Is a Google executive’s vision of a digital afterlife feasible or a fantasy?
Living forever while staying healthy and vital has always been one of the oldest quests of mankind. Some wish it would happen, and some say they’ll make it happen. The director of engineering at Google, Ray Kurzweil, is one of them. His
unbounded self-confidence and unshakable faith in science and technology propelled him into the spotlight as a leading figure in the transhumanist movement, which is seeking to transcend our perishable, earthly bodies into immortal beings.
Kurzweil dreams of a future where we transfer our consciousness to robots, thus
shedding the mortal coil of our biological bodies. He claims that the digital immortality is going to be a reality over three decades. By 2045, to be precise.
Beyond any shadow of a doubt, Kurzweil is a very smart man. His inventions include the first flatbed scanner and a text-to-speech reading machine for the blind. But is his latest ambitious and controversial target raising the bar too high?
The idea in and of itself is not bad. Why deal with a body that gets all sorts of diseases, deteriorates and dies when it could be possible to live without any of these biological constraints?
To get a more accurate insight into some of his thought-provoking and entertaining
ideas, let’s look at his public statements. Kurzweil begins his each of his lectures
by charting some of the successes of artificial intelligence (AI), citing IBM’s
Jeopardy-winning Watson computer and Google’s self-driving car as examples of
impressive leaps of technology. But is this really the sort of AI envisioned by
pioneers like Marvin Minsky who, as the head of MIT’s Artificial Intelligence Lab,
proclaimed in 1968: “Within a generation we will have intelligent computers
like HAL 9000 in the film, 2001”. Minsky was optimistic about AI for a long while but now admits that, despite important progress, computers have yet to even imitate the commonsense of a baby.
Rather, what Kurzweil so proudly talks about are examples of neural networkbased
systems with strong pattern recognition capabilities that are nowhere near the original concept of AI in terms of human-like intelligence. Kurzweil claims that he has a very
good grasp of the working of the brain, and argues that it employs Hidden
Markov Models – a fundamental unit of computation composed of a few dozens
of neurons – for its myriad of computations. Curiously enough, Kurzweil never
says why this is true; he just states that it is. His beliefs seem to be based on his
innate sense of how things must be. There’s almost no evidence that the brain
uses Hidden Markov Models as its fundamental platform.
In my own field of expertise, brain imaging, he also states that imaging is revolutionising our understanding of the human brain by mapping it in incredible detail, with its spatial resolution doubling every year as a result of Moore’s law: the observation that computing power doubles approximately every 18 months. Magnetic resonance imaging (MRI) has a resolution in the millimetre-to-submillimetre range. Extending MRI to finer resolution has faced enormous challenges as a result of its inherently low sensitivity. Kurzweil cherry-picks from the vast literature to justify his view that Moore’s law applies to all technology, including brain imaging, as evidenced by his own data on the resolution enhancement of MRI. It’s simply naïve to think that current technological trends will continue to accelerate. In the case of MRI technology,
employing higher energies would cook the brain like a microwave instead of imaging it. Without a major breakthrough in the physics, it’s highly improbable to expect much finer detail from MRI scans.
But that doesn’t seem to be an issue for Kurzweil, who envisions billions of nanobots the size of human blood cells, or even smaller, travelling through every brain capillary and scanning every relevant feature from up close. This alternative to scanning the brain from outside is reminiscent of the cold fantasy of cryonics, which similarly gives transhumanists hope of repairing each destroyed cell within our bodies from within using nanobots. Many experts doubt that Moore’s law will hold up indefinitely.
Apart from the economic reasons that Moore himself was worried about, there are also ultimate limits set by the thermodynamics itself (thermal noise) and quantum physics. However, Kurzweil doesn’t seem at all bothered by these questions as he thinks that as soon as Moore’s law’s comes to an end it’ll be replaced by a new paradigm.
Setting aside this and many similar constraints implied by the physics, no matter at how high a resolution, imaging the brain in incredible detail may get us no closer to fully understanding it. Even knowing all the individual elements and how they are connected together doesn’t necessarily grant the ability to understand how they function together as a whole. As Nobel laureate in physics, Robert Laughlin,
explained to the Library of Economics and Liberty : Rigidity of matter. Made of little atoms. Obviously absurd, but also experimentally wrong, because if you take a small bunch of atoms and do nanoscience on them you’ll find they are not rigid... So in fact rigidity is a law of nature that is like a pointillist painting, like a painting of Monet, get up close it’s meaningless little dots; back more, it makes more and more sense, becomes perfect.... The whole has properties that are not visible from the parts.
So, what if the brain is indeed an emergent structure? Can we ever understand how billions of individual neurons woven together to form an incredibly complex network of
connections act collectively to give rise to this emergence?
But Kurzweil is not interested in a coalition of neurons creating a conscious state. That’s why it is nothing but the mystery of consciousness that Kurzweil keeps brushing under the carpet. He argues boldly that self-awareness is not
necessary in his model of accommodating intelligence. Fair enough, but do we even
know what intelligence is?
As Michael Hauskeller of The University of Exeter pointed out in the International Journal of Machine Consciousness, how likely is it that a software model of my mind, uploaded to a computer, will really be my mind?
Kurzweil’s claim is that Jack, a person he asks us to imagine, will be the same
person after he transitions gradually from an organic body to a machine. Hauskeller,
however, counters that this argument will always be fallacious as it denies the reality of change.
Assume a heap of sand from which one single grain is removed each time. It
will still be a heap, but we are not able to pinpoint the exact moment when, after
the removal of yet another grain, the heap ceases to be a heap. At some point we will
all agree that it is no longer one. Similarly, when a person gradually changes
we will at some point begin to doubt whether it is the still the same person.
And the nagging question remains: what is it exactly that needs to be digitised
or mapped so that we can transfer ourselves to a machine and carry on
existing? How deep does one need to dive to be able to accurately capture the core
element of who we are?
As a corollary, Kurzweil also falls short of explaining what and how he is going to digitise exactly. Even if all the technical hurdles could be overcome and the core thing that makes each individual self –feelings, memories, intellect and identity – can be transferred onto a machine,is it really going to be your conscious self or just a digital knock-off that claims to be you?
And who’s going to keep your digital self alive and safe against all kind of threats
such as viruses and software or hardware failures? Perhaps this will become a moot
point as our “sophisticated future technologies” will be able to handle all this –
so why worry now?
I don’t think the human body is sacred, and if getting rid of our mortal bodies benefits humanity I’m all for it. But I can’t help wondering if a biological body is an intrinsic aspect of being human. Can we really give up eating, touching, sleeping and still be human? It seems that the old immortality story where the human soul lives on after the earthly physical body dies is now being replaced by a transhumanist techno-theological narrative. Death remains the last taboo in our modern western culture, leaving a large cultural void that transhumanism is trying to fill, promising a new hope, a new paradise where death will be defeated altogether.
But until then, instead of avoiding it, I’ll carry on pondering my own mortality, which provides one hell of a silver lining: living each day to the fullest.
This article was originally published as cover story in the November 2015 issue of Australasian Science Magazine.