Image Hosted by  

Post-Brunch Intelligencer

Midmorning ramblings on the state of the species

The Last Question

Posted by Nath at 10:03 PM
Like most people, I sometimes wonder about the so-called Big Questions. Why are we here? Why do some people eat marmite? Those are reasonable questions to ask, but I think it's more interesting to think about how we're here, rather than why. The concept of 'why' doesn't really mean anything when applied to natural phenomena.

Of course, we do not completely understand the mechanisms that caused us to exist. For instance, we don't know what species lie on the evolutionary chain between human beings and the early hominids. However, we've found several possibilities consistent with what we know about evolution. Like in maths, knowing that a solution exists is often more interesting than knowing precisely what the solution is. The existence of a solution does not prove that our understanding of the universe is correct, but at least tells us that our understanding is consistent with reality.

In this limited sense, at least, I know the answers to the Big Questions. My understanding of the universe is consistent with what I know of reality; in other words, a solution exists. The specifics of the solution will change as I find out more about the universe, but the fundamentals are pretty stable. I haven't had to change them much, because they've been consistent with reality in every way.


There's one thing I simply don't understand, and haven't been able to fit into what I understand about reality. It's kind of a vague problem, so I'm not sure that I can communicate it clearly, but here goes.

Let's think of the human nervous system as a computing machine, for a moment. It takes input, processes it, and gives output. These machines emerged through millions of years of natural selection. Our consciousness is encoded in the state of these computers. All our thoughts are bits in memory. This is all well and good, and I can imagine a universe where this would happen.

The problem is that I can observe this happening. Right now, I can observe part of the state of my brain. I can observe the recent input of my senses, and some previous experiences recorded in my memory. I can observe part of the state of the computer called 'Nath' at a particular point in time. Why can I observe this particular computer, at this particular time? Who am I, the observer? I can understand that the computer called 'Nath' might incorrectly believe that there is an observer; the observer might be a useful abstraction created by this computer. 'Nath' might behave identically whether or not there really is such an observer. But I, the observer, know that I exist; I cannot a figment of the computer's imagination, because I observe myself observing the computer. (Though I suppose the computer would be thinking these thoughts and typing these words whether or not this was the case.)

This is what I cannot account for within my current model of how the universe works; I do not know how there is an observer.


Posted by Blogger Revealed at 30 October, 2007 00:53:
Loved it. Maybe though the computer imagines that it can see itself. Maybe the computer imagines another itself who can then observe it, thereby providing a sense of security in existing. Cos if the computer didn't have anyone knowing it existed, it might not exist. No?

Posted by Blogger Nath at 31 October, 2007 04:16:
I don't know. It is possible that the computer imagines an observer, but the observer would just be data encoded in neurons. I could see how imaginary observers might be useful, but I know that there exists at least one non-imaginary observer. It observes, therefore it is.

Or something.

Posted by Blogger Revealed at 12 November, 2007 18:31:
Shady territory. If it observed only because it was a projection of me, it would still exist but its existence would directly depend on my existence. Which means that it doesn't technically have an existence of its own. Does it exist?

Posted by Blogger Nath at 13 November, 2007 01:57:
Probably not; that's why I think there'd need to be at least one observer that wasn't a projection of a computer.

Then again, conversations about things like this are sufficiently hand-wavey that I'm not entirely sure what I'm disagreeing with.

Posted by Blogger Revealed at 13 November, 2007 22:25:
Hahahah. Yeah, I know what you mean. Halfway through typing out the comment I was like waitaminute, what am I saying again?

(Psst: What are you studying?)

Posted by Blogger Nath at 14 November, 2007 19:58:
I'm in the giant evil robot business.

(Actually, I teach computers to see patterns in large amounts of data. But that doesn't sound as interesting.)

Posted by Blogger Revealed at 14 November, 2007 21:49:
Yeah, go with the giant evil robot thing. It'll make people ask you your opinion on I, Robot. Such fun, no?

Posted by Anonymous Mustafa Fanaswala at 07 May, 2008 20:22:
Hey Aniruddh!

This is bizarre but I just came across your comment on my ages old blog which I'd abandoned to the powers that be but by some freak chance I just happened to visit it today. Anyhow, yes I'm the same Mustafa at ADIS. I believe we went to 10th grade together right? I see that your posts have a lot of machine learning flavour. Coincidentally, I'm into the same field albeit from a more statistical pattern recognition viewpoint. I added you on facebook...


Powered by 


Something broken? Let us know.