The likes of Stross and MacLeod often use backups and “digital people” as plot devices and characters, but Egan really gets into the concept: discussing the identity and self-perception of the virtual humans, and how they all relate to virtual minds in virtual realities. As always Egan’s view of the future is highly realistic. A key part of the plot involves the fact that “copies” of humans can only “run” at 1/19th the rate of normal flesh-and-blood human beings. This turns the usual belief that virtual human minds will run faster than normal humans on its head, and makes for some interesting observations.
There is also the usual philosophical debate: if someone is an indentical virtual copy of someone else are they that person? From my Kurzweil-oriented point of view, I believe that what is key in this situation is the pattern of information. From the millisecond the copy and the original data construct (say: a flesh and blood human) diverge then they become two different people. What is key is the pattern of data, not the substrate in which that data (or information) exists, hence a digital person, a virtual person, an “analogue” F&B person, or a person that is described by the actions of a weak Turing-complete system (i.e. a machine or computer that can perform any computational task).
Terry Pratchett explores these ideas in The Fifth Elephant. At the end of the book Sam Vimes is presented with an axe by the dwarf King. The dwarf observes that if Vimes’ ancestors were to replace the blade, and then their ancestors were to replace the handle – then could it be said to be the same axe?
It is a widely-believed fact that every atom in the human body is replaced every seven years, and the atoms of the material we think of as being most intimately “us” – the brain, CNS, nerves etc – are replaced every few months. This demonstrates that it is not the lumpen matter that matters when it comes to defining a person (without unnecessarily invoking supernatural irrelevences like the “soul”, or weird quantum “stuff” a la Roger Penrose). It also demonstrates that people are dynamic: I wasn’t the same person I was a few seconds ago. I’m not the same person I’ll be in a year.
I feel I can be forgiven for being squeamish: when the technology to upload your brain or a part of your brain into software (and run it at = to or > baseline speeds) becomes available I would rather tip-toe into the swimming pool, rather than dive in all at once.
By this I mean I would choose to model a small area of my brain, then devise an item of hardware that can respond in exactly the same way to stimuli as the area I have scanned does. The item would contain a computer running a simulation of the area of the brain it is designed to replace, and sufficient hardware to interact with the surrounding areas of the brain in the same manner as the original area.
This “hardware” is likely to require extremel complex devices. I don’t know enough neurobiology to be specific, though I imagine that simulating hormones, neurons, and the intricacies of the human brain are likely to be difficult. From this point of view it would be much easier to simply render the entire structure in virtual reality. Then you wouldn’t have to worry about the interface problem.
Over time I would gradually “build up” the area of my brain running in synthetic substrates, until the whole thing is wholly synthetic. Assuming, of course, that such a state does not compromise my health, mental or otherwise. The likely benefits: including the ability to “learn” new skills by reinforcing the appropriate mental pathways and speeding up my perception of time will hopefully, eventually outweigh the possibly downsides.