Uploading your mindSaturday, 8 August, 2009
Mind uploading is a radical form of human enhancement, whereby the human mind is transferred from the vulnerable organic medium of the brain to a computer system of some kind. With such an upload comes the benefit of being able to use your consciousness at electronic speeds (allowing for hours of thought processes to take place in mere seconds), and essentially be in multiple locations at the same time. In addition, the human brain vulnerable to many of the frailties of the human body whereas a computer system does not suffer from age or strigent bodily requirements for blood and nutrients (though computers do get viruses, and they do require electricity and ventilation). Finally, a computerised version of the brain is theoretically much easier to enhance or alter than the biological version.
Firstly, there is the continuity problem, which is by far the most difficult to solve because it is largely philosophical rather than practical. This problem is based on the fact that the process of ‘uploading’ does not actually move a files, but rather involves making a copy of the file on the server (and deleting the old files). Therefore if I upload my mind, I may not have transferred my mind to the computer but instead merely made another copy of my mind on the computer. And the last thing you want is to have two copies of the same mind, because it will be impossible to tell the copies apart as both will literally feel as if they are the original (hence why I said two copies, and not the original and the copy). And in fact, a computerised copy of your mind would be far easier to copy than the organic version, and therefore somebody could split (or be split) into several indistinguishable copies of themself.
The proposed solution to avoid copying a mind is to keep information flowing between the organic and inorganic portions of the mind. Because the mind could be viewed as a process rather than an entity, it would therefore be necessary to keep the process continuous between the biological mind and the computerised mind during the process of uploading. As soon as the copies can no longer transfer information between each other, they will diverge into seperate entities (another interesting topic is whether minds can be merged by allowing full information transfer, such that two people could become one in a much more real way than any human relationship in the past).
So any practical solution must not only feature the ability to read information from the brain, but also to send information from the computerised mind back into the uncopied portions of the mind to maintain continuity.
There are just so many philosophical issues here, but fortunately we have a lot of time to ponder them as there are a great deal of technical problems with mind uploading.
As the mind is (essentially) produced by the brain, mind uploading requires the ability to emulate the entire brain. This is not an easy task, as the human brain contains a hundred billion neurons with trillions of connections between them. And even an individual neuron is such a very complex cell that it cannot be emulated fully. Further, neurons are not the only cells in the brain that matter – glial cells, which are just as numerous as neurons, also have a very important functional role. It’s not theoretically impossible, but any emulation will be a poor one indeed until vast advances in both neuroscience and computer power are made.
Importantly, to create a viable approach to mind uploading (as opposed to ‘merely’ an artificial intelligence) it is not enough to merely be able to emulate a human brain, but it is essential to emulate a particular human brain. If I want to upload my brain, the resulting simulation has to be my brain and not just any human-like brain. This would require a very high resolution and instantaneous snapshot of my brain. Analagously, this is like capturing an image of a metropolis like New York down to the level of millimetres, capturing both the position and the velocity of every vehicle, person and object. While we may have high resolution photography that can capture one person or object to the required resolution, doing the entire city at the same instant is exponentially harder. And likewise even if we can scan a brain to a very high resolution if we focus on just one small slice of a brain region, but to do the entirety of the brain at one instant would be a far greater task. And we have to capture not just the structure of each cell but also the current state of that cell and the cerebrospinal fluid surrounding each cell. It’s a mammoth task indeed, and one that may not be completed until far after a passable emulation of a human brain is accomplished.
And finally we don’t only have to be able to read the brain in sufficient detail, we also need to manipulate it in similar fidelity to maintain continuity. Information in the brain is stored in neuronal architecture, cellular structure and activity and concentrations of molecules in and around neurons. So if we have a portion of the brain simulated on a computer, and if a simulated neuron tries to send a signal to an organic neuron, that organic neuron would have to be stimulated at the appropriate time. Or if a simulated hormone is drifting toward a portion of the brain not uploaded yet, then a hormone would have to be released into the corresponding organic portion of the brain. This would require a completely perfect brain-computer interface, perhaps an even greater technical feat than a brain emulation.
I think I have listed enough technical problems for now, some of which may prove, with increasing knowledge of the brain, to be less (or more) of a problem than I’ve made out. Regardless, mind uploading seems like a very very distant technology to me, and therefore I would rather focus on achieving the goals of longer-lived bodies and enhanced minds using genetic enhancements and primitive brain-computer interfaces, and focus on the political, ethical and philosophical dilemmas that arise from more near-term issues in human enhancement.