|
|
|
|
|
The
seminar is held in hybrid
format, in person (Múzeum
krt. 4/i Room 224) and
online at the following
link:
|
| 4
December (Friday) 4:15
PM Room 224 + ONLINE
|
|
Dániel
Kodaj
|
|
Department of General
Philosophy, Institute of
Philosophy
Eötvös Loránd University,
Budapest
|
| |
| AI
vs.
incompleteness/inexpressibility
|
Going back to John
Lucas and Roger Penrose, a number
of people have used Gödel’s first
incompleteness theorem to argue
that the mind is not a mechanism
and AI cognition is in some sense
inferior to human cognition. I
will argue that this demarcation
argument is extremely limited even
if it is sound; among other
things, it presupposes an older,
pre-LLM paradigm. On a more
constructive note, I will defend
two different but loosely related
cases against AI cognition. The
first appeals to Rice’s theorem
while the second is based on a
theorem about the inability of
formal languages to express both
their semantic and syntactic
properties.
|
|
|
|
|