Considering the Details of Replacing the Brain

Of the billions presently alive, some fraction will go on to live for thousands of years. The age of rejuvenation therapies is just around the corner, and new approaches to medicine will enable the old to be made young again. This will happen within a few decades, perhaps soon enough for those in middle age today in wealthier regions, perhaps not. Whatever the timeline turns out to be - and we have the opportunity to accelerate it - the fundamental forms of cellular and molecular damage that cause aging will become just another set of chronic medical conditions that are kept under control with regular treatments: periodically repaired, so as to maintain youth and indefinitely postpone age related disease.

In this future people will still die, however. The current mortality rate due to fatal accidents, if maintained, would give an ageless, disease-free person a life expectancy of a few thousand years. If you want to live longer than that, then you require either (a) the means of greatly reducing the occurrence and severity of accidents, or (b) to ability to change yourself to be less vulnerable. Those people alive today who are still alive ten thousand years from now, and some will be, will most likely have altered themselves dramatically, abandoning flesh and the human form in favor of far more robust machinery.

It should not be terribly controversial to suggest that a hundred years from now replacing your body with an artificial chassis will be a very feasible, cost-effective option. The manufacturing and design technologies of that era will involve mature artificial intelligence and precise atomic construction. An artificial body should be a simple undertaking by that point, and there's more than enough time to wait for that technology if you survive today's first hurdle of living long enough to benefit from the first wave of rejuvenation biotechnologies.

When it comes to transforming yourself into an entity likely to survive for longer than recorded human history to date, the body is a trivial matter, however, hardly worth putting much thought into at this point. Almost any easily replaceable, mobile, and very robust machinery will do. The more interesting questions relate to the brain and the self: how can you switch out the intricate biology of the brain for more durable machines without killing yourself in the process? All that is you is encoded as data in the fine structure of brain tissue. Making a copy of your mind to run as software seems like a feasible undertaking, something that can be envisaged even today: it's possible to speculate in a useful fashion as to how that might be accomplished within the next few decades. But a copy of you is its own entity, not you, and there are many other questions and doubts relating to the continuity of the self associated with an intelligence running as software.

The best approach to replacing the brain while retaining your self is a slow process of replacing each neuron with machinery that serves exactly the same purpose and integrates with the rest of the brain in exactly the same way as the neuron did. The brain creates and loses neurons on an ongoing basis already - though a plausible replacement methodology would run many times faster than that process, and would replace neurons that are normally never replaced. Some of those cells last a lifetime.

This gradual neural replacement is a fine thing to talk about in abstract, but how would it even work in practice? How would a neuron machine be constructed? How do you assure continuity of the self when doing this for real? Some people have put a fair amount of thought into this topic, even though it is a far future prospect and we still have to sort out the first step of not aging to death in the bodies and brains we have now. Over at the Rational Argumentator you'll find an eleven part series on the important parts of the path to replacing the brain with machinery. There's quite a lot of reading material there, and I make no warranty as to the quality and rigor of the work, but I think you'll find it interesting.

  1. The Moral Imperative and Technical Feasibility of Defeating Death
  2. Immortality: Material or Ethereal? Nanotech Does Both!
  3. Concepts for Functional Replication of Biological Neurons
  4. Gradual Neuron Replacement for the Preservation of Subjective-Continuity
  5. Wireless Synapses, Artificial Plasticity, and Neuromodulation
  6. Mind as Interference with Itself: A New Approach to Immediate Subjective-Continuity
  7. Neuronal 'Scanning' and NRU Integration
  8. Squishy Machines: Bio-Cybernetic Neuron Hybrids
  9. Choosing the Right Scale for Brain Emulation
  10. Maintaining the Operational Continuity of Replicated Neurons
  11. Immortality: Bio or Techno?
If we seek to replace biological neurons with artificial equivalents, once we have a simulation of a given neuron in a computer outside the body, how is that simulated neuron to communicate with the biological neurons still inside that biological body, and vice versa? My solution was the use of initially MEMS (micro-electro-mechanical systems) but later NEMS (nano-electro-mechanical-systems) to detect biophysical properties via sensors and translate them into computational inputs, and likewise to translate computational output into biophysical properties via electrical actuators and the programmed release of chemical stores (essentially stored quantities of indexed chemicals to be released upon command). While the computational hardware could hypothetically be located outside the body, communicating wirelessly to corresponding in-vivo sensors and actuators, I saw the replacement of neurons with enclosed in-vivo computational hardware in direct operative connection with its corresponding sensors and actuators as preferable.

I didn't realize until 2010 that this approach - the use of NEMS to computationally model the neurons, to integrate (i.e., construct and place) the artificial neurons and translate to biophysical signals into computational signals and vice versa - was already suggested by Kurzweil and conceptually developed more formally by Robert Freitas, and when I did, I felt that I didn't really have much to present that hadn't already been conceived and developed.

However, since then I've come to realize some significant distinctions between my approach and Brain-Emulation, and that besides being an interesting story that helps validate the naturality of Immortalism's premises (that indefinite longevity is a physically realizable state, and thus technologically realizable - and what can be considered the "strong Immortalist" claim: that providing people the choice of indefinite longevity if it were realizable is a moral imperative), I had several novel notions and conceptions which might prove useful to the larger community working and thinking on these topics.