A Survey of ILE (Indefinite Life Extension) Technologies -- draftYan King Yin
Background: Philosophy of Mind
In order to understand physical immortality we need a basic understanding of what is life and death, which leads to the question of what is consciousness. A current theory of consciousness is functionalism, which posits that consciousness emerges out of information-processing systems, and that “experience is invariant across systems with the same fine-grained functional organization” [Chalmers 1995]. The space-scale, time-scale, and substrate of the systems are unessential. A computer simulating the brain can run on silicon instead of neurons. It may also execute only one instruction per a million years and still experience a consciousness similar to that of humans.
If consciousness is information processing, then the collapse of this process can be construed as death. With the advent of new technologies such as cryonics, brain emulation, and “mind uploading”, the distinction between life and death will be blurred. We can imagine that, in a post-upload world, substrate-hopping intelligence will become the norm.
Moore’s Law: The “Prophecy”
If we estimate the human brain’s computing power at 100-100,000 teraflops per second, it follows from Moore’s Law, computers will reach that level around 2008-2024 [Dani Eder, internet forum]. (Since we don’t fully understand the brain’s information-processing capacities, this estimation may be off by a bit.) If uploading does become feasible, the human race may face an identity crisis. Biological reproduction may not be the most competitive, or desirable, choice for the continuation of our species. Technological progress is likely to continue after this point, driven by human desires. Post-humans will probably continue to compete for limited resources on earth, and the evolutionary perspective will apply to this future scenario.
Is Evolution Against Immortality?
From the evolutionary point of view, not all desires are equal. There are “detrimental” desires that will decrease an organism’s chance of self-preservation or proliferation; over time, the evolutionary process would winnow these organisms out [as Robin Henson suggested during an ImmInst chat]. The various forms of physical immortality that scientists have proposed can be characterized as resistance to change. Evolution dictates that any successful organism must continually change itself to adapt to its environment. An organism cannot remain unchanged for a very long time. Sexual reproduction (which is actually the basis of the life cycle and thus of aging) provides an important mechanism for adaptivity (DNA recombination). Physical immortality eliminates the need of sexual reproduction for an individual’s prolonged existence (sexual reproduction is optional), so we must find a substitute for adaptivity, in the form of continual self-modifications. In fact, nature itself has also evolved an epigenetic form of behavior modification: learning.
Due to the complexity of life, it is very difficult to come up with an objective measure of adaptiveness; though this problem may be tackled by statistics and probability, chaos theory, game theory, or by historical analogies.
Two Major Approaches to Immortality: Uploading and Brain Rejuvenation
I’ll only consider the brain which is the most complex organ in the body and also the most important one.
1. Uploading
Uploading offers a lot of advantages over brain rejuvenation, such as the ability for an upload to increase its intelligence greatly and self-modify its cognitive structure. This will enable unprecedented variations and rates of evolution that will be superior over biological life forms. Uploading requires:
I) A scanning method that can extract:
1. Connectivity and morphology
2. Synaptic efficacy
3. Molecular / gene activation states
II) Modeling (brain emulation):
1. Electrophysiology
2. Neuromodulation
3. Plasticity
The major obstacle to uploading is limitation of computing power, and secondly the elucidation of neurochemical pathways. Scanning usually destroys the brain in the process but it may be possible to replace the brain gradually (“neuroprosthesis”), which requires sophisticated neural interfacing.
For more technical details please visit MURG (mind uploading research group):
www.minduploading.org
.
2. Brain Rejuvenation
Current theories of aging and a lot of experimental facts suggest that brain aging is extremely hard to stop or reverse. Aging of the brain involves myriad molecular processes, most of which are poorly understood, including neurodegenerative diseases such as Alzheimer’s, accumulation of pigments in neurons, cell death, demyelination, subtle changes in electrical properties and plasticity, and much more. Therefore the most practical strategy might be to replace rather than repair neurons in the brain.
Main requirements for neural replacement include:
1. A source of neurons for implantation (such as stem cells);
2. Migration / injection of neurons to appropriate regions;
3. Growth of neurites (axons and dendrites) from new neurons to their appropriate targets;
4. Formation of functional synapses;
5. Expression of appropriate genes and thus physiological functions.
The biggest challenge is the innervation of target tissues (3, 4 & 5) which requires axonal / dendritic guidance mechanisms. The neonatal brain expresses various chemical cues that specify the wiring pattern of the brain, but which may be lost in the adult. It is extremely difficult to re-create this developmental “niche” (or something to that effect) [Guillaume & Zhang 2003].
The Role of Artificial Intelligence
In the coming decades increase in computing power will drastically alter the technological landscape. One predictable consequence is the replacement of low-complexity work by AIs. As pertains to uploading, experimental research in neurobiology will be accelerated by automated biochemical assays, brain mapping (scanning), genome and proteome analysis, and so on. DNI (direct neural interface) capable of memory transfer may be designed with the aid of computers. On the other hand, brain rejuvenation techniques may be advanced by automated experimentation.
In the posthuman future, humans must face the question of continuation. Besides uploading and rejuvenation, one may choose to “adopt” humanoid babies created through WBE (whole brain emulation), or adopt “AI babies” that do not closely mimic the human brain. Some of these choices may be more economical (cost-effective) than others. As I have argued, there is always the conflict between desires and adaptiveness.
Reference
Chalmers, David (1995). Absent Qualia, Fading Qualia, Dancing Qualia, Conscious Experience (ed Thomas Metzinger) Imprint Academic.
Guillaume DJ, Zhang, SC (2003). Neuronal Replacement by Transplantation, Neural Stem Cells (ed JE Bottenstein) Kluwer Academic Publishers.
Thanks to Ed Minchau from the MURG mailing list for proofreading and suggestions.
