Re: Feasibility: 100% Bayesian systems

From: Nick Hay (
Date: Mon Oct 06 2003 - 00:04:00 MDT

Metaqualia wrote:
> Would the set of all bayesian probability data necessary to replicate a
> 100IQ human be prohibitively large and therefore impossible to store on a
> modern supercomputer? (exclude the visual cortex, just think common sense
> reorganized in a bayesian fashion)
> Would the lookup time for a moderately complex thought be too long in such
> system? (imagine that all probabilities are stored on a hard disk).

My impression is yes. Without simplifying assumptions like using cognition to
understand and manipulate reality (ie. cognition as being equivalent to
approximating so intractable Bayesian calculation), explicit Bayesian
calculations will be far too expensive.

> Programming tricks, compression, anything goes to reduce the size - but
> after nothing more can be done and it's just a matter of rough storage
> space and speed, are 100% bayesian human equivalent AIs theoretically
> possible or impossible to implement with present off the shelves
> technology?

The programming tricks necessary involve creating a mind.

> Is there any use for alternatives to bayes a cognition paradigms?

The use of Bayesian probability is it's a model of truth and rationality. One
that seems to cover everything (see Jaynes for starters). As such, any mind
that actually seeks truth is in fact approximating some Bayesian calculation.
I believe you can get stronger statements even that this. It does not mean
you explicitly calculate Bayes theorem, or run through the mathmatical
calculations statisticians typically use, or anything that looks like maths
at all. It's more of a framework.

- Nick

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT