I fully expect to be able to achieve functional immortality within my lifetime. Do not read that as plain immortality—I have little attachment to my own biology. Many argue against this on the basis of biological immortality and philosophical or physical concerns. Over-population, they say, would ruin us; immortality is unnatural the nay-sayers might quip. They might argue that no living creature is meant to live forever, despite several excellent examples of biological creatures that do live forever in a practical sense. Our life span, barring unforeseen events such as cancerous mutations or sudden impacts with a bus, is not fixed at 100 years of age. In 1900 the average life-span was just under a mere 50 years! With technology, we can do better, and have been improving steadily over time.
Let me back up for a moment and claim a specific, and I feel, realistic goal: I plan on uploading my consciousness—the sum total of my experiences, knowledge, and patterns of thought—into a computer system by 2029. I’m not the only one who believes this is possible.
There are a number of avenues being explored to extend our existence, potentially extending for as long as we wish to be around. I, for one, would prefer to choose the time and place of my exit from consciousness. These avenues range from the complete simulation of our physical brain (by far the longest off in terms of practical use) to shorter-term life extension and suspended animation. All of these approaches boil down to three philosophies: bottom-up, top-down, and avoidance.
Avoidance is the easiest to describe. Freezing yourself solves no immediate problem unless you suffer from a life-threatening ailment. While you’re frozen, you’re not thinking. You are dead. The goal of this approach is to enable you to be unfrozen at some point in the future when the problem that landed you in the popsicle machine has been solved. It’s not very interesting to me, especially as we have not successfully unfrozen anybody; you are not just waiting for your particular problem to be solved, but the problem of reanimation. It’s an unproven technology only really tested on animals, and even then usually animals with an existing, proven ability to be un-frozen, like amphibians. It’s a high-risk gamble, but people do it due to the high reward if they’re right: they can live again. With that out of the way come the two really interesting approaches.
Top-down takes the functional approach towards simulating human beings. We don’t need to know how the wiring itself works if we can accurately model the resulting behaviour. In this way we may eventually be able to create advanced cybernetic intelligences that, at a fundamental level, have senses, process those senses, perform abstract thought, learn, and behave as we do. For all intents and purposes, we would be able to create living machines. Note my careful avoidance of the term artificial intelligence; what I’m proposing is actual intelligence, just in a form we have created instead of biologically evolved. (With an aside that the word artificial is inherently prejudiced; artificial can not be real by definition.) With a working simulation framework that would take orders of magnitude less processing power than the bottom-up approach, we could feed it everything we have experienced (the recording of which is a whole separate problem, also being worked on) and create an approximation of ourselves as we are.
I’m actively working towards this first step through the combination of data acquisition, neural network modelling against that data, and sensory perception processing pipelines. I have nearly every bit of significant digital information I have ever touched since 2000 or so, totalling around 15 TiB, a trained neural network on the textual data from that corpus which is capable of forming simple word association patterns, and the beginnings of a parallel processing pipeline, using Stackless Python, for realtime data such as the feeds from cameras and the multi-channel audio from multiple microphones. Each of these deserves its own dedicated and detailed article. (I call this my Exocortex, and I use it daily to retrieve information.)
The second step, and logical conclusion once successful, is to understand and simulate the fundamental building blocks of the brain (neurons, synapses, electrical and chemical communication channels, etc.) If we can do this, and it’s only a matter of time, we could then image your brain, specifically the functional connectome, and run it in simulation. Currently this imaging is destructive, but like digital computer technologies, brain imaging technologies are doubling, too. And timeframe of each doubling is becoming shorter.
You might scoff at the possibility of simulating the human brain on computers. Admittedly, the brain is a complex device. But we’re already doing it. If, as several top researchers suggest, our neocortex is primarily a memory storage and reality simulation device, then the future is closer than you might think.
Technology will transform us, invade our lives, be fully self-aware, and show emotion, all before we begin to incorporate our own intelligences. Oh, wait, they already have transformed us, invaded our lives, show self-awareness and emotion. The future will give us more, better, faster, cheaper, and more complex.
We can already simulate, accurately, a complete cat brain (and rat brain) and a single neural stack (here’s an article about both).
Currently we’re using these technologies to understand ourselves, and improve our technologies in a feedback loop. We even have a roadmap for where to go from here. In the end, when these technologies combine and we master the structure of thought, our brains, we will have the opportunity to become immortal.
You should watch the movie The Man from Earth -- somewhat relevant to this discussion.