From: Elaine and Andy (brannen@lookingglass.net)
Date: Sun Jan 04 2004 - 16:44:36 MST
Ben G:
<comment snipped to:>
For some of us, expressing thoughts in mathematics is habitual, even in
cases where the tools of math aren't useful for proving anything.... I
assume everyone on this list is skeptic enough to understand that ...
Me:
Assumption noted.
Ben G:
I do believe, however, that for superhuman AI's in the future, mathematics
WILL be a useful tool for moral philosophy. It's rarely useful for humans
in this context, simply because our ability to make precise calculations
regarding moral situations is generally very poor (usually due to data
gathering issues more than calculation-related issues).
Me:
Given the financial fraud based on "Generally Accepted Accounting
Principles" both in the US and Europe and the fact this was done
using arithmetic ... I hereby invoke previous assumption. ;-)
Also I would say the "data gathering issues" may prove
exceedingly difficult. And even when one can gather the data,
mapping to mathematical structures is going to be another
challenge. One of the distinquishing characteristics
of qualitative data is its resistence to numerical analysis.
I think we can both agree that morality is "qualitative" with
a vengence.
Ben G:
In fact, I can easily envision a future argument between super-smart AI's
regarding the optimal value of some parameter much like my "c" parameter....
Perhaps the outcome of the argument would be an adjustment to the amount of
time the human population gets to spend hooked up to the AI-created
Orgasmotron ;-)
Me:
As I am rapidly approaching the age wherein such things do not
have the same *uh* 'urgency' I will leave that to those
having a more *ah* 'interest' in the calculation. :-)
Andy
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:43 MDT