Keith Wiley and Randal Koene on gradual versus instantaneous whole brain uploading on personal identity and personal survival

Carboncopies.org is a nonprofit organiaation with a goal of advancing the reverse engineering of neural tissue and complete brains, Whole Brain Emulation and development of neuroprostheses that reproduce functions of mind, creating what we call Substrate-Independent Minds (SIM).

Carbon copies is led by Randal Koene.

Randal and his team keep very close track of all developments related to Whole Brain Emulation progress and Keith Wiley and he have written or co-written many important papers related to Brain emulation.

The human brain is composed of 100 billion nerve cells, called neurons; each neuron may connect to thousands of other brain cells. The trillions of connections are necessary for the integrative power of the brain. The brain can be divided into three areas: brain stem, cerebellum, and cerebrum.

Arxiv – The Fallacy of Favoring Gradual Replacement Mind Uploading Over Scan-and-Copy by Keith Wiley and Randal Koene This Arxiv version is not the final version of this paper. The final paper is slightly different from the version offered here, has been accepted for publication, but due to the copyright restrictions of the journal involved, cannot be offered through public servers such as arxiv until 2018. The version offered here is a reasonably close approximation of the final version

Mind uploading speculation and debate often concludes that a procedure described as gradual in-place replacement preserves personal identity while a procedure described as destructive scan-and-copy produces some other identity in the target substrate such that personal identity is lost along with the biological brain. This paper demonstrates a chain of reasoning that establishes metaphysical equivalence between these two methods in terms of preserving personal identity.

This view is often colloquially stated as “Even if scan-and-copy did work, it still wouldn’t be me, just a copy.” We restate this position in somewhat more formal terms in the detractor claim:

Mind uploading via slow gradual in-place replacement will preserve the identity of the person associated with the biological brain and therefore represent their personal survival. However, mind uploading via destructive scan-and-copy, even assuming technical efficacy in that it produces exactly the same post-operative material brain as the gradual in-place replacement procedure, a brain of corresponding neural function and a person of corresponding psychological continuation, will nevertheless generate an entirely new identity, leaving the person from the biological brain “behind”, and will therefore represent death of the person who preceded the procedure.

This article only addresses this precise detractor claim. It does not address questions such as whether either procedure may be impractical on technical grounds or whether certain considered procedures are more prone to imperfections or errors that would result in flawed duplications of neural function and corresponding psychological qualities. Note that the detractor claim is purely metaphysical in nature. Both the colloquial phrasing and the formal statement above grant total procedural success while nevertheless making differing claims about identity between two physical procedures. In this way, the claim can only be metaphysical, namely on the nature of identity (or alternatively survival, as we will discuss). Consequently, considerations of procedural conceivability, practicality, or flaws are outside the scope of this paper.

Personal identity and personal survival

Sometimes it is useful to speak in terms of personal identity and other times in terms of personal survival. We must therefore disambiguate these two concepts. One could conceivably view them orthogonally, thereby offering four possible judgments to a destructive mind uploading procedure:

1. Identity is preserved and the person survives.
2. Identity is preserved but the person does not survive (i.e., they die).
3. Identity is not preserved, thereby producing some other identity, but the person survives anyway.
4. Identity is not preserved and the person does not survive.

Of the four possible judgments, the second one is fairly bizarre upon closer consideration, so we can dismiss it (what popular theory of mind or identity would work in this way?), but the other three are all possible. Crucially, the third option is possible, in which by some definitions, identity may be lost without sacrificing the notion of survival. This idea may seem equally bizarre at first glance, but actually it reflects more formal definitions of identity.

Neural Prostheses

The concepts behind the terms neural prostheses, whole brain emulation and mind uploading are related, and in that order their objectives are of increasing complexity. A neural prosthetic is a replacement for or augmentation of a function or component of the nervous system in general, and of the human brain in particular. Currently, a number of neural prostheses exist that replace or improve specific functions that are most commonly sensory functions. Examples are cochlear and retinal implants, but also implanted electrodes that inhibit seizures and research toward the development of a hippocampus prosthesis.

A neural prosthesis can be implemented as a hardware component, as software in a general computing environment or as a mixture of both.

Equating Destructive Scan-and-Copy with Instantaneous Replacement

We begin by showing the equivalence of scan-and-copy and instantaneous replacement. In the former, the brain is frozen, sectioned, and scanned, and the scan data is used to build an artificial brain composed of billions of prosthetic neurons of perhaps electronic or optical design (or even some more exotic mechanism yet to be developed). In the latter, the brain is infused with billions of prosthetic neurons that travel to assigned biological neurons in a one-to-one relationship, passively observe the neurons to model their behavior (their functions), and then instantaneously replace all the neurons at the flip of a master switch. For the purpose of the thought experiment, these two procedures produce the exact same physical product, as if the scan-and-copy method had frozen the brain for subsequent sectioning at the very same moment in time that the instantaneous replacement procedure would have otherwise flipped the master switch

Some readers may feel that the distinction between instantaneous replacement and scan-and-copy lies in the likelihood that scan-and-copy involves a greater spatial translation of neural function from the biological brain to the new substrate, perhaps even on the belief that any in-place replacement procedure requires no spatial translation at all. In these debates it is sometimes asked how function and identity could possibly move through space from one brain to another.

To state the alternative, we favor the notion that minds are entirely nonphysical and nonspatial, and are merely instantiated by physical brains. Being nonphysical in nature, minds are not so much spatially colocated with brains as they are associated with brains, since they are not capable of spatial locations to begin with. If such wording strikes the reader as dualistic, bear in mind that it implies property dualism at most, certainly not substance dualism.

Wiley and Koene argue that gradual in place replacement involves distance translation from micrometers instead of centimeters or more.

Wiley and Koene discusses replacement rate.

Variants of the temporal argument have been proposed before, such as Chalmers’ chapter in Blackford and Broderick [2014]. What we offer in this paper is a previously unstated connection between the spatial and temporal examples, thus closing a transitive relation from slow replacement to instantaneous replacement and then to scan-and-copy.

A total procedure time of 100 days, operating around the clock, would require replacing 10,000 neurons every second nonstop. To complete the procedure in a twenty-four hour period would require continually replacing one million neurons every second for an entire day. Putting aside whether a procedure of practical duration would actually be meaningfully slow, incremental replacement is nevertheless how gradual replacement is generally presented and in this way it differs from scan-and-copy, which equates to global instantaneous (or at least discontinuous) replacement.

Nextbigfuture view on having constant brain replacement or addition rate with computer uploads a rate slower than a mild stroke

The brain is constantly changing. Learning causes new neural connections. We have determined that identity is preserved so long as their is non-coerced learning. Brain washing can alter and lose the prior identity.

We have viewed that a mild stroke can leave the personality intact as the brain can repair and correct for lost sections.

I think adding new brain substrate (sequential addition of neural prostheses that have the same or higher connections to the rest of the existing biological brain) that fully communicates to the biological brain but the addition or replacement rate is slower than a series of mild strokes with time for recovery. It seems to me would definitely allow time for the brain and mind to incorporate any additions or losses into a constantly preserved identity. This would be using the proven capability of the brain to preserve personality by withstanding mild strokes and by proxy preserving identity.

A perfect copying of neurons and synapses and microtubules could preserve the entire state of a brain including memory and personality. However, there does not seem to be any reason not to go for a slow transition other than the risk of the biological brain dying before the transition is complete.