Jump to content
Objectivism Online Forum

Hal

Regulars
  • Posts

    1212
  • Joined

  • Last visited

Everything posted by Hal

  1. I dont think he's talking about necessity as opposed to contingency in the contexts you mention, although he does use it in that way as well at points. I generally agree with Peikoff that the necessary/contingency distinction is flawed, but Kant accepted it and it does feature in his work quite often, particularly when he discusses mathematics and logic. A being a necessary condition for B means that B cannot occur without A. For instance, the television being turned on is a necessary condition for you watching the basketball game. When Kant talks about the "necessary conditions" for experience, he means those things that have to be true in order for us to actually have the experiences we have. For instance, in order for it be possible for someone to see a table, he must have some kind of receptivity to sense-data. In order for it to be possible for an object to understand arithmetic, it must have a mind (sorry, these examples are pretty bad but I couldnt think of any better ones). Kant wants to talk about what is necessary for the possibility of experience in general - ie what conditions have to be satisfied in order for us to perceive the world in the way which we do. Its a fact that we DO perceive certain things in certain ways (for instance we see objects AS objects rather than as series of discrete images and we also have clear knowledge about concepts such as casuality, which Hume showed could not have arisen from experience alone), and Kant tried to investigate what this implies about the structure of our cognition, and about knowledge in general.
  2. Why is there any reason to believe humans can solve the Halting Problem? Does the following program terminate? : int i=4; bool loop=true; WHILE (loop) { IF FALSE (i is expressible as the sum of 2 prime numbers) THEN loop=false; i=i+1; } end;[/code] Remember, being able to solve the Halting Problem means that you can have a decision procedure for checking whether ANY arbitrary program terminates, not just a particular one. reason for editing: I forgot to add the i=i+1 statement.
  3. An analogy is just an analogy. The hardware/software distinction provides a decent metaphor for a certain model of consciousness (mind produced by events in the brain), but that's about it. People generally try to explain the unknown by reference to the known, and it so happens that a lot is known about computers, and very little about the workings of consciousness. I don't really think this is relevant to the AI debate though; the issue isnt whether the brain 'is' a computer, but whether something akin to the mind/brain could be replicated on a computer. Even if someone managed to produce a fully conscious/volitional/whatever robot tomorrow, it would no way imply that the brain 'is' a computer or that the mind 'is' a computer program.
  4. What do you mean by running a computer simulation? You could certainly learn to fly a plane by use of a computer simulation in a virtual reality environment, and I believe that pilots are often trained that way. There would be no phenomenological difference whatsoever between using a sufficiently advanced computer simulation, and flying an actual plane. If I've missed your point then please clarify what you mean. But there's no difference from the point of view of the person as he is experiencing it. When the imagining is sufficiently vivid (such as during a dream or hallucination) there is nothing to distinguish between what is purely a product of the mind, and what is real. It is only after one wakes up that the difference can be appreciated - ie only when viewed from the outside. I meant the relations to which the parts stand in to each other. The world consists of relations as well as objects, and I was using 'structure' to describe the former. When I say that a computer could replicate the structure I mean that it could duplicate the relations using different 'non-material' objects. For instance a computer running a flight simulation would have to duplicate all the structural relations that flying involves (such as the sky being above the plane, the location of controls in the cockpit, and so on). The computer wouldnt be using 'real' objects (the 'sky' in the program obviously isnt the actual sky), but the relations between parts would be identical. Likewise when a child plays with a model airport, the structural relations of the toy duplicates the structural relations of a 'real' airport. The model plane is on the model runway, and the model pilot is in the model cockpit.
  5. I don't have a copy of the Critique handy, but I'll do the best I can off the top of my head. Yes, in a sense the Critiques were about judgements, but more about the necessary conditions for these judgements to be possible. Kant was primarilly interested in HOW we are able to do what we do - ie what epistemic conditions must be satisfied in order for the possibility of judgements to even exist. No I think this is too narrow; it sounds more like logical positivism than Kant. Kant subdivided logic into several categories. What he called 'pure' logic was essentially the traditional interpretation of Aristotlean syllogistic logic, which was concerned with the abstract study of forms - ie what is left after abstracting away all the particular characteristics of objects. He believed that 'pure' logic. as a science, was essentially complete. However he also wanted to develop what he called "Transcendental Logic" in order to address the questions which were of interest to him. Transcendental logic does not abstract all qualities from objects in the sense which 'pure' logic does; it abstracts away all their empirical qualities, but leaves behind what Kant called their 'transcendental' qualities (ie the necessary form they must have in order for our experiences of them to be possible) The closest thing Kant recognised to Rand's "non-contradictory identification" would be what he called "applied logic". I'm