> Aaron Davidson <ajd@ualberta.ca> wrote:
> It just occured to me (regardless of whether my thought is original
> or not) that nanites could easily be used to create very real virtual
> realities.
It has to be pretty original because we have never had the data before to know if it is doable or only a fantasy.
> Instead, nanobots could reside within sensory neurons controlling touch,
> balance, motion, vision, smell, and hearing. These nanobots could all be
> controlled and coordinated by some external computer sending wireless
> signals.
They may not actually reside "in" the neurons, but sit in the synapses. Nanobots, I beleive, don't have the power to interfere with the electrical signals, but can wrestle with the chemical messages. The "wireless" part gets very tricky for 2 reasons: (a) You can't receive RF very far into the body [covered in Nanomedicine], so you would need a skin surface antenna; and (b) potential bandwidth problems. While I don't see a problem with sound and probably taste, there may be problems with visual/smell/touch. As I recall the estimates for a "realistic" visual environment (for say flight simulators) are in the vicinity of a TeraOp. If you have to individually stimulate each photoreceptor in the eye that is a lot of data. You might be able to do it if you move downstream where the data is more compressed. It would take sitting down with Nanomedicine and some texts on visual cortex processing to come up with the right way to do this. Someone I'm sure will point out that you could transmit in compressed data. The problem then becomes where you locate the computers to do the decompression and potential heat problems as you put more processing into the brain (you have to remember your body is a radiator for your brain).
My guess is that it is doable, but it would require a non-trivial amount of thought for the best architecture.
> The nanobot could inhibit a sensory neuron from firing naturally or excite
> a neuron when the VR simulation demands it.
Yep, the nanobots can suck up the neurotransmitters or release them on queue.
> I wonder why no one has discussed this alternative.
Because very few people think at the micro/nano-scale.
Without the data to back it up most people would think of it as a kooky idea.
> I can imagine the difficulties would lie in the rapid communication and
> coordination to make the virtual reality coherent.
Yep. Since nanobots are going to normally coordinate with ultrasonic
sound, you will have to worry about the sound density (and power to
produce the sound) heating things up too much. You can probably lay
down a fiber network in the brain to eliminate some of this problem.
Of course pretty soon, you start to beg the question of why keep the
neurons at all? (Since they are such poor computers.) Robert didn't
address this in Volume I (probably to keep people from running screaming
into the aisles), but I think it may be included in Volume II or III.
> I cannot remember what
> you said in regards to what sort of communication method would be used
> between an external device and cellular nanobots. Is it possible to achieve
> real time communications without frying the brain?
Physicians would probably use pulsed ultrasound (its got a higher bandwidth than things like touch or normal sound frequencies). If you lying on a ultrasound "grid" bed, the bandwidth could get pretty high since you can make the grid size quite small (probaly down to a few square microns). You can always go to a full fiber interface (but then you are "jacked" in). [The Matrix scenario is pretty realistic all things considered.]
As far as frying the brain goes, I don't think Nanomedicine goes through and adds up all the power requirements for various senses or deals directly with full VR bandwidth requirements. The data you need to get a rough feel for the ease or difficulty I believe is present. Of course since the book will be available in a month or two, you will have plenty of time to extract and summarize this data for your Extro5 talk: "Nanobots for VR: A new approach to entertainment". :-)
Robert