AI from the Perspective of Cognitive Science

topsalmonΤεχνίτη Νοημοσύνη και Ρομποτική

23 Φεβ 2014 (πριν από 3 χρόνια και 5 μήνες)

829 εμφανίσεις

AI from the Perspective of Cognitive Science

Study Questions for Meeting #3

Cognitive Science and AI


Stillings et al., Chapter 1 (“What is Cognitive Science”)


1.

At the beginning of section 1.2, the authors say that understanding cognitive
science requires
one to understand “what it means to say that something is a
computational system” (p. 2). After reading the chapter, can you answer this
question (say what it means)?

2.

Some key concepts to look out for in section 1.2 (what do they mean?):

a.

Cognitive informa
tion processing has
semantic content
, is
intentional

(
about

the world), is
representational

(these concepts are closely
related).

b.

Cognitive representational schemes have
syntactic structure
, are
combinatorial, generative, productive

(these concepts are syn
onyms),
and have
compositional semantics
.

c.

Cognitive processes are
algorithmic
. and algorithms are
formal

procedures.

3.

What is the relationship between these concepts and the
physical symbol system
hypothesis

of cognition (Chapter 2.3)?

4.

What are the three ma
in levels at which cognitive information processing can be
analyzed (section 1.3)? Why is each level important? Which level seems
most/least important?

5.

What are the five core disciplines of Cognitive Science? What aspects do you
think each discipline might

contribute? What is the role of computers and
Computer Science in Cognitive Science?

Stillings et al., Chapter 2.3 (“A Global View of the Cognitive Architecture”)


You can focus on pages 21


25.


1.

How much of the “cognitive architecture” depicted in Figur
e 2.1 do you think is
mere “intellectual convenience” and how much is “real.” What does it mean to be
“real”?

2.

What is the relation between a Turing machine, a Universal Turing machine, the
Von Neumann architecture, the “classical view” of the cognitive arc
hitecture, and
the physical symbol system hypothesis? In what ways does the human physical
symbol system (with its cognitive architecture) clearly differ from the von
Neumann architecture (i.e., standard digital computer)?

3.

What classic model of memory seem
s to underlie the first full paragraph on page 24?

4.

Does the discussion of
structure
-
based computation

and
productivity

in this
section help you understand the earlier discussion of these concepts in Chapter 1?
What is the meaning of the additional concept,

systematicity
?

Rapaport, “Cognitive Science” (offprint)


1.

What are the different views presented about the nature of Cognitive Science as a
research discipline?

2.

What is the distinction between the
strong

and
weak

claims about the relationship
between cogni
tion and computation? (cf. strong and weak AI in G&N)

3.

What are some of the advantages of testing cognitive theories by writing and
testing computer programs?

4.

What is the Representational Theory of Mind (RTM)? What is its relationship to
the Physical Symbol

System Hypothesis (PSSH)?

5.

How do the PSSH and RTM “
re
solve” the mind
-
body issue?

6.

What is the difference between a system that
has

representations, beliefs, and
intentions, and one which can only be conveniently
described as if

it has such
entities (cf. q
uotes by Winograd and Dennett on p. 4).

7.

Which assumptions of the symbolic approach to Cognitive Science (and AI) are
shared by the connectionist approach, and which are contested? What are some
ways of reconciling the two approaches?

8.

Look at the examples
that are listed in section 4.4 of work in Cognitive
Psychology that has been influential on Cognitive Science. Why do you think
these specific examples were chosen?

9.

What are the two major open issues for the future of Cognitive Science that are
presented i
n section 5?

Goldsmith & Nevo (Sections 13.3


13.4)

See study questions embedded in the chapter.