Continuing the discussion from Technophobia goes off the Depp end in Transcendence [Movie Review]:
You can download your body into a grave, but you cannot upload consciousness.
I can see why a lot of people might agree with that, if they’re basing that call on what they’ve come across in fiction rather than an abiding interest in the notion, and the attendant pondering…
Because what I’ve seen in fiction often misses or sidesteps the question of preserving and transferring awareness; Transcendence is no exception (The Prestige and Be Right Back from Black Mirror sort of are)…
Never mind the staggering infeasibility of copying a brain - how is copying one supposed to transfer its awareness? The copy may well come into existence remembering your past and feeling like you, and that you were transferred and survived, but you you? There’s no escaping your fate by this means.
But I’m inclined to think there is a possible means. Given the existence of a computing substrate capable of emulating a brain (not to mention generating a matrix for it), which all these scenarios depend on (I don’t think that’s at all impossible, although it probably is currently), it seems to me that what you need to upload your awareness is a fat pipe - between your brain and the new substrate, wide and fast enough for your process to be running on both.
Once you’re running on modern hardware rather than the ancient legacy gear we’re so familiar with, you’d be in a whole new realm; presumably the nature of thinking and learning would be free to shift dramatically - it’d be like a ‘superhero discovers and masters his powers’ bit, only for real. But perhaps the most interesting question here is what would happen to your awareness if you were to disconnect while inhabiting both substrates…
As a former AI researcher and developer I can point out a few things; the plausability of the first premise (strong AI in normal computers) is so low as not even to begin to beg any questions. We indeed need a quantum leap (hehe) before we can start talking about any of this being close to doable.
Why’s that necessarily the case? If you want hardware that can work like a brain natively without emulation, that’s true, but I can’t see why we can’t just throw enough digital computing power at the problem to solve it. I gather the human brain crunches roughly 20 petaflops, about the same number as the most recent ubercomputer? Obviously if we’re emulating, we’re going to need at least another order of magnitude or two more grunt to do the job, but Moore’s law makes with one of those once or twice a decade, regardless of substrate (if you go back it holds for valves and mechanical computers).
So I’m not sure the hardware is ever going to be a deal-breaking problem, or even figuring out how to program that hardware, if a future of increasing intellectual augmentation sees us unconsciously programming hardware as tendrils of our awareness activate it… in the end, transcendence might be a relatively seamless event, with nobody able to actually put their finger on the moment they realised they didn’t need their body anymore.