On Tue, Mar 5, 2013 at 5:43 PM, Craig Weinberg <whatsons...@gmail.com> wrote: > > > On Tuesday, March 5, 2013 8:39:37 AM UTC-5, telmo_menezes wrote: >> >> Hi Craig, >> >> On Tue, Mar 5, 2013 at 1:44 PM, Craig Weinberg <whats...@gmail.com> wrote: >> > On Monday, March 4, 2013 11:27:21 PM UTC-5, Pierz wrote: >> >> >> >> Really Craig? It invalidates mechanistic assumptions about eyes? I'm >> >> sure >> >> the researchers would be astonished at such a wild conclusion. All the >> >> research shows is brain plasticity in interpreting signals from unusual >> >> neural pathways. How does that invalidate mechanism? >> > >> > >> > It's not that wild of a conclusion. The experiment shows that we cannot >> > assume that vision is the result of a passive process that relies on a >> > one-way path leading from light to eye to optic nerve to brain. >> >> No, it just shows that we cannot assume that the eye has to be >> connected to the optic nerve specifically. > > > Yes, but I think that's only part of what it shows. It also shows that the > brain and spinal cord have the general intelligence to recognize and > integrate the eye as an eye. It's an active system. It's not just a matter > of saying that you can receive mail at more than one address, it is that > your mail will figure out where you are living without having instructed the > post office.
Agreed, it does show that. I'm not going to claim that we have algorithms as good as the brain as doing this sort of thing, but we certainly do have algorithms that show a glimpse of this type of possibility, and they have been improving steadily. Unlike consciousness, there seems to be a viable way forward in terms of self-organising, decentralised systems. We know how to build them, to a degree, and we are getting better at it. We have (possibly incomplete) view of the first principles. One simple example is genetic programming. You give the algorithm a set of primitives to build programs towards a certain goal, and we've seen, over and over again, that the algorithm is capable of finding uses for these primitives that were not anticipated by the human designer. > >> >> > The brain >> > actively shows that there is a path leading the other way as well, as >> > the >> > whole organism seeks to see through the eye. >> >> The brain is always looking for patterns in its inputs that could be >> useful. > > > a computer is always scanning its ports and slots for activity also. That > doesn't mean you can just solder a DRAM card somehwere on the motherboard > and expect to use it. Sure, but that's just a restriction imposed by the von Neumann model of computation. Other models are possible where this is possible. > >> >> >> > This shows that there is >> > sensory-motor activity going on within the micro-level of the tadpole as >> > the >> > rather under-signifyingly termed "plasticity" knows exactly what the >> > eyeball >> > is, and finds a way to use it. >> >> Or, the brain is just capable of recognising old patterns from a new >> source. > > > When you say that the brain is "just capable of recognizing", that is > already sense. You're not saying that this capability is just luck or > telepathy, you are saying that there is a particular sense interaction in > which neural tissue initiates ephaptic or other contact. It's not like the > patterns are leaking out of the eye in some formless way, it has to be > recognized that this organ as something which can be used as a sense organ > before it can get any signals out of it. This brings up the question also of > 'why have sense organs at all'? If the brain can just recognize old patterns > from new sources, why not just use anything you can touch as an eye or an > ear? Because specialisation is needed to attain things like light or sound sensitivity. >> >> >> > Try that with your computer. See what happens when you try plugging a >> > microphone into a DRAM slot or listening to your car radio through the >> > transmission. >> >> We know of many algorithms (possibly equivalent) that could be used to >> achieve something like that. They could require human assistance -- is >> this what you want me to do? -- but so do humans. This, of course, >> provided you are willing to disregard interface incompatibilities that >> are outside of the control of a normal computer. But I can't see why >> hardware without such incompatibilities could not be built. It's just >> that there isn't any incentive to do it at the moment. > > > Sure, yes. The assumption of mechanism however, should lead us to expect > that primitive organisms would be like the early machines that humans have > built so far. Just as we have no incentive - why would biology have any > different incentive? The opposite seems to be the case - human machines are > founded on a rigid, unambiguous ontology, whereas biological organisms are > founded on flexibility and ambiguous relation between generality and > specialization. I don't find this surprising at all. Human-created systems are typically built from the top down with macroscopic tools and processes, and they are designed from the beginning with a certain goal in mind. Biological systems are grown from the bottom up with nanotechnology and no goal. They accumulate adaptations as the environment itself changes. >> >> Notice that I'm not attacking your theory, I don't grok it well enough >> for that. I'm just objecting to this specific argument, because I find >> there are simple explanations within the realms of we already know >> about the brain. For example, we know that entire sectors of the brain >> can be repurposed after an injury. > > > A brain repairing itself is a little easier to explain computationally > (basically like RAID 5 drive rebuild.. data is stored in such a way that it > can be reconstructed through triangulation of a missing drive). The idea of > a new drive being inserted into some random slot directly and having the > computer begin using it is a little different. The eye could be a tumor or > some foreign object, it would have to have some way of recognizing what it > is first before it could be useful. This doesn't make sense under the > assumption of visual sense as a passive machine in which photons strike the > retina, signals, travel to the brain, brain interprets. The brain has to be > interpreting the eye itself, as a organ which it collaborates with, not just > an abstract source of signals, as it would be in a computer. I think you're taking this analogy too far. The lack of certain capabilities in commodity computers is not a proof of the impossibility of recreating those capabilities artificially. It is certainly possible to imagine an OS that is capable of adaptation, by using already known algorithms like artificial neural networks, evolutionary computation and so on. Cheers, Telmo. > Thanks, > Craig > >> >> Best, >> Telmo. >> >> > >> > Craig >> > >> > -- >> > You received this message because you are subscribed to the Google >> > Groups >> > "Everything List" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> > an >> > email to everything-li...@googlegroups.com. >> > To post to this group, send email to everyth...@googlegroups.com. >> > Visit this group at >> > http://groups.google.com/group/everything-list?hl=en. >> > For more options, visit https://groups.google.com/groups/opt_out. >> > >> > > > -- > You received this message because you are subscribed to the Google Groups > "Everything List" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to everything-list+unsubscr...@googlegroups.com. > To post to this group, send email to everything-list@googlegroups.com. > Visit this group at http://groups.google.com/group/everything-list?hl=en. > For more options, visit https://groups.google.com/groups/opt_out. > > -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscr...@googlegroups.com. To post to this group, send email to everything-list@googlegroups.com. Visit this group at http://groups.google.com/group/everything-list?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
