|
|
|
|
| 5
March (Friday) 4:15 PM
ONLINE |
| György
Turán |
Department of
Mathematics, Statistics,
and Computer Science
University of Illinois
Chicago
|
| |
Explanation
from Specification
|
Machine learning is
often required to produce
comprehensible results for the
user, but recent methods, in
particular deep learning, are
problematic from this point of
view. Explainable AI (XAI) is the
subfield of AI trying to deal with
this issue. We give an overview of
the area and discuss the approach
of explanation from specification,
where the type of explanation
required is specified by the user.
We also discuss connections to a
notion of explanation in the
philosophy of science. (Joint work
with Harish Naik.)
|
|
| 12
March (Friday) 4:15
PM ONLINE |
Zhiwei
Gu
|
|
Department
of Philosophy,
Central European
University, Vienna
Institute of
Philosophy, Eötvös
University, Budapest
|
| |
Anomalous
Disjunctivism
|
The causal argument
from hallucination leads to the
screening-off problem, namely
that the non-relational account
of hallucinatory experience will
hijack the naďve realist account
of perceptual experience,
rendering the ordinary objects
and properties explanatorily
redundant. To address the
screening-off problem,
disjunctivists usually deny that
hallucinatory experience and
perceptual experience are of the
same fundamental psychological
kind (the common-kind thesis),
and further deny the generative
theory, which claims that the
underlying neurophysiological
state is sufficient to bring
about the alleged experience.
I offer a new disjunctivist
solution—anomalous
disjunctivism—to the problem. I
show that anomalous
disjunctivism is compatible with
the generative theory and the
common-kind thesis. According to
anomalous disjunctivism,
perceptual experience and
hallucinatory experience are not
produced in the same way: the
presence of the seen thing and
the failure of seeing play
essential conditions for causing
seeing and hallucination. I
provide several analogies to
show how the external causal
conditions make a difference on
the internal mechanisms
underlying perception and
hallucination. Thanks to the
difference of the essential
causal conditions, whatever
account of hallucination need
not apply to perception. Thus,
the screening-off problem will
not arise.
|
|
| 26
March (Friday) 4:15 PM
ONLINE |
| András
Jánossy |
John Wesley
College, Budapest
|
| |
Four
Worlds' Citizen
|
In this talk, I will
revisit Karl Popper's theory of of
three ontological domains or
worlds - first put forward in his
Objective Knowledge (1972), then
explicated in more detail in his
Three Worlds lecture (delivered in
1978) - and while sustaining my
sympathy towards his concerns, I
will employ ontological
distinctions which much differ
from his.
My main tenets:
Popper rejects the well-known
philosophical point of view -
called 'monism' in his terminology
- which considers only individual
material objects existent, and
maintains that any other objects
which seem to have different
nature are nothing else but
compositions of such objects, or
interactions of such objects. I
prefer to call this philosophical
position 'reductionism'.
Popper admits that if we so wish,
we can subdivide the world of
material objects into the world of
non-living physical objects and
into the world of living things,
of biological objects; but does
not consider this distinction
important. For me, this
distinction is sharp because some
reductionists postulate or presume
that the existence and operations
of living things can be deducted
from the existence and workings of
non-living physical objects.
Popper considers the realm of
(animal or human) mental states a
world on its own - whereas I
believe that it can be assimilated
into the world of biological
objects and phenomena. On the
other hand, I believe that the
realm of (animal or human) social
phenomena - not studied separately
by Popper - does form world on its
own, and its objects cannot be
deducted from individual
biological objects.
Popper puts the contents of human
thoughts - for example, theories
developed by scientists - into a
world own its own in which the
world of the products of the human
mind, or in other words: 'abstract
objects' dwell. I will argue that
abstract objects - at least the
ones truly considered so - exist
independently from human discovery
or creation.
I also claim that cognition of
objects of these four worlds
requires different capabilities;
and that man is well equipped with
all four. That is, he/she walks
homely and with confidence in each
and every of the four worlds.
Nevertheless, it is false to
consider him/her as being in all
four places at the same time.
|
|
|
|
|
