COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.require "let.pub" source
C00005 ENDMK
C⊗;
.require "let.pub" source
∂AIL Harry Shershow↓Editor, Computer-Chess Department↓Personal Computing
↓1050 Commonwealth Avenue↓Boston, MA 02215∞
.<<(617)232-5470: subscription costs $14, 26, 38>>

Dear Mr. Shershow:

	The two photostats you sent me don't have the quotation from
Dondis reprinted in I.J. Good's reply.  Good is certainly right that
for a computer to move the pieces in its memory is no more cheating
than for a human to move them it his memory.

	As to whether computers should be barred from tournaments, it
seems to me that this is a matter of expediency.  When the human
chess players no longer find it interesting to play against machines,
then they should be barred.  Neither the programs themselves nor
the programmers have any special rights in the matter.  It seems to
me that when machines win most of the time, the chess players will
want to bar them or at least restrict them.  It would be interesting
to try to formulate a restriction on the amount of computing the
program is allowed to do to decide on a move, but it wouldn't be easy
in the present state of computer science.  Artificial intelligence
would benefit from the ability to compare how clever a program is
apart from how fast it is.

	I have no commercial interest in Kalah, and I repeat that
it is not a computer game per se.  I don't even know if it is still
being marketed.  Our programs were written in the early 60s for the
now defunct PDP-1 computer in machine language, and I can't readily
lay my hands on them.  Enclosed are some copies of some memos about
the program.

.sgn
.<<Please find early AI memos about Kalah and enclose copies with SHERSH.LE0↓_sp4(