Tuesday, November 18, 2008

things humans have trouble dealing with

The human mind has serious trouble dealing with a couple of inherently difficult things:
  • the exponential function.
  • higher spatial dimensions.
  • self-awareness or self-reference.
  • ugly infinity (the non-countable kind).
Does this have anything to do with the cost of emulating those specific algorithms?
Is anything that cant be run in O(n) time on the brain inherently tough?
I guess if we reduce the domain enough, a lot of things end up being run with O(n) time on the brain... at least with enough practice...
Anyone else want to point another shortcoming? Or any another idea regarding the issue?

6 comments:

Mario Bergés González said...

Probability theory?

hobbes' consultant said...

I thought about it, but wasn't sure specifically what part (or parts) to choose? What about other key concepts, not exactly fields of study..

MartianEgg said...

Metacognition?

hobbes' consultant said...

Maybe self-reference is the dificulty behind Metacognition, Metaprogramming, and generaly most meta-models. Although abstract algabra and group theory have eased us through some of the parts; self-reference is still intrinsically dificult for us... (i think)

marydaly said...

I'd say the acknowledgement of a complete stranger's egocentric reality. We write him off as “people”, part of an abstract concept of human equality and kinship we could care less about. Until he becomes friend and suddenly he’s not an object anymore. It blows my mind to sit and try perceiving reality through that other person, a culmination of moments that got her here to see me as I see her, an object, a prop in my reality.

hobbes' consultant said...

md:
that gets me to thinking about classifiablility...

what if one accepts a neural network (or hierarchical neural network) interpretation of brain processes, then things with an overabundance of salient qualities is hard to manage.

still, of the other phenomena, the exponential function is arguably quite classifiable... hmmm, this gets me onto formal language translation and the amount of neurons required to create homomorphisms between these abstractions and the brain...

?