unsure how to best explain the distinction Kant makes between pure and applied logic, but I think the best way is so say that 'pure' logic is ENTIRELY general - it is not restricted to humans, but would apply to ANY being in the universe capable of thought. Applied logic on the other hand would be logic 'for humans' - ie one that takes into account the specifics of human consciousness and mental processes, such as their memory and other cognitive functions. In a sense, applied logic would be an outline of the best way for humans to reason, although I'm not sure that Kant would have used this terminology. I think that Rand's definition of logic would almost certainly have been subsumed by Kant under 'applied logic', but applied logic wasnt something Kant was particularly concerned with in the first critique and I havent read any other of his works, so I could be wrong. In addition to all this, Kant believed that every particular science, has its own specialised logic which is concerned with the specific objects which it studied. Pure logic would be, in a sense, a generalisation of these 'specialised' logics, although again I doubt Kant would have used this terminology. In summary: GENERAL LOGIC Pure logic - abstracts away all conditions of objects Applied logic - deals with the cognitive function of humans; "how should we reason" PARTICULAR LOGIC The logic of a particular science. Will vary depending on the objects investigated y that science TRANSCENDENTAL LOGIC Abstracts away all empirical qualities of objects, but leaves behind those contributed by the understanding (their transcendental qualities)[/code] He also went on to make several more subdivisions, such as 'analytic' vs 'dialectic' logic, although I believe that this was the convention of the time rather than being an innovation of Kant's (not 100% sure about this). Whatever else can be said about Kant, he certainly had a methodological mind. Again, Kant tended to subdivide Reason into different types, each limited to its own sphere. The Critique of Pure Reason outlined what he described as 'pure' reason, and the Critique of Practical Reason does the same for his 'practical' reason. I seem to remember him defining both terms in the preface to the first critique but as I said I don't have a copy handy; I'll dig it out when I get home. I would recommend Henry Allison's "Kant's Transcendental Idealism" if you were interested in reading Kant (I'd probably recommend it instead of reading the actual Critique to be honest; Kant's prose is appalling). It's not an outline of the Critique but rather an interpretation, and one which I found a lot more defensible than most traditional works of Kant "scholarship". Yes I think that's correct. although I don't speak German either so I'm not 100% sure. Often they get translated into English still capitalised, which is why oddities such as 'Being' and 'the Other' tend to feature a lot in continental philosophy.
  6. You could certainly argue that the specific details regarding how humans actually go about synthesising sensations in practice is an issue for biology and child psychology, but Kant's concern with the necessary conditions for the possibility of experience in general do seem to fall firmly into the philosoophical category since they are intimately tied to epistemology and the philosophy of mind. For instance, one of Kant's insights was that our understanding of concepts (I use the word 'concept' here in the Kantian sense, which is not directly equivalent to the sense in which Rand used the term) such as 'object' or 'entity' must in some sense be a priori - they cannot be derived from experience alone, because unless we already had the concept of 'object', we couldn't experience them at all - we would only experience unrelated sensations. There seems to be no way to move from a series of disjoint images into the perception of an object as a unity unless you already possess, in some way, the idea of an object. Similar arguments apply to ideas such as the 'self' - it doesnt make sense to say that we derive our concept of the self purely from experience (although our explicit formation of it certainly comes from analysing what we implicitly do anyway) since as Hume pointed out, the 'self' is not given in experience, only a series of disjoint mental states. The fact that we DO manage to form an idea of the self despite this (ie the fact that passage from the sensual level to the perceptual level is actually possible) seems to imply certain things about the structure of human knowledge and the mind, and this formed the basis for Kant's attack on naive Humean empiricism. Issues such as these are definitely philosophical and I can't think of how a scientific investigation of them would even be possible.
  7. I don't think this is entirely true; they both used the word reason to describe different things, but this is because they were attacking different problems (and remember that Kant's use of the term 'reason' is a translation from his German, and was the standard usage of the term at that point in history, which was largely dominated by the philosophy of Leibniz and Wolff). Kant was concerned with the 'low level' problem of how experiences are possible, including the necessary conditions for having them. Therefore, he used reason to refer to the faculty which humans to pass from receiving sensations ('intuitions') to perceiving actual objects, via the process of synthesis (ie the bringing of sensible intutions under the categories of the understanding). From what I've read of Rand she doesnt seem concerned with this issue - she agreed that it occurred somehow, but wasn't interested in the specifics (in Rand's terminology, I think Kant's 'synthesis' would be described as "the passage from the 'sensory level' to the 'level of perception'"). Since I havent found any work by Rand in which she writes about this at any length, its understandable that she wouldnt have needed the concepts which Kant used when analysing it in great detail. Likewise Kant wrote very little about the areas of epistemology with which Rand was concerned (namely how humans form and use concepts at the day-to-day level), which is why he wouldnt have used the concepts Rand used when discussing this. Different areas of philosophy (and philosophers) generally use different vocabularies, especially when they are seperated by several centuries.
