COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	EPISTEMOLOGICAL PROBLEMS OF ARTIFICIAL INTELLIGENCE
C00006 ENDMK
C⊗;
EPISTEMOLOGICAL PROBLEMS OF ARTIFICIAL INTELLIGENCE

Making fully intelligent computer  programs will require solving  problems
hitherto only studied by philosophers, namely - what "partial" information
about  the  world  can  be  known  by  a  person  or  machine  with  given
opportunities to observe and how can it be expressed in a formal language.
It also requires establishing the general laws of the common sense  world,
representing them as computer data  structures, and writing programs  that
use them to solve problems.   The concreteness of artificial  intelligence
problems  nicely  complements   the  generality   hitherto  preferred   by
philosophers.

*********************

HOME COMPUTER TERMINALS

The real  "computer  revolution" hasn't  happened  yet, because  very  few
people use computers in  their daily non-professional  lives.  It is  just
becoming  cost-effective   to   use  computer   terminals   for   personal
correspondence, for receiving and paying bills, for reading anything  from
daily newspapers  to any  book  in the  Library  of Congress  to  anything
required to  be  public under  the  Freedom of  Information  Act.   Beyond
convenience, public  life will  be de-monopolized  by making  authors  and
journalists independent of  publishers and making  available an  effective
right of immediate reply to  the hit-and-run journalists and  politicians.
Some relevant experiments  and experience at  Stanford University will  be
described.

*********************

THE SCIENTIFIC BASIS FOR TECHNOLOGICAL OPTIMISM

Many people  now  believe  that  the high  American  material  and  energy
consumption cannot long continue.  The  problems of energy, food,  mineral
resources and pollution  are examined,  and we give  reasons for  optimism
that ten times the world's present population support itself at five times
the American consumption until the sun fails and the thoriam decays.  Some
gaps in the evidence are pointed out.