Cryonics, AD, Neurobiology, Twitter, and Agents

Cryonics, as defined in the 1960s, seeks to preserve physical-biological information at death, with the goal of recreating personal identities and physical brains and bodies when the technology becomes available. This goal may prove too difficult, and misdirected compared to the probable capabilities and needs of future personal identity.

Since cryonics was defined as an endeavor, we've learned many things:

• We've learned something about the mechanics of memory. The brain is plastic: it will grow new neurons, axons, dendrites, and it will restructure synapses. Neurobiologist Gary Lynch claims memory is stored as nano-structural changes in these synapses. The area and electrochemical "resistance" of synapses changes because of L.T.P., Long Term Potentiation, with structural proteins built and destroyed in the dendritic spines to change their contact area and the number of active receptors. This fast-acting mechanical change reprograms the analog response of portions of the neuronal net, which we perceive as memory. Further, we know that the neocortex contains the oldest cells in the body; we can't replace the cells by normal cell reproduction without erasing memory. And we can't add large numbers of neurons; if nothing else, there is finite room in the skull.

• We've learned something about the genetics and mechanics of brain aging. The APOE4 - AD connection is strong. Scott Small's lab at Columbia has developed genetically modified mouse models for human AD, showing that AD may propagate as prion-like structural modifications of the Tau protein, along neuronal paths, from the entorhinal cortex to the rest of the brain. While we may find chemical interventions that can slow or stop this process, sophisticated nanomachinery may be required to restore the neurons. There is no natural healing process that will do so. If information is destroyed, restoration will necessarily be incomplete.

• Dementia costs are skyrocketing. We've learned to delay cardiovascular death; people are living longer. But we've learned little about treating dementia. Old people with AD, vascular dementia, and stroke damage are filling "memory care" facilities, and the cost of warehousing these unfortunate people may dominate our economy. The economic and personal costs, while devastating, may be nothing compared to the moral costs of making tough decisions about limiting the resources we devote to this - we may subject millions of people to mindless pain, followed by bureaucracy-mandated euthanasia. A society with this hardened attitude is terrifying.

• We've developed massive digital/electronic information capabilities. In 1962, Ettinger published "The Prospects of Immortality", and transistors cost approximately a dollar. In 2012, a transistor costs approximately 1E-10 dollars, and a bit on a triple-redundant hard disk array costs approximately 1E-11 dollars. A human neocortex contains around 1E14 (analog) synapses. If we assume 10 bits per synapse to encode an analog channel, and another 20 bits for addressing/location, storing the results of "digitizing a brain" (if we knew how) might cost $300K today. If current trends continue (a factor of 10 cost reduction every 5 years), then by 2062 the cost of storing a brain will be 0.003 cents.

• We are entering the age of the extended mind. Many younger people are spending hours per day texting and tweeting, blogging and facebooking, sharing their conscious thoughts with others, constructing interpersonal networks and public descriptions of themselves. If this information is preserved and searchable, there will come a time in these people's lives when more of their valued memories are online than in their brain.

• We are learning to connect directly to our machines. Implantable interfaces exist today. Cochlear implants replace damaged ears. A few victims of 'lock in syndrome' have experimental implanted electrode arrays, which they are using to move cursors, steer wheelchairs, and communicate with others for the first time in years. Bioarrays, scanners, and gene sequencing provide huge volumes of data about our real-time state for computational analysis. As we learn about our detailed biochemistry, we are learning to individualize treatments for diseases; soon we will be able to choose our internal chemical environments. We are learning to construct artificial biological replacement organs; soon we will be inventing new organs and cells, denovo, improving old biological processes and creating new ones.

• We are entering the age of the personal digital agent. Google's search engine is a collective digital agent - we can set it to work on "remembering" and it will provide a list of memories for us to choose from. Google's "agency", however, consists of scanning the web, and our collective searches, and reordering its databases to make searches more germane. On a personal level, augmented reality applications for smart phones annotate our environment with associations. It may not be too long before many of us spend our days looking through digital goggles, with our visual fields remapped with helpful hints, embellishments, and deletion of irrelevancies. Multi-player video games, RSS feeds, spam filters, 23andMe genome analysis, and other services are all embryonic personal digital agents. Twenty years from now, a single human personality will be associated with a huge posse of interacting digital agents, performing routine tasks and freeing the human mind to concentrate on priorities.