  8. I'm not entirely sure what you mean here. It is certainly possible at the present time to create AIs which operate according to non-classical logics, but you're still limited to the physical properties of the hardware itself. No matter how advanced your logic is, it's still going to be translated into basic 'if then else' statements of machine code, which are going to be executed on a CPU composed of deterministic transistors and other electrical components. The only way to get around this limitation is to either a) construct a computer which is somehow non-deterministic at the hardware level, if this is even possible, or to find a way in which these 'IF-THEN-ELSE' statements can somehow transcend determinism as a result of structural complexity, in the same way which matter in the human brain does according to some models.
  9. The way I see it, there are two possibilities; either a) the physical world is entirely deterministic, yet still manages to produce volitional entities, in which case there is no reason why a fully deterministic computer program would not be able to do likewise, or the physical world is not fully deterministic, in which case there is no reason why a computer would have to be fully deterministic. Either option leaves open the possibilility for volitional computing. You seem to be wanting to say that, although nature is deterministic, the fact that computers are also deterministic means they are incapable of 'choice' or consciousness. I think that this is intrinsically self-contradictory. No I don't. I think that if the human brain can be considered as a sufficiently complex deterministic program of some sort, then a sufficiently complex deterministic structure is somehow able to produce creatures with true volition. I've no idea how this is possible and it's something that hopefully science will clarify in the future, but I find it no more 'strange' than the idea that deterministic physical 'matter' is somehow able to produce creatures capable of perceiving other pieces of matter and having intentional experiences. I'm not sure what you mean. A lot of physical phenomonen can be simulated by computer, and I can't think of any reason not to assume that this would also apply to the brain. If a program is capable of fully modelling a physical system then this would entail that 'lines of code' would be capable of replicating the physical structure of the human brain. Since I believe that consciousness, including volition, arises from this structure rather than from the properties of the physical material itself, this means that I generally refuse to rule out the possibility of computers with volitional concsiousness(although how we could ever know that a particular computer has actually obtained consciousness, let alone volitional consciousness, is an entirely different question...)
  10. It's not a bad idea in theory. Most of the arguments against anarchy scale up to the national level; a world government functioning as an ultimate arbiter between countries would serve the same purpose as the state which arbitates between individuals. Obviously the UN is a terrible attempt at world government, and has probably turned quite a lot of people against the idea altogether, but one that was based on more proper principles would be less objectionable.
  11. Hi, just thought I'd post a brief introduction. I'm a philosophy student from the UK and I first read Fountainhead and Atlas Shrugged a few years ago. Although I enjoyed them, I didn't follow it up with any Objectivist non-fiction. A friend recommended OPAR to me about a month ago however, and I've been working my way through it along with Introduction to Objectivist Epistemology. I found this forum while googling for an Objectivist board so that I had somewhere to come if I had any questions about the material, since I don't really know anyone who has read Ayn Rand in any great depth.
  12. This isn't true; Feynmann showed in 1982 that a computer designed to exploit quantum principles would be capable of modelling phenomena that couldn't be modelled classically, and his ideas were later extended into a idea of a general quantum turing machine. As far as I know, the question regarding whether quantum algorithms could get around the Church-Turing thesis is an open questions within complexity theory at present. There was an (in)famous paper published by Tien Kieu a few years ago where claimed to have found an algorithm capable of solving a classically undecidable problem (Hilbert's tenth), and although some errors were later found in his paper causing him to tone down some particular claims (I believe he switched to 'such an algorithm may be possible'), the debate rages on. Most experts do agree with you, but there are certainly dissenters, and to completely write off the possibility of new computing models seems a bit hasty - people have only been investigating QTMs for around 20 years, so I would be sceptical of anyone claiming to have definitive conclusions at the present time. In any case, I'm not sure why it matters. The fundamental particles of nature may turn out to operate according to fully deterministic laws, but this doesn't prevent them from forming systems that are capable of volitional action (ie humans). I can think of no strong reason to assume that a "sufficiently complex" computer program would be unable to produce the same result, even assuming that it was entirely deterministic at the level of individual code lines. I'm not prepared to believe that theres something magical about the material making up our brains that makes it capable of implementing structures which can't be modelled abstractly, unless someone can show me some pretty solid evidence.
×
×
  • Create New...