First Person Singular

While reading this post, you should be listening to Let the Good Times Roll by Layo and Bushwacka!

I tend to think that the Singularity is merely the ne plus ultra of technotopianism, a trend which must be combated wherever it is found, but I avoid talking about it for fear of drawing down the wrath of Eliezer deliberately. However singularity thinking does force us to ask difficult questions about almost every aspect of human endeavour, and I like that. Probably a bit too much.

One such aspect is the question of what it really means to be human, intelligent and conscious, with the prospect of various combinations of biology, chemistry and engineering about to radically alter the way the human does business. You need to read a bit of background before the following will make any sense, and understand that one of the key tenets of singularity thinking is that - at some point in the near future, as computer processing power increases - it will be possible to upload a human mind to a computer.

This raises all sorts of questions, but in the course of arguing for the singularity on 3quarksdaily, Jesse M makes the following point:

Of course a simulated brain would require inputs like those from a real body or it would probably go crazy or become comatose, like a person locked in a sensory deprivation tank forever.

As I burnt my tongue on the pizza I’d just made, this set me thinking. What does it mean to be crazy or comatose? In essence it means that you involuntarily apply an abnormal filtering process on the inputs your brain receives. So being without inputs wouldn’t just be a cause of being crazy or comatose - it would be the result of being crazy or comatose as well. The causative chain doesn’t just go in one direction here, which makes it hard to disentangle which came first without being able to go inside and have a look - and that’s the one thing that currently we can’t do.¹

This made me think some more - what would it mean to have a brain without inputs? This is where you have to separate out mind from brain (sorry dualists): let’s say that you have a brain in a tank - how could you tell if there was anything going on inside that brain? From outside, how would you know a mind existed, if it wasn’t open to inputs against which you could test its responses as per the Turing test? From inside, how would you know you were a mind if you weren’t open to inputs against which you could define yourself (bearing in mind that even the Cogito requires external experiences, even if only to doubt them)?

And then I had the flash of insight which made me forget about my burnt tongue². A brain might exist in a purely physical sense, but without inputs it makes no sense to make the assertion that a mind exists. If there are no inputs, for all practical (and possibly for all philosophical) purposes there is no mind. It’s therefore entirely reasonable to assert that the mind exists only insofar as it interacts with an external world.

If this is the case, is it sensible to draw a hard line between our minds and that external world? (No.) Does the external world need us as much as we need it? (Possibly.) Does my tongue still hurt? (Yes, but I’ll be alright.) In the context of the ongoing singularity conversation, if you upload your mind to a machine, and then close that machine off to all inputs, what are the implications for the continued existence of that mind? Perhaps the Singularity doesn’t provide all the answers to the question of mortality after all.

The good news is that this does help to confirm my own prejudices that (a) The internet already constitutes a machine intelligence - just as these machines are becoming an extension (extrusion?) of our intelligence, we’re an extension of their intelligence - and (b) God is an emergent property of a universe with intelligence in it, in a similar way as consciousness is an emergent property of a body with a mind in it.³

1. Of course with a simulated brain it may well be possible to go inside and have a look (although I tend to doubt it, for reasons which are irrelevant to this discussion but have to do with levels of complexity). Interesting ethical question - should it be considered torture to deliberately inflict this type of condition on a conscious entity, even if it’s artificial? I strongly suspect that this may form part of the case that Eliezer makes in his AI-Box Experiment. [↩]
2. I’m sure there’s some lesson in there as well, incidentally [↩]
3. Note to self: this also creates problems for God, but he’ll have to take care of that on his own. Literally. [↩]