and extended to the world of everyday life.As a result of failing to ask what a world is, five years of stagnation in AI was mistaken for progress.
Drefus brings attention to the way in which machine knowledge tends to exist in isolated islands, or small unconnected ‘micro-worlds’; which are in turn in no way linked to any higher purposes, goals, relationships or possible future happenings in the biggest sense. Once again, the machine is not conscious of itself, the universe, and humans; and/or of the myriad of relations present between the two natures (natural and technological) and their intimate affects/influences with respect to the needs of humans (for whom they were created to serve).
Hubert L. Drefus again:
My thesis, which owes allot to Wittgenstein, is that whenever human behaviour is analysed in terms of rules, these rules must always contain a ceteris paribus, i.e., they apply ‘everything else being equal’ and ‘what everything else’ and ‘equal’ means in a specific situation can never be spelled out without a regress…The ceteris paribus condition points to a background of practices which are the condition of possibility of all rule-like activity. In explaining our actions we must always sooner or later fall back on our everyday practices and simply say ‘this is what we do’. Thus in the last analysis all intelligibility and all intelligent behaviour must be traced back to our sense of what we are, which is, according to this argument, necessarily, on pain of regress, something we can never explicitly know…. Only our general sense of what is typical can decide here, and that background understanding by definition cannot be situation-specific.
Put simply, we need machines that: explore who, what and where I am, and not (only) what I know. We need programming that creates computers that consider human objectives; that know for what (ultimate) purposes they were created; and take responsibility for their actions.
According to Buckminster Fuller, what defines humanity is the ability to identify the operating principles/systems present in the universe; and key is the ‘feeling’ of spatial phenomena. It seems that intelligence must be situated; and made relative to the human/machine body itself; and to the world in which we (and the machines) live inside of. Machines must ‘align themselves’ with the human point of view, and behave accordingly.
It is mankind’s ability to model his immediate and real-world surroundings, and to create plans for the future, plans that are in close accordance to universal laws and also in harmony with his own goals; that separates him from the animals/machines. Humans create solutions using the laws of physics, chemistry, biology; and using all kinds of models, concepts and knowledge about what are the possibilities and limitations of future action. We employ principles, theory and understanding to ‘construct’ the future.
Buckminster Fuller created his own theory of how to understand systems in transformation, which he called Synergetics. According to Fuller, since systems are identifiable at every scale from the quantum level to the cosmic, and humanity both articulates the behaviour of these systems and is composed of these systems. Synergetics is a very broad discipline, and embraces a broad range of scientific and philosophical studies including: tetrahedral and close-packed- sphere geometries, thermodynamics, chemistry, psychology, biochemistry, economics, philosophy, and theology etc. Synergetics was an attempt to differentiate and relate all aspects of reality including: the ideal and the physically realised, the container and the contained, the one and the many, the observer and the observed, the human microcosm and the universal macrocosm.
Buckminster Fuller said of conceptuality/experience:
The greatest of all the faculties is the ability of the imagination to formulate conceptually. Conceptuality (understanding of principles) is subjective; realisation is objective. Conceptuality is metaphysical and weightless, reality is physical… Conceptuality operates experimentally, independent of size… Conceptuality implodes, becoming increasingly more simplified… Experience is the raw material of science. It is the nature of all our experiences that they begin and end. They are packaged, both physical and metaphysical, and are always special case.
At once we glimpse the reason why robots and machines are so different from humans. It is because they do not form and interrelate concepts; based on experience. They do not reflect on their experiences, and on the nature of the world.
Machines do not have operating principles of which they themselves are aware. Robots do not possess a conceptual language with which to record, transcribe and evaluate happenings; and they know nothing of the past, present or future. Thinking requires flexibility of analysis. The thinker must be free to adopt a variety of propositions, and to test them in accordance with reality; and to perceive possibilities. All this is required to understand reality in any way whosoever. The thinker experiments and tests propositions to see if they fit reality; and using overlaid and multiple analogous concepts which are by no means an exact match to the specific item under investigation.
And all knowledge has a social element, provided by the open-thoughts of other people. In Fuller’s terms: I am willing to accredit the experiences of other men when I am convinced by my experiences that they communicate to me faithfully; that is I am able to enlarge my experiences by the experience of other.
And as mentioned by Kim Veltman, we all dream; and have beliefs and wishes, and/or feeling-tones that strongly influence our behaviour. Human knowledge is no mere dead, ossified conglomeration of thought atoms; but is alive with: situating viewpoints, opinions, feelings, experiences, intuitions, experiments, discoveries and productions etc.
As Fuller stated:
Evolution pivots on the conscious, selective use of cumulative human experience and not on Darwin’s hypothesis of chance adaptation to survival nor on his assumption of evolution independent of individual will and design.
A human being is alive to all the possibilities present in the universe; whereas a robot, we must concede, knows nothing of anything; and because it cannot form concepts, analogies and identify principles or generate adaptive plans for action.