• We are learning about limits of terrestrial life support. The good news is that human population will probably peak at 10 billion, sometime in the middle of the century. In Whole Earth Discipline, ecologist/polymath Stewart Brand points out that most of the world lives in cities now, 80% will by 2040, and urban populations do not reproduce at replacement rate. The worrisome news is that energy and resource use is increasing, rapidly, as the developing world climbs inside the velvet ropes of the developed world. Even if we cut developed world energy use in half (most unlikely, it is still increasing, with energy-intensive manufacture exported to China, helping us pretend to conserve domestically), the demand for energy and resources will increase by at least 5x by the end of the century, with much of those resources devoted to or controlled by computation. Meanwhile, it is becoming clear that the planet cannot support even a fraction of the current developed world's resource demand. A careful study of so-called "sustainable" technologies reveals a frightening ignorance of their actual damage to the biosphere. We must get better at managing our planet, and shout down the fashionable forces of ignorant technophobia. But even in the ideal case, there may not be room on earth for many resurrected humans.

• Space solar energy is superabundant, and amenable to Moore's law scaling (see http://server-sky.com) but is difficult to deliver to the ground. The best place to use this energy is space, but space is hostile to organic life. There is energy and resources for quintillions of digital entities in space.

Implications for Personality Resurrection

Our brains change every day. We learn, we synthesize, we confabulate, and we forget. For active people, every day is new. For growing minds, the possibilities of each new day can exceed the sum of all previous days. Aging, decaying brains can record valuable experiences, even as older experiences slip away. So there is a difficult question to answer: which version of the mind do we hope to resurrect? The mind at the peak of capability? The peak of memory? The last capable mind? Or some combination of all of these, which never co-existed at one particular time?

And what do reanimators do about mistaken information stored in our biological brains? The average 2012 American believes many incorrect factoids (and we can argue about what those are). Do reanimators restore the brain with factual mistakes? How about emotional defects, irrationalities, bad habits? And if reanimators decide to edit, how do they objectively and morally choose what to edit? How much awareness of inconsistency and mistakes will they allow their patients to have?

Much of that mistaken information will be subjective - for example, in 2012, an individual may self-identify as a Roman Catholic, but most Roman Catholics have beliefs, examined or not, that are quite different from the beliefs of theologically trained Catholics.

If a mind is brought back, repopulated with preserved information where available, educated speculation where it is not, can you say it is the same mind?

The future in which we hope to reappear will be very different from the times we live in now. Individuals will be cooperatives of biological and computational resources. Development from infancy to maturity will be accompanied by an accumulation of digital capabilities that will, over time, grow far past the biological starting point. In 2012, we connect to our digital tools by our senses, speech, and muscular movements. In 2062, we will connect directly from our brains, through massively parallel invasive sensors, with terabit-per-second bandwidth. The fastest path from brain to external manipulation will not be neocortex to motor cortex to spinal cord to nerve to muscle, but electronically, neocortex to sensor to computer to electric motor, whether that motor is 10 centimeters or a million kilometers away.

Even if we were able to restore a cryopreserved brain to perfect health in a young body, what would that resurrected human be able to do in this new world? The resurrected brain will start changing, accumulating, forgetting. This new/old mind will saturate the capabilities of a finite brain in a finite skull, in decades or centuries, a tiny fraction of truly immortal deep time. The resurrected human will be a curiosity, a zoo creature, and a digitally-connected globe may have little use for zoos. Meanwhile, those women who choose to have biological children may resent the loss of opportunities posed by cryonauts, should they return to the flesh in significant numbers to displace birthing opportunities in a limited world.

For both the original-born and successfully-integrated cryonic-reborn, aging bodies and minds will become increasingly small fractions of growing "cybrain" (cyborg-brain, digital-biological) composite personalities. There will come a time in every cybrain existence when the body and biobrain becomes optional.

The sun puts out 380 trillion terawatts. Almost all of it blasts out unused into empty intergalactic space. The human brain consumes 100 watts, both as a computation device and for biological maintenance. If we capture 3.8E26 watts at 50 astronomical units, and use it for analog/digital computation at 60 Kelvin, we can arguably emulate a human brain with less power, while simulating a slowly changing digital "environment" for that mind with only a little more power. There may be room out there for a Dyson sphere of 1E24 such minds (http://server-sky.com/DigitalImmortality). If the earth births 100 million biological minds a year, and all those minds eventually upload into the Dyson sphere and "live" forever, then over the remaining 10 billion years of the solar system the Dyson sphere will accumulate "only" 1E18 minds.

