Volume 1, Issue 4 
4th Quarter, 2006


Scope and Resolution in Neural Prosthetics and Special Concerns for the Emulation of a Whole Brain

Randal Koene, Ph.D.

page 4 of 5

The question is - is it even possible for a human who receives neural prosthetics that amount to whole brain emulation to tell us if personal identity is preserved? And the larger question remains: Is it imperative that the subjective experience is unaltered?

Koene
Image 7: Preservation of Personal Identity

The medical procedure for whole brain emulation needs to deal with the distributed nature of memory, its presence in all loci affected by experience. We may even consider a memory external when our actions change the environment, such as when we write in a journal. Of course, the question of scope here usually refers to the human body. 

And on the other side, just as the development of prosthetic hands seeks fine motor control and the development of retinal prostheses seeks resolution, resolution is the other objective of neural prosthetics. Characteristics of behavior need to be retained and neither the ability to retrieve, nor the resolution of specific memories should be affected. And, the quality of the experience in the new substrate cannot be a washed-out copy. In other words, the granularity of the totality of reproduction must be such that its embodiment of personal identity remains strong and clear.

There is another dimension to all of this, as the work of Bruce McCormick and Henry Markram clearly shows. Whole brain emulation is a tool for exploration in neuroscience as much as it is a potential medical procedure. As an optimal model of the human brain with complete access to all its parts, whole brain emulation can greatly facilitate exploration of intelligence, psychology and humanity. 

At this point, there are obvious ethical questions that need to be resolved, since a conscious emulated mind might rightfully be considered sentient, to a large degree the equivalent of an aware human being. Nevertheless, careful exploration may well fall within the realm of our current scientific procedures, since a plethora of regulations already exists that govern work with human subjects.

Now I am moving on to the topic of resolution, which is really all about expected function.

The biological implementation of brain functions depends on biophysical mechanisms, such as the precise location of synapses between axonal and dendritic fiber, the delays of signal propagating between neurons, specific synapse type, and the strength of a connection as expressed by receptor field sizes. 

Despite this, implemented function does not depend critically on the performance of any one individual element. It is group activity that is paramount. Implementations in a neural network are strong, self-correcting, redundant, and homeostatic. A large number of unreliable components are linked in a very precise manner that utilizes surprising detail of biophysical properties to establish robust function. 

My first hypothesis is that group effects are relevant. There’s an acceptable range of responses, or variance, that is due to differences in elemental responses at different occasions. My second hypothesis is that as long as any changes in the elemental details responsible for group function keep the group effect within the acceptable range of responses, those changes are inconsequential.

As we must know the group function and its range of responses, we may try to decode those effects directly during analysis of the subject. Alternatively, we can replicate the element performance as rigorously as possible, thereby seeking to incorporate any effects, even those that we do not know how to elicit and measure. 

A problem for the second approach is that, unlike physics, where increasing the number of particles simplifies predicting the behaviors of the group; in biology, an increase in the number of components increases the theoretical complexity of their group behavior. 

 

 

1 2 3 4 5 Next Page>