Memo from To:

JOSHUA LEDERBERG Dr P Naegeli
I

FEB 17 1974
res "Artificial Intelligence" -
Thank you for sending me this note, which

echoes many of my own concerns. about expecta-
tions from blatantly a-theoretical methods of
pattern analysis. Many contemporary efforts in
computerized medical diagnosis-are equally
liable to grave faults from uncritical excess
of enthusiasm,

However, I hope you might still be interes-
ted in totally different approaches that are
embraced by the phrase "artificial intelligence"
(a cacophemism that I would happily replace")
as illustrated in the enclosed reprint,

I wonder what is the significance of a

. Predictor of parity that scores 6/6? It may of
course just be a statistical fluke (26 is only
64); or there may be some other "epincidence"
in the relationship of names to structures
that would be worth knowing --but not very much.
(consider that of the alkanes from Cc, to Clos
80% have the opposite parity from the composition),

Jrcer/y
~ G-¢ .
os arlAna “ta
~ FRoresson Josnua Leverserc :

\

t
spartment of Genctics

School of Medicine

Stanford Universit t

Stanford, California 94305