Re: META: List allowables & cults (was DirkJOIN)

From: Jeff Medina (
Date: Tue Mar 07 2006 - 12:49:26 MST

On 3/7/06, Eliezer S. Yudkowsky <> wrote:
> But if you're intelligent you also get the intrinsically valuable fun of
> figuring things out.

This fun is not intrinsic. I know many who find no joy in solving
problems. Not every person expending his or her time & money on booze,
sports, prostitutes, or what have you, instead of on intelligence and
knowledge improvement (books, conferences, schooling, etc.), does so
because they're less intelligent than those who find figuring things
out more fun.

> It is more fun to do things through your own wits
> than to do them via genie bottle.

Again, I know many who would *much* rather have a genie bottle.

> So the good of intelligence equals
> all the good it brings us, plus the good of using our intelligence to
> get it.

Given two people who equally enjoy the feeling of figuring things out,
one with twice the g of the other (g1 and g2, where g2 = 2g1), it is
plausible that Mr. g1 and Ms. g2 would have available to them the same
volume of Fun Space, though those volumes be differently located. Why?
Because there may be infinitely many problems of difficulty-g1 (such
that a g1 figure-outer would find them challenging in a fun way) and
infinitely many *other* problems of difficulty-g2 (ditto). This also
works for finitely-many-but-enough-such-that-you'd-never-run-out
quantities of challenging problems at a given difficulty level.

That such problems might require variations on similar themes, or even
the same problem with different content (e.g., x-2=5 vs. 2x-3=15), I
wouldn't be so sure this repetition would bore everyone the way I
expect you expect it would. How many games of Go must one play before
Go itself becomes boring? Crossword puzzles? Sudoku? There are many
intelligence-based challenge games that people seem to never grow
tired of. Maybe they would eventually -- I couldn't say -- but I
certainly suggest we don't just take it as an axiom that this or that
would be boring, and hence that only an increase in the complexity of
one's problem domain will bring one new fun.

Your (and my, and others') personal experience, becoming bored with
progressively more complex challenges, is duly noted. But that may
have been the result of our becoming more intelligent rather than
being a reflection of a truism about limited quantities of fun being
present in a given activity relative to a fixed intelligence.

> Furthermore, many intrinsic goods can only take place within a
> matrix of supporting intelligence. Compare chimpanzee Fun Space to
> human Fun Space. Many things we don't think of as stereotypically
> "intelligent" fun would be difficult to do without human intelligence.

And many things chimpanzees find Fun, we find boring. The rate of
funcrease (Fun-gained minus Fun-lost / delta-intelligence) may well be
positive, and perhaps even exponential. But without more data to
support this, equivalent Fun Space magnitudes across different
intelligences seems to remain possible.

Jeff Medina
Community Director
Singularity Institute for Artificial Intelligence
Relationships & Community Fellow
Institute for Ethics & Emerging Technologies
School of Philosophy, Birkbeck, University of London

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT