Re: Metarationality (was: JOIN: Alden Streeter)

From: Gordon Worley (
Date: Fri Aug 30 2002 - 07:00:49 MDT

I've been discussing rationality more with Eliezer the past few days.
He's not off his rocker, as I first though, just looking deeper than I
had been. Eliezer wants me to point out that I don't fully understand
and that anything I write on the topic of rationality does not
necessarily reflect his understanding, but I'm far enough along the path
that I can at least try to explain some things.

On Thursday, August 29, 2002, at 07:25 PM, Samantha Atkins wrote:

> Eliezer S. Yudkowsky wrote:
>> No, some sorts of creative achievement seem to go along with rampant
>> internal nondeliberation, perhaps even nondeliberative activities that
>> create warped deliberative thought processes as side effects. It
>> doesn't mean that the nonaccidental successes are rationally
>> incomprehensible - that it is not possible for a Bayesian observer to
>> understand how they work, and why they work, and conceivably employ a
>> process of that type deliberately in order to arrive at the same
>> answer.
> OK. That is better. Yes, sometimes some of it can be "stepped-down" to
> the level of a Bayesian observer. But I don't think all of it can and
> I do not believe the parts that can't are thereby without value.

`Stepped-down' is not really the right term. `Reduction' is more

All work is BPT-reducible (I'd say that it's all Tao-reducible, but I'll
avoid adding confusion for now). This does not mean, of course, that
every aspect of all work is BPT-reducible, only that all work depends on
the BPT. This is like the halting problem is computer science. Many
problems look solvable, until you discover that they reduce to the
halting problem (i.e. your problem would solve the halting problem if it
weren't impossible). In the same way, work may look accidental, but if
it can be reduced to the BPT (i.e. accomplishing the work at some point
required BPT modelable events), it is nonaccidental.

Irrational aspects of cognition are important to the way humans think.
Despite what some of you might think, neither Eliezer nor I nor anyone
else we know is fully rationally intelligent. Yet, human thought can be
reduced to the BPT. If you don't believe me, we can always go down to
the physical brain and look at neurons. Neurons function by a simple,
rational process to accomplish work. They receive electrical signals
and when the neuron has a specific charge, it fires. You don't even
need something as fancy as the BPT to understand the process that is
going on, though you can still use it. The charge level at which to
fire may be set by an irrational process as may be the reason a
particular neuron received signals, but as you can see rationality is
involved in all human cognition. It also exists at higher levels, but
this example illustrates how rational processes are fundamental to any
nonaccidental success.

It is acknowledged that irrationality plays a role in human cognition
and is very important for many discoveries. Yet without some degree of
rational intelligence, your mind would not be able to get anywhere on
purpose and no one would find your randomly firing mind very interesting
except maybe as a pretty house ornament.

Gordon Worley                     `When I use a word,' Humpty Dumpty            said, `it means just what I choose                it to mean--neither more nor less.'
PGP:  0xBBD3B003                                  --Lewis Carroll

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT