X-Message-Number: 14914 From: "Gary Tripp" <> Subject: reality and abstraction Date: Mon, 13 Nov 2000 21:33:23 -0500 This is a multi-part message in MIME format. ------=_NextPart_000_0068_01C04DB9.5968C1E0 Content-Type: text/plain; charset="iso-8859-1" Thomas raises an interesting point when he says: "However there IS another way of looking at identity and its survival: just what attributes of identity at time t0 must be preserved at time t1 > t0 for us to say that the identity is the same? Yes, I am different from the person I was yesterday, but are those differences important or minor? While I personally would say that in the sense of this previous question we ourselves are abstract beings, it most certainly does not follow that we can be treated as if we need not take any special form ie. biological, for instance. Even a little reading on just how neurons work makes me wonder whether a computer in the present sense could really imitate us at all well. Yes, they're nice analogies, but the real question is whether or not they really match us closely when we learn everything about how our brains work. If nothing else, the assumption that we work like Turing machines looks very faulty..." A digital computer may simulate an analogue process to any desired degree of precision but perhaps eventually the original and the simulated version would diverge - OR would they? I believe that close scrutiny of the notion of identity would compel us to concede that we are more than a collection of bones and guts. As Thomas points out, the thought that we can strictly equate identity with our immediate physical state is untenable because we are constantly changing from one moment to the next. Yet there is a unifying thread of coherence over these changes and a sort of homeostasis about our purpose. So in this wider sense a digital simulation may not perform identically at every nanosecond but it would eventually converge to the same outputs given the same inputs in much the same way that a dampened iteration may find a solution to a system of equations faster than a simple iteration yet both would converge to the same limit. On that note, the myriad little mental subprocesses that ultimately coalesce, combine and then percolate to the surface of our consciousness as a thought might be streamlined so that they achieve their contributing affects in a more efficient way. The end result would be the same and perhaps the feeling would be the same but the precise details of these machinations would be different. How far could we go with this? We could continue to make adjustments in subtle ways until we've radically changed the underlying structures for optimal efficiency yet preserved the outward behaviour of the system. I feel that we must redefine identity in a way that accounts for these possibilities. We cannot be mere guts & bones, nor, it appears, can we be a fixed algorithm; rather, it would appear that we are an equivalence class of algorithms. I speculate that digital computers would be entirely adequate for our simulation. /gary ------=_NextPart_000_0068_01C04DB9.5968C1E0 Content-Type: text/html; [ AUTOMATICALLY SKIPPING HTML ENCODING! ] Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=14914