Done, your profile is created.Finish your profile by filling in the following fields
Forgot Password Earn Money,Free Notes
Password sent to your Email Id, Please Check your Mail
Updating Cart........ Please Wait........
and Cognitive Systems
Joanna J. Bryson
University of Bath, United KingdomFrom Last Week
Combinatorics is the problem, search is the only
The task of intelligence is to focus search.
Called bias (learning) or constraint (planning).
Most `intelligent’ behavior has no or little real-
time search (non-cognitive) (c.f. Brooks IJCAI91).
For artiﬁcial intelligence, most focus from design.
What kinds of parts does the system need?
How should those parts be put together?
How exactly is the whole thing arranged?
Like reactive planning, the term cognitive
architecture doesn’t quite mean what its
component words do.
People have been looking for a generic plan
for building “real” (human-like) AI.
This used to be a popular area of research,
now gets fewer publications.
Nevertheless, evolutionary history tells us
something about what worked & didn’t.What Worked
The past does not
necessarily predict the
future, particularly in
Changes in hardware
and other tech change
what is possible.Cognitive Architecture
Where do you put the cognition?
Really: How do you bias / constrain /
focus cognition (learning, search) so it
works?Basic Unit– Production
From sensing to action (c.f. Skinner;
conditioning; Witkowski 2007.)
These work basic component of
The problem is choice (search).
Require an arbitration mechanism.
Expert Systems: allow choice of
policies, e.g. recency, utility, random.
SOAR: problem spaces (from GPS),
impasses, chunk learning.
ACT-R: (Bayesian) utility, problem
spaces (reluctantly, from SOAR/GPS.)Expert Systems
Idea: Encode the knowledge of a
domain expert as productions, replace
them with AI.
Big hype in 1980s, do still exist e.g. for
checking circuit boards, credit / fraud
detection, device driver code.
Problem: Experts don’t know why they
do what they do, tend to report novice
knowledge (last explicit rules learned.)General Problem Solver
GPS, written by Newell, Shaw & Simon
(1959, CMU), ﬁrst program that separated
speciﬁc problem (coded as productions)
from reasoning system.
Cool early AI, but suffered from both
combinatorial explosion and the Markov
Soar was Newell’s next try.
operate on a
chunkSoar has serious
Contributing Soar Major Example Implementation
Ideas Version Results Systems
Goal Substate MOUTBOT
Decision Cycle Soar8 - 1999 SGIO
Dependency Coherence QuakeBot
Soar” is a
Improved TacAir-Soar TCL/Tk
Soar7 - 1996
Interfaces RWA-Soar Wrapper
Soar6 - 1992 C
paper (Laird &
Single State Soar5 - 1989
Soar4 - 1986
1996) – admits
Chunking Soar3 - 1984
Preferences Subgoals Soar2 - 1983
Subgoaling Dypar-Soar Lisp
Production Universal XAPS 2
Soar1 - 1982 Toy Tasks
Systems Weak Method Lisp
Symbol Heuristic Problem
Systems Search Spaces
← One problem: main ap / funding
is war games for US military.Architecture Lessons
An architecture needs:
action from perception, and
further structure to combat
Dealing with time is hard (Soar 5).
Learns (& executes)
For arbitration, relies
Call utility “implicit
Replicate lots of
See if the brain
does what you
think it needs to.
(from CMU Ψ)
Architectures need productions and
Real-time is hard.
Grounding in biology is good PR, may
be good science too.
Being easy to use can be a win.
Neural Arch.; Maes
from senses and
from goals through
net of actions.
action acts.Spreading Activation
brain-like (priming, action potential).
Still inﬂuential (Franklin & Baars 2010,
Can’t do full action selection:
Don’t scale; don’t converge on
comsumatory acts (Tyrrell 1993).Tyrrell’s Extended
Rosenblatt & Payton
Consider all information & all possible
actions at all times.
Favour consumatory actions by system
Also weight uncertainty (e.g. of memory,
temporal discounting).Tyrrell (1993)
= small negative activation
= zero activation
= small positive activation
= positive activation
in Den Clean
= large positive activation
For Mates Clean
R. Den this Sq
P. Den R. Den
P. Mate Rand. Dir All Dirs
Mate in Sq
Mate in Sq
N NE E SE S SW W NW
Extended Rosenblatt and Payton Free-Flow Hierarchy