Von Neumann Minds: A Toy Model of Meaning in a Natural World by Jochen Szangolies

Dear Conrad,

I don't know how I missed your reply earlier---sorry for that. And thank you for your kind words!

I agree that representationalism isn't necessarily the only way to get meaning out of some system; one could, for instance, also think in terms of subsymbolic approaches. Representationalism's main virtue, to me, is that if it works, it's completely clear how---by simply having some vehicle standing in place of some object or state of affairs. But of course, this direct route is blocked by the homunculus; hence, my attempt to patch things up. If that turns out not to work, it might be necessary to abandon representationalism altogether, and move on to something else; but since, to me, this seems to entail a certain loss of intuitiveness and clarity, I'm going to keep on digging on this ground until I'm absolutely certain I'll never strike gold.

I'll certainly have a look at your essay; maybe I'll find something interesting to say about it.

However, a point of clarification: I don't understand my model as being mainly computational; in fact, I'm skeptical of computational models. I know that usually CAs are thought off as a computational system, but that just means that they are systems that can be used to compute, not that they are intrinsically computational. To me, what's more important is the pattern, which is a physically real thing (an analogy to the pattern of neuron firings in a brain), and its properties. The meaning I see is the semantic information the pattern contains about itself, and about the environmental conditions. But that's not a point I wanted to put too much emphasis on in the present essay.

Anyway, thanks again for your comment!

Cheers,

Jochen

Dear Dizhechko Boris,

sorry for not replying earlier. Thank you for appreciating my essay; I will have a look at yours---however, I must confess I am somewhat skeptical that it needs a new physics in order to make sense of intentional, goal-directed behavior. But I will try and form an unbiased opinion of your work.

Cheers,

Jochen

Dear Christi,

thanks for the kind words! Yes, I think that even if there's some germ of truth in my model, it'll be a long way yet before it'll be clear whether it actually solves the problems it sets out to solve. I think things are looking somewhat hopeful at the current stage, and the main virtue is that it provides a relatively concrete, well-defined model to play with; so I think there's a justifiable hope that even if things ultimately don't work out, we'll get out some useful pointers regarding what not to do.

Cheers,

Jochen

Dear Miles,

I'm glad you found something of value in my essay! You're right, I think of this model as a kind of 'hands-on' test bed for my ideas; and as you point out, there's still lots to tinker with.

Regarding the question of identity, I'm afraid I don't have an answer. In a sense, it's analogous to the question of when a speciation event occurs---when was the first little bundle of feathers clawing its way out of an egg no longer a dinosaur, but a bird?

I'm not sure the question is very meaningful, at least in that case: the boundary between 'bird' and 'dinosaur' is ultimately as arbitrary, and as man-made, as the boundaries between nations on a map. But is there more meaning to the question in the case of brains/minds? Lots of people, from Hume to the Buddha, didn't think so. I, for myself, am just going to continue tinkering with my model for the moment.

Cheers,

Jochen

Dear Robert Groess,

thank you for the kind words! Yes, it's certainly a testimony to the genius of von Neumann that his work continues to influence and direct modern ideas---he ought to be rated much higher on the list of all-time greatest minds than he usually is.

I'll try and take a look at your essay, too!

Cheers,

Jochen

Ah, forgot to comment regarding the two-tiered dynamics of quantum mechanics. At present, I'm not sure if one can really formalize a parallel, but just on the level of analogy and metaphor, these things may not be too far away from one another---there have often been attempts to link quantum mechanics and self-reference (one prominent exponent of this view being John Wheeler), and of course, the bipartite structure of von Neumann's replicators is exactly due to the problems of self-reference (which makes a self-scanning mechanism impossible). So well, maybe?

Dear Donald,

thank you for taking the time of reading and commenting on my essay. Regarding intentionality, I agree that the concept is treated somewhat vaguely in much philosophical literature, but I'd say my level of rigor is par for the course, at least---compare my definition: "Mental content exhibits the curious property of intentionality--of being directed at, concerned with, or simply about entities external to itself.", and that of the Stanford Encyclopedia of Philosophy: "Intentionality is the power of minds to be about, to represent, or to stand for, things, properties and states of affairs." Both essentially follow Brentano, who introduced the concept into modern discourse (saving it from the old scholastics).

Again, I can understand---and to some degree, sympathize with---wanting a more thorough definition, but sometimes one also risks getting embroiled in petty turf wars when trying to clarify every last definitional issue ('rigour mortis'). So rather than spending most of the allocated room on such definitions, I chose to introduce my model instead, hoping that this would help clarify lingering issues---apologies if it didn't.

Regarding Richard's paradox, the ordering itself isn't really so important; one merely needs an unambiguous way of referring to certain elements (either of English phrases corresponding to natural numbers, or of behaviors of a given automaton). Any association between these elements and natural numbers will do fine, since then, you can refer to the nth element, which picks out a unique one; then, you can use the diagonalization trick by creating a new element that wasn't part of the original association.

But since that was claimed to be complete (a list of all the English sentences describing real numbers/a theory of all behaviors of the automaton), we arrive at a contradiction.

Regarding self-replication, you raise a good point---natural replication is indeed never exact. This doesn't necessarily address the infinite regress, though: if the parent needs to have access to a plan of the child in order to construct the next generation, we still get a regress, even if all of the plans are allowed to differ. Furthermore, when replication is inexact, we start getting into issues of vagueness: when is a construct a 'child'? How similar do parent and child-generation have to be in order to constitute an example of self-replication? If a stone, rolling down a hill, breaks off another, is that an example of self-replication?

Lastly, you're dead-on regarding perception: it's indeed a far more active process than my caricature gives it credit. But whether the outside world is just faithfully projected onto an internal screen, or whether a sort of virtual internal reality, perhaps only loosely 'inspired by actual events', is created, doesn't make a difference for the conceptual point: both implicitly presume some central meaner (as Dennett calls the homunculus) using the internal representation as pertaining to the outside world. And with that, we're already off the rails as far as a theory of representation goes.

Again, thanks for your thoughtful comments!

Cheers,

Jochen

Dear Stefan,

thanks for your comment. I have to say that I feel somewhat narrow in focus in this competition---most people seem to propose entire cosmologies, while I just play around with cellular automata!

I'm glad, though, that some people seem to find some value in my ideas nevertheless. I'll have a look at your essay!

Cheers,

Jochen

Dear Torsten,

thanks for your comment. Glad you found something to like about my work!

Regarding evolution, I'm unsure if I would really say that mutation adds information---in a sense, mutation merely creates an ensemble of possible signals; selection then chooses among these. The ensemble of messages doesn't really carry information, but choosing one of the options then at the very least carries information about the entity making the choice---in my case, the environment, as mediated by the cellular automaton. But that's maybe something for another day to ponder.

I had actually already read your essay, and found it very intriguing; although I apparently didn't add a comment (I find it hard to keep track of conversation threads in this forum). Thanks for your good wishes, and right back to you!

Cheers,

Jochen

Dear Marc,

thanks for your kind words! Although I have to say, I'm still a little humbled by the breadth and depth of ideas and concepts presented in this contest. I mean, of course I have my own ideas about what the world, deep down, is like, but I'm not sure I'll ever consider them well-developed enough to risk airing them in such a forum---so all the more props to those who do!

Von Neumann truly was a thinker of rare accomplishment; I'm happy enough if I can help popularizing some of his ideas.

Thanks for the well wishes!

Cheers,

Jochen