These are incredibly extreme speculations; however, the capabilities necessary to reconstruct a functional biological brain from a frozen one are at least as formidable as the capabilities necessary to build these digital ones. Indeed, simulated digital brains may be an essential step in the process of successful biological reanimation. It is unlikely that we can just pour magic nanotechnology goop on a frozen human brain, weld it onto a generic body, and have anything but a silently screaming paralytic as a result. The actual process will be individualized, massively coordinated over the meter scale, and dependent on external consistency checks heavily dependent on externally-supplied corroborative information.

Evidence-based Cryo-Resurrection

It will not be easy to produce a functional, coherent, sane mind given the scraps of information available to future technicians. It will be far harder to demonstrate that the mind produced is in fact the same individual as the suspension candidate, by the criteria that the candidate would have agreed to. So what constitutes evidence of cryo-resurrection success?

The end goal is NOT preserving brain cells, or producing functional brains. We already know how to do that, the old fashioned way. The goal is not even to demonstrate conformance to some global definition of identity between cryonics candidate and resurrectee, even if we had such a definition and the ability to perform the work.

The goal is to continue a process, begun with a naive pre-cryonics individual, of education, data capture, expectation management, and organizational preparation, that ends up continuing into the far future in a way that more-or-less meets the expectations of an educated cryonics candidate.

We can't just engage in a complicated and expensive series of technological procedures, and call that a reanimation. Future re-animation teams should be able to prove their success, by the best objective criteria available. We can help them by capturing information now.

Implications for Cryo-Preservation

The reconstruction process will require every scrap of information we can afford to gather. Of course the physical artifact of a properly cryo-preserved brain will contain essential information. But it will not be enough to construct an exact and viable replica of the original. Reanimation teams must fill in a lot of missing information to produce a functional, coherent mind. That information will be synthesized from external records, global and individual, physical and social. It is incumbent on us to preserve as much of that recorded environment as is practical, in parallel with biological cryo-preservation.

However, the complete external recording of internal mental state is impossible, and a partial job requires prioritization of what to preserve and what let slip away. Survival of preserved information over decades, when a mistaken "rm -rf *" can send all of it to digital oblivion, requires more diligence than we apply to the ephemeral information we use every day. You may have 10 gigabytes of gmail, but what happens when Google follows Montgomery Ward into the dustbin of corporate history, sooner or later? Corporations have surprisingly short lifespans these days.

The cryonics community includes many computer geeks. Some of us have preserved our personal digital memories as stacks of old hard drives, but most of that is irrelevant to our personal identities. One copy of /bin will do for all of us, one copy of /etc might do for each of us, and daily copies of our working documents, preserved as a series of small changes, will take some effort to create but will be easier to store as a record of our changing lives.

Meanwhile, many of us store our memories on Google and Facebook and Twitter, one corporate policy change away from oblivion. This information is how we present ourselves socially to others. The most objective measurement of a successful cryo-resurrection will be how this recorded behavior matches the behavior of the resurrected mind. If we care about quality of reanimation, we must find some way of extracting and preserving this social network information if we hope to objectively measure the success of the reconstruction effort.

In addition, we should work on educating cryonics candidates to develop realistic expectations of the results of cryo-resurrection, based on the efforts they are willing to put in now to maximize outcomes in the future. For that education to work, we must learn what is reasonable to expect from candidates, and make sober estimates of what is realistically possible for future reanimation teams. We know we will have vast capabilities in the future, but we also know we cannot create accurate information out of nothing.

We must also develop ways of providing future resurrection teams with the goals of individual cryonics candidates. While it is impossible in 2012 to predict what options will be possible in 2062 or 2112 or later, those teams will be better able to choose an option for the patient if the patient can provide the some general instructions now, and update those instructions as both the individuals and the preservation teams learn more.

All of this I.T. work will be an impossible distraction for a biopreservation organization. Cryonics organizations have computer geek clients with the appropriate skills, who can organize independently, only loosely affiliated with the biology geeks. Indeed, there may be value for a separate organization that preserves digital identities only, as a service for non-immortalists who nonetheless feel they have knowledge and experience to contribute to the future. Digital immortalists may evolve into bio-immortalists, as well as expand the information environment for reanimation teams.

We are among the last generations of humans who must suffer from permanent, involuntary death. Our experiences, communicated either "in person" or as recorded documents, may be prized by few at any given time and place, but those few will number in the quintillions over billions of years and across billions of stars. Our "vain, selfish" choice to preserve as much of ourselves as we can will be a great blessing to those few, and our diligence and accuracy will be appreciated as long as minds exist.