Seite - 169 - in Cyborg Mind - What Brain–Computer and Mind–Cyberspace Interfaces Mean for Cyberneuroethics

Cyberneuroethics • 169 Practical Challenges Before attempting to grasp the process of mind uploading and brain simula- tion, it is necessary to appreciate how the human brain transits and processes information. As already mentioned, neurons are cells that transmit electrical nerve impulses, carrying and processing information from one part of the body to another. Their spider-like shape of a central body with spindly legs, formed by axons and dendrites, is crucial to their function. The legs branch out repeatedly until they create up to 10,000 endings, which reach out and make contact, at a point known as a synapse, with other parts of the same neuron or with other neurons. A fully developed healthy human brain has around one trillion (1012) neurons. If each of these has 10,000 synaptic contacts, this gives the brain 10 quadrillion (1016) possible connections. Through the ageing process, some of these will be lost so that by adulthood, an individual would only have about one quadrillion connections. Though there remains much debate upon the exact figures at stake, engaging in the process of simply counting these con- nections would be a task that any computer would find impossible to achieve either at present or in the near future. If this was not difficult enough, synaptic connections are also constantly forming, strengthening, weakening and dissolving. This permanent state of flux helps create a complex web of connections that clearly challenges any replicating procedure. Kurzweil estimates that the brain’s billions of interconnected neurons can perform 1016 calculations per second (cps).310 In order to capture in detail all the connections between neurons required to successfully upload a human mind into a computer, represented by a single binary number (0 or 1), called a bit, Kurzweil boosts his estimates to 1019 cps.311 Thus, with the eager anticipation of being able, one day, to successfully copy a human mind into a computer, he proposes that 1018 bits should suffice to represent all the inter- neuronal connections required. The scale of the numbers is fantastic, but Kurzweil perceives these to be achievable based upon the law of accelerating returns, whereby he predicts that supercomputers will eventually match the computational power of the human brain.312 However, in order to reach Kurzweil’s goal, engineers will require not just the ability to make machines that think, but think as well as humans.313 This requires the software of human thought to be mastered, which is something that has only just begun to be considered through advances in computational power. In addition, it is possible to ask whether simply matching the human brain’s neuronal network and computational power is the only challenge to copying a human mind into a machine. What about perceptions, memories, This open access edition has been made available under a CC-BY-NC-ND 4.0 license thanks to the support of Knowledge Unlatched. Not for